|
Citation |
- Permanent Link:
- http://ufdc.ufl.edu/AA00054911/00001
Material Information
- Title:
- Dendritic cell development and distribution in Type I diabetes overcoming genetic defects and establishing tolerance via pharmacotherapeutic manipulation of dendritic cell populations in vivo
- Alternate title:
- Overcoming genetic defects and establishing tolerance via pharmacotherapeutic manipulation of dendritic cell populations in vivo
- Creator:
- Bahjat, Keith S
- Publisher:
- University of Florida
- Publication Date:
- 2001
- Language:
- English
- Physical Description:
- viii, 91 leaves : ill. ; 29 cm.
Subjects
- Subjects / Keywords:
- Antigens ( jstor )
Cells ( jstor ) Cultured cells ( jstor ) Cytokines ( jstor ) Dendritic cells ( jstor ) Diabetes ( jstor ) Immatures ( jstor ) Lymph nodes ( jstor ) Molecules ( jstor ) Type 1 diabetes mellitus ( jstor ) Dendritic Cells -- genetics ( mesh ) Dendritic Cells -- immunology ( mesh ) Dendritic Cells -- pathology ( mesh ) Department of Pathology, Immunology, and Laboratory Medicine thesis, Ph.D. ( mesh ) Diabetes Mellitus, Type I -- etiology ( mesh ) Diabetes Mellitus, Type I -- genetics ( mesh ) Islets of Langerhans ( mesh ) Killer Cells, Natural ( mesh ) Mice, Inbred NOD ( mesh ) Research ( mesh ) ZZ : Dissertations, Academic -- College of Medicine -- Department of Pathology, Immunology, and Laboratory Medicine -- UF ( mesh )
Notes
- Thesis:
- Thesis (Ph.D)--University of Florida, 2001.
- Bibliography:
- Bibliography: leaves 79-90.
- General Note:
- Typescript.
- General Note:
- Vita.
- Statement of Responsibility:
- by Keith S. Bahjat.
Record Information
- Source Institution:
- University of Florida
- Holding Location:
- University of Florida
- Rights Management:
- The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
- Resource Identifier:
- 027101114 ( ALEPH )
70334040 ( OCLC )
Aggregation Information
- UFIR:
- Institutional Repository at the University of Florida (IR@UF)
- UFETD:
- University of Florida Theses & Dissertations
- IUF:
- University of Florida
|
Downloads |
This item has the following downloads:
|
Full Text |
DENDRITIC CELL DEVELOPMENT AND DISTRIBUTION IN TYPE I DIABETES: OVERCOMING GENETIC DEFECTS AND ESTABLISHING TOLERANCE VIA PHARMACOTHERAPEUTIC MANIPULATION OF DENDRITIC CELL POPULATIONS IN VIVO
By
KEITH S. BAHJAT
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
2001
ACKNOWLEDGMENTS
I would first like to thank my mentor, Dr. Michael Clare-Salzer, for his encouragement, thoughtful discussions, and professional guidance over the past 4 years. Through the peaks and valleys of my graduate career, he has always been there to encourage me to continue, to lead me to the next critical experiment when I may get off the path, and to offer personal guidance for dealing with the various curveballs life may throw my way. I know I would not be in the position I'm in today without his incredible insight and professionalism, and I will be grateful for the remainder of my scientific career becasue he taught me so much in a short amount of time.
I thank the members of my committee, Drs. Mark Atkinson, Raul Braylan, Peter J. Hansen, and Eric Sobel for their time and insightful comments during the last 4 years.
Of course, I'd also like to thank my wife, Rena, who has put up with me moving all around the country over the last 8 years in search of professional satisfaction. Her support and helpful scientific discussions over the dinner table have been key in my progression through graduate school. She encouraged me when nothing seemed to be working correctly.
No graduate career could succeed without knowledgeable and helpful peers to get you through the rough spots, and mine has been no different. I extend my gratitude to Donna Whittaker, who was my lab buddy and spiritual guide for the first 3 years of graduate school. Donna traversed the rough road of graduate school with uncommon ease, juggling both family and military responsibilities. Any time I wanted to complain about having too much work, I simply had to look across the lab and realize that I didn't have it so bad. I know I wouldn't have made it this far without Donna there to buffer things for me those first 3 years.
ii
iii
I also thank Dr. Parker Small, who discussed science, graduate school, and life with me on a daily basis. His daily doses of sanity helped sequester the chaos around me and gave me the strength to push through to the end.
I'd like to thank Dr. Peter Kima for his insightful discussions, encouraging
remarks, and professional guidance over the last part of my graduate career. He gave me confidence that I could succeed as a scientist, and he generated insightful discussions on science outside my tiny area of specialization.
I also thank Sara Eckenrode, who spent much time teaching me all the molecular techniques I never took the time to learn.
I thank everyone who worked in Dr. Clare-Salzler's lab during my time here, including Jessica Elf, Sally Litherland, Yiyu Li, Vikas Dharnidharka, Yancy VanPatten, and Kristie Grebe.
TABLE OF CONTENTS
ACKNOW LEDGM ENTS.................................................................................................. ii
ABSTRACT ...................................................................................................................... vii
INTRODUCTION ..............................................................................................................1
REVIEW OF LITERATURE...............................................................................................5
Dendritic Cell Biology ....................................................................................................5
M yeloid Dendritic Cells..............................................................................................5
Lymphoid Dendritic Cells...........................................................................................9
Central Tolerance: The Role of Dendritic Cells............................................................10
Peripheral Tolerance: Myeloid and Lymphoid Dendritic Cell Interactions..................11
Regulatory Lymphocyte Populations............................................................................12
Diabetes Onset and Progression in the NOD M ouse....................................................13
Defects in Pancreatic (3 Cells....................................................................................14
Contributions of M HC Class II.................................................................................14
Defects in Activation Induced Cell Death.................................................................15
IDENTIFICATION AND CHARACTERIZATION OF MURINE DENDRITIC CELL POPULATIONS.................................................................................................................17
Background....................................................................................................................17
M aterials and M ethods..................................................................................................18
M ice ...........................................................................................................................18
Creation of Single-Celled Suspension from Lymph Node .......................................18
Analysis of Surface Phenotype.................................................................................19
Results............................................................................................................................20
Ex Vivo characterization of Dendritic Cell Populations............................................20
Five Distinct Dendritic Cell Populations in Ex Vivo Preparations............................20
Variations in Dendritic Cell Populations Between Anatomic Sites..........................26
Variations in Dendritic Cell Populations by Strain...................................................27
Summ ary and Conclusions ...........................................................................................30
DYSFUNCTIONAL DEVELOPMENT AND MATURATION OF MYELOID DENDRITIC CELLS FROM NOD M ICE........................................................................35
Background ....................................................................................................................35
iv
V
M aterials and M ethods..................................................................................................37
M ice ...........................................................................................................................37
Generation of Myeloid Dendritic Cells.....................................................................37
Analysis of Surface Phenotype.................................................................................38
Purification of Hematopoetic Stem Cells..................................................................39
Purification of CD4+ T cells.....................................................................................39
Proliferation Assessment with CFSE........................................................................39
Uptake of Fluorescent Labeled Proteins or Particles................................................40
ELISA for Cytokine in Culture Supernatants...........................................................40
R esu lts............................................................................................................................40
Identification of Protective Dendritic Cell Populations ............................................40
R evisiting an O ld Friend.......................................................................................40
Metrizamide Gradients Isolate Mature Dendritic Cells.........................................42
NOD Myeloid Dendritic Cells Do Not Develop, Mature, or Function Properly.......42
Atypical Development of Myeloid Dendritic Cells from NOD Bone Marrow
C ultures..................................................................................................................43
NOD Dendritic Cells Do Not Respond To Maturation Stimuli............................44
MHC Instability Does Not Affect NOD Dendritic Cell Phenotype......................44
NOD Dendritic Cells With Immature Phenotype Take Up Particulate Antigen .... 45 Cytokine Dosage Does Not Affect NOD Dendritic Cell Development................47
NOD Dendritic Cells are Poor Stimulators of Allogeneic T Cells .......................47
Investigating Causes of NOD Maturation Abnormalities .........................................47
Microscopy of NOD Dendritic Cell Populations..................................................47
Soluble Factors Do Not Direct NOD Dendritic Cell Development......................49
T, B, and NK-T Cells Do Not Affect NOD Dendritic Cell Development..............50
A Non-lymphocytic Cell Population Leads to Generation of Pseudo-Mature Dendritic C ells............................................................................................................. 50
NOD Dendritic Cells Become Arrested During Maturation.....................................51
Genes Within IddlO Affect NOD Dendritic Cell Maturation...................................51
Screening of NOD Congenic Strains ....................................................................51
IddlO Affects Dendritic Cell Maturation ..............................................................52
IddlO Affects Dendritic Cell Cytokine Production Capacity ...............................52
Sum m ary and D iscussion............................................................................................ 53
a-GALACTOSYLCERAMIDE ACTIVATES AND EXPANDS NK-T CELLS AND RECRUITS MYELOID DENDRITIC CELLS SPECIFICALLY WITHIN THE PANCREATIC DRAINING NODE OF NOD MICE..................................................... 57
B ackground .................................................................................................................. 57
M aterials and M ethods................................................................................................. 59
M ice .......................................................................................................................... 59
Treatment with a-Galactosylceramide ................................................................... 59
a-Galactosylceramide In Vitro Recall...................................................................... 59
ELISA for Cytokine in Culture Supernatants......................................................... 60
Analysis of Surface Phenotype................................................................................ 60
R esults ........................................................................................................................... 6 1
vi
Treatment of NOD Mice with a-Galactosylceramide Expands NK-T Cells Specifically W ithin the Pancreatic Draining Node.............................................................. 61
Increased Functional Responses Ex Vivo Following a-Galactosylceramide Treatment ............................................................................................................................ .......6 2
Increased Numbers of Myeloid Dendritic Cells in the Pancreatic Draining Node
After a-Galactosylceramide Treatment .................................................................. 64
Summary and Discussion............................................................................................ 65
SUM M ARY AND CONCLUSIONS............................................................................... 72
Hypothesis of Dendritic Cell Polarization ................................................................... 75
Summation .....................................................................................................................77
REFERENCES ..................................................................................................................79
BIOGRAPHICAL SKETCH ............................................................................................91
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
DENDRITIC CELL DEVELOPMENT AND DISTRIBUTION IN TYPE I DIABETES:
OVERCOMING GENETIC DEFECTS AND ESTABLISHING TOLERANCE VIA
PHARMACOTHERAPEUTIC MANIPULATION OF DENDRITIC CELL POPULATIONS IN VIVO
By
Keith S. Bahjat
August 2001
Chairman: Michael J. Clare-Salzler, M.D. Major Department: Pathology, Immunology, and Laboratory Medicine
As the most potent antigen-presenting cell of the immune system, the dendritic cell has the unique capacity to both initiate and regulate antigen-specific immune responses. In the normally functioning immune system, the combination of costimulatory molecules and cytokines presented with MHC-peptide complexes on dendritic cells determine the penultimate function of naive antigen-specific T cells. Thus, scientist's have great interest in the function of dendritic cells in autoimmune disease, where self-reactive T cells are activated in the periphery and regulation of this response is deficient. This study demonstrates abnormalities of dendritic cell subsets within secondary lymphoid tissues of the autoimmune-prone NOD mouse and demonstrates a genetic basis for abnormalities of these cells in vivo. Studies where hematopoeitic progenitors were cultured with GM-CSF and IL-4 showed that NOD stem cells have a limited capacity to differentiate to fully mature dendritic cells. Analysis of several congenic mouse strains have identified this as a genetically controlled phenomenon, and further
vii
Viii
indicates potential in vivo differentiation and functional abnormalities that may lead to a dysfunctional immune system. Finally, activation of a lymphoid cell population known to interact with dendritic cells, the NK-T cell, is capable of limiting pancreatic beta cell destruction, apparently by initiating the emigration of dendritic cells from the pancreas to the draining lymph node, thus decreasing islet inflammation. Together, these findings establish a crucial role for abnormal dendritic cell development in the initiation and resolution of beta cell destruction in Type I diabetes.
INTRODUCTION
Dendritic cells are the most potent antigen presenting cells known, with a 10-100 fold greater capacity to stimulate naive T cells than other antigen-presenting cells (Steinman et al., 1997a). The dendritic cell's overall potency is dependent upon high levels of MHC-peptide complexes, membrane bound costimulatory molecules such as CD80 and CD86, and production of soluble cytokines such as IL-12 and IL-10 (Banchereau and Steinman, 1998).
All but the latest work in dendritic cell biology has held that two dendritic cell states exist, the immature dendritic cell and the mature dendritic cell (Banchereau and Steinman, 1998). It is now apparent that the dendritic cell goes through defined phases while maturing, with dendritic cells at each phase stimulating a different T cell outcome than that seen when the dendritic cell is fully mature (Langenkamp, 2000). Variations in maturation stimuli may modify the developmental pathway of these dendritic cells. The capacity to drive a naive T cell to activation and proliferation is related to the quantity of MHC-peptide complexes on the surface of the cell, in combination with T cell costimulatory molecules such as CD86 (Shortman and Maraskovsky, 1998). While the immature dendritic cell has low levels of these molecules, and thus has little effect on naive T cells, maturing and fully mature dendritic cells constitutively express high levels and thus are potent stimulators of naive T cells.
The dendritic cell produces two main cytokines that are involved in the polarization of T cell responses: IL-12 and IL-10 (Banchereau and Steinman, 1998; Kronin et al., 2000a; Shortman and Maraskovsky, 1998; Whittaker et al., 2000). What determines which of these cytokines is produced by the dendritic cell is a function of its maturation state upon encounter with the T cell (Langenkamp et al., 2000). As the dendritic cell begins
1
2
maturation, it enters a phase of high IL- 12p70 production. A T cell interacting with the dendritic cell at this point results in those responding T cells assuming a Thl profile, producing cytokines such as IFN-y which activate components of the cell-mediated immune response. But if the dendritic cell fully traverses the maturing phase to the fully mature state, the dendritic cell will secrete large amounts of IL-10, resulting in a Th2 profile for responding T cells, producing cytokines such as IL-4 and driving humoral type immune responses (Langenkamp, 2000) (Perez et al., 1995). Factors such as PGE2 appear to move the dendritic cell toward an IL-10 producing state, leading to the priming of Th2 type responses. It is not clear if the dendritic cell moves more rapidly through its IL- 12 phase, or if this is bypassed altogether (Kalinski et al., 1997). These two functional T cell subsets are frequently used to classify responses to pathogens, as well as diseases of immune dysfunction. Type I diabetes has been frequently characterized as a disease where Th I cells promote diabetes, while Th2 type cells prevent disease (Bradley et al., 1999).
Several pieces of evidence indicate a role for dendritic cells in the pathogenesis of Type I diabetes, the most obvious being their ability to stimulate and polarize T cell responses. Key experiments detailing the location and functional capacity of these cells during the progression of disease support a role for these cells in initiation of and protection from autoimmune-mediated destruction of pancreatic beta cells.
Dendritic cells are the first leukocytes infiltrating the pancreas of both the NOD mouse and the diabetes-prone BB rat (Delemarre et al., 1999; Rosmalen et al., 2000a). These cells accumulate within the pancreas during the course of disease, forming reticular networks with infiltrating T cells, and are presumed to locally prime autoreactive T cells (Jansen et al., 1994; Ludewig et al., 1998). The accumulation of antigen-presenting cells and T cells and the formation of lymphoid structures within the islets of Langerhans suggests there are factors which inhibit the resolution of inflammation in this tissue. The retention of dendritic cells within the islet would promote the activation of lymphocytes and thus promote chronic inflammation in the islet. Since the dendritic cell typically
3
upregulates the chemokine receptor CCR7 and travels to the draining node early in the maturation process, dendritic cells capable of priming naive T cells are not typically seen in the periphery. For this reason, we suspect dysfunctional dendritic cell development and maturation as a possible cause of the dendritic cell accumulation within the pancreatic islets seen in NOD mice (Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Rosmalen et al., 2000a; Rosmalen et al., 1997).
It has also been established that dendritic cells from the pancreatic draining lymph node are capable of preventing diabetes when transferred to juvenile NOD mice (ClareSalzler et al., 1992). These experiments established that dendritic cells from this anatomic location were programmed to downregulate responses toward pancreatic 3 cell antigens, and provided one of the first examples of the tolerance-inducing capacity of the dendritic cell.
More recent studies in the NOD mouse showed that in vitro generated myeloid dendritic cells transferred to juvenile NOD mice also prevent diabetes (Feili-Hariri et al., 1999; Papaccio et al., 2000). Additionally, splenic dendritic cells pulsed ex vivo with IFN-y proved protective against diabetes (Shinomiya et al., 1999). This data further reinforces the tolerizing capacity of the dendritic cell, however, many questions remain regarding the role of the dendritic cell in the NOD mouse. Principal among these is how do dendritic cells appear to play a dichotomous role in the development of diabetes? As the dendritic cell population in the islet are myeloid dendritic cells, it is presumed that there may be a primary defect in this dendritic cell population leading to activation of autoreactive T cells. At the same time, recent studies strongly support myeloid dendritic cells playing a tolerigenic role (Feili-Hariri et al., 1999; Papaccio et al., 2000; Shinomiya et al., 1999). Why are dendritic cells within the pancreas itself priming autoreactive responses, while those same cells within the draining node are protective?
We hypothesize that dendritic cell maturation and migration play crucial roles in the pathogenesis of diabetes in the NOD mouse. These studies investigate in vivo
4
dendritic cell populations, development and maturation of myeloid dendritic cells in vitro, and the interaction of dendritic cells in vivo with the regulatory NK-T cell population.
REVIEW OF LITERATURE
Dendritic Cell Biology
Dendritic cells are divided into two lineages: myeloid and lymphoid. How these lineages function to initiate responses toward exogenous antigen or regulate responses to antigens is not clear at this point. Regardless, several studies have been published concerning the development and function of these two dendritic cell populations. Myeloid Dendritic Cells
Myeloid dendritic cells are found in all tissues except brain. Within the peripheral immune system, myeloid dendritic cells are found in the marginal zone of spleen, and T cell areas of LN and Peyer's patches. Within tissues and organs, myeloid dendritic cells are present in an immature state. Studies of myeloid dendritic cell development reveal two distinct immature myeloid dendritic cell populations, both yielding a common mature myeloid dendritic cell. The first pathway yields an epidermal Langerhans' cell, while the second generates the immature dendritic cell seen in other tissues and organs (Figure 2-1) (Banchereau and Steinman, 1998).
Immature myeloid dendritic cells are highly endocytic, bearing high levels of Fcy and Fces receptor, high levels of complement receptors CD11 lb and CD1 c, and high intracellular levels of MHC Class II (Cella et al., 1997). Accessory molecules important for T cell activation, such as CD80, and CD86, are absent or expressed at low levels on immature dendritic cells. In addition, actin cables, used in dendritic cell cytplasmic extension, elongation, and migration, are visible within the cell. This mobile immature dendritic cell patrols tissues, continuously sampling the surrounding fluid environment,
5
6
Red Platelet blood .
cells t
c I Eosinophil Megakaryocyte
Erythroblast Eosinophil
Pro myeloid- Basophil /i \ Basophil Lymphoid precursor Progenitor Myeloid Neutrophil / \oProgenitor T progenitor progenitor
NK ell
NK-T
O Gel Macrophage Myeloid
T Lprgeni o id Bedii prgniollncr
den t c dendritic cell CellMphrphad Myeloid
Figure 2-1. Dendritic cell developmental pathways.
via pinocytosis, as well as taking up particulate matter recognized by specific pattern recognition receptors (complement receptors, Toll receptors, Fc receptors, mannose receptor, integrins).
Our understanding of myeloid dendritic cell maturation has progressed significantly over the past few years. It now appears that once the dendritic cell has received a maturation signal, it begins to traverse a preset program of changes, involving modulation of chemokine receptors (i.e., CCR6, CCR7), cytokine production (i.e., IL-12, IL-10), density of MHC-peptide complexes on the cell surface, and T cell costimulatory molecules. IL-12 consists of two subunits, the p35 and p40. In an immature state, a dendritic cell
7
will produce the p40 subunit at a constituitive level while producing very little p35. Biologically active IL-12 requires a heterodimer of the p35 and p40 subunits. While p40 monomers and p40 homodimers are also secreted from the cell, these do not transduce signals through the IL-12 receptor, and may in fact inhibit this signalling Where the dendritic cell is in the maturation process when it encounters a T cell largely determines the functional polarization of antigen-specific T cell responses (Figure 2-2) (Langenkamp et al., 2000).
Paramount to peripheral tolerance induction is the uptake, processing, and presenCytokines[ IL-12p70 CytokIL-1
[ CCR6
Chemokine CCR6
Receptors CCR7 [MHC Class 1i
Surface CD80
Antigens
CD86
Immature Maturing 0 Mature Outtcome of Uibat
interation it pi meniv
naive Tcells T celLocation PeripheralS
Figure 2-2. Phenotypic and functional changes in myeloid dendritic cells during maturation.
tation of apoptotic cells and locally acquired soluble self antigens by APCs. As a cell undergoes apoptosis, typically due to irreparable DNA or mitochondrial damage, it loses the ability to maintain the integrity of its phospholipid bilayer. This leads to exposure of residues such as phosphatidyl serine on the surface of the cell, marking this cell for uptake and removal from the peripheral pool.
8
Phagocytosed cells are processed in the endocytic pathway like any extracellular pathogen or protein. The effect of apoptotic cell uptake on the state of dendritic cell maturation is still a topic of debate. While some studies report immature dendritic cells do not undergo maturation after uptake of apoptotic cells, most reports indicate that dendritic cell maturation takes place after interaction with apoptotic cells (Gallucci et al., 1999; Albert et al., 1998a; Albert et al., 1998b; Inaba et al., 1998; Mevorach et al., 1998; Rovere et al., 1998; Wells et al., 1999). Since uptake of apoptotic cells means the dendritic cell will be taking up and processing large amounts of self-antigen, how the dendritic cell responds, and in turn, the final outcome of interaction with naive T cells is of key importance. Previous hypotheses, such as those put forth by Matzinger, held that inherent danger signals exist which trigger dendritic cell maturation, and in turn a pro-inflammatory response (Matzinger, 1994; Matzinger, 1998). We now understand that dendritic cell maturation does not always lead to inflammation, and that the triggering of dendritic cell maturation by apoptotic cells is a necessary event, as the immature dendritic cell has little influence over naive T cells, and no capacity to play a role in peripheral tolerance (low MHC, costimulatory molecules, and cytokines). This mature dendritic cell can now interact with antigen-specific T cells in a manner that may induce apoptosis, anergy, or the induction regulatory T cell populations.
In response to exogenous pathogens, the dendritic cell will receive maturation signals such as LPS, TNF, bacterial CpGs, viral mRNA, or PGE2 (Lanzavecchia, 1999). These factors trigger the dendritic cell to progress through the previously described maturation states (Figure 2). But in this case, the number of T cells specific for these exogenous antigens is much greater than for self antigen, and the dendritic cell receives strong CD40 signals very early. This locks the dendritic cell into an IL-12 producing phenotype, where it primes Th 1 type T cell responses against these exogenous antigens (Langenkamp et al., 2000a).
9
Lymphoid Dendritic Cells
Lymphoid dendritic cells are distinguished from myeloid dendritic cells based on their phenotype, as lymphoid dendritic cells express high levels of the CD8a (though those in thymus do express low levels of CD8 3), low levels of CD1 Ib, and high levels of DEC-205. The lymphoid dendritic cells have been shown to specifically express high levels of self-peptides within MHC complexes, further highlighting the possible role of these cells in tolerance to self-antigen (Banchereau and Steinman, 1998; de St Groth, 1998; Steinman et al., 1997b). Increasing the number of lymphoid dendritic cells has also been shown to improve oral tolerance induction. Using Flt3L, dendritic cell numbers within peripheral lymphoid tissues can be dramatically increased (Viney et al., 1998). This increase correlates with enhanced tolerance induction to orally delivered antigen, allowing low doses of antigen to limit specific T cell proliferation as effectively as high doses.
Unlike myeloid dendritic cells, it is difficult to generate lymphoid dendritic cells in vitro, thus much less is known about developmental and functional properties of this lineage of dendritic cell. Lymphoid dendritic cells were first described as residents of the thymic medulla responsible for presenting self-antigen to maturing T cells, inducing apoptosis of autoreactive T cells (Ardavin et al., 1993). Lymphoid dendritic cells develop from a T cell precursor, and reside in the thymus, as well as within T cell areas of spleen, lymph node, and Peyer's patches (Wu et al., 1996). In contrast to the migratory myeloid dendritic cell, this dendritic cell is quite sessile, residing only within the lymphoid tissues, taking up antigens contained within the lymph, as well as that carried in by other cells. Immature lymphoid dendritic cells exist on the edge of the T cell areas, moving into the central regions as they mature. As the tissue associated myeloid dendritic cell homes to the secondary lymphoid tissues, it requires certain survival signals, such as CD40L on activated T cells upon entry to the lymph node. If these signals are not received, the myeloid dendritic cell may undergo apoptosis and be taken up by the lymphoid dendritic
10
cells, and its antigens presented in a tolerizing manner. Likewise, a mature myeloid dendritic cell is susceptible to Fas-mediated apoptosis if not receiving survival signals through CD40 (which upregulates bcl-2) (Bjorck et al., 1997). The lymphoid dendritic cell appears specially suited to take up these apoptotic cells within the lymph node and present the acquired peptides in high quantities on the surface with costimulatory molecules. Inaba and Steinman demonstrated that lymphoid dendritic cells were uniquely capable of capturing, processing and presenting antigen transferred from apoptotic cells captured in the lymph nodes (Inaba et al., 1998). They went on to suggest that transfer of antigen from dying myeloid dendritic cells to lymphoid dendritic cells leads to peripheral tolerance of the T cell repertoire to self.
Central Tolerance: The Role of Dendritic Cells
Together, lymphoid and myeloid dendritic cells work to orchestrate the complex interactions of the cell-mediated immune response, including tolerance induction and activation of specific T cells to eliminate pathogens. How might the dendritic cell system control these two diverse functions?
Lymphoid dendritic cells are the sole dendritic cell type found in the thymic medulla (Banchereau and Steinman, 1998; Fairchild and Austyn, 1990). The thymic medulla plays host to migrating thymocytes that have undergone positive selection in the thymic cortex and have been sent forward in the thymic selection process toward their final CD4+ or CD8+ destination. At this stage of maturation, any strong signal through the T cell receptor will trigger apoptosis. In other words, recognition of self-antigens in the context of MHC by the TCR triggers elimination of the autoreactive T cell. It is not clear whether a lymphoid dendritic cell, thymic epithelial cells in the medulla, or thymic macrophages must provide this signal. Matzinger made the assertion that dendritic cells mediate this process by demonstrating that splenic dendritic cells were capable of triggering deletion of thymocytes (Matzinger and Guerder, 1989). However, these studies failed to address whether lymphoid or myeloid dendritic cells mediated this T cell
11
deletion. The threshold required to trigger this process is lower than that required in the periphery for a nalve T cell to differentiate, ensuring that weakly reactive T cells will be eliminated (Dautigny et al., 1999).
Peripheral Tolerance: Myeloid and Lymphoid Dendritic Cell Interactions
Most potentially autoreactive T cells are eliminated in the thymus, but some do escape to the periphery. To control potentially deleterious responses, mechanisms of peripheral tolerance have been proposed. Since the thymus is partially sequestered from pathogens (unless systemically present), it could be argued that only self-antigens present within the vascular compartment have the opportunity to reach the thymus. In the periphery, the immune system functions in a manner that allows self-antigen to be distinguished from the milieu of pathogenic antigen likely to be present. Though theories have been proposed involving repetitive motifs on bacterial cell walls and inflammatory cytokines signaling to the dendritic cell that an antigen is indeed foreign, this system is still not well characterized. How and why one antigen results in a pro-inflammatory response while another results in tolerance remains a central question in immunology today.
Four possible outcomes from a T cell receptor recognizing a peptide-MHC complex in the secondary lymphoid tissues include ignorance, anergy, activation, or activation-induced cell death. Ignorance is a complete lack of recognition of MHC-peptide complex by the T cell, or non-presentation of the antigen (sequestration). Assuming that all potentially autoreactive T cells were deleted in the thymus, this would be the response any time self-antigen was presented in the periphery. Anergy may occur if the combination of primary (TCR) signal and secondary (CD28) signals does not reach an activation threshold, best exemplified by T cells receiving Signal 1 through the TCR-CD3 complex without engagement of CD28 (Signal 2) on the T cell (Wells et al., 1997). Anergy may also be induced when CTLA-4 engagement supersedes CD28 engagement (Karandikar et al., 1996; Walunas et al., 1996). CTLA-4 shares its ligands with CD28,
12
but CTLA-4 has a much higher affinity for these ligands, and effectively competes for, and binds CD80 and CD86 on the APC. CTLA-4 engagement directly inhibits phosphorylation of the TCR-CD3 complex, limiting T cell responsiveness (McAdam et al., 1998; van der Merwe et al., 1997). While unable to proliferate, anergic T cells may be capable of providing effector cytokines, and must be recognized as possible participants in autoimmune disease (McAdam et al., 1998; Mondino et al., 1996; Salojin et al., 1998; Van Parijs and Abbas, 1998). Furthermore, cytokines from anergic T cells, such as IFN-y, may polarize T cell-dendritic cell interactions toward a Th, profile.
Activation of a naive T cell occurs when the TCR and CD28 signals are of
appropriate strength to initiate IL-2 production and proliferation, but not so strong that the dividing population becomes fratricidal. Activation induced cell death occurs after recurrent activation via large numbers of TCR molecules. While activation-induced cell death involves an initial proliferative phase, this is followed by the Fas-mediated apoptosis of the dividing cell population. Activation-induced cell death is essential for limiting peripheral immune responses to foreign antigen, as well as to eliminate autoreactive T cells. Activation-induced cell death is dependent upon Fas-FasL interactions on CD4+ T cells, and TNF-TNFRp55 interactions on CD8+ T cells, and both humans and mice defective in either Fas or FasL suffer from multiple autoimmune disorders, due to a predominance of autoreactive T cells in the periphery (Ucker et al., 1989).
Regulatory Lymphocyte Populations
Rather than directly modulating effector T cells, dendritic cells may assert their effects by promoting regulatory T cell populations. Several regulatory T cell populations have been identified in the NOD, and most appear to function by steering T cell responses toward the Th2 phenotype, which is associated with protection from diabetes.
One type of regulatory T cell, the NK-T cell, expresses both a TCR with invariant a and 3 chains as well as NK receptors NK1.1 and Ly49. These cells appear specially suited to respond to self-peptides presented on CD 1 d, a primitive MHC class I type
13
molecule (Chiu et al., 1999). When engaged via the TCR, these cells release preformed intracellular stores of either 1L-4 or IFN-y, skewing the local immune response toward Th or Th2 type responses. NOD mice have reduced numbers and are functionally deficient in these cells when compared to nonautoimmune strains, giving possible explanation to reduced Th2 responses and 1L-4 in the NOD (Gombert et al., 1996a). Additionally, increasing the number of these cells either via cytokines or transgenic TCR prevents diabetes in the NOD (Gombert et al., 1996b; Lehuen et al., 1998).
The Th3 cell appears to predominantly secrete IL-4 and TGF-P. It was shown that these cells are induced by nasal and oral administration of self-antigens, and appear to mediate resistance to autoimmune diease (Han et al., 1997; Maron et al., 1999).
A third T cell population, the CD38+CD45Rb-low T cell, do not secrete 1L-10 or TGF-13, but have been functionally characterized as inhibiting anti-CD3-mediated T cell responses in vitro (Read et al., 1998). In addition, they are induced following CFA treatment of NOD mice, which delays the onset of diabetes (Martins and Guas, 1999).
It is difficult to summarize the diverse literature on regulatory T cell populations. While the function of a regulatory T cell in prevention of autoimmune disease has been demonstrated, mechanisms accounting for ineffective generation of these cells in the NOD mouse have not been defined. If dendritic cell function is impaired in the NOD, this could result in an inability to generate effective numbers of regulatory T cells. Myeloid dendritic cells were reported to generate Th2 type responses in vitro, thus an inability to derive fully competent myeloid dendritic cells may lead to inadequate generation of regulatory T cells.
Diabetes Onset and Progression in the NOD Mouse
The NOD mouse exhibits lymphocytic infiltration of the peri-islet areas around 4 weeks of age. Subsequently, intra-islet infiltration and 0 cell destruction occurs in all NOD males and females, and diabetes develops in 70-80% of females, and 25-50% of males by 30 weeks of age. The infiltrates contain both CD4 and CD8 T cells expressing
14
a variety of VP chains. Macrophages or dendritic cells appear to be the earliest entrants into the pancreas, as they are detected in the peri-islet areas one week after transfer to irradiated NOD females (O'Reilly et al., 1991). P cells close to the infiltrating cells in these areas express higher levels of class I, but do not express class II or CD54 (O'Reilly et al., 1991).
Both CD8+ and CD4+ T cells play a role in initiation of insulitis and P3 cell
destruction. The initial insulitis appears to be caused by CD8+ T cells, as Class I-deficient NOD mice do not develop diabetes or insulitis. CD4+ T cells, in particular Th I cells, appear responsible for triggering actual destruction of 3 cells, as Class II KO mice suffer limited islet damage, but still have significant insulitis (Forster and Lieberam, 1996).
Factors that lead to the eventual T cell activation and expansion in the NOD
mouse can be broken down into three basic categories; underlying products within the target tissues, atypical MHC class II molecule, and deficiencies in T cell activation and apoptosis. None of these alone are sufficient to cause autoimmune disease, but together lead to the 75% occurrence of diabetes in female NOD mice. Defects in Pancreatic 3 Cells
In juvenile NOD mice, an underlying genetic defect leads to death of P cells,
leading to upregulation of chemokines and adhesion molecules, and infiltration of macrophages and dendritic cells (Faveeuw et al., 1994a). A similar infiltration and interaction is seen within the salivary glands, which are also destroyed in NOD mice (Faveeuw et al., 1994a). These intrinsic defects in the target tissues of autoimmune disease are thought to be the factors that initiate and focus autoreactivity toward these tissues. Contributions of MHC Class II
The major predisposing factor in Type I diabetes in both humans and mice is
the MHC class II molecule. The NOD mouse carries the atypical I-Ag7 molecule, which is necessary for the onset of diabetes in the mouse. Congenic NOD mice carrying I-Ab
15
of the C57B1/10 mouse do not develop diabetes (Wicker et al., 1992). In humans, HLADQP1-0201, -0302, 0101 are found in 85-95% of patients with Type I diabetes, while only possessed by 20% of the Caucasian population (McDevitt, 1998). Both the NOD and human class II alleles share a homozygous lack of Asp at position 57 in the class II P3 chain. Substitution of Asp at position 57 protects NOD mice from diabetes (QuarteyPapafio et al., 1995). This susceptibility is most likely due to a combination of an inherent instability of the peptide-MHC complex, and an increased ability to bind and present 13 cell specific antigens.
Defects in Activation Induced Cell Death
In addition to the atypical MHC class II molecule, defects in the ability of naive T cells to undergo activation and proliferation, which are necessary for the elimination autoreactive T cells via activation induced cell death, contribute to expanded numbers of autoreactive T cells in the periphery of the NOD mouse.
For example, production of IL-2, a cytokine critical for activation-induced cell death, is decreased in the NOD mouse (Serreze et al., 1989). In addition, Delovitch and colleagues have noted proximal defects in the intracellular signaling cascade occurring after T cell receptor ligation (Jaramillo et al., 1994). Finally, treatment of NOD mice with recombinant IL-2 or treatment of NOD mice with non-deleting anti-CD3 antibody markedly reduces diabetes incidence (Serreze et al., 1989; Vallera et al., 1992).
We hypothesize that these factors alone are not solely responsible for disease. Central to all of these is the dendritic cell. It is the most potent antigen presenting cell known, the most prominent presenter of MHC class II-peptide complexes, and is found in the pancreas of very young NOD mice. Transfer of dendritic cells from the pancreatic draining lymph node was sufficient to protect NOD mice from developing diabetes (Clare-Salzler, 1991). This study establishes that these cells are capable of inducing long lasting peripheral tolerance to self-antigens, but are unable to do so in their endogenous location. Our hypothesis suggests that dendritic cell development, maturation, and migra-
16
tion could impair the ability to either eliminate autoreactive T cells or generate regulatory T cell populations, and as such contribute significantly to the autoimmune predisposition of the NOD mouse.
IDENTIFICATION AND CHARACTERIZATION OF MURINE DENDRITIC CELL POPULATIONS
Background
Initially described in 1973 as a single population of antigen presenting cells with an extremely high potential for T cell stimulation, the dendritic cell is now recognized as the most potent of the antigen presenting cells (Knight et al., 1983; Steinman and Cohn, 1973). Techniques for purification of these cells would not come until many years later, with recognition that these were more than a single population of dendritic cells not occurring until the late 1980's (Crowley et al., 1989; Steinman et al., 1979).
We now know that dendritic cells from the mouse originate from one of two
distinct lineages, myeloid or lymphoid (Shortman and Maraskovsky, 1998). While cells of myeloid lineage maintain high levels of surface antigens such as CD1 lb, characteristic of all myeloid cells, lymphoid dendritic cells express the antigen CD8a, typically seen only on Class I-restricted T cells (Shortman, 2000). Whether these lineages have unique functions, or are even exclusive from one another, remains a debated topic.
In the late 1990s, several investigators attempted basic characterization of dendritic cell populations in murine secondary lymphoid tissues (Salomon et al., 1998; Shortman, 2000; Vremec et al., 2000). This revealed that both myeloid and lymphoid dendritic cells existed in the secondary lymphoid tissues of the mouse, and that several key phenotypic differences separated the two populations.
The NOD mouse develops severe insulitis and 13 cell destruction leading to
diabetes beginning at 6 to 12 weeks of age (Bowman et al., 1994). The first cells seen infiltrating the pancreas are myeloid dendritic cells and macrophages (Shinomiya et al.,
17
18
2000). Studies have shown this APC infiltration may be due to defects with the 3 cells themselves, and may be enhanced by the atypical expression of chemokines and adhesion molecules (Faveeuw et al., 1994a; Rosmalen et al., 2000a).
Conversely, dendritic cells isolated from the pancreatic-draining node protect juvenile NOD mice from diabetes (Clare-Salzler et al., 1992). Why dendritic cells from the pancreatic draining node protect against disease while those within the islets themselves may promote disease was a key question that arose from these studies. To begin to address this issue, we performed enumeration and characterization of dendritic cell subtypes from the uninvolved inguinal node and the pancreatic draining node, to ascertain if obvious discrepancies in dendritic cell number or activation state were present. To properly perform such an analysis, we first had to describe all dendritic cell populations seen within the nodes in a manner appropriate for our future studies.
Materials and Methods
Mice
Female C57BL/6, NOD/LtJ, Balb/c, and CBA mice between 4-16 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center.
Creation of Single-Celled Suspension from Lymph Node
Lymph nodes were harvested from 3 to 6 mice into ice cold Hank's Buffered Salt Solution (HBSS) without calcium or magnesium (Mediatech, Herndon, VA). Nodes were then resuspended in a solution of lx HBSS with calcium and magnesium and 100 U/mL collagenase D (Boehringer Mannheim, Indianapolis), transferred to a petri dish on ice, and needle dissected using 1.5 inch 27-gauge needles. Free cells and tissue fragments were then transferred to a 15 mL tube and 4 mL of 400 U/mL collagenase D in HBSS was added. The solution was mixed, then incubated at 370 for 30 minutes. The cell suspension
19
was then pipetted up and down 10 to 20 times with a disposable plastic pipette to dissociate all cell fragments and clusters. The cell suspension was immediately transferred to 40 mL of ice cold HBSS without calcium or magnesium containing 1 mM EDTA and 1% BSA. Cells were washed twice in this solution and resuspended in 1 mL of the same solution with 4% BSA. Cells were counted on the hemacytometer and lx106 cells were aliquoted to each tube for staining.
Analysis of Surface Phenotype
Surface markers were analyzed by flow cytometry. Briefly, cells were blocked in 100 gL of HAB (Hanks Buffered Salt Solution without calcium or magnesium, with 1 mM EDTA, 0.01% sodium azide, and 1% bovine serum albumin (Calbiochem) with 4% BSA+1 gg anti-CD16/32 (purified clone 2.4G2) and stained with the appropriate surface antibodies. Antibodies used were: CD I c (clone HL3, Pharmingen), CD40 (clone 3/23, Pharmingen), CD54 (clone 3E2, Pharmingen), CD80 (clone 16-10A1, Pharmingen), CD86 (clone GL-1, Pharmingen), I-A/I-E (clone 2G9, Pharmingen), I-Ak (A-k) (clone 10-3.6, Pharmingen), which reacts with I-Ag7 of the NOD mouse. Biotinylated antibodies were detected with either streptavidin-PE (Coulter-Immunotech, Miami, FL), or streptavidin-APC (Molecular Probes, Eugene, OR). The dye 7-AAD (7- aminoactinomycin D, Molecular Probes, Eugene, OR) was added at 1 gg per mL to some samples to exclude dead cells from analysis. After establishing that >90% of cells were viable in all preparations, we substituted additional fluorochrome labeled monoclonal antibodies for the viability dye. Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry, San Jose, CA).
20
Results
Ex Vivo characterization of Dendritic Cell Populations
In order to characterize dendritic cell populations in diabetes-prone mice, we
first needed to recognize all dendritic cell populations present within murine secondary lymphoid tissues.
Utilizing techniques that result in the least amount of manipulation, and the best representation of the dendritic cells present in vivo, nodes were extracted and single cell suspensions created. These cells were stained and analyzed by flow cytometry to characterize and enumerate the dendritic cell populations within the lymph nodes. Five Distinct Dendritic Cell Populations in Ex Vivo Preparations
By creating a suspension containing all the cells within the lymph node, we are more likely to retain all the dendritic cell populations present in vivo than if we used an ex vivo manipulation to enrich these cells. The flow cytometer can easily do the work of isolating our population of interest for analysis, without the risks of losing certain cells through ex vivo manipulation.
In the murine system, only dendritic cells are CD I1 Ic+. While preparations from spleen have large numbers of autofluorescent macrophages that impair the ability to discriminate other cell populations, few macrophages are found in the lymph nodes, yielding a clear phenotypic picture for analysis. Further more, the dendritic cell populations within the spleen vary greatly from those within the lymph nodes.
Utilizing antibodies against Class II MHC and CD1 c, a two-parameter plot
reveals 3 distinct CD1 lc+ dendritic cell populations (Figure 3-1, upper panel). The first of these, labeled as population A, has very high class II expression, with a wide range of CD1 lc positivity. The second, labeled population B, has slightly lower expression of class
21
c
6A
B cells
00 11 12 10 104 Class II FITC
A B C
0 200 400 600 800 1000 0 200 400 600 000 1000 0 200 400 000 800 1000 Fonard Sca.ler Fowad Scatter Foward Scatter
B cells
0 200 400 600 00 1000
FoIrwad Scatter
Figure 3-1. Three populations of CD11c+ dendritic cells in murine lymph node preparations. After creation of a single-celled suspension as described in materials and methods, cells were stained with antibodies to class II MHC and CD11 c. A minimum of 100,000 events were acquired for each sample. Top panel represents class II MHC versus CD 1 c staining after gating on high forward scatter events. This allows visualization of our populations using a contour plot. Lower four panels represent the scatter characteristics of those populations indicated in the upper panel. The scatter of the B cell population is given as a point of reference.
II, but a higher median level of CDI lc. The third, population C, has very low expression of class II, and an intermediate level of CD1 Ic staining.
Adding antibodies against CD8a and CD1 lb, which discriminate lymphoid and myeloid dendritic cells respectively, allows further classification of these three dendritic cell populations.
Examination of population A reveals that these are very large cells, evidenced by their high forward scatter characteristics (Figure 3-1, lower panel). These cells have
22
moderate amounts of CD 11 lb on their surface, and no significant CD8ca, indicating that these dendritic cells are of myeloid lineage (Figure 3-2).
Dendritic cells within population B have lower forward scatter than those within population A, but still higher than typical lymphocytes (Figure 3-1, lower panel). Attempting to classify these as myeloid or lymphoid reveals a clearly bimorphic population. While a portion of these cells exhibit high CD1 lb and low CD8X expression, similar to the cells within population A, the other fraction shows lower CD1 lb expression with high expression of CD8a, indicating these cells are of lymphoid origin (Figure 3-2).
Population C is quite distinct in its complete absence of CD1 lb expression. While a subset of population C does express CD8a, a significant number of these cells express neither CD1 lb nor CD8ct (Figure 3-2). Scatter characteristics of population C were also quite unique, with a majority of the cells falling near the low forward scatter range typically associated with larger lymphocytic cells (Figure 3-1). As population C had yet to be described in the literature, the possibility that these were not dendritic cells existed. Running a panel of monoclonal antibodies targeting lineage specific antigens such as CD3, CD19, F4/80, and CD4, it was not possible to detect expression of any additional lineage specific molecules on these cells. Demonstration of these 5 dendritic cell populations, A, B(CD8+), B(CD8-), C(CD8+), and C(CD8-), has been repeated in our lab in at least 30 independent experiments.
Having characterized lineage origins of these various dendritic cell populations, the next goal was to characterize other molecules expressed by these different populations. Examination of the molecules CD40, CD54 and CD86 revealed much higher levels of expression on cells within population A. This coincides with the elevated class II MHC expression on these cells relative to the myeloid dendritic cells found within population B to indicate that the myeloid dendritic cells within population A are mature myeloid dendritic cells. Population C showed little expression of CD80 or CD86, indicating the immaturity of this dendritic cell population (Figure 3-2).
23
T A B C
o=% o-. o-%
- 0
-1 4 4 4C
CD11 b APC CD11lb APC CD11lb APC
O O-M O=
ma- Cc
o 101 0 014 o 10 101 014 O10
CC
1 10 10 10 10 10 10 10 10 10 10 10 10 10
CD11 APC CD 1b APC CD11bAPC
A. ( B .2. 1
0- 0- 0D- 0 a- A
01
240 CD40APC CDI .APC U 1U 4 10 1 1
1 1 1 1 10 10 10 10 10 10 10 10 10 10
CD40 APC CD40 APC CD40 APC
Figure 3-2. Phenotype of 5 dendritic cell populations. Samples were stained with class IH [-C FI[TC and CD11]c PE, as well as the indicated antibodies from these plots. The
3 dendritic cell populations described in Figure 3-1 were gated, and cells within each region were displayed for analysis of fluorochromes within FL3 and FL4. All samples displayed here are from NOD mice, but similar phenotypic profiles are generated from C57BL/6 lymph nodes. Within pancreatic draining node (PLN) preparation, some cells from population B overlapped with population A due to decreased class II expression on population A, leading to the appearance of some CDga positive events within these clots.
A Nr B C
a.4 APC CD4 APC a. APC
CIMHC F I an CD 11cPaCela teidctdatb Die rmteeposh
3 dendritic l ouain ecie nFgr eegtd n el ihnec
region were dipae for anlyi of 0loohoe wihi 4L n .lape displaye her are fro NO mie but1 simla ph1otpi prflsarieeatdfo
C57BLCD4 lymp nodes WihnpnraiAriignd PLN prpaaton somecel
frmopltin B vrapdwt oatio A u oderae casIepeso on4)C pouainAlaigt teapaace ofsm C oitv vet ihi hs plots.
24
A B C
0a-o fl- .%2
CD54 APC CD54 APC CD54 APC a ... o ,
Cc CU -CU -! o- 0o-
"2"""' 4 6,0....I .....2l.6 '3....-4 60, ";1 ....2--"03 '-4
1 10 10 104 10 10 10 1 104 10 10 10 10 10
CD54 APC CD54 APC CD54 APC a. o a.c:.a o
A B C
m O -"O -IO ,-: 0- 00 0 c
"~4 Q 0
-10 ... 1 "6 .....= .. ... ..041 0- 0 --02 4 -6o ..0 .0 -0 1 4
-I10 10 10 10 10 10 10 10 10 10 10 10 10 10 10
CD54 APC CD80 APC CD5480 APC
A B C
a 0 0 o
Co- Co- C o-c 10 10 10 10 10 10 10 10 1010 10 10 10 104
CD80APC CD80 APC CD80 APC
Fiur -o. ni. 00- 00.00
010
60 ...... lu 1 u16 10 1'10 10 10
CD80 APC CD80 APC CD80 APC
Figure 3-2 Continued
PLN Inguinal PLN Inguinal
CD8a PerCP CD8a PerCP CD8a PerCP CD8a PerCP
100 101 102 103 104 100 101 102 103 104 100 101 102 103 104 100 101 102 103 104
000 0
> a
0 '0 0 0
CD8a PerCP CD8a PerCP CD8a PerCP CD8a PerCP
S100 01 102 103 104 100 01 102 103 104 100 10 10 102 103 104 .100 101 102 103 104
0 0
mo mo 0 0 0 0
I~ ....... .... ... . ......
CD~~~Oa PeC Da eC D0 eC C8 eC o o8 0- 0 o o
0 0 0 0
CD8a PerCP CD8a PerCP CD8a PerCP CD8a PerCP
. 100 10 2 103 104 ..100 102 34 101 102 103 104 o100 101 102 103 104
C 0 ....
08 0O 08 08
0- 00
4--) CD" C08' '
oo) M8 o >.,>
>J >. TJ
26
We next stained the cells with a host of other antibodies, and found that not only did the lymphoid component of population B have significant amounts of CD38 on its surface, but population C expressed equally high amounts of this molecule suggesting a relationship between cells within population C and the CD8+ component of population B (Figure 3-3)
Variations in Dendritic Cell Populations Between Anatomic Sites
Although not formally addressed, the assumption amongst immunologists has been that all lymph nodes are equivalent in their cellular makeup. Since some of our original hypotheses revolved around the properties of cells within the pancreatic draining lymph node, we wished to determine whether pancreatic draining node dendritic cell composition was similar to that in other nodes. In addition, as the pancreatic draining node represents a lymph node draining an inflammatory site, effects secondary to inflammation could occur in this compartment.
The first striking difference between nodes from these different locations was the amount of MHC class II found on the surface of the dendritic cell (Figure 3-4). In the mouse, B cells constitutively express high levels of class II on their surface, and make a good reference within each node. While population A dendritic cells from peripheral nodes, such as the inguinal node, showed levels of class II which exceeded that found on B cells, population A dendritic cells from mesenteric nodes, including the pancreatic draining nodes, all have lower levels of class II. While all of the dendritic cell populations described show reduced class II expression in mesenteric nodes, the greatest differences were seen on dendritic cells within population A, the mature myeloid population (Figure 3-4). T cell costimulatory molecules such as CD86 were also decreased on dendritic cells from the pancreatic-draining node (relative to the inguinal node) (Figure 3-2). This was seen on all dendritic cell populations from lymph nodes within the mesenteric cavity.
Differences in the percentage of each dendritic cell population also varied greatly between these sites. The proportion of myeloid versus lymphoid dendritic cells within
27
population B was significantly different between the inguinal and mesenteric nodes, with a greater proportion of lymphoid dendritic cells in the inguinal nodes (1:1 in the inguinal node in contrast to 2:1 myeloid versus lymphoid in the pancreatic draining node). Variations in Dendritic Cell Populations by Strain
Inguinal and pancreatic draining lymph nodes were extracted from C57BL/6 and NOD mice, dissociated, and dendritic cell populations characterized. No overt signs of dendritic cell activation, such as elevated class II MHC or costimulatory molecule expression, were noted in NOD lymph nodes in comparison to C57BL/6 controls. Population C, which appear to be the least mature dendritic cells, was significantly elevated in the non-inflamed inguinal nodes of NOD mice relative to C57BL/6 controls (Figure 3-5).
A B C
0 10 10 0
M -C- 0-co' 00 o
a a.-
0
0 1010 4 10 10 110 10 1 4 1 0 10 1
CD69 PE CD69 PE CD69 PE
(2 !..- C, U C )
F4/80 PE F4/80 PE F4/80 PE0
oD 0'oD 0 dhhG
_10 .10 10210 10 4 -10 .10 10-10 10 4 110 11 010
F4/80 PE F4/80 PE F4/80 PE
Figure 3-3. Additional phenotypic analysis of dendritic cell populations. Cells were isolated, stained, and gated on populations A, B, and C. Additional markers used include CD25, CD38, CD69, and F4/80. Note that population C routinely contains some macrophages, which are highly autofluorescent. This is the positive population seen in the isotype plot above. There is no way to eliminate these cells from analysis, but the dendritic cells of population C are clearly discernable from these macrophages, so they do not impeed our ability to interpret this data.
28
A B C
- oI-i.0 o
2 02
O O- O ao 0 0.,
0 1000 0 1000 : 1000
Forward Scatter Forward Scatter Forward Scatter
0: o 0
0 0O O
S10 1010 10 10 1 1 10 1010 101010 10" I10"
IgG2b PE IgG2b PE IgG2b PE
co =- oa.o 0L a.nQo 0 o,- (-,
Go Do 0o 0- 00
0 0
a ,,. i ""'"'2 "' 3' 4 n o o i 2 6 4 ,,,,, ,,,, ,,, .,
10 10 102 1010 0 10 10 10 1 10 10 102 10 104
CD25PE CD25PE CD25 PE
Mm% a.m
5(010 10 10 10 0""
CD38 PE CD38 PE CD38 PE
Figure 3-3 Continued
Co-Co- Go0-~0 o *
4 O 6 '--1"'2- '3 4 0 14
1 10 1 1010 10 10 10 10 1 10 1 01 10 10 1
CD38 PE CD38 PE CD38 PE
Figure 3-3 Continued
29
- Inguinal Node ........ Pancreatic Draining Node
Population A
0 1 2 4.... 100 101 102 103 104 Class II Fluorescein
Population B CD8a-+ 8
100 10 102 10o3 104 Class I Fluorescein
Population B
CD8aco l ... ... .. ..... ""6 34
100 10 10 102 103 10 Class II Fluorescein
Population C 3
100 101 12 13 10 Class II Fluorescein
Figure 3-4. MHC class II variations between anatomic compartments. Cells were prepared and stained as before, then the three dendritic cell populations were gated and their class II expression displayed on a histogram for easier interpretation. The largest differences occur within the myeloid dendritic cell population, both within population A and population B. Population C was not broken into CD8 positive and negative components due to the lack of clear separation between these populations.
Additionally, population A dendritic cells made up a significantly lower proportion of the total dendritic cell milieu in pancreatic draining nodes from the NOD. We initially performed these studies segregating results based on the age of the mice sampled, which showed no differences with respect to age, and subsequent analyses were performed with all data grouped together for increased statistical power.
Because of the described interactions between NK-T cells, we also examined the molecule CD Id, an MHC-like lipid binding molecule responsible for presentation of gly-
30
35 colipid antigen to NK-T cells, on den- 30 dritic cells. Expression of CDl1d was
25
5 elevated on lymphoid dendritic cells
020
15- from the pancreatic draining node of
S10
10- NOD mice (Figure 3-6). While all dena.0 50 dritic cell populations in the NOD had
0
o, 4' 0 "' 0 higher basal levels of CD1d than their
Strain and Population Inguinal Node C57BL/6 counterparts, the level found
*
45- on lymphoid dendritic cells from the
40
35 .pancreatic draining node far exceeded
+ 30
E 25 that seen on other populations.
0
S20
S15 Summary and Conclusions
810
(D
I I We felt a necessary first step to V. rb 1 0rC
4' o 4' 0' 0o understanding dendritic cells in our disStrain and Population Pancreatic Draining Node
ease model was to obtain a complete Figure 3-5. Composition of dendritic cell populations between strain and profile of all dendritic cell populations anatomic location. C57BL/6 and NOD nodes were prepared as described and then within murine tissues. analyzed by flow cytometry. The com- Working with previously position of the total pool of CD 11 c positive events was broken down by each described protocols for tissue dissociaof the three distinct dendritic cell populations when analyzed by class II MHC tion, we modified these procedures to versus CD1 c. Significant differences are give the greatest cellular yield and viaindicated by black overhanging bars. Five mice were pooled for each experiment. bility with the least amount of manipThe data here represents the average of 12 independent experiments. indicates ulation (Salomon et al., 1998). These p<0.01 as determined by paired t-test. techniques consistently give preparatechniques consistently give preparations of greater than 90% viability and
consistent differential counts between experiments for all dendritic cell populations
described.
31
We observed five distinct dendritic cell populations, which coincide with various lineages of dendritic cells previously described (Shortman and Maraskovsky, 1998). Population A has high expression of class II and costimulatory molecules, lacks CD8a, and bears the molecule CD 1 b, indicating this population represents mature myeloid dendritic cells. Population B contains two separate dendritic cell populations, a CD8at positive component corresponding to the lymphoid dendritic cell population, and a Pancreatic
Inguinal Draining Node Draining Node
C3 12.10 CD 10.56
C57BL/6
al 12.96 11.31
co 28.88596 o To 1 1 0 104 10 10 10 1 104
0 10 i288 10M~ 0 0 1
CDld APC
Figure 3-6. CD1d is elevated on dendritic cells from the NOD pancreatic draining node. Cells stained and gated for population B were then analyzed for expression of CD8a and CDld. Number next to each population indicates the median fluorescence intensity for that population during this experiment. Myeloid dendritic cells
of population A stained in a similar manner to those shown here
(CD8 negative). Population C showed no staining for CD1d.
CD8a negative component which represents a more immature myeloid dendritic cell population. Studies with in vitro derived myeloid dendritic cells confirm that CD1 lb
32
staining decreases with maturation, supporting the lineage of the CD8ct negative component of population B as less mature myeloid dendritic cells (Shortman, 2000).
Population C has very low surface expression of class II MHC, and is devoid of
lineage markers besides CDl ic. Several avenues suggest this cell population is composed of immature lymphoid dendritic cells. First, a recent work described isolation of a low class II dendritic cell population from spleen which, when matured, led to mature lymphoid dendritic cells (Kronin et al., 2000b). Second, we observed very high levels of CD38 on this population. CD38 has traditionally been associated with immature cells of lymphoid origin (Malavasi et al., 1994). Within population B, the lymphoid component maintains high expression of CD38, implying that the CD8a population of cells in population B were derived from cells within population C. However, population C does not uniformly express the lymphoid lineage marker CD8a. When purifying and maturing this population, Shortman and colleagues observed that all cells developed into lymphoid dendritic cells (Kronin et al., 2000b; Vremec et al., 2000). We hypothesize that even those cells within the CD8a negative component of population C will eventually upregulate CD8ot and become lymphoid dendritic cells, based on these previous studies and the uniformly high expression of CD38 (Kelly et al., 2001; Vremec et al., 2000). Alternatively, population C may represent a pool of dendritic cells that develop into both mature myeloid and lymphoid dendritic cells.
The characterization of dendritic cell populations in different secondary lymphoid tissues is of considerable interest, as little work has assessed whether all lymph nodes contain these dendritic cell populations (Salomon et al., 1998). Our data clearly shows lower levels of MHC class II and T cell costimulatory molecules on dendritic cells from within the peritoneal cavity. Previous literature has correlated decreased class II MHC expression with decreased dendritic cell maturity and stimulatory capacity. The epidermal Langerhan's cell, which patrols the skin and migrates to the inguinal node, may have a different phenotype than those myeloid dendritic cells found in other tissues, leading to
33
the increased class II MHC seen on dendritic cells from peripheral nodes. In contrast, we show that lymphoid dendritic cells from the pancreatic draining node of the NOD mouse express significantly higher levels of the glycolipid-presenting molecule CD1d. The increase in CD 1 d was seen only in the pancreatic draining node and not the mesenteric lymph node suggesting that this molecule may be upregulated in lymph nodes draining inflammatory sites.
Finally, to follow up on previous studies demonstrating a protective effect when dendritic cells from the pancreatic draining node were transferred, but not those from the inguinal nodes, an attempt was made to phenotypically characterize dendritic cells of the pancreatic draining node and those of peripheral tissues. We compared dendritic cells within secondary lymphoid tissues of the diabetes-prone NOD mouse and the control C57BL/6 strain, as well as other controls. This revealed an expanded population C within the inguinal nodes of the NOD mouse, and a decreased number of mature myeloid dendritic cells within the pancreatic-draining node.
It was not clear why myeloid dendritic cells were reduced in the pancreatic draining node of NOD mice in comparison to the inguinal lymph node or the control C57BL/6 pancreatic draining node. Decreased myeloid dendritic cells within the pancreatic-draining node may relate to dendritic cells trapped within the inflamed pancreatic islets (Papaccio et al., 1999a; Shinomiya et al., 2000). Immature dendritic cells express high levels of CCR6, which chemoattracts the immature dendritic cell to areas of inflammation (Caux et al., 2000). The islets of the NOD mouse express constitutive levels of chemokines and adhesion molecules that serve to attract immature dendritic cells to the area (Faveeuw et al., 1994a). If these myeloid dendritic cells are unable to exit the inflamed pancreas, we postulate that this would result in a decline in myeloid dendritic cells within the pancreatic draining node. Alternatively, increased levels of cell death within the dendritic cell population, or an inability to develop and proliferate from precursors may lead to a generalized decrease in these myeloid dendritic cells as well.
34
The increased number of dendritic cells within population C of the inguinal node in the NOD mouse suggests a defect in maturation of dendritic cells within both the myeloid and lymphoid lineages. This proved fortuitous for these experiments, as it allowed visualization of this population while our isolation techniques were still quite crude. If mature myeloid and lymphoid dendritic cells play a role in the uptake and presentation of antigen in a tolerizing manner, the inability to progress to a fully mature state could partially explain the predisposition of the NOD mouse toward autoimmune disease.
This study provides a framework for future experiments in murine dendritic cell biology, and provides potential avenues for exploring the role of dendritic cells in the pathogenesis of not only Type I diabetes, but other diseases of immune tolerance in the mouse.
DYSFUNCTIONAL DEVELOPMENT AND MATURATION OF MYELOID DENDRITIC CELLS FROM NOD MICE
Background
Effective control of the immune system involves the coordinated movement of
both cells and molecules to defined locations at precise times. Immature myeloid dendritic cells, which exist in peripheral tissues, must move to secondary lymphoid tissues after receiving signals, whether endogenous or exogenous, to most efficiently interact with nafve T cells, and thus initiate an immune response. Similarly, proteins within the dendritic cell must traffic to their appropriate cellular locations in order for the dendritic cell to bring about the desired outcome, whether that is immunity or tolerance (Steinman, 1991; Steinman et al., 1997a).
The NOD mouse develops insulitis by 6 weeks of age and T cell dependent autoimmune diabetes by 8-12 weeks of age. As in humans, the NOD mouse requires the presence of an atypical MHC class II molecule lacking aspartic acid at position 1357 (Corper et al., 2000; Quartey-Papafio et al., 1995). The NOD H-297 is necessary, but not sufficient for initiation of diabetes in this model (McDevitt, 1998; Ridgway et al., 1998; Wicker et al., 1992). Additional genetic loci contributing to defective T cell activation in the thymus and periphery, have been described in this strain (Arreaza et al., 1997; Rapoport et al., 1993b). However the MHC, Iddl0, Idd3, and Idd5 appear to be most critical and can reconstitute almost full diabetes susceptibility in congenic mice (Cordell et al., 2001).
Though defects in T cell signaling could allow the survival of autoreactive cells by inhibiting AICD or generation of regulatory T cells in the periphery, a role for
35
36
APC regulation of these potentially deleterious cells can be established. Early studies demonstrated that dendritic cells transferred from pre-diabetic mice were capable of limiting disease onset in the recipient. Only cells from the lymph nodes draining the pancreas were capable of this protection, suggesting that cells in this compartment either carry specific antigens, or are exposed to certain cytokines or other factors that render them tolerizing upon transfer (Clare-Salzler et al., 1992). Subsequent studies ascertained that NOD splenic dendritic cells stimulated ex vivo with IFN-y provided similar protection as those from the pancreatic draining node (Shinomiya et al., 1999). These data suggest that endogenous NOD dendritic cells require additional stimuli to develop functions required for prevention of autoimmune disease. Additional studies demonstrated a protective effect after transfer of in vitro bone marrow derived myeloid dendritic cells, and again did not appear to require a diabetes related antigen to provide a protective effect (Morel et al., 1999; Papaccio et al., 2000).
Dendritic cells are among the earliest leukocytes found in the pancreas of diabetes prone animals, and accumulate throughout the course of disease. These dendritic cells form networks with T cells similar to those found within secondary lymphoid tissues, where they locally prime autoreactive T cells (Ludewig et al., 1998).
Macrophages from the NOD mouse are limited in their capacity to differentiate in response to M-CSF and respond to IFN-y (Leiter and Serreze, 1992; Serreze et al., 1993a; Serreze et al., 1993b). Myeloid dendritic cells develop from the same precursor as the macrophage, and may manifest similar defects to those seen in development of this myeloid lineage. To investigate this possibility, we examined lymphoid tissues from the NOD mouse and found they harbored increased numbers of immature dendritic cells (relative to the C57BL/6 control strain). We then derived myeloid dendritic cells from NOD bone marrow, and compared these with dendritic cells generated from various control strains for phenotype, maturation, and function. Our data demonstrates that defects within NOD hematopoeitic cells lead to generation of atypical dendritic cell populations
37
which are highly resistant to typical maturation stimuli (e.g. LPS, anti-CD40). We have additionally identified a region on chromosome 3 near the Iddl0 locus that contributes approximately 50% in the defective maturation of NOD dendritic cells. We postulate that as a result of this defect, when encountering typical maturation signals and self-antigen in the pancreas, these cells are retained in the islets of Langerhans' instead of becoming activated and traveling to the draining lymph node. Within the tissue these activated dendritic cells are capable of priming autoreactive T cells in a manner that leads to the destruction of 3 cells. As a corollary, these dendritic cells, which mature and migrate to the pancreatic draining node, acquire tolerizing function and help to regulate the inflammatory process by promoting regulatory cells as previously described (Clare-Salzler et al., 1992).
Materials and Methods
Mice
Female C57BL/6, NOD/LtJ, Balb/c, CBA, NOD.B 10-H2b, NOD.IddlO, and C57BL/6.c 17 mice 4-12 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center. Generation of Myeloid Dendritic Cells
Tibias and femurs were extracted and excess tissue removed by scraping with a scalpel blade. Ends of the bones were cut, and bones flushed with 2-5 mL of 40 C RPMI-C (RPMI-1640 (Mediatech, Herndon, VA) with 10% FCS (Mediatech) and lx PenStrep-Neo (Life Technologies, Rockville, MD). Red cells were lysed using a standard ammonium chloride lysis buffer (0.15M NH4C1, 1.0 mM KHCO3, 0.1 mM Na2EDTA, pH 7.2). Cells were resuspended at lx106 cells per mL of RPMI-C. Murine recombinant GM-CSF (R&D Systems, Minneapolis, MN) and IL-4 (Pharmingen, San Diego, CA)) were added at 500U and 1000U per mL, respectively. Cells were plated in 24 well culture
38
dishes (Coming Inc., Corning, NY) at 1 mL per well. After 48 hours, 500 gL of media was aspirated from each well, and replaced with 500 gL of RPMI-C containing fresh cytokines.
After 96 hours, cells were matured by adding either 1 gg of LPS (Sigma Chemical, St. Louis), 5 gg anti-CD40 (clone HM40-3, no azide, low endotoxin, Pharmingen), or a combination of LPS and 1000U mouse IFN-y (R&D Systems) to each well. Cells were matured for 24-48 hours and harvested for further analysis. Analysis of Surface Phenotype
Surface markers were analyzed by flow cytometry. Briefly, cells were harvested from culture and washed in HAB (calcium, magnesium, and phenol free HBSS with 1% BSA, 0.1% NaN3, and 1 mM EDTA)A total of 5x105 cells were then aliquoted to tubes for staining. Cells were blocked in 100 gL of HAB with 4% BSA+I.g antiCD 16/32 (purified clone 2.4G2) and stained with the appropriate surface antibodies. Antibodies used were: CD1 Ic (clone HL3, Pharmingen), CD40 (clone 3/23, Pharmingen), CD54 (clone 3E2, Pharmingen), CD80 (clone 16-10A1, Pharmingen), CD86 (clone GL-1, Pharmingen), I-A/I-E (clone 2G9, Pharmingen), I-Ak (A-k) (clone 10-3.6, Pharmingen), which reacts with I-Ag7 of the NOD mouse. Biotinylated antibodies were detected with either streptavidin-PE (Coulter-Immunotech, Miami, FL), or streptavidin-APC (Molecular Probes, Eugene, OR). 7-AAD (7- aminoactinomycin D, Molecular Probes, Eugene, OR) was added at 1 gig per mL to all samples to exclude dead cells from analysis. Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry, San Jose, CA). Purification of Hematopoetic Stem Cells
For all column based cell enrichments, products from StemCell Technologies
(Vancouver, BC, Canada) were utilized, unless noted. Briefly, bone marrow was harvested and red cells lysed as described above, then resuspended at 5x107 cells per mL in HBSS
39
with 5% FCS (MediaTech, Herndon, VA) and 5% rat serum (Sigma, St. Louis, MO). A biotinylated lineage cocktail consisting of clones Ly-1 (CD5), B220 (CD45R), Mac-1 (CD1 Ib), Gr-1 (Myeloid Differentiation Antigen), and TER 119 (RBCs) was added to the cell suspension. An anti-biotin anti-dextran tetramer was then added, followed by dextran coated magnetic particles. The suspension was then run through the magnetic column, and effluent washed, counted, and plated as described for whole marrow cultures above. Purification of CD4+ T cells
For the allogeneic mixed leukocyte reaction (MLR), CD4+ T cells were purified from CBA spleens utilizing magnetic based negative selection. Enrichment was performed exactly as that used for stem cell enrichment described above, with the following exceptions. Spleens were taken from CBA mice, smashed to release cells, and red blood cells lysed. An antibody cocktail containing clone Mac-1 (CDI l b), B220 (CD45R), CD8, Gr-1 (Myeloid Differentiation Antigen), and TER119 (RBCs) was utilized (StemCell Technologies). Effluent was washed, counted, and verified to be >98% pure CD4+ T cells by flow cytometry.
Proliferation Assessment with CFSE
To assess the proliferation of CD4+ T cells cocultured with various dendritic cell populations, T cell were resuspended at 1xl07 cells per mL of PBS, labelled with 2 ptL per mL of cell suspension using a 5 mM stock of the dye CFSE (5-(and 6)-carboxyfluorescein diacetate, succinimidyl esther, Molecular Probes, Eugene, OR). Following MLR, cells were resuspended in HAB, stained with CD4 biotin followed by a streptavidin-APC secondary, and finally resuspended in a solution containg the viability dye 7-AAD. CD4+ viable cells were gated for analysis. Regions were placed around peaks of CFSE intensity, and events within those regions enumerated. Number of mitotic events was computed using the techniques and formulae of Wells (Wells et al., 1997).
40
Uptake of Fluorescent Labeled Proteins or Particles
To assess uptake of particulate antigen, standard techniques utilizing fluorescent labeled dextran, ovalbumin, and BSA were utilized. Dendritic cells were suspended in
1 mL of RPMI-C with 5 mM HEPES, and 1 mg of Rhodamine Green-Dextran, DQOvalbumin, or DQ-BSA (Molecular Probes, Eugene, OR). Cells were incubated for 6 hours at either 4o or 370, and then analyzed for fluorescence. Both DQ-OVA and DQ-BSA are self-quenched BODIPY dyes that must be cleaved in order for fluorescence to be seen. Additionally, both the Rhodamine green and BODIPY dyes are pH insensitive. Thus, no signal is lost due to the pH of the lysosomal compartment. ELISA for Cytokine in Culture Supernatants
Cytokine concentrations in culture supernatant were analyzed using products from Pharmingen (San Diego, CA) unless specified otherwise. The following antibody pairs were utilized for cytokine detection, listed as capture and detection respectively; IL-2 (JES6-1A12 & JES6-5H4), IL-4 (BVD4-1D 11 & BVD6-24G2), IFN-y (R4-6A2 & XMGl.2), IL-10 (JES5-2A5 & SXC-1), IL-12p40 (C15.6 & C17.8), IL-12p70 (9A5 & C 17.8), A streptavidin-HRP secondary was added, followed by addition of a TMB substrate. The reaction was stopped using IN H2SO4 and read at 450 nm on a 3550-UV microplate reader (Bio-Rad, Hercules, CA).
Results
Identification of Protective Dendritic Cell Populations
Revisiting an Old Friend
In the early 1990's, our lab performed experiments involving adoptive transfer of dendritic cell populations to juvenile NOD mice. These dendritic cells migrated from the foot pad to the popliteal lymph node where they stimulated cellular proliferation and
41
After protected against developUnmanipulated metrizamrnide ment of diabetes At the gradient
time, we were unaware
0 of dendritic cell subsets Z W:
.C oand the intricacies of den, dritic cell maturation. To
.o 1 1 o oo 1 2 .. so,10 characterize the protecClass 11 FITC Class II FrITC
tive population(s), we
. repeated dendritic cell
i .t isolation from lymph 1 nodes as it was initially
-o ..... ... ; o k .... 0 ; "o ..... .. ... ...;
o1 o io< o 4 10 101 102 10 104
Class II APC Class II APO performed, and analyzed Figure 4-1. Metrizamide gradients enrich mature
isolated cells for phemyeloid dendritic cells. Single cell suspensions from lymph nodes and in vitro derived myeloid dendritic notype and maturation cell cultures were layered onto 14.5% metrizamide gra- state. This analysis demdients and centrifuged at 600 xg for 10 minutes. Cells at the interface were removed and either stained for onstrated that the pancreflow cytometery (top right), or incubated for 2 hours atic draining node denatic draining node denwith rhodamine green dextran and stained for surface expression of class II and CD1 Ic. Both cases showed dritic cell, which had prothe majority of the cells found at the interface after cen- tested NOD mice from tected NOD mice from
trifugation to be mature myeloid dendritic cells. These results were repeated in three independent experiments, diabetes, were in fact Data from female NOD mice is shown. mature myeloid dendritic
cells (Figure 4-1).
To better understand the function of the pancreatic draining node dendritic cells, we isolated dendritic cells from both the inguinal node and the pancreatic draining node
and stimulated them in vitro with LPS and anti-CD40. We then assayed the supernatant for soluble IL-12p40, IL-12p70, and IL-10. Both IL-12p70 and IL-10 were undetectable
with our techniques. IL-12p40 was secreted in significantly higher concentrations from
42
p<0.001 myeloid dendritic cells from the inguiI I
2000. nal nodes compared to those of the
1800
1600 pancreatic-draining node (Figure 4-2). o 1400S1200- Metrizamide Gradients Isolate Sloo00o- Mature Dendritic Cells .o
n 800600
S400- Recent publications have
- 200'
0 described transfer of in vitro bone
+LPS +'LPS
Untreated +CD40 Untreated +CD40
+D40 +CD4 marrow-derived myeloid dendritic Inguinal Pancreatic
Node Draining cells to NOD mice, and their potential Node
Figure 4-2. PLN derived dendritic cells to prevent disease (Papaccio et al., produce significantly less IL-12p40 than those from the inguinal node. Myeloid 2000). Interestingly, although these dendritic cells were purified from female dendritic cells were generated in vitro, NOD mice over metrizamide gradients as described, then cultured for 48 hours in the authors chose to "clean up" their complete RPMI with the indicated stimuli. Supernatants were harvested and assayed preparations prior to transfer using by ELISA. IL-12p70 and IL-10 were unde- metrizamide gradients similar to those tectable. Represents the mean of three independent experiments, analyzed by paired used during our ex vivo transfer expert-test.
iments. We examined in vitro bone
marrow-derived myeloid dendritic cell populations before and after passing over a metrizamide gradient, and found that only mature myeloid dendritic cells remain at the
interface (Figure 4-1). This indicates all previous experiments implicating dendritic cells
in protection from diabetes have utilized mature myeloid dendritic cells.
NOD Myeloid Dendritic Cells Do Not Develop, Mature, or Function Properly
With mounting evidence implicating mature myeloid dendritic cells as regulators
of autoimmune diabetes, we generated dendritic cells in vitro from C57BL/6 and NOD
mice, and then examined the phenotype and function of these dendritic cells.
43
(a) + GM-CSF and IL-4 Only + LPS + anti-CD40
o. 0
C57BL/6 W 2 0 1
0o co 0
0 2 3, ro, ;1;1 3 104 oo -';2 3 1,4
10 10 10 10 10 100 10 10 10 10 10 10 10 10 Class II FITC Class II FITC Class II FITC
0. %N
NOD
10 10 1 1;2 103 14 0 0 101 102 13 104 10 101 102 103 10 Class11 FITC Class II FITC Class II FITC
(b) *
*
80
70 S60 ~50
2 40
S30
( 20
10
0
Figure 4-3. NOD dendritic cells do not respond to maturation stimuli. (a) Bone marrow derived dendritic cells from C57BL/6 and other control strains give a 40-50% combination of mature to immature dendritic cells when generated, increasing to 80-90% after addition of maturation stimuli such as LPS or anti-CD40. Dendritic cells from NOD cultures develop as almost completely immature cells, and remain resistant to these maturation stimuli in vitro. (b) Cummulative data from four experiments showing the percent mature dendritic cells found in culture. indicates p<0.001 as determined by paired t-test.
Atypical Development of Myeloid Dendritic Cells from NOD Bone Marrow Cultures
Culture of bone marrow from non-autoimmune mouse strains (C57BL/6, CBA,
Balb/c, etc.) with GM-CSF and IL-4 yields a highly heterogeneous culture after 96
hours composed of equivalent numbers of mature and immature myeloid dendritic cells.
Identical cultures from NOD mice yield vastly different results. NOD cultures yield only
44
15-25% mature myeloid dendritic cells after initial culture. In addition, a third population of dendritic cells, expressing high levels of costimulatory molecules, but lacking high surface MHC class II expression, were found in NOD cultures (Figure 4-3a). We refer to this atypical third population as pseudo-mature, for reasons that will be demonstrated later.
NOD Dendritic Cells Do Not Respond To Maturation Stimuli
To assess whether these excess immature NOD dendritic cells would respond to
maturation stimuli, cultures were treated with either 1 gg/mL LPS, or 5 gg/mL anti-CD40 and incubated an additional 24 to 48 hours. Cells were then assessed for percent mature dendritic cells in each treatment. This showed significant differences between C57BL/6 and NOD cultures for all maturation stimuli tested, with NOD cultures undergoing significantly less maturation than their C57BL/6 counterparts (Figure 4-3). The pseudomature population was not affected by any of these stimuli. Dendritic cells generated from additional non-autoimmune strains, Balb/c and CBA, showed a phenotype and maturation pattern identical to C57BL/6 cultures (data not shown). MHC Instability Does Not Affect NOD Dendritic Cell Phenotype
Based on several reports of the inherent instability of the NOD I-AF7 molecule, we questioned if this could lead to the maturation resistance, or the presence of the psuedomature population seen in NOD dendritic cell cultures. Cultures were established using NOD.B10-H-2b congenic mice, which carry I-Ab from the C57BL/10. This strain showed no improvement in dendritic cell maturation, and maintained the psuedo-mature population seen in stock NOD cultures (data not shown). These experiments were repeated no less than 10 times.
45
a
0- M
0
o ... i 1 ... -2 ... 3 4 --lO ... 1 ... -2 I3
10 10 10 103 104 00 101 10 103 104 Class II APC CD86 APC
b
6007 12000 8000 7000
500- 10000 70
*6000
-400- 5 8000 E
18000
-300 0 6000 400024 3000200 000,- 2000100 2000 10001 0 0,
Figure 4-4. Phenotypically immature dendritic cells are functionally immature. (a) After 4 days of culture, NOD dendritic cells were incubated with rhodamine green dextran for 2 hours. Cells were then stained with antibodies to Class II MHC, CD1 c, and CD86. A gate was set on CD1 lc+ events to eliminate debris, and those cells displayed. M are mature dendritic cells, I immature, and P-M are pseudomature; (b) cytokines present in culture supernatants from C57BL/6 and NOD dendritic cells treated with maturation stimuli for 24 hours. indicates p<0.01 as determined by by paired t-test.
NOD Dendritic Cells With Immature Phenotype Take Up Particulate Antigen
Though cells we described as phenotypically immature fit previously published
descriptions of immature dendritic cells (CD 11 c+, moderate surface class II MHC, low
or no costimulatory molecule expression), it was necessary to verify whether these cells
functionally resembled immature dendritic cells. Cells from culture were incubated with
Rhodamine Green-Dextran for 4-6 hours, and then stained for MHC class II, CD 11 c, and
CD86. This demonstrated that cells classified phenotypically as immature dendritic cells
maintained high levels of antigen uptake, consistent with previous reports of immature
dendritic cell function. These studies also revealed that the atypical third dendritic cell
population, which appears phenotypically mature with the exception of surface class II
46
-1- **
200
i180
av) 1600140
120
c 100a 80
60
S40
C 200
0
500 10 100 10
LJL V J 0 C/) 0
CD 0 C
j _j
(0 + C) + +
to + X4 C4
ZO ZOZO Z 0 6
* ** z 0 0
* *.._____z 0 Z *Z Z
Fiueo-.o O dnrii cell ar porsiuors ofalgnioocls 5B/
3000 _500- 700 700
yd ir400 C 220 0- 2 00
a p 300m (400FIi 400g 0106 a, 300 .300
Sw ooo 0 200 200
Z500 Z100 Z7 Z100
0 -0
i W 4. D 0 C 0 0 q W p0 V 0adO bddt Cw o d F Cabeled aloa A T c
(0+ N a + 0( T NC + ( + C'JC + 0 + 0 +
of a ratio typcy pdoducell the c es, we coude tht + fewer NOD denditi
COsNCO CrC4CO Cundero Ci mCOaCn
as 0 a I -4)0 z (0 z z0z 0 z z z z
Figure 4-5. NOD dendritic cells are poor stimulators of allogeneic T cells. C57BL/6
and NOD.H2b dendritic cells were cocultured with CFSE labeled allogeneic CBA T cells
at a ratio of 1:100 (dendritic cells per T cell). At 96 hours, proliferation was assessed
by flow cytometry and cytokine concentrations were measured by ELISA. All conditions
yielded significant differences between C57BL/6 and NOD cultures. This data represents
the mean of three independent experiments. indicates p<0.01, ** indicates p<.Oit as
determined by paired t-test.
expression, does not take up antigen in vitro. This leads to our classification of these cells
as pseudo-mature (Figure 4-4).
Supernatants from dendritic cell cultures matured with a variety of stimuli were
assayed for cytokine concentrations at 24 hours, and supported the assertion that NOD
dendritic cells were less mature. NOD cultures produced significantly less IL- l2p70 and
EL-l10 when compared to C57BL/6 (Figure 4-4). As cells progressing through the stages
of maturation typically produce these cytokines, we conclude that fewer NOD dendritic
cells are undergoing maturation.
47
Cytokine Dosage Does Not Affect NOD Dendritic Cell Development
We noted early in our experiments that high doses of GM-CSF were capable of limiting maturation of C57BL/6 dendritic cells prior to addition of maturation stimuli (data not shown). The possibility existed that NOD stem cells could be more or less responsive to the combination of GM-CSF and IL-4 than their C57BL/6 counterparts. NOD bone marrow cultures were established with concentrations of GM-CSF and IL-4 ranging from 100U/mL (1/5 the standard GM-CSF dose and 1/10 the standard IL-4 dose) up to 5000U/mL (10x the standard GM-CSF dose and 5x the standard IL-4 dose). Cytokine concentrations were checkerboarded so multiple combinations were realized. Cultures were analyzed before and after addition of maturation stimuli, and no significant increase in maturation of NOD dendritic cells was seen with any increase or decrease in cytokine concentration (data not shown).
NOD Dendritic Cells are Poor Stimulators of Allogeneic T Cells
As further support, NOD dendritic cells were cultured with MHC mismatched T cells and cytokine production and proliferation assessed. Again utilizing the NOD.H-2b congenic, we cultured C57BL/6 and NOD.H-2b dendritic cells with CBA CD4+ T cells, and at 96 hours, the previously established point of greatest effector cytokine concentration, measured cytokines in the culture supernatants. C57BL/6 dendritic cells stimulated significantly higher levels of proliferation, IL-2, IFN-y, and 1L-4 from CBA T cells than did similar NOD.H-2b dendritic cells (Figure 4-5). Investigating Causes of NOD Maturation Abnormalities
Microscopy of NOD Dendritic Cell Populations
To better understand the pseudo-mature dendritic cell population, we sorted and stained cells from NOD cultures and examined them by light microscopy. This revealed that morphologically, the pseudo-mature dendritic cells resemble typical mature dendritic
48
C57BL/6 Mature Population NOD Mature Population
NOD Pseudo-Mature Population NOD Pseudo-Mature DC
Figure 4-6. Microscopic analysis of dendritic cell populations. In upper panels and lower left panel, cells from dendritic cell cultures were cultured, stained and sorted, then cytospins were prepared from the sorted cell populations. Slides were stained with Wright's stain and viewed at 100x. Lower right panel represents cells unique to NOD dendritic cell cultures. Cells are stained for class II MHC (green), CD86 (red), and DNA (DAPI), then visualized by deconvolution microscopy.
49
(a) C57BL/6 NOD dendritic cells from both C57BL/6 and
S 55% 9 21% NOD cultures (Figure 4-6). Additionally, we
" 8 &adhered unsorted cells from C57BL/6 and on0 .o 0 o7o, 10 1; 1o NOD dendritic cell cultures to alcian blue
Class II FITC Class 11 FITC
(b) +G4 + anti-CD40 coverslips, stained the cells with antibodies S~30% 0 29%
U against class II and CD86, then visualized
0 on the deconvolution microscope. On all
Z
o..... -01,0 ,ato to '" slides from NOD cultures, we visualized a + (C) C57BL/6 NOD
It population of cells expressing large amounts
0 58% 5
U D of CD86, but little class II. These cells were
Z
Never visualized on slides from C57BL/6 bone marrow cultures. We concluded that Figure 4-7. Persistent developmental defects of the NOD dendritic cell lin- these must be the psuedo-mature dendritic eage. Dendritic cell cultures were established using (a) transwell systems, (b) cell population. These cells harbor little bone marrow from NOD.SCID mice, and
(c) p rifiedrC +m c noe, od intracellular class II, indicating the lack of
(c) purified CD34+ stem cells, none of which lead to proper development of surface class II expression on these cells NOD dendritic cells. Number indicates percentage CD86 high dendritic cells, cal- is related to impaired transcription and/or culated as the mean of three replicate translation of class II genes, and not an experiments.
inability to bring class II peptide complexes
to the cell surface (Figure 4-6).
Soluble Factors Do Not Direct NOD Dendritic Cell Development
To assess the role of soluble factors in dendritic cell development, NOD and
C57BL/6 dendritic cell cultures were established either above or below a 0.2 mm
membrane, fed and matured with LPS as before, then analyzed for phenotype on day
5 of culture. NOD dendritic cells cultured either above or below C57BL/6 dendritic
cells maintained a mostly immature phenotype. Additionally, the pseudo-mature phenotype was not diminished. Conversely, C57BL/6 dendritic cells cultured with NOD den-
50
1400- dritic cells showed no inhibition of mat12000
'" uration, nor induction of the pseudoeoo
4000 mature dendritic cell population (Figure
2 0
4-7a).
CS7BU6 NOD
T, B, and NK-T Cells Do Not Affect NOD Dendritic Cell Development
E 4000 1400 S2500
300 l,000
2000M =o We established cultures using
a15004 c
soo 200 bone marrow from the NOD.SCID
u v mouse, which lacks T, B, and NK-T
C57BlU6 NOD C578U6 NOD
Figure 4-8. NOD dendritic cells undergo a cells. Cultures derived from NOD.SCID prolonged "maturing" phase. Spontaneously mature dendritic cells from day 4 GM-CSF mice were similar phenotypically to and IL-4 cultures were harvested and purified standard NOD cultures. These cells over 14.5% metrizamide gradients. Mature dendritic cells were then cultured with the were slightly, but not significantly, more indicated activating stimuli. After 48 hours, supernatants were harvested and cytokines responsive to maturation with LPS. The measured by ELISA. NOD dendritic cells pro- pseudo-mature dendritic cell population duced significantly lower levels of IL-10 and significantly higher levels of IL-12p70 than was unaffected (Figure 4-7b), establishC57BL/6 controls. Data represents three replicates. indicates p<0.01, ** indicates p<0.001 ing that direct interactions with T, B, as determined by paired t test. or NK-T cells are not responsible for
limiting the ability of NOD dendritic
cells to fully mature, and play no role in induction of the psuedo-mature dendritic cell
population.
A Non-lymphocytic Cell Population Leads to Generation of Pseudo-Mature Dendritic Cells
CD34+ stem cells were enriched by negatively selecting lin+ cells, and then
cultured with GM-CSF and IL-4 as before. Though the ability to respond to LPS
or anti-CD40 did not improve in NOD cultures, the pseudo-mature population was
completely eliminated (Figure 4-7c). This indicates either that another cell population
51
influences development of the pseudo-mature dendritic cell population, or the population itself is derived from a cell type eliminated during stem cell enrichment. NOD Dendritic Cells Become Arrested During Maturation
Although the majority of NOD dendritic cells remained immature, a small population did mature without addition of maturation stimuli. We questioned whether these "mature" dendritic cells would have similar function to mature dendritic cells from C57BL/6 cultures. Utilizing the previously described property of the metrizamide gradient to isolate mature myeloid dendritic cells, we ran day 4 dendritic cell cultures from NOD and C57BL/6 over gradients to isolate those dendritic cells that had acquired the mature phenotype. These dendritic cells were plated and treated with the combination of LPS and anti-CD40 for 48 hours, which provides the maximual activation/maturation stimulus.
Data revealed that mature dendritic cells from C57BL/6 cultures produced cytokines consistent with the fully mature phenotype previously described, producing low levels of IL-12p70 and high levels of IL-10 (Figure 4-8). In contrast, "mature" dendritic cells from NOD cultures produced cytokines characteristic of the "maturing" dendritic cell, with very high levels of IL-12p70 and low levels of 1-10 (Figure 4-8). Genes Within IddlO Affect NOD Dendritic Cell Maturation
Screening of NOD Congenic Strains
We are fortunate to have a large number of NOD congenic strains available to us, allowing rapid screening of several reported diabetes susceptibility intervals in the maturation of myeloid dendritic cells. This preliminary screen revealed the IddlO interval on chromosome 3 as a candidate for further study.
52
IddlO Affects Dendritic Cell Maturation
We established dendritic cell cultures utilizing NOD, NOD.IddlO, and
C57BL/6.cl17 (I-Ag7), giving us identical MHCs on all three strains to be tested. These
cultures were assessed for phenotype, cytokine production, and stimulation of allogeneic
T cells. NOD.IddlO cultures had higher baseline numbers of mature dendritic cells, and
responded significantly better to maturation stimuli than NOD dendritic cells (Figure 4-9).
IddlO Affects Dendritic Cell Cytokine Production and T Cell Stimulatory Capacity
Analysis of supernatant from these cultures revealed that NOD.IddlO dendritic
cells produce significantly higher levels of IL-12p70 and IL-10 than the NOD following
addition of LPS. Culture of these dendritic cells with CD4+ CBA T cells showed that
90*
08 600 16
76* 70500 El
6 1h
* 400 *
000
3 '6 200
22
1 0 -100 2 10 0-2 50
S_ z L 2 oo a
ZOO zO zO 0+0 + 0
z0 + +
Z ~ ~ u0 o 26 :Z0a
etbih ui ND No nis w n 1400 z z
'6 ; Co 000
600d .1200 w 3000 EI5oo* Eoo 2r500 8C:O
.*400 800 020000 *
~1500 ___~300- 600-_
C111
2oo0 0,,o ,.00
100-.: 20 500
0 0 0 3
+0 + .0+ + + +.o ,0 + .a + 0 -+ 0 0+ .0 + + 0 0+00)+ + +0 0+0 + to zo!V; (U ZL Z 0-v
zo Z Zo z z z
Figure 4-9. Genes within IddlO affect NOD dendritic cell maturation. Cultures were established using NOD, NOD.IddlO, and B6.C17 mice so no allotypic differences existed between the strains. Dendritic cells were assessed for percent mature dendritic cells, and cocultured in an allogeneic MLR with CBA CD4+ T cells at 1:100, and cytokines measured in the supernatants. Graphs represent the mean of three independent experiments. indicates p<0.01, ** indicates p<0.001 as determined by paired t-test.
53
NOD.IddlO dendritic cells were significantly better at stimulating cytokine production from allogeneic T cells, and in turn produced more cytokines in response to the activated T cells (Figure 4-9).
Summary and Discussion
The NOD suffers a polygenic immune dysfunction, manifesting itself as a propensity for development of autoimmune diseases such as diabetes and Sjogrens-like syndrome. This predisposition to develop autoimmune disease has previously been attributed to two factors; atypical T cell activation and resistance to AICD within the T cell compartment, and the structure of the unusual Class II MHC carried by the NOD lacking aspartic acid at position (357 (Corper et al., 2000; Jaramillo et al., 1994; Kanagawa et al., 1998; Quartey-Papafio et al., 1995; Rapoport et al., 1993a; Rapoport et al., 1993b; Zipris et al., 1991). In this paper, we add a third avenue, that antigen presentation in the NOD mouse is insufficient for effective peripheral tolerance due to inadequate dendritic cell development and maturation.
Dendritic cells play a vital role in the initiation of immunity to pathogens in
peripheral tissues. Immature dendritic cells are found in all peripheral tissues, and do not interact with large numbers of T cells, as they lack significant levels of MHC-peptide complexes, costimulatory molecules, and adhesion molecules required for these interactions. Following antigen uptake and receipt of some activating signal(s), these immature dendritic cells undergo a well-orchestrated maturation and subsequent migration to the draining lymph node. Within the node, the dendritic cell is able to interact with large numbers of naive T cells and rapidly initiate an appropriate clonal T cell response.
In the pancreas of juvenile NOD mice, dendritic cells are found within the islets (Faveeuw et al., 1994a). Dendritic cell accumulation within the islet may attract T cells and prime them from within the pancreas itself (Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Shinomiya et al., 2000). A similar infiltration and interaction is
54
seen within the salivary glands, which are also destroyed in NOD mice (Faveeuw et al., 1994a).
We have also demonstrated that NOD dendritic cells are refractory to typical maturation stimuli in vitro. Myeloid dendritic cells normally produce high levels of IL-12p70 while maturing, but down regulate production once fully mature. The fully mature dendritic cell produces large amounts of IL-10, but little or no IL-12p70 (Langenkamp, 2000). This correlates with an initial priming of Th 1 type responses, followed by priming of Th2 type responses as IL-12 production ceases and IL-10 production increases. As the dendritic cell reaches maturity, it has upregulated CCR7 and traversed the afferent lymphatics to the draining node, and has started secreting large amounts of IL- 12p70, capable of priming Th 1 type responses in the draining lymph nodes. After a period of time, the details of which are only now being elucidated, these dendritic cells switch to an IL-10 producing phenotype (Langenkamp, 2000). In the NOD mouse, which is predisposed to developing a Thl -mediated autoimmune pathology, this lack of transition results in impaired migration and continued IL-12 secretion, and may prove critical in propogating islet inflammation while limiting generation of regulatory cell populations within the pancreatic draining lymph node.
In juvenile NOD mice, we presume an underlying genetic defect leads to death of
3 cells, leading to upregulation of chemokines and adhesion molecules, and infiltration of dendritic cells (Faveeuw et al., 1994a). Normally, these dendritic cells would enter the pancreas, endocytose antigen from dead and dying cells, receive maturation signals from the damaged 0 cells, and then mature and migrate to the draining lymph node where they mediate tolerizing interactions with autoreactive T cells. This is seen to some extent by the fact that cells from the pancreatic draining lymph node do not transfer disease, and dendritic cells from that node are protective when transferred to juvenile NOD mice (Clare-Salzler et al., 1992). In our model of diabetes in the NOD mouse, dendritic cells within the islet that should mature and migrate to the draining node instead enter an
55
extended maturing state. They may be unable to upregulate CCR7 and exit the pancreas, and continue to produce high levels of IL-12p70 and little or no IL-10. These dendritic cells, now effectively trapped within the pancreas by defective maturation, may release chemokines, such as MIP-33, that further attract lymphocytes and immature dendritic cells to the area (Sallusto et al., 1999). From this location, they may prime naive autoreactive T cells, and perpetuate P cell destruction (Ludewig et al., 1998).
Authors of previous papers have argued that dendritic cells from NOD mice
are more effective at stimulating T cells than those from control strains (Feili-Hariri et al., 1999; Morel et al., 1999; Papaccio et al., 2000). As has been published, however, increasing T cell activation in NOD mice actually prevents disease (Arreaza et al., 1997; Forster and Lieberam, 1996; Serreze et al., 1989). In addition, T cells from the pancreatic draining lymph node in the NOD mouse show no signs of overt activation, the disease process takes 12-20 weeks to manifest itself, and the memory/na'ive phenotypes seen are not indicative of typical ongoing immunostimulation (D'Adamio et al., 1993; Fabien et al., 1995; Forster and Lieberam, 1996).
We argue that in light of these previous experiments, as well as those included in this paper, the more acceptable hypothesis is that maturation of NOD dendritic cells is inadequate, thus they fail to migrate appropriately to draining lymph nodes, and do not activate and expand antigen-specific regulatory cell populations. Our studies have clearly shown that bone marrow derived dendritic cells from NOD mice fail to mature in response to LPS and anti-CD40, and develop dendritic cell populations with atypical phenotypes in comparison to control strains. Studies by Delovitch demonstrated that treating NOD mice with anti-CD28 antibodies was sufficient to prevent disease (Arreaza et al., 1997), suggesting that antigen-presenting cells responsible for expressing CD28 activating molecules such as CD80 and CD86 are dysfunctional in NOD mice. The only cells capable of presenting MHC-peptide complexes along with high levels of CD80/86 to T cells in the absence of exogenous stimuli are dendritic cells. Salomon and
56
Bluestone demonstrated that blocking CD28 signaling exacerbated diabetes in NOD mice, impairing homeostasis of CD4+CD25+ regulatory T cells. This finding suggests that some endogenous capacity to deliver signals through CD28 remains intact in the NOD mouse (Salomon et al., 2000). Work from Stockinger showed that ligation of additional TCRs was sufficient to prevent diabetes in NOD mice (Fossati et al., 1999). Since we know ligation of CD28 decreases the number of TCRs required for T cell activation, we can relate the capacity of APC to transduce sufficient signal to the T cell with the prevention of autoimmune disease (Gudmundsdottir et al., 1999; Schrum et al., 2000; Wells et al., 1997).
The question of which function these dendritic cells fail to carry out is of considerable interest. It is well documented that diabetes in the NOD mouse is dominated by Thl type cytokines such as IFN-y (Bradley et al., 1999; Sarvetnick, 1997). Treatment of NOD mice with Th2 type cytokines, such as IL-4, prevents disease (Arreaza et al., 1997; Cameron et al., 1997b; Gombert et al., 1996b; Maron et al., 1999; Mueller et al., 1997; Rapoport et al., 1993a; Teros et al., 2000). Some treatments that prevent disease, such as the previously mentioned anti-CD28 therapy, are ineffective when anti-IL-4 antibodies are administered, indicating a substantial role for IL-4 induction in these prevention strategies (Arreaza et al., 1997). Several studies have shown that before a T cell can produce IL-4, it must undergo several cell divisions (Bird et al., 1998; Lederer et al., 1996; Schweitzer et al., 1997). This requires both the ligation of large numbers of TCRs and simultaneous costimulation. An additional requirement is to be where the action is, i.e., the T cell areas of the secondary lymphatics. As the myeloid dendritic cell fits these requirements, it then follows that the myeloid dendritic cell is the responsible (or irresponsible) party in this scenario.
a-GALACTOSYLCERAMIDE ACTIVATES AND EXPANDS NK-T CELLS AND
RECRUITS MYELOID DENDRITIC CELLS SPECIFICALLY WITHIN THE
PANCREATIC DRAINING NODE OF NOD MICE
Background
NK-T cells represent a bridge between the innate and adaptive immune systems. Though they express a T cell receptor and the molecule CD3, they lack subclass specific markers such as CD4 and CD8. These cells also express a wide array of factors typically associated with NK cells, such as perforin and granzyme, and surface antigens such as NKl.1 and Ly49C (Bendelac et al., 1997).
The invariant TCR of the NK-T cell is unique in that it interacts with the MHC Class I-like molecule CDld (Bendelac et al., 1995). The CDld molecule is found on many cell types, but its expression on dendritic cells is most crucial for activation of NK-T cells. First, dendritic cells appear to express higher levels of the CD1d molecule, and are well suited to process and present antigen in the context of this molecule. Second, the NK-T cell must receive costimulatory signals through CD28, just as a naive T cell would (Bendelac et al., 1997).
The invariant TCR specifically recognizes glycolipid antigens within the context of CD1d (Brossay et al., 1998a; Brossay et al., 1998b). The synthetic glycolipid a-galactosylceramide binds to CDI d and acts as a ligand for the invariant TCR, and treatment of mice with this glycolipid leads to activation and expansion of NK-T cells (Brossay et al., 1998b).
Once encountering its ligand, the NK-T cell rapidly secretes large amounts of the cytokines IFN-y and IL-4. Upon restimulation by its ligand, the NK-T cell orients to an IL-4 secretion pattern (Chen and Paul, 1997; Chiu et al., 1999; Hammond et al.,
57
58
1999). Many studies have shown the positive effects of IL-4 treatment on Type I diabetes (Berman et al., 1996; Cameron et al., 1997a; Fox and Danska, 1997). The NOD mouse has been shown to have an intrinsic defect which limits the number and function of NK-T cells in the periphery (Falcone et al., 1999; Gombert et al., 1996a). Increasing the number of NK-T cells has proven to protect against development of disease (Hammond et al., 1998; Lehuen et al., 1998).
Previous studies support that classical T cells provide signals to the dendritic cell, e.g. CD40L, which promote dendritic cell activation. However, it is not known whether the NK-T cell-dendritic cell interaction leads to a similar activation of the dendritic cell. As NK-T cells recognize the glycolipid-CD 1 d complex presented by dendritic cells and are activated, how might this affect the dendritic cell? A recent study has shown that dendritic cells from mice treated with a-galactosylceramide do in fact show increased IL-12 production (Kitamura et al., 1999).
In previous studies, we have shown that myeloid dendritic cells from the NOD mouse do not develop and mature properly in vitro. We also demonstrated an elevated level of the molecule CD1d on dendritic cells specifically within the pancreatic draining node. Recent studies also demonstrated increased numbers of NK-T cells in this lymph node in contrast to other lymph nodes draining non-inflamed sites (Laloux et al., 2001). Many studies have reported the early entrance and persistence of dendritic cells within the pancreatic islets (Dahlen et al., 1998; Faveeuw et al., 1994a; Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Papaccio et al., 1999b; Rosmalen et al., 2000a; Rosmalen et al., 1997; Rosmalen et al., 2000b; Shinomiya et al., 2000). Results from our collaboration with Dr. Brian Wilson (Beth Israel Deaconess Medical Center, Boston, MA) suggest that there are few NK-T cells in the islets of NOD mice in contrast to genetically similar, disease resistant strains (NOR). We hypothesized that if activated NK-T cells exited the pancreatic draining node, their entry into inflamed islets may provide maturation signals to resident dendritic cells that promote their migration to the
59
pancreatic draining node. If so, this would suggest a novel role not yet described for NK-T cells, that of promoting dendritic cell emigration and resolution of inflammation.
Materials and Methods
Mice
Female NOD/LtJ mice 6-8 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center. Treatment with a-galactosylceramide
Female NOD/LtJ mice were injected with a-galactosylceramide (KRN7000, Kirin Brewery, Gunma, Japan) or cx-mannosylceramide (AGL595, Kirin Brewery, Gunma, Japan) at a concentration of 10 gg/mL in 0.5% Tween/ PBS. Two i.p. injections of 250 pgL each were given, the first on day 0 and the second on day 5. Lymph nodes were extracted from groups of 3 mice on either day 10 or day 20. Single cell suspensions were created as described previously (Chapter 3 materials and methods). a-Galactosylceramide In Vitro Recall
Mice were treated, nodes harvested and single suspensions created as previously described, but cells were washed in HBSS without calcium and magnesium and 1% BSA, but ImM EDTA was not added to preserve complete function of the isolated cells. Cells were plated at 2 x106 cells per mL in RPMI complete media with a lx concentration of a-galactosylceramide. The exact concentration of u-galactosylceramide was not known, but was titered in the lab to give optimal results, and that optimum concentration was then considered to be l x. After 48 hours, supernatants were harvested and assayed for cytokine concentration.
60
ELISA for Cytokine in Culture Supernatants
Cytokine levels in culture supernatants were analyzed using products from
Pharmingen (San Diego, CA) unless specified otherwise. The following antibody pairs were utilized for cytokine detection, listed as capture and detection respectively; IL-2 (JES6-1A12 & JES6-5H4), IL-4 (BVD4-1D11 & BVD6-24G2), IFN-y (R4-6A2 & XMG1.2), L1-10 (JES5-2A5 & SXC-1), IL-12p40 (C15.6 & C17.8), IL-12p70 (9A5 & C 17.8), A streptavidin-HRP secondary was added, followed by addition of a TMB substrate. The reaction was stopped using IN H2SO4 and read at 450 nm on a 3550-UV microplate reader (Bio-Rad, Hercules, CA). Analysis of Surface Phenotype
Surface markers were analyzed by flow cytometry. Briefly, cells were washed
in HAB (calcium, magnesium, and phenol free HBSS with 1% BSA, 0.1% NaN3, and 1 mM EDTA), 5x 105 cells were then aliquotted to tubes for staining. Cells were blocked in 100 gL of HAB with 4% BSA+I4g anti-CD16/32 (purified clone 2.4G2), and then stained with the appropriate surface antibodies. Antibodies used were: CD 11 c (clone HL3, Pharmingen), I-Ak (A-k) (clone 10-3.6, Pharmingen), CD8a (clone Ly-2, Pharmingen), CD 1 d (clone Ly-38, Pharmingen), CD4 (clone RM4-5, Pharmingen), TCR-b (clone H57-597, Pharmingen), CD19 (clone 1D3, Pharmingen). Biotinylated antibodies were detected with streptavidin-APC (Molecular Probes, Eugene, OR). Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry, San Jose, CA).
61
Results
Treatment of NOD Mice with a-Galactosylceramide Expands NK-T Cells Specifically Within the Pancreatic Draining Node
Mice treated in vivo with o-galactosylceramide were sacrificed and inguinal and pancreatic draining nodes were extracted. Single cell suspensions were created as before, and cells were stained with antibodies against CD4, CD8, T cell receptor (TCR), and CD 19. Since NOD mice are non-reactive with the antibody against NK1.1, the population of TCR+, CD4-CD8- cells is the only means to approximate NK-T cell number in these mice.
Upon opening the peritoneal cavity of the treated mice, the pancreatic draining
lymph node is grossly enlarged. While the inguinal node remains unremarkable compared to untreated mice, the pancreatic draining node has almost doubled in cell number (Figure 5-1). No significant difference in cell number was seen between treated and untreated mesenteric lymph nodes p
25000000 P
Phenotypic analysis of these
200000000 cells reveals significant increase in the
S0o0o0oo number of TCR+, CD4-CD8- events
5000000 from the treated pancreatic draining
0
.5 .- node, with no detectable increase in a these cells within the inguinal nodes Figure 5-1. Increased cellular yield from (Figure 5-2). As stated before, this is a-galactosylceramide treated pancreatic nodes. Mice were treated as described with the only technique of phenotypically two injections of a-galactosylceramide assessing these cells in NOD mice, (a-GC), then inguinal and pancreatic draining nodes (PLN) were harvested, dissoci- as they lack the NK1.1 antigen found ated, and cells counted on a hemacytomein other strains. These results have
ter. This data represents five separate experiments, p value calculated by paired t-test.
62
10 Days After Initial Treatment 20 Days After Initial Treatment been confirmed by our coli5- 5C 4 4 laborator Dr. Wilson, who Z 3 has demonstrated increased
o 21 2
4 d I 0 W numbers of NK-T cells by
Vehicle -manosyl -galactosyl Vehie n-manosyl a-galacosyl
S Only ceramide ceramide Only ceramide ceramnide quantitative mRNA analysis
+o 05 5. quantitative mRNA analysis
a: 2 4 4o of the invariant T cell recepC 2 tor. 0,,0
S Vehi myl dyl 0 Vehi -manosal a y Increased Functional
Only cm.d cmid Only camide cmid. Responses Ex Vivo FollowFigure 5-2. Expansion of double negative T ing o-Galactosylceramide cells within the pancreatic draining node after Treatment c-galactosylceramide treatment. Mice were treated with two injections, as described, nodes harvested, dissociated, counted and stained, then the number To better assess the of CD4 negative CD8 negative TCR positive events effect of a-galactosylwere enumerated. CD19 was also included to eliminate B cells from any of these analyses. Results rep- ceramide on in vivo NK-T resent the mean of five replicate experiments. indicates p<0.01 as determined by paired t-test. cell number and function, the functional characteristics
of NK-T cells from NOD
mice treated with this lipid were assessed in vitro by recall response.
Following in vivo treatment with doses of a-galactosylceramide or vehicle alone, pancreatic draining and inguinal lymph nodes were extracted, dissociated, and single cell suspensions were plated with or without a-galactosylceramide. After 96 hours, supernatants were harvested and soluble cytokines measured by ELISA.
A 5-10 fold increase in IL-4 and IFN-y production by cells from the pancreatic draining lymph node, but not from the inguinal node, were observed in response to in vitro a-galactosylceramide recall (Figure 5-3). These increases were only noted in mice previously primed in vivo with ca-galactosylceramide, indicating that NK-T cells were indeed activated and/or expanded within the pancreatic draining lymph node by a-galactosylceramide treatment.
63
V=Vehicle only
(a=treated in vitro with a-galactosylceramide
1X=2.5mg a-galactosylceramide 2x=5.0mg a-galactosylceramide,
350 1600S300- 1400250 1200250.
1000
200- o
0
= 800
%. 150-
S 600
100
Z 400-.
50- 2000 0 I
Invivo V 1X 2X V 1X 2X V 1X 2X V lX 2X Invivo V 1X 2X V lX 2X V 1X 2X V 1X 2X In vitro V V V a t a v V V a a In vitro V V V a C a V V V c t a
Inguinal PLN Inguinal PLN
600- 1200S500- 1000
400- 800
o
S300- 0 600S200- 400v ol
100 7 200
0 00
Invivo V 1X 2X V 1X 2X V 1X 2X V IX 2X Invivo V 1X 2X V IX 2X V 1X 2X V 1X 2X In vitro V V V a a a V V V a a a In vitro V V V a (a a V V V a a a
inguinal PLN Inguinal PLN
Figure 5-3. In vitro recall response to in vivo administered a-galactosylceramide.
Mice were treated as described with vehicle or doses of C-galactosylceramide. On day 10, lymph nodes were harvested, dissociated, and 2x106 cells cultured per well in 24 well plates for 96 hours with the indicated stimulus. Supernatants were harvested and cytokine concentration measured by ELISA. Data from one representative experiment of 3 replicates.
In addition, we observed a dramatic increase in production of IL-10, with a dose
dependent decrease in GM-CSF production after treatment with a-galactosylceramide
(Figure 5-3). Secretion of these cytokines after a-galactosylceramide treatment (presumably directly from, or induced by NK-T cells) has not been previously described in the
literature. Secretion of high amounts of GM-CSF is of particular interest in the context of
its ability to limit dendritic cell maturation in vitro (see chapter 4).
64
Increased Numbers of Myeloid Dendritic Cells in the Pancreatic Draining Node After a-galactosylceramide Treatment
To determine if these expanded and primed NK-T cells would have any effect
on dendritic cell emigration from the islets, dendritic cell number within the pancreatic
draining and inguinal nodes was assessed both with and without a-galactosylceramide
treatment. We also stained sections of pancreas for CD 1 c to assess the number of
dendritic cells within the inflamed islets after a-galactosylceramide treatment.
Knowing that dendritic cells located within the pancreas are myeloid dendritic
cells (lymphoid dendritic cells are not found in peripheral tissues), the percent and
absolute number of myeloid and lymphoid dendritic cells was determined, considering all
of population A and the CD8a negative component of population B as myeloid dendritic
cells. This analysis showed a significant increase in the percentage of myeloid dendritic
cells within the pancreatic draining node, and a greater than 2 fold increase in the
Inguinal Nodes Pancreatic Draining Nodes
400000 400000350000 350000O
30000ooo0 300000oooo(1) 250000 2500000
200000 200000'2150000 1500000
100000 10000050000 500000 Jm0Fc o o mo d 5 t p a o . k .. a. z
E E E E E : E
Vehicle a-ManCer c-GalCer Vehicle ca-ManCer ca-GalCer
Figure 5-4. Accumulation of myeloid dendritic cells within the pancreatic draining node following a-galactosylceramide treatment. Following treatment with vehicle, a(-mannosylceramide, or az-galactosylceramide, lymph nodes were harvested on day 10 and the persentage of myeloid and lymphoid dendritic cells was enumerated as previously described. Numbers here represent the absolute number of dendritic cells present within each node, and is the mean of four independent experiments. indicates p
65
700 absolute number of myeloid dendritic
600- cells (Figure 5-4). The percentage
E500
5 oo
S"and absolute number of myeloid
S400
C0- 00- dendritic cells within the inguinal
_ 200 node remained constant with all treat100" ments administered. Histologically,
. 0 () -_ 0 (D 0 A? (0 2 () we observed a complete loss of
.2 .2 0.9 Q) 9 .25 .2 0 .2 0
:E E- E .. .. CD1 c staining within the islets after
Ing PLN Ing PLN Ing PLN
Untreated + LPS +LPS+a-CD40 treatment with a-galactosylceramide Figure 5-5. Increased numbers of
(slides not shown).
myeloid dendritic cells do not lead to increased IL-12 secretion. Whole nodes To determine the functional were dissociated and plated with nothing, LPS, or the combination of LPS and capacity of these dendritic cells, anti-CD40, the maximum stimulus for single cell suspensions from treated IL-12 in our hands. After 48 hours, supernatants were harvested and soluble cytokine and untreated nodes were cultured measured by ELISA. IL-12p70 and IL-10 were undetectable. Represents the mean of with and without LPS and anti-CD40. three replicate experiments. After 48 hours, supernatants were harvested and soluble IL-12 was measured. EL-12p70 and IL-10 were undetectable in these preparations despite the two-fold increase in dendritic cell number.
IL-12p40 was detectable in all cultures, and showed no significant increase
between the a-galactosylceramide treated and vehicle treated pancreatic draining node, despite the significant increase in the percentage of myeloid dendritic cells present in the (-galactosylceramide treated cultures (Figure 5-5). This is in sharp contrast to the inguinal node, where large amounts of IL-12p40 are produced by dendritic cells.
Summary and Discussion
Prior to starting this discussion, it is important to bring some additional data
into this picture. These studies were performed as part of a collaboration with Dr. Brian
66
100
Om ----NO D/veh
e ---NOD/aGal
(1) 80- A- NOD/CDKO/veh
- # -NOD/CD1 KO/caGal
.0 60
'-- 40
0
U 0
c 16 20 24 28 Age (weeks)
Figure 5-6. Treatment of 4 week old NOD mice with
a-galactosylceramide delays the onset of diabetes. NOD and CD ld deficient NOD females were treated beginning at 4 weeks of age with weekly injections of either vehicle or a-galactosylceramide. This data and graph were generated by the laboratory of Dr. Brian Wilson, Beth Israel Deaconess Medical Center, Boston Massachusettes, and printed
here with his permission.
Wilson and Dr. Yuri Nuamov, and as such, some of their experimental results are critical to understanding the overall significance of these findings.
Treatment of 4-week-old mice with a-galactosylceramide significantly decreased the incidence of diabetes (Figure 5-6). This protection was CDl1d dependent, as treatment of NOD CD1d KO mice with a-galactosylceramide demonstrated no protection from diabetes.
As mentioned earlier, previous studies have demonstrated the persistence of dendritic cells within the inflamed pancreatic islets of the NOD mouse (Jansen et al., 1994; Ludewig et al., 1998). Our previous data have shown that dendritic cell development and maturation from NOD bone marrow is impaired in vitro, and that these cells appear to go through a prolonged maturing phase, with extended high level IL-12 production.
67
That a-galactosylceramide decreases disease incidence should come as no surprise. It's known that this glycolipid binds CD1d and interacts with NK-T cells, leading to their activation and expansion. Since previous studies have demonstrated increasing NK-T cell number and function prevents diabetes, our studies are consistent with these previously published reports (Lehuen et al., 1998). Of particular interest was the significant decrease in survival after treatment with x-mannosylceramide. a-mannosylceramide binds CD1d, but is not an activating ligand for the invariant TCR of the NK-T cell. This suggests that a-mannosylceramide blocks endogenous CD1 d-NK-T cell activation, indicating that some residual NK-T cell activity in the NOD mouse provides some protection from diabetes.
Since we make the assertion that the source of increased dendritic cells in the pancreatic draining node is derived from the population within the pancreas, a logical experiment would be to look for CD1 lc+ dendritic cells within the pancreas with and without treatment. Our lab has performed these experiments, and the results were quite dramatic. While vehicle treated mice showed large numbers of CD 11 c+ cells within the islets. These cells were almost completely lost following a-galactosylceramide treatment. Semi-quantitative RT-PCR was also performed, and showed a loss of IL-12 transcripts within the islets after u-galactosylceramide treatment. While not providing definitive proof, the observation that dendritic cells decreased within the islet infiltrates and increased within the draining node suggests that these cells migrated to the pancreatic draining lymph node.
How activated NK-T cells affect dendritic cell maturation and migration is still unknown, as is the mechanism for how these cells regulate inflammation and autoimmunity. We observed high levels of GM-CSF in the supernatants of pancreatic draining node cells from vehicle treated mice that were treated in vitro with -galactosylceramide. However, when mice were treated in vivo with c-galactosylceramide a majority of this GM-CSF production was lost. Earlier cytokine titration experiments showed that high
68
levels of GM-CSF limited the ability of myeloid dendritic cells to mature. This high level of GM-CSF production may not only limit the capacity of dendritic cells within the islet to mature and migrate to the draining node, but may also limit the capacity of dendritic cells resident within the pancreatic draining node itself to become functionally mature, limiting tolerance induction within this compartment. To date, no other labs have described the production of GM-CSF by NK-T cells, leaving us with little data with which to compare our results. If substantial numbers of NK-T cells are developing in the bone marrow and the liver, but are not receiving sufficient signals via TCR and CD28, these cells may not progress past the point of high GM-CSF production, and may impair myeloid dendritic cell maturation, a condition that is exaggerated within the inflamed microenvironment of the pancreatic draining node.
During in vitro recall experiments, we saw that in vivo primed cells from the
pancreatic-draining node produced high levels of IFN-y. This cytokine has been shown to synergize with other signals to increase dendritic cell maturation, cytokine secretion, and T cell stimulatory capacity, and has been shown to impart tolerizing capacity upon normally non-toleregenic splenic dendritic cells (Shinomiya et al., 1999). The increase in IFN-y production within the microenvironment of the pancreatic draining node and the inflamed islets may be sufficient to drive dendritic cell maturation and allow emigration from the islets.
Studies examining the anti-tumor effects of NK-T cells revealed that activation
with o-galactosylceramide lead to increased IL-12 production by dendritic cells, and that continued IFN-y release by the NK-T cell was dependent upon TCR and IL- 12R-mediated signals from the dendritic cell, as well as CD40L from the NK-T cell (Kitamura et al., 1999). While these studies used CD40 blocking antibodies to establish this latter interaction, no studies have shown the presence of CD40L on the surface of an NK-T cell. These were the first experiments to demonstrate feedback from the NK-T cell to the dendritic cell. We know the myeloid dendritic cell moves from its immature state to a
69
state of the "maturing" dendritic cell, where it produces high levels of IL-12, followed by its final "mature" phase, where it secretes high amounts of IL-10 (Langenkamp et al., 2000). Can the NK-T cell provide signals to the dendritic cell to move it from one phase to another? Our data suggests that after expansion and activation of NK-T cells within the pancreatic draining node, myeloid dendritic cells emigrate from the islets, and 13 cell destruction is halted. How might this effect take place?
The first possibility is that the expanded NK-T cells in the pancreatic-draining node either secrete chemokines themselves, or induce chemokine secretion from other cells in the node, leading to recruitment of those dendritic cells from within the islets to the lymphoid tissues. We know that when the dendritic cell begins to mature, the chemokine receptor CCR7 will be upregulated and this should guide the dendritic cell to the draining node (Langenkamp et al., 2000). But in the NOD, the islets themselves aberrantly express several chemokines and cell adhesion molecules, such as MIP-l a, Lymphotoxin-a, and SLC, that may counteract this migration (Faveeuw et al., 1994a; Faveeuw et al., 1994b; Papaccio et al., 1999a; Cameron et al., 2000; Hjelmstrom et al., 2000). The activated NK-T cells may act within the draining node to increase secretion of chemokines such as MIP-30 (ELC) and 6Ckine (SLC), driving the balance of chemokines in favor of migration to the draining node (Caux et al., 2000).
Alternatively, NK-T cells may migrate out of the pancreatic draining node and home to the inflamed pancreas, where they directly interact with dendritic cells and trigger maturation and upregulation of CCR7, leading to a state capable of leaving the islet. Our collaborators did note a significant increase in invariant T cell receptor transcripts within the islets after a-galactosylceramide treatment. This hypothesis still lacks key points necessary for a full understanding of this process. First, do NK-T cells preferentially home to sites of inflammation, such as the islets or the pancreatic draining node? We see NK-T cells expanded specifically within the inflamed pancreatic draining node, but is this due to NK-T cell homing to the pancreatic draining node, or the local
70
expansion stimulated by dendritic cells within that node (high CD1d expression)? If these cells do home to inflammatory sites, they may play a key role in resolving islet inflammation by promoting dendritic cell emigration. This hypothesis dramatically expands the role for NK-T cells in the immune response and autoimmune diseases where unresolved inflammation leads to tissue destruction.
For protection from diabetes when treating 6-week-old NOD mice with
a-galactosylceramide, treatments must be repeated monthly. This indicates that once islet damage has exceeded a certain threshold, dendritic cells will simply reinfiltrate the area due to the amount of cellular damage once treatment stops. More recent experiments have shown that treating 4-week-old NOD mice with weekly injections of oa-galactosylceramide gives an incidence of diabetes lower than 20%. Again, treatments must be continued for protection from disease. But is this dependance on continued treatment necessarily deleterious? First, it shows that protection of mice from diabetes is a direct effect of the a-galactosylceramide treatment, not some anomaly involved with handling the mice, a shortfall of many previous studies, or a contaminant in the lipid preparation (Atkinson and Leiter, 1999). Second, it supports the previous dendritic cell transfer data. In those experiments, dendritic cells were tolerizing after transfer, but not while within their existing environment within the pancreatic draining node, seen by the development of diabetes in untreated NOD mice. Current data supports the pancreatic draining node as a site of tolerance induction, but one that becomes inadequate on its own to overcome intrinsic defects that maintain the inflammatory state within the pancreas (Kurts et al., 1997). Our data shows that even when we can reduce the islet infiltrate in 6-week-old mice, tolerance may not be regained and the islet infiltrate reaccumulates. Together these studies indicate that within the environment of the pancreatic draining lymph node, dendritic cells that would normally induce tolerance to self-antigen are unable to exert their affects. This holds true even as we increase the number of dendritic cells within that compartment following treatment with a-galactosylceramide. Since
71
moving dendritic cells to another location (as in the previous dendritic cell transfer studies) allows tolerance induction, this implicates the local microenvironment of that compartment (the pancreas and its draining node) in limiting tolerance induction.
This work suggests that techniques to affect dendritic cell activation, maturation, and migration using intrinsic NK-T cells to target these effects to inflamed tissues has great promise for manipulation of dendritic cell populations without unwanted systemic effects. Its applicability to other inflammatory diseases remains to be tested, but the presence of CDl1d expressing dendritic cells and the possibility to expand NK-T cells locally should be considered as logical contributors for resolution of the inflammatory state.
SUMMARY AND CONCLUSIONS
This work has focused on the role of dendritic cells in the pathogenesis of Type I diabetes. Since there is little published work in this area, many gaps in the current knowledge exist. This study attempts to fill a few of those holes, creating a new framework for future understanding of dendritic cell biology in the NOD mouse, and the role of interactions with NK-T cells in autoimmune and inflammatory processes.
Our first studies involved basic characterization of dendritic cell populations
within the secondary lymphoid tissues. We were able to visualize dendritic cell populations for the first time, and generate better descriptions of others. This allowed the characterization of these populations from lymph nodes of the diabetes-prone NOD mouse. The molecule CDld was significantly elevated on lymphoid dendritic cells from the pancreatic draining lymph node of the NOD, which lead to questions surrounding the role of these cells as activators of NK-T cells in this location.
Mature myeloid dendritic cells were decreased in number in the node that drains the pancreas, and a surplus of immature lymphoid dendritic cells existed in the peripheral inguinal lymph node. This data, coupled with the previously cited reports of dendritic cells within the pancreas, led us to question the functional aptitude of both myeloid and lymphoid dendritic cells in the NOD mouse.
A key fact discovered at this point was identification of the exact dendritic cell population transferred during the protection experiments performed in the early 1990's. We found that metrizamide gradients selectively enrich mature myeloid dendritic cells, and that these isolated dendritic cells appear to be of the "exhausted" phenotype described by Langenkamp as stimulating Th2 type responses (Langenkamp et al., 2000). This
72
73
means the fully mature myeloid dendritic cell in some fashion imparted the tolerance in those early studies.
The next study focused on generation of myeloid dendritic cells from murine bone marrow, using the autoimmune-prone NOD mouse. We found major deficiencies in the capability of NOD bone marrow-derived stem cells to differentiate to fully mature dendritic cells. No maturation stimuli tested were capable of generating the number of mature dendritic cells seen in cultures from other strains. The NOD consistently maintained at least twice as many immature dendritic cells as control strains, and developed an atypical third dendritic cell population with high levels of costimulatory molecules, but low levels of class II MHC. We referred to this population as pseudo-mature, as its functional characteristics (antigen uptake) closely resemble those of typical mature dendritic cells. The functional significance of these cells is unclear now, but the genetic implications are quite significant. That this population can develop in such large numbers from NOD cultures indicates the existence of serious defects in the differentiation pathways of dendritic cells in this strain. We confirmed that NOD derived dendritic cells are indeed functionally congruent with what is known about immature dendritic cells, as they take up high levels of antigen, produce low levels of cytokine, and are poor stimulators of T cells.
Next we demonstrated that NOD dendritic cells appear to undergo a prolonged phase of maturation, becoming ensnared at the developmental point of high IL-12p70 production. This extended period of time in the IL-12 producing phase of maturation increases the chances of Thl responses to the antigens presented. In a mouse that develops a Th 1-mediated autoimmune disease, this implicates the myeloid dendritic cell as a major player in the process.
We identified a genomic interval on chromosome 3, previously described as IddlO, that to some degree affects the maturation of NOD dendritic cells. Dendritic cells from
74
NOD mice with the IddlO interval from C57BL/6 mice were more mature phenotypically, produced more cytokine, and were better stimulators of allogeneic T cells.
The final phase of our study involved attempts to induce maturation or migration of the dendritic cell mass from the islets to the draining node. Knowing that dendritic cells within the islet appear to promote destruction, while those which traverse to the draining node are protective, we hypothesized that simply moving these dendritic cells from the pancreas might protect these mice from P cell destruction.
While agents such as LPS, TNF, and CFA have proven effective in inducing
dendritic cell maturation in vivo, these act systemically and have much greater potential for lethality than curing disease. We chose to use knowledge from our previous studies and other publications to accomplish the same goals locally.
We know that CDld is abnormally elevated on dendritic cells within the pancreatic draining node. CD1d is a class I like molecule that presents glycolipid antigen. The NK-T cell is a regulatory cell population which specifically recognizes this CD 1 dglycolipid combination, and as a result becomes activated and secretes large amounts of cytokine. Most studies have focused on what the dendritic cell can do for the NK-T cell. We hypothesized that the NK-T cell could also deliver signals to the dendritic cell, possibly affecting their maturation state.
We found that treatment of NOD mice with a-galactosylceramide expanded NK-T cells specifically within the pancreatic draining node, and recruited myeloid dendritic cells from the pancreas to that node. This halted 3 cell destruction, but was insufficient to convey long-term tolerance to P cell antigens. Mice treated with a-galactosylceramide had significantly delayed onset of disease, while those treated with the CDld binding lipid a-mannosylceramide (which binds CDld but does not interact with NK-T cells) had increased disease incidence.
This work establishes that dendritic cell developmental defects in the NOD mouse do exist. These defects are genetically linked, and most likely contribute to the population
75
of dendritic cells seen within the pancreatic islets of the NOD mouse. Myeloid dendritic cells are responsible for tolerance induction in previously performed transfer experiments, and an inability to progress to the fully mature dendritic cell in vivo most likely leads to the breakdown of peripheral tolerance to P cell antigens. But if we use pharmacologic agents to induce dendritic cell migration from the islets, we can stop additional 0 cell destruction from occurring. This proof of potential has significant implications for the treatment of humans recently diagnosed with Type I diabetes.
Hypothesis of Dendritic Cell Polarization
For the past ten years, immunologists have struggled to understand how the
immune system makes decisions regarding how it will respond to which stimuli. While others have focused on delineating exogenous from endogenous antigens, and innate "danger" signals present within certain bacteria and viruses, these decisions may in fact come down to a simple numbers game. Basic mathematics.
Recent insights from our lab, as well as others abroad have shown that the
immature and mature dendritic cell are not the only two states for this cell type. In fact, the dendritic cell progresses through a "maturing" phase, where it produces high levels of the Th I polarizing cytokine IL-12p70. It then proceeds to its final maturation state of low IL-12p70 and high IL-10 production, which favors Th2 type responses.
What this model has in its favor is redundancy, many checks and balances that
must be overcome for an antigen-specific response to become pro-inflammatory in nature.
What type of response will a given dendritic cell induce? Every dendritic cell, once it receives a maturation signal, will progress through the previously described stages of the maturing dendritic cell (pro-Th 1) to the mature dendritic cell (pro-Th2). At this point, the dendritic cell has constitutively high levels of both T cell costimulatory molecules such as CD86, as well as the molecule CD40, ready to receive signals back from antigen-specific T cells. When the dendritic cell receives signals via CD40 from antigen-specific T cells that exceed the predetermined threshold, it becomes locked into
76
that phase of maturation and primes naive T cells accordingly (Langenkamp, 2000). If this occurs early, the dendritic cell will prime Th I type responses, if it occurs later, Th2 type responses.
So what variables affect the time frame in which a dendritic cell will receive sufficient CD40 signals? First is thymic selection. The majority of T cells specific for self-antigens are eliminated in the thymus, and never make it to the periphery. Those specific for non-self antigens are released to the periphery. This significantly increases the chance that exogenous peptide will be recognized by a T cell versus those of endogenous origin. The affinity of the TCR for an exogenous peptide-MHC complex would also be of significantly greater strength, leading to more rapid upregulation of CD40L on the responding T cell.
The greater the number of T cells in the periphery specific for a certain antigen, the shorter the time before the dendritic cell interacts with antigen-specific T cells, increasing the chances that the dendritic cell will be locked into a state of high IL-12 production. Since a majority of autoreactive T cells are eliminated in the thymus of nonautoimmune-prone mice and humans, the circulating T cell pool contains very few TCRs specific for self antigen. It is in this manner that a dendritic cell taking up self does not promote inflammatory Th I type responses toward the host tissues.
The most important factor in this model from the standpoint of this study is the maturation time of the dendritic cell. The longer the dendritic cell takes to traverse the 'maturing' phase to arrive at the final state of exhaustion, the greater the chance of it receiving super-threshold CD40 signals while still in the IL-12 producing phase. In the NOD mouse, we see dendritic cells that spend a significantly longer period as 'maturing' dendritic cells than non-autoimmune strains.
What this model offers is an explanation for polarization of immune responses without imparting a 'consciousness' upon the immune system to delineate endogenous from exogenous antigen. It can explain why tolerance to tumor antigen can be overcome
77
by systemic treatment with soluble CD40L. There is no doubt active uptake of tumor antigen taking place at all times, and the ligation of CD40 on these dendritic cells locks them into a Th 1 promoting phase.
In the NOD mouse, we have poor thymic tolerance leading to increased numbers of potentially autoreactive T cells in the periphery (Gerling et al., 1994; Kanagawa et al., 1998; Rapoport et al., 1993b). When coupled with the prolonged maturation time for the myeloid dendritic cell, we start to understand how this mouse has such a high incidence of autoimmune disease. The affinity of specific antigens for MHC also play a role in how well T cells will respond, and it is well established that the atypical MHC, I-Ag7, of the NOD mouse is well suited to tightly bind certain antigenic components of the pancreatic p cell (Corper et al., 2000). The target tissues themselves likely play a role in the initiation of the disease as well, but T cell and dendritic cell defects that lead to Th 1 priming dendritic cells, coupled with the appropriate MHC haplotype, are what likely leads to total ablation of the target organ as opposed to resolution of inflammation.
Summation
While diagnosis of Type I diabetes has steadily improved over the past 10-15
years, our ability to circumvent the inevitability of complete P cell destruction has made no progress. This study shows that if we mobilize dendritic cells from the pancreatic islets, we save intact 13 cells from destruction. Strategies to affect dendritic cell maturation and migration specifically within the pancreas and its draining nodes must be explored. a-galactosylceramide is currently in clinical trial for treatment of inflammatory bowel disease, so trials utilizing it in recently diagnosed Type I diabetics would not be difficult to initiate. The use of recombinant chemokines or the introductions of viral vectors capable of delivering cytokine signals to dendritic cells within the pancreas also exist as potential avenues for future exploration. Now that we've begun to understand that resolution of the inflammatory state within the islets will spare intact beta cell mass, and that we can resolve this inflammation by targeting dendritic cells within the pancreas,
78
novel strategies for targeting these cells in Type I diabetes, as well as other autoimmune diseases of chronic inflammation should emerge.
REFERENCES
Adler, A. J., Marsh, D. W., Yochum, G. S., Guzzo, J. L., Nigam, A., Nelson, W. G.,
and Pardoll, D. M. (1998). CD4+ T cell tolerance to parenchymal self-antigens
requires presentation by bone marrow-derived antigen-presenting cells, J Exp Med
187, 1555-64.
Albert, M. L., Pearce, S. F., Francisco, L. M., Sauter, B., Roy, P., Silverstein, R. L.,
and Bhardwaj, N. (1998a). Immature dendritic cells phagocytose apoptotic cells via
alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes, J
Exp Med 188, 1359-68.
Albert, M. L., Sauter, B., and Bhardwaj, N. (1998b). Dendritic cells acquire antigen from
apoptotic cells and induce class I- restricted CTLs, Nature 392, 86-9.
Ardavin, C., Wu, L., Li, C. L., and Shortman, K. (1993). Thymic dendritic cells and T
cells develop simultaneously in the thymus from a common precursor population,
Nature 362, 761-3.
Arreaza, G. A., Cameron, M. J., Jaramillo, A., Gill, B. M., Hardy, D., Laupland, K. B.,
Rapoport, M. J., Zucker, P., Chakrabarti, S., Chensue, S. W., et al. (1997). Neonatal
activation of CD28 signaling overcomes T cell anergy and prevents autoimmune
diabetes by an IL-4-dependent mechanism, J Clin Invest 100, 2243-53.
Atkinson, M. A., and Leiter, E. H. (1999). The NOD mouse model of type 1 diabetes: as
good as it gets?, Nat Med 5, 601-4.
Banchereau, J., and Steinman, R. M. (1998). Dendritic cells and the control of immunity,
Nature 392, 245-52.
Bendelac, A., Lantz, O., Quimby, M. E., Yewdell, J. W., Bennink, J. R., and Brutkiewicz,
R. R. (1995). CD1 recognition by mouse NKI+ T lymphocytes, Science 268, 863-5.
Bendelac, A., Rivera, M. N., Park, H.-S., and Roark, J. H. (1997). Mouse CDI-Specific
NK1 T Cells: Development, Specificity, and Function, Annual Review of Immunology 15, 535-62.
Berman, M. A., Sandborg, C. I., Wang, Z., Imfeld, K. L., Zaldivar, F. J., Dadufalza, V.,
and Buckingham, B. A. (1996). Decreased IL-4 production in new onset type 1
insulin-dependent diabetes mellitus, J Immunol 157, 4690-6.
79
80
Bird, J. J., Brown, D. R., Mullen, A. C., Moskowitz, N. H., Mahowald, M. A., Sider, J. R.,
Gajewski, T. F., Wang, C. R., and Reiner, S. L. (1998). Helper T cell differentiation
is controlled by the cell cycle, Immunity 9, 229-37.
Bjorck, P., Banchereau, J., and Flores-Romo, L. (1997). CD40 ligation counteracts Fasinduced apoptosis of human dendritic cells, Int Immunol 9, 365-72.
Bowman, M. A., Leiter, E. H., and Atkinson, M. A. (1994). Prevention of diabetes in the
NOD mouse: implications for therapeutic intervention in human disease, Immunol
Today 15, 115-20.
Bradley, L. M., Asensio, V. C., Schioetz, L. K., Harbertson, J., Krahl, T., Patstone, G.,
Woolf, N., Campbell, I. L., and Sarvetnick, N. (1999). Islet-specific Thl, but not
Th2, cells secrete multiple chemokines and promote rapid induction of autoimmune
diabetes, J Immunol 162, 2511-20.
Brossay, L., Chioda, M., Burdin, N., Koezuka, Y., Casorati, G., Dellabona, P., and Kronenberg, M. (1998a). CDld-mediated recognition of an alpha-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution, J Exp
Med 188, 1521-8.
Brossay, L., Naidenko, O., Burdin, N., Matsuda, J., Sakai, T., and Kronenberg, M.
(1998b). Structural requirements for galactosylceramide recognition by CD1restricted NK T cells, J Immunol 161, 5124-8.
Cameron, M. J., Arreaza, G. A., Grattan, M., Meagher, C., Sharif, S., Burdick, M. D.,
Strieter, R. M., Cook, D. N., and Delovitch, T. L. (2000). Differential expression
of CC chemokines and the CCR5 receptor in the pancreas is associated with
progression to Type I diabetes, J Immunol 165, 1102-10.
Cameron, M. J., Arreaza, G. A., Zucker, P., Chensue, S. W., Strieter, R. M., Chakrabarti,
S., and Delovitch, T. L. (1997a). IL-4 prevents insulitis and insulin-dependent
diabetes mellitus in nonobese diabetic mice by potentiation of regulatory T helper-2
cell function, J Immunol 159, 4686-92.
Cameron, M. J., Arreaza, G. A., Zucker, P., Chensue, S. W., Strieter, R. M., Chakrabarti,
S., and Delovitch, T. L. (1997b). IL-4 prevents insulitis and insulin-dependent
diabetes mellitus in nonobese diabetic mice by potentiation of regulatory T helper-2
cell function, J Immunol 159, 4686-92.
Caux, C., Ait-Yahia, S., Chemin, K., de Bouteiller, O., Dieu-Nosjean, M. C., Homey, B.,
Massacrier, C., Vanbervliet, B., Zlotnik, A., and Vicari, A. (2000). Dendritic cell
biology and regulation of dendritic cell trafficking by chemokines, Springer Semin
Immunopathol 22, 345-69.
81
Celia, M., Engering, A., Pinet, V., Pieters, J., and Lanzavecchia, A. (1997). Inflammatory
stimuli induce accumulation of MHC class II complexes on dendritic cells [see
comments], Nature 388, 782-7.
Chen, H., and Paul, W. E. (1997). Cultured NK1.1+ CD4+ T cells produce large amounts
of IL-4 and IFN-g upon activation by anti-CD3 of CDI, J Immunol 159, 2240-9.
Chiu, Y. H., Jayawardena, J., Weiss, A., Lee, D., Park, S. H., Dautry-Varsat, A., and
Bendelac, A. (1999). Distinct subsets of CD ld-restricted T cells recognize selfantigens loaded in different cellular compartments, J Exp Med 189, 103-10.
Clare-Salzler, M. J., Brooks, J., Chai, A., Van Herle, K., and Anderson, C. (1992).
Prevention of diabetes in nonobese diabetic mice by dendritic cell transfer, J Clin
Invest 90, 741-8.
Cordell, H. J., Todd, J. A., Hill, N. J., Lord, C. J., Lyons, P. A., Peterson, L. B., Wicker,
L. S., and Clayton, D. G. (2001). Statistical Modeling of Interlocus Interactions
in a Complex Disease. Rejection of the multiplicative model of epistasis in type 1
diabetes, Genetics 158, 357-67.
Corper, A. L., Stratmann, T., Apostolopoulos, V., Scott, C. A., Garcia, K. C., Kang, A. S.,
Wilson, I. A., and Teyton, L. (2000). A structural framework for deciphering the link
between I-Ag7 and autoimmune diabetes, Science 288, 505-11.
Crowley, M., Inaba, K., Witmer-Pack, M., and Steinman, R. M. (1989). The cell surface
of mouse dendritic cells: FACS analyses of dendritic cells from different tissues
including thymus, Cell Immunol 118, 108-25.
D'Adamio, L., Awad, K. M., and Reinherz, E. L. (1993). Thymic and peripheral apoptosis
of antigen-specific T cells might cooperate in establishing self tolerance, Eur J
Immunol 23, 747-53.
Dahlen, E., Dawe, K., Ohlsson, L., and Hedlund, G. (1998). Dendritic cells and macrophages are the first and major producers of TNF-alpha in pancreatic islets in the
nonobese diabetic mouse, J Immunol 160, 3585-93.
Dautigny, N., Le Campion, A., and Lucas, B. (1999). Timing and casting for actors of
thymic negative selection, J Immunol 162, 1294-302.
de St Groth, B. F. (1998). The evolution of self-tolerance: a new cell arises to meet the
challenge of self-reactivity, Immunol Today 19, 448-54.
Delemarre, F. G., Simons, P. J., de Heer, H. J., and Drexhage, H. A. (1999). Signs of
immaturity of splenic dendritic cells from the autoimmune-prone biobreeding rat:
consequences for the in vitro expansion of regulator and effector T cells, J Immunol
162, 1795-801.
82
Fabien, N., Bergerot, I., Maguer-Satta, V., Orgiazzi, J., and Thivolet, C. (1995). Pancreatic
lymph nodes are early targets of T cells during adoptive transfer of diabetes in NOD
mice, J Autoimmun 8, 323-34.
Fairchild, P. J., and Austyn, J. M. (1990). Thymic dendritic cells: phenotype and function,
Int Rev Immunol 6, 187-96.
Falcone, M., Yeung, B., Tucker, L., Rodriguez, E., and Sarvetnick, N. (1999). A defect
in interleukin 12-induced activation and interferon gamma secretion of peripheral
natural killer T cells in nonobese diabetic mice suggests new pathogenic mechanisms for insulin-dependent diabetes mellitus, J Exp Med 190, 963-72.
Faveeuw, C., Gagnerault, M. C., and Lepault, F. (1994a). Expression of homing and adhesion molecules in infiltrated islets of Langerhans and salivary glands of nonobese
diabetic mice, J Immunol 152, 5969-78.
Faveeuw, C., Gagnerault, M. C., and Lepault, F. (1994b). Modifications of the expression
of homing and adhesion molecules in infiltrated islets of Langerhans in NOD mice,
Adv Exp Med Biol 355, 137-42.
Feili-Hariri, M., Dong, X., Alber, S. M., Watkins, S. C., Salter, R. D., and Morel, P. A.
(1999). Immunotherapy of NOD mice with bone marrow-derived dendritic cells,
Diabetes 48, 2300-8.
Forster, I., and Lieberam, I. (1996). Peripheral tolerance of CD4 T cells following local
activation in adolescent mice, Eur J Immunol 26, 3194-202.
Fossati, G., Cooke, A., Papafio, R. Q., Haskins, K., and Stockinger, B. (1999). Triggering
a second T cell receptor on diabetogenic T cells can prevent induction of diabetes,
J Exp Med 190, 577-83.
Fox, C. J., and Danska, J. S. (1997). IL-4 expression at the onset of islet inflammation predicts nondestructive insulitis in nonobese diabetic mice, J Immunol 158, 2414-24.
Gallucci, S., Lolkema, M., and Matzinger, P. (1999). Natural adjuvants: endogenous
activators of dendritic cells, Nat Med 5, 1249-55.
Gerling, I. C., Atkinson, M. A., and Leiter, E. H. (1994). The thymus as a site for evaluating the potency of candidate beta cell autoantigens in NOD mice, J Autoimmun
7, 851-8.
Gombert, J. M., Herbelin, A., Tancrede-Bohin, E., Dy, M., Carnaud, C., and Bach, J. F.
(1996a). Early quantitative and functional deficiency of NK1+-like thymocytes in
the NOD mouse, Eur J Immunol 26, 2989-98.
83
Gombert, J. M., Tancrede-Bohin, E., Hameg, A., Leite-de-Moraes, M. C., Vicari, A.,
Bach, J. F., and Herbelin, A. (1996b). IL-7 reverses NKI+ T cell-defective IL-4
production in the non-obese diabetic mouse, Int Immunol 8, 1751-8.
Graser, R. T., DiLorenzo, T. P., Wang, E, Christianson, G. J., Chapman, H. D., Roopenian,
D. C., Nathenson, S. G., and Serreze, D. V. (2000). Identification of a CD8 T cell
that can independently mediate autoimmune diabetes development in the complete
absence of CD4 T cell helper functions, J Immunol 164, 3913-8.
Gudmundsdottir, H., Wells, A. D., and Turka, L. A. (1999). Dynamics and requirements of
T cell clonal expansion in vivo at the single-cell level: effector function is linked to
proliferative capacity, J Immunol 162, 5212-23.
Hammond, K. I. L., Poulton, L. D., Palmisano, L. J., Silveira, P. A., Godfrey, D. I.,
and Baxter, A. G. (1998). alpha/beta-T cell receptor (TCR)+CD4-CD8- (NKT) thymocytes prevent insulin-dependent diabetes mellitus in nonobese diabetic (NOD)/Lt
mice by the influence of interleukin (IL)-4 and/or IL-10., J Exp Med 187, 1047-56.
Hammond, K. J., Pelikan, S. B., Crowe, N. Y, Randle-Barrett, E., Nakayama, T., Taniguchi, M., Smyth, M. J., van Driel, I. R., Scollay, R., Baxter, A. G., and Godfrey, D.
I. (1999). NKT cells are phenotypically and functionally diverse, Eur J Immunol
29, 3768-81.
Han, H. S., Jun, H. S., Utsugi, T., and Yoon, J. W. (1997). Molecular role of TGF-beta,
secreted from a new type of CD4+ suppressor T cell, NY4.2, in the prevention of
autoimmune IDDM in NOD mice, J Autoimmun 10, 299-307.
Heath, W. R., and Carbone, F. R. (2001). Cross-presentation, dendritic cells, tolerance and
immunity, Annu Rev Immunol 19, 47-64.
Hjelmstrom, P., Fjell, J., Nakagawa, T., Sacca, R., Cuff, C. A., and Ruddle, N. H. (2000).
Lymphoid tissue homing chemokines are expressed in chronic inflammation, Am
J Pathol 156, 1133-8.
Inaba, K., Turley, S., Yamaide, F., lyoda, T., Mahnke, K., Inaba, M., Pack, M., Subklewe,
M., Sauter, B., Sheff, D., et al. (1998). Efficient presentation of phagocytosed
cellular fragments on the major histocompatibility complex class II products of
dendritic cells, J Exp Med 188, 2163-73.
Jansen, A., Homo-Delarche, F., Hooijkaas, H., Leenen, P. J., Dardenne, M., and Drexhage,
H. A. (1994). Immunohistochemical characterization of monocytes-macrophages
and dendritic cells involved in the initiation of the insulitis and beta- cell destruction
in NOD mice, Diabetes 43, 667-75.
Jaramillo, A., Gill, B. M., and Delovitch, T. L. (1994). Insulin dependent diabetes mellitus
in the non-obese diabetic mouse: a disease mediated by T cell anergy?, Life Sci
55, 1163-77.
84
Kalinski, P., Hilkens, C. M., Snijders, A., Snijdewint, F. G., and Kapsenberg, M. L.
(1997). IL-12-deficient dendritic cells, generated in the presence of prostaglandin E2, promote type 2 cytokine production in maturing human naive T helper cells,
J Immunol 159, 28-35.
Kanagawa, O., Vaupel, B. A., Xu, G., Unanue, E. R., and Katz, J. D. (1998). Thymic
positive selection and peripheral activation of islet antigen- specific T cells: separation of two diabetogenic steps by an I-A(g7) class II MHC beta-chain mutant, J
Immunol 161, 4489-92.
Karandikar, N. J., Vanderlugt, C. L., Walunas, T. L., Miller, S. D., and Bluestone, J.
A. (1996). CTLA-4: a negative regulator of autoimmune disease, J Exp Med 184,
783-8.
Kelly, K. A., Lucas, K., Hochrein, H., Metcalf, D., Wu, L., and Shortman, K. (2001).
Development of dendritic cells in culture from human and murine thymic precursor
cells, Cell Mol Biol (Noisy-le-grand) 47, 43-54.
Kitamura, H., Iwanabe, K., Yahata, T., Nishimura, S.-i., Ohta, A., Ohmi, Y., Sato, M.,
Takeda, K., Okumura, K., Van Kaer, L., et al. (1999). The Natural Killer (NKT)
Cell Ligand alpha-Galactosylceramide Demonstrates Its Immunopotentiating Effect
by Inducing Interleukin (IL)-12 Production by Dendritic Cells and IL-12 Receptor
Expression on NKT Cells, J Exp Med 189, 1121-7.
Knight, S. C., Mertin, J., Stackpoole, A., and Clark, J. (1983). Induction of immune
responses in vivo with small numbers of veiled (dendritic) cells, Proc Natl Acad
Sci U S A 80, 6032-5.
Kronin, V., Hochrein, H., Shortman, K., and Kelso, A. (2000a). Regulation of T cell
cytokine production by dendritic cells, Immunol Cell Biol 78, 214-23.
Kronin, V., Wu, L., Gong, S., Nussenzweig, M. C., and Shortman, K. (2000b). DEC-205
as a marker of dendritic cells with regulatory effects on CD8 T cell responses [In
Process Citation], Int Immunol 12, 731-5.
Kurts, C., Kosaka, H., Carbone, F. R., Miller, J. F., and Heath, W. R. (1997). Class
I-restricted cross-presentation of exogenous self-antigens leads to deletion of autoreactive CD8(+) T cells, J Exp Med 186, 239-45.
Laloux, V., Beaudoin, L., Jeske, D., Carnaud, C., and Lehuen, A. (2001). NK T cellinduced protection against diabetes in V alpha 14-J alpha 281 transgenic nonobese
diabetic mice is associated with a Th2 shift circumscribed regionally to the islets
and functionally to islet autoantigen, J Immunol 166, 3749-56.
Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000). Kinetics of
dendritic cell activation: impact on priming of TH1, TH2 and nonpolarized T cells,
Nat Immunol 1, 311-6.
85
Lanzavecchia, A. (1999). Dendritic cell maturation and generation of immune responses,
Haematologica 84, 23-5.
Lederer, J. A., Perez, V. L., DesRoches, L., Kim, S. M., Abbas, A. K., and Lichtman, A. H.
(1996). Cytokine transcriptional events during helper T cell subset differentiation,
J Exp Med 184, 397-406.
Lehuen, A., Lantz, O., Beaudoin, L., Laloux, V., Carnaud, C., Bendelac, A., Bach, J.
F., and Monteiro, R. C. (1998). Overexpression of natural killer T cells protects Valphal4- Jalpha281 transgenic nonobese diabetic mice against diabetes, J Exp
Med 188, 1831-9.
Leiter, E. H., and Serreze, D. V. (1992). Antigen presenting cells and the immunogenetics
of autoimmune diabetes in NOD mice, Reg Immunol 4, 263-73.
Ludewig, B., Odermatt, B., Landmann, S., Hengartner, H., and Zinkernagel, R. M. (1998).
Dendritic cells induce autoimmune diabetes and maintain disease via de novo
formation of local lymphoid tissue, J Exp Med 188, 1493-501.
Malavasi, F., Funaro, A., Roggero, S., Horenstein, A., Calosso, L., and Mehta, K. (1994).
Human CD38: a glycoprotein in search of a function, Immunol Today 15, 95-7.
Maron, R., Melican, N. S., and Weiner, H. L. (1999). Regulatory Th2-type T cell lines
against insulin and GAD peptides derived from orally- and nasally-treated NOD
mice suppress diabetes, J Autoimmun 12, 251-8.
Martins, T. C., and guas, A. P. (1999). A role for CD45RBlow CD38+ T cells and
costimulatory pathways of T-cell activation in protection of non-obese diabetic
(NOD) mice from diabetes, Immunology 96, 600-5.
Marzo, A. L., Lake, R. A., Lo, D., Sherman, L., McWilliam, A., Nelson, D., Robinson, B.
W., and Scott, B. (1999). Tumor antigens are constitutively presented in the draining
lymph nodes [In Process Citation], J Immunol 162, 5838-45.
Matzinger, P. (1994). Tolerance, danger, and the extended family, Annu Rev Immunol
12, 991-1045.
Matzinger, P. (1998). An innate sense of danger, Semin Immunol 10, 399-415.
Matzinger, P., and Guerder, S. (1989). Does T-cell tolerance require a dedicated antigenpresenting cell?, Nature 338, 74-6.
McAdam, A. J., Schweitzer, A. N., and Sharpe, A. H. (1998). The role of B7 costimulation in activation and differentiation of CD4+ and CD8+ T cells, Immunol
Rev 165, 231-47.
86
McDevitt, H. O. (1998). The role of MHC class II molecules in susceptibility and
resistance to autoimmunity, Curr Opin Immunol 10, 677-81.
Mevorach, D., Zhou, J. L., Song, X., and Elkon, K. B. (1998). Systemic exposure to
irradiated apoptotic cells induces autoantibody production, J Exp Med 188, 387-92.
Mondino, A., Khoruts, A., and Jenkins, M. K. (1996). The anatomy of T-cell activation
and tolerance, Proc Natl Acad Sci U S A 93, 2245-52.
Morel, P. A., Vasquez, A. C., and Feili-Hariri, M. (1999). Immunobiology of DC in NOD
mice, J Leukoc Biol 66, 276-80.
Mueller, R., Bradley, L. M., Krahl, T., and Sarvetnick, N. (1997). Mechanism underlying
counterregulation of autoimmune diabetes by IL-4, Immunity 7, 411-8.
O'Reilly, L. A., Hutchings, P. R., Crocker, P. R., Simpson, E., Lund, T., Kioussis, D.,
Takei, F., Baird, J., and Cooke, A. (1991). Characterization of pancreatic islet cell
infiltrates in NOD mice: effect of cell transfer and transgene expression, Eur J
Immunol 21, 1171-80.
Papaccio, G., De Luca, A., De Luca, B., Pisanti, F. A., and Zarrilli, S. (1999). Detection
of dendritic cells in the non-obese diabetic (NOD) mouse islet pancreas infiltrate is
correlated with Th2-cytokine production, J Cell Biochem 74, 447-57.
Papaccio, G., Nicoletti, F, Pisanti, F. A., Bendtzen, K., and Galdieri, M. (2000). Prevention of spontaneous autoimmune diabetes in NOD mice by transferring in vitro
antigen-pulsed syngeneic dendritic cells, Endocrinology 141, 1500-5.
Perez, V. L., Lederer, J. A., Lichtman, A. H., and Abbas, A. K. (1995). Stability of Thl
and Th2 populations, Int Immunol 7, 869-75.
Quartey-Papafio, R., Lund, T., Chandler, P., Picard, J., Ozegbe, P., Day, S., Hutchings, P.
R., O'Reilly, L., Kioussis, D., Simpson, E., and et al. (1995). Aspartate at position 57 of nonobese diabetic I-Ag7 beta-chain diminishes the spontaneous incidence of
insulin-dependent diabetes mellitus, J Immunol 154, 5567-75.
Rapoport, M. J., Jaramillo, A., Zipris, D., Lazarus, A. H., Serreze, D. V., Leiter, E. H.,
Cyopick, P., Danska, J. S., and Delovitch, T. L. (1993a). Interleukin 4 reverses T
cell proliferative unresponsiveness and prevents the onset of diabetes in nonobese
diabetic mice, J Exp Med 178, 87-99.
Rapoport, M. J., Lazarus, A. H., Jaramillo, A., Speck, E., and Delovitch, T. L. (1993b).
Thymic T cell anergy in autoimmune nonobese diabetic mice is mediated by
deficient T cell receptor regulation of the pathway of p21 ras activation, J Exp Med
177, 1221-6.
87
Read, S., Mauze, S., Asseman, C., Bean, A., Coffman, R., and Powrie, F. (1998). CD38+
CD45RB(low) CD4+ T cells: a population of T cells with immune regulatory
activities in vitro, Eur J Immunol 28, 3435-47.
Ridgway, W. M., Ito, H., Fasso, M., Yu, C., and Fathman, C. G. (1998). Analysis of
the role of variation of major histocompatibility complex class II expression on nonobese diabetic (NOD) peripheral T cell response, J Exp Med 188, 2267-75.
Rosmalen, J. G., Homo-Delarche, F., Durant, S., Kap, M., Leenen, P. J., and Drexhage, H.
A. (2000a). Islet abnormalities associated with an early influx of dendritic cells and
macrophages in NOD and NODscid mice, Lab Invest 80, 769-77.
Rosmalen, J. G., Leenen, P. J., Katz, J. D., Voerman, J. S., and Drexhage, H. A. (1997).
Dendritic cells in the autoimmune insulitis in NOD mouse models of diabetes, Adv
Exp Med Biol 417, 291-4.
Rosmalen, J. G., Martin, T., Dobbs, C., Voerman, J. S., Drexhage, H. A., Haskins, K.,
and Leenen, P. J. (2000b). Subsets of macrophages and dendritic cells in nonobese
diabetic mouse pancreatic inflammatory infiltrates: correlation with the development
of diabetes, Lab Invest 80, 23-30.
Rovere, P., Vallinoto, C., Bondanza, A., Crosti, M. C., Rescigno, M., Ricciardi-Castagnoli, P., Rugarli, C., and Manfredi, A. A. (1998). Bystander apoptosis triggers
dendritic cell maturation and antigen- presenting function, J Immunol 161, 4467-71.
Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999). Two subsets of
memory T lymphocytes with distinct homing potentials and effector functions [see
comments], Nature 401, 708-12.
Salojin, K. V., Zhang, J., Madrenas, J., and Delovitch, T. L. (1998). T-cell anergy and
altered T-cell receptor signaling: effects on autoimmune disease, Immunol Today
19, 468-73.
Salomon, B., Cohen, J. L., Masurier, C., and Klatzmann, D. (1998). Three populations of
mouse lymph node dendritic cells with different origins and dynamics, J Immunol
160, 708-17.
Salomon, B., Lenschow, D. J., Rhee, L., Ashourian, N., Singh, B., Sharpe, A., and
Bluestone, J. A. (2000). B7/CD28 costimulation is essential for the homeostasis
of the CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes,
Immunity 12, 431-40.
Sarvetnick, N. (1997). IFN-gamma, IGIF, and IDDM [editorial], J Clin Invest 99, 371-2.
Schrum, A. G., Wells, A. D., and Turka, L. A. (2000). Enhanced surface TCR replenishment mediated by CD28 leads to greater TCR engagement during primary stimulation, Int Immunol 12, 833-42.
88
Schweitzer, A. N., Borriello, F., Wong, R. C., Abbas, A. K., and Sharpe, A. H. (1997).
Role of costimulators in T cell differentiation: studies using antigen- presenting
cells lacking expression of CD80 or CD86, J Immunol 158, 2713-22.
Serreze, D. V., Gaedeke, J. W., and Leiter, E. H. (1993a). Hematopoietic stem-cell defects
underlying abnormal macrophage development and maturation in NOD/Lt mice:
defective regulation of cytokine receptors and protein kinase C, Proc Natl Acad Sci
U S A 90, 9625-9.
Serreze, D. V., Gaskins, H. R., and Leiter, E. H. (1993b). Defects in the differentiation and
function of antigen presenting cells in NOD/Lt mice, J Immunol 150, 2534-43.
Serreze, D. V., Hamaguchi, K., and Leiter, E. H. (1989). Immunostimulation circumvents
diabetes in NOD/Lt mice, J Autoimmun 2, 759-76.
Shinomiya, M., Fazle Akbar, S. M., Shinomiya, H., and Onji, M. (1999). Transfer
of dendritic cells (DC) ex vivo stimulated with interferon- gamma (IFN-gamma)
down-modulates autoimmune diabetes in non-obese diabetic (NOD) mice, Clin Exp
Immunol 117, 38-43.
Shinomiya, M., Nadano, S., Shinomiya, H., and Onji, M. (2000). In situ characterization
of dendritic cells occurring in the islets of nonobese diabetic mice during the
development of insulitis, Pancreas 20, 290-6.
Shortman, K. (2000). Burnet oration: dendritic cells: multiple subtypes, multiple origins,
multiple functions, Immunol Cell Biol 78, 161-5.
Shortman, K., and Maraskovsky, E. (1998). Developmental options [comment], Science
282, 424-5.
Steinman, R. M. (1991). The dendritic cell system and its role in immunogenicity, Annu
Rev Immunol 9, 271-96.
Steinman, R. M., and Cohn, Z. A. (1973). Identification of a novel cell type in peripheral
lymphoid organs of mice. I. Morphology, quantitation, tissue distribution, J Exp
Med 137, 1142-62.
Steinman, R. M., Kaplan, G., Witmer, M. D., and Cohn, Z. A. (1979). Identification of
a novel cell type in peripheral lymphoid organs of mice. V. Purification of spleen
dendritic cells, new surface markers, and maintenance in vitro, J Exp Med 149,
1-16.
Steinman, R. M., Pack, M., and Inaba, K. (1997a). Dendritic cell development and
maturation, Adv Exp Med Biol 417, 1-6.
Steinman, R. M., Pack, M., and Inaba, K. (1997b). Dendritic cells in the T-cell areas of
lymphoid organs, Immunol Rev 156, 25-37.
89
Teros, T., Hakala, R., Ylinen, L., Liukas, A., Arvilommi, P., Sainio-Pollanen, S., Verajankorva, E., Pollanen, P., and Simell, O. (2000). Cytokine balance and lipid antigen
presentation in the NOD mouse pancreas during development of insulitis, Pancreas
20, 191-6.
Ucker, D. S., Ashwell, J. D., and Nickas, G. (1989). Activation-driven T cell death. I.
Requirements for de novo transcription and translation and association with genome
fragmentation, J Immunol 143, 3461-9.
Vallera, D. A., Carroll, S. F., Brief, S., and Blazar, B. R. (1992). Anti-CD3 immunotoxin
prevents low-dose STZ/interferon-induced autoimmune diabetes in mouse, Diabetes
41,457-64.
van der Merwe, P. A., Bodian, D. L., Daenke, S., Linsley, P., and Davis, S. J. (1997).
CD80 (B7-1) binds both CD28 and CTLA-4 with a low affinity and very fast
kinetics, J Exp Med 185, 393-403.
Van Parijs, L., and Abbas, A. K. (1998). Homeostasis and self-tolerance in the immune
system: turning lymphocytes off, Science 280, 243-8.
Viney, J. L., Mowat, A. M., O'Malley, J. M., Williamson, E., and Fanger, N. A. (1998).
Expanding dendritic cells in vivo enhances the induction of oral tolerance, J Immunol 160, 5815-25.
Vremec, D., Pooley, J., Hochrein, H., Wu, L., and Shortman, K. (2000). CD4 and CD8
expression by dendritic cell subtypes in mouse thymus and spleen, J Immunol 164,
2978-86.
Walunas, T. L., Bakker, C. Y., and Bluestone, J. A. (1996). CTLA-4 ligation blocks CD28dependent T cell activation [published erratum appears in J Exp Med 1996 Jul
1;184(1):301], J Exp Med 183, 2541-50.
Wells, A. D., Gudmundsdottir, H., and Turka, L. A. (1997). Following the fate of
individual T cells throughout activation and clonal expansion. Signals from T cell
receptor and CD28 differentially regulate the induction and duration of a proliferative response, J Clin Invest 100, 3173-83.
Wells, A. D., Li, X. C., Li, Y., Walsh, M. C., Zheng, X. X., Wu, Z., Nunez, G., Tang, A.,
Sayegh, M., Hancock, W. W., et al. (1999). Requirement for T-cell apoptosis in the
induction of peripheral transplantation tolerance, Nat Med 5, 1303-7.
Whittaker, D. S., Bahjat, K. S., Moldawer, L. L., and Clare-Salzler, M. J. (2000).
Autoregulation of human monocyte-derived dendritic cell maturation and IL-12
production by cyclooxygenase-2-mediated prostanoid production, J Immunol 165,
4298-304.
90
Wicker, L. S., Appel, M. C., Dotta, F., Pressey, A., Miller, B. J., DeLarato, N. H., Fischer,
P. A., Boltz, R. C., Jr., and Peterson, L. B. (1992). Autoimmune syndromes in major
histocompatibility complex (MHC) congenic strains of nonobese diabetic (NOD)
mice. The NOD MHC is dominant for insulitis and cyclophosphamide-induced
diabetes, J Exp Med 176, 67-77.
Wu, L., Li, C. L., and Shortman, K. (1996). Thymic dendritic cell precursors: relationship
to the T lymphocyte lineage and phenotype of the dendritic cell progeny, J Exp
Med 184, 903-11.
Zipris, D., Lazarus, A. H., Crow, A. R., Hadzija, M., and Delovitch, T. L. (1991).
Defective thymic T cell activation by concanavalin A and anti-CD3 in autoimmune nonobese diabetic mice. Evidence for thymic T cell anergy that correlates with the
onset of insulitis, J Immunol 146, 3763-71.
BIOGRAPHICAL SKETCH
Keith S. Bahjat was born and raised in the small Oklahoma town of Ponca City where his parents, Dhari and Jeanette Bahjat, still live today. Following graduation from Ponca City Senior High School, he received his Bachelor's of Science degree in medical technology from Oklahoma State University, which included a one year internship at Valley View Regional Hospital in Ada, Oklahoma. Keith is licensed by the American Society of Clinical Pathologists as a Medical Technologist. He was married to his lovely wife Rena on July 2nd, 1993. Following receipt of his undergraduate degree, Keith spent time in South Bend, Indiana and Chicago, Illinois working as a medical technologist in the areas of clinical hematology and flow cytometry.
91
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
t; *-.' i* I
Michael J. Clare-Salzler, Chair Associate Professor of Pathology, Immunology, and Laboratory Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
Mark Ainson
Professor of Pathology,
Immunology, and Laboratory Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
Raul Braylan /
Professor of Pathology,
Immunology, and Laboratory Medicine
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
Peter J. 11sen
Professor of Animal Sciences
I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.
gC_- &UIC
Eric Sobel
Associate Professor of Medicine
|
Full Text |
PAGE 1
DENDRITIC CELL DEVELOPMENT AND DISTRIBUTION IN TYPE I DIABETES: OVERCOMING GENETIC DEFECTS AND ESTABLISHING TOLERANCE VIA PHARMACOTHERAPEUTIC MANIPULATION OF DENDRITIC CELL POPULATIONS IN VIVO By KEITH S. BAHJAT A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2001
PAGE 2
ACKNOWLEDGMENTS I would first like to thank my mentor, Dr. Michael Clare-Salzer, for his encourage ment thoughtful discussions, and professional guidance over the past 4 years. Through the peaks and valleys of my graduate career, he has always been there to encourage me to continue, to lead me to the next critical experiment when I may get off the path and to offer personal guidance for dealing with the various curveballs life may throw my way. I know I would not be in the position I'm in today without his incredible insight and professionalism, and I will be grateful for the remainder of my scientific career becasue he taught me so much in a short amount of time I thank the members of my committee, Drs Mark Atkinson, Raul Bray Ian, Peter J. Hansen, and Eric Sobel for their time and insightful comments during the last 4 years. Of course, I'd also like to thank my wife, Rena, who has put up with me moving all around the country over the last 8 years in search of profe s sional satisfaction. Her support and helpful scientific discussions over the dinner table have been key in my progression through graduate school. She encouraged me when nothing seemed to be working correctly. No graduate career could succeed without knowledgeable and helpful peers to get you through the rough spots, and mine has been no different. I extend my gratitude to Donna Whittaker who was my lab buddy and spiritual guide for the first 3 years of graduate school. Donna traversed the rough road of graduate school with uncommon ease juggling both family and military responsibilities. Any time I wanted to complain about having too much work, I simply had to look across the lab and realize that I didn't have it so bad. I know I wouldn't have mad e it this far without Donna there to buffer things for me those first 3 years. 11
PAGE 3
iii I also thank Dr. Parker Small, who discussed science, graduate school, and life with me on a daily basis. His daily doses of sanity helped sequester the chaos around me and gave me the strength to push through to the end. I'd like to thank Dr. Peter Kima for his insightful discussions, encouraging remarks, and professional guidance over the last part of my graduate career. He gave me confidence that I could succeed as a scientist, and he generated insightful discussions on science outside my tiny area of specialization. I also thank Sara Eckenrode, who spent much time teaching me all the molecular techniques I never took the time to learn. I thank everyone who worked in Dr. Clare-Salzler's lab during my time here, including Jessica Elf, Sally Litherland, Yiyu Li, Vikas Dharnidharka, Yancy VanPatten, and Kristie Grebe.
PAGE 4
TABLE OF CONTENTS ACKNOWLEDGMENTS ......................................... ............................. .. ..... .. .. ...... .......... ii ABSTRACT .. .......................... .. .......... .... ....... ......................... ....... ..... .. .......................... vii INTRODUCTION ...................... ................................ ..... ...... ...................................... .. .. 1 REVIEW OF LITERATURE ............................................................................................... 5 Dendritic Cell Biology ..................................................................................................... 5 Myeloid Dendritic Cells ............................................................................................... 5 Lymphoid Dendritic Cells ............................................................................................ 9 Central Tolerance: The Role of Dendritic Cells ............................................................. 10 Peripheral Tolerance: Myeloid and Lymphoid Dendritic Cell Interactions ................... 11 Regulatory Lymphocyte Populations ............ .............. ....................... ......................... 12 Diabetes Onset and Progression in the NOD Mouse ............... ...... ....... ..... ...... ............ 13 Defects in Pancreatic Cells ....... ....................................... ........ ............. .... ... ......... 14 Contributions of MHC Class Il ........ ..... ..................... ..... ............ ....... .. ..... ................ 14 Defects in Activation Induced Cell Death .................... .............. .. .............................. 15 IDENTIFICATION AND CHARACTERIZATION OF MURINE DENDRITIC CELL POPULATIONS .................................................................... ..... ........................................ 17 Background .................................................................................................................... 17 Materials and Methods ................................................................................................... 18 Mice ........................................................................................................................... 18 Creation of Single-Celled Suspension from Lymph Node ........................................ 18 Analysis of Surf ace Phenotype .................................................................................. 19 Results ...... ..................................................................................................................... 20 Ex Vivo characterization of Dendritic Cell Populations ..................... ..... ..... .. ........... 20 Five Distinct Dendritic Cell Populations in Ex Vivo Preparations ... .......... .............. 20 Variations in Dendritic Cell Populations Between Anatomic Sites ..... ............ ........ 26 Variations in Dendritic Cell Populations by Strain ................ ............. .............. ...... 27 Summary and Conclusions ............................................................................................ 30 DYSFUNCTIONAL DEVELOPMENT AND MATURATION OF MYELOID DENDRITIC CELLS FROM NOD MICE ........................................................................ 35 Background .................................................... ........................ ......... ........ .... ........ ........ 35 IV
PAGE 5
V Materials and Methods ........................ .. .... .. .............. ..... ...... ...................... .... ............. 37 Mice ........................................................................................................................... 37 Generation of Myeloid Dendritic Cells ...................................................................... 37 Analysis of Surface Phenotype .................... ............... ............................................. 38 Purification of Hematopoetic Stem Cells ...................................... ............. ..... .... ..... 39 Purification of CD4+ T cells ................................................................................. ..... 39 Proliferation Assessment with CFSE ............. ............. ..................... ....................... 39 Uptake of Fluorescent Labeled Proteins or Particles ................................................ .40 ELISA for Cytokine in Culture Supernatants ........................................................... .40 Results ............................................................................................. ....... .. .... .. .............. 40 Identification of Protective Dendritic Cell Populations ............................................ .40 Revisiting an Old Friend .................................................. ..... ..... .... ........................ 40 Metrizamide Gradients Isolate Mature Dendritic Cells ........................................ .42 NOD Myeloid Dendritic Cells Do Not Develop, Mature, or Function Properly ...... .42 Atypical Development of Myeloid Dendritic Cells from NOD Bone Marrow Cultures ...... .. ..... .................. ....... ................................ .... .......... ......... ................... 43 NOD Dendritic Cells Do Not Respond To Maturation Stimuli ...................... ...... .44 MHC Instability Does Not Affect NOD Dendritic Cell Phenotype ...................... 44 NOD Dendritic Cells With Immature Phenotype Take Up Particulate Antigen ... .45 Cytokine Dosage Does Not Affect NOD Dendritic Cell Development.. .............. .47 NOD Dendritic Cells are Poor Stimulators of Allogeneic T Cells ....................... .47 Investigating Causes of NOD Maturation Abnormalities ......................................... .47 Microscopy of NOD Dendritic Cell Populations .................................................. .47 Soluble Factors Do Not Direct NOD Dendritic Cell Development.. ..................... 49 T, B, and NK-T Cells Do Not Affect NOD Dendritic Cell Development .............. 50 A Non-lymphocytic Cell Population Leads to Generation of Pseudo-Mature Dendritic Cells .............................................. ...................................... ........... .... .... .. .. 50 NOD Dendritic Cells Become Arrested During Maturation ...................................... 51 Genes Within Iddl O Affect NOD Dendritic Cell Maturation ................... ................ 51 Screening of NOD Congenic Strains ..................................................................... 51 lddlO Affects Dendritic Cell Maturation ............................................................... 52 lddlO Affects Dendritic Cell Cytokine Production Capacity ................................ 52 Summary and Discussion .............. ....... .... ................................ ...... .... ......... .. ............. 53 a-GALACTOSYLCERAMIDE ACTIVATES AND EXPANDS NK-T CELLS AND RECRUITS MYELOID DENDRITIC CELLS SPECIFICALLY WITHIN THE PANCREATIC DRAINING NODE OF NOD MICE ............................... ....................... 57 Background........ ....... ............................. .. .. ........ ..... ......................... ... ........... ..... .. ...... 57 Materials and Methods ................................... .... ... ...................................... ...... ......... 59 Mice .......................... ................................................. .............. .. ........................ ... 59 Treatment with a-Galactosylceramide .................................................................... 59 a-Galactosylcerarnide In Vitro Recall ....................................................................... 59 ELISA for Cytokine in Culture Supernatants .......................................................... 60 Analysis of Surface Phenotype ................................................................................. 60 Results ............................................... ................... ........................................... ..... ...... 61
PAGE 6
VI Treatment of NOD Mice with a-Galactosylceramide Expands NK-T Cells Specifically Within the Pancreatic Draining Node ............................................. .... . ....... 61 Increased Functional Responses Ex Vivo Following a-Galactosylceramide Treatment .... ............ ................. ......... ........................................... .... .... ........... ..... ...... 62 Increased Numbers of Myeloid Dendritic Cells in the Pancreatic Draining Node After a-Galactosylceramide Treatment ....... .......................................................... 64 Summary and Discussion ............................................................................................. 65 SUMMARY AND CONCLUSIONS ............................................ ................. ...... .......... 72 Hypothesis of Dendritic Cell Polarization ............................................................. ...... 75 Summation .... .... . ... ... . . ..... ....... . . ...... . . ... . ... .... . ... ... . .... ........... 77 REFERENCES ............. ....... ......... . ..... ... ... .... ..... . . ...... .... .... ... ....... . .... ...... 79 BIOGRAPHICAL SKETCH .... . . ...... ...... . ........ . . . ......... ........... ...... ..... ......... 91
PAGE 7
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy DENDRITIC CELL DEVELOPMENT AND DISTRIBUTION IN TYPE I DIABETES: OVERCOMING GENETIC DEFECTS AND ESTABLISHING TOLERANCE VIA PHARMACOTHERAPEUTIC MANIPULATION OF DENDRITIC CELL POPULATIONS IN VIVO By Keith S Bahjat August 2001 Chairman : Michael J. Clare-Salzler, M D Major Department: Pathology, Immunology, and Laboratory Medicine As the most potent antigen-presenting cell of the immune system, the dendritic cell has the unique capacity to both initiate and regulate antigen-specific immune responses. In the normally functioning immune system, the combination of costimulatory molecules and cytokines presented with MHC-peptide complexes on dendritic cells determine the penultimate function of naive antigen-specific T cells. Thus scientist's have great interest in the function of dendritic cells in autoimmune disease, where self-reactive T cells are activated in the periphery and regulation of this response is deficient. This study demonstrates abnormalities of dendritic cell subsets within second ary lymphoid tissues of the autoimmune-prone NOD mouse and demonstrates a genetic basis for abnormalities of these cells in vivo. Studies where hematopoeitic progenitors were cultured with GM-CSF and IL-4 showed that NOD stem cells have a limited capacity to differentiate to fully mature dendritic cells. Analysis of several congenic mouse strains have identified this as a genetically controlled phenomenon, and further Vll
PAGE 8
Vlll indicates potential in vivo differentiation and functional abnormalities that may lead to a dysfunctional immune system. Finally, activation of a lymphoid cell population known to interact with dendritic cells, the NK-T cell, is capable of limiting pancreatic beta cell destruction, apparently by initiating the emigration of dendritic cells from the pancreas to the draining lymph node, thus decreasing islet inflammation. Together, these findings establish a crucial role for abnormal dendritic cell development in the initiation and resolution of beta cell destruction in Type I diabetes.
PAGE 9
INTRODUCTION Dendritic cells are the most potent antigen presenting cells known, with a 10-100 fold greater capacity to stimulate na'ive T cells than other antigen-presenting cells (Stein man et al., 1997a). The dendritic cell's overall potency is dependent upon high levels of MHC-peptide complexes, membrane bound costimulatory molecules such as CD80 and CD86 and production of soluble cytokines such as IL-12 and IL-10 (Banchereau and Steinman, 1998). All but the latest work in dendritic cell biology has held that two dendritic cell states exist, the immature dendritic cell and the mature dendritic cell (Banchereau and Steinman, 1998). It is now apparent that the dendritic cell goes through defined phases while maturing, with dendritic cells at each phase stimulating a different T cell outcome than that seen when the dendritic cell is fully mature (Langenkamp 2000). Variations in maturation stimuli may modify the developmental pathway of these dendritic cells. The capacity to drive a na'ive T cell to activation and proliferation is related to the quantity of MHC-peptide complexes on the surface of the cell, in combination with T cell costimula t ory molecules such as CD86 (Shortman and Maraskovsky 1998). While the immature dendritic cell has low levels of these molecules, and thus has little effect on na'ive T cells, maturing and fully mature dendritic cells constitutively express high levels and thus are potent stimulators of na'ive T cells The dendritic cell produces two main cytokines that are involved in the polariza tion of T cell responses: IL-12 and IL-10 (Banchereau and Steinman, 1998 ; Kronin et al. 2000a; Shortman and Maraskovsky, 1998; Whittaker et al., 2000). What determines which of these cytokines is produced by the dendritic cell is a function of its maturation state upon encounter with the T cell (Langenkamp et al., 2000). As the dendritic cell begin s 1
PAGE 10
2 maturation, it enters a phase of high IL-12p70 production. AT cell interacting with the dendritic cell at this point results in those responding T cells assuming a Th 1 profile, producing cytokines such as IFN-y which activate components of the cell-mediated immune response. But if the dendritic cell fully traverses the maturing phase to the fully mature state the dendritic cell will secrete large amounts of ILI 0 resulting in a Th2 profile for responding T cells, producing cytokines such a s IL-4 and driving humoral type immune responses (Langenkamp, 2000) (Perez et al., 1995). Factors such as PGE2 appear to move the dendritic cell toward an IL-IO producing state, leading to the priming of Th2 type responses. It is not clear if the dendritic cell moves more rapidly through its IL-12 phase, or if this is bypassed altogether (Kalinski et al., 1997). These two functional T cell subsets are frequently used to classify responses to pathogens, as well as diseases of immune dysfunction. Type I diabetes has been frequently characterized as a disease where Th 1 cells promote diabetes, while Th2 type cells prevent disease (Bradley et al., 1999) Several pieces of evidence indicate a role for dendritic cells in the pathogenesis of Type I diabetes, the most obvious being their ability to stimulate and polarize T cell responses. Key experiments detailing the location and functional capacity of these cells during the progression of disease support a role for these cells in initiation of and protection from autoimmune-mediated destruction of pancreatic beta cells. Dendritic cells are the first leukocytes infiltrating the pancreas of both the NOD mouse and the diabetes-prone BB rat (Delemarre et al. 1999; Rosmalen et al., 2000a). These cells accumulate within the pancreas during the course of disease forming reticular networks with infiltrating T cells, and are presumed to locally prime autoreactive T cells (Jansen et al., 1994; Ludewig et al., 1998). The accumulation of antigen-presenting cells and T cells and the formation of lymphoid structures within the islets of Langerhans suggests there are factors which inhibit the resolution of inflammation in this tissue. The retention of dendritic cells within the islet would promote the activation of lymphocytes and thus promote chronic inflammation in the islet. Since the dendritic cell typically
PAGE 11
3 upregulates the chemokine receptor CCR7 and travels to the draining node early in the maturation process, dendritic cells capable of priming nai"ve T cells are not typically seen in the periphery. For this reason, we suspect dysfunctional dendritic cell development and maturation as a possible cause of the dendritic cell accumulation within the pancreatic islets seen in NOD mice (Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Rosmalen et al., 2000a; Rosmalen et al., 1997). It has also been established that dendritic cells from the pancreatic draining lymph node are capable of preventing diabetes when transferred to juvenile NOD mice (Clare Salzler et al., 1992). These experiments established that dendritic cells from this anatomic location were programmed to downregulate responses toward pancreatic cell antigens, and provided one of the first examples of the tolerance-inducing capacity of the dendritic cell More recent studies in the NOD mouse showed that in vitro generated myeloid dendritic cells transferred to juvenile NOD mice also prevent diabetes (Feili-Hariri et al., 1999; Papaccio et al., 2000). Additionally, splenic dendritic cells pulsed ex vivo with IFN-y proved protective against diabetes (Shinomiya et al., 1999). This data further reinforces the t olerizing capacity of the dendritic cell, however, many questions remain regarding the role of the dendritic cell in the NOD mouse. Principal among these is how do dendritic cells appear to play a dichotomous role in the development of diabetes? As the dendritic cell population in the islet are myeloid dendritic cells, it is presumed that there may be a primary defect in this dendritic cell population leading to activation of autoreactive T cells. At the same time, recent studies strongly support myeloid dendritic cells playing a tolerigenic role (Feili-Hariri et al., 1999; Papaccio et al., 2000; Shinomiya et al., 1999). Why are dendritic cells within the pancreas itself priming autoreactive responses, while those same cells within the draining node are protective? We hypothesize that dendritic cell maturation and migration play crucial roles in the pathogenesis of diabetes in the NOD mouse. These studies investigate in vivo
PAGE 12
4 dendritic cell populations, development and maturation of myeloid dendritic cells in vitro, and the interaction of dendritic cells in vivo with the regulatory NK-T cell population.
PAGE 13
REVIEW OF LITERATURE Dendritic Cell Biology Dendritic cells are divided into two lineages: myeloid and lymphoid. How these lineages function to initiate responses toward exogenous antigen or regulate responses to antigens is not clear at this point. Regardless, several studies have been published concerning the development and function of these two dendritic cell populations. Myeloid Dendritic Cells Myeloid dendritic cells are found in all tissues except brain. Within the peripheral immune system, myeloid dendritic cells are found in the marginal zone of spleen, and T cell areas of LN and Peyer's patches. Within tissues and organs myeloid dendritic cells are present in an immature state. Studies of myeloid dendritic cell development reveal two distinct immature myeloid dendritic cell populations, both yielding a common mature myeloid dendritic cell. The first pathway yields an epidermal Langerhans' cell, while the second generates the immature dendritic cell seen in other tissues and organs (Figure 2-1) (Banchereau and Steinman, 1998). Immature myeloid dendritic cells are highly endocytic, bearing high levels of Fey and Fee receptor, high levels of complement receptors CD 11 b and CD 11 c, and high intracellular levels of MHC Class II (Cella et al., 1997). Accessory molecules important for T cell activation, such as CD80, and CD86, are absent or expressed at low levels on immature dendritic cells. In addition, actin cables, used in dendritic cell cytplasmic extension, elongation, and migration, are visible within the cell. This mobile immature dendritic cell patrols tissues, continuously sampling the surrounding fluid environment 5
PAGE 14
Lymphoid dendritic cell 6 Figure 2-1. Dendritic cell developmental pathways. via pinocytosis, as well as taking up particulate matter recognized by specific pattern rec ognition receptors (complement receptors, Toll receptors, Fe receptors, mannose receptor, integrins). Our understanding of myeloid dendritic cell maturation has progressed signifi cantly over the past few years. It now appears that once the dendritic cell has received a maturation signal, it begins to traverse a preset program of changes, involving modulation of chemokine receptors (i.e., CCR6, CCR7), cytokine production (i.e IL-12, IL-10), den sity of MHC-peptide complexes on the cell surface, and T cell costimulatory molecules. IL-12 consists of two subunits, the p35 and p40. In an immature state, a dendritic cell
PAGE 15
7 will produce the p40 subunit at a constituitive level while producing very little p35. Biologically active IL-12 requires a heterodimer of the p35 and p40 subunits While p40 monomers and p40 homodimers are also secreted from the cell, these do not transduce signals through the IL-12 receptor, and may in fact inhibit this signalling Where the dendritic cell is in the maturation process when it encounters a T cell largely determines the functional polarization of antigen-specific T cell responses (Figure 2-2) (Langenkamp et al., 2000) Paramount to peripheral tolerance induction is the uptake processing, and presenc,,.,; ., [ IL-12p70 IL -10 CCR6 Chemokine [ Receptors CCR7 [ MHCClass II Surface CDSO Antigens CD86 Immature O'utcomiof interact! ,~ "1 ; "p~, Th1 type Th2 type %1 responses responses nail!! Location Secondary Lymphatics Figure 2-2. Phenotypic and functional changes in myeloid den dritic cells during maturation. tation of apoptotic cells and locally acquired soluble self antigens by APCs. As a cell undergoes apoptosis, typically due to irreparable DNA or mitochondrial damage it loses the ability to maintain the integrity of its phospholipid bilayer. This leads to exposure of residues such as phosphatidyl serine on the surface of the cell, marking this cell for uptake and removal from the peripheral pool.
PAGE 16
8 Phagocytosed cells are processed in the endocytic pathway like any extracellular pathogen or protein. The effect of apoptotic cell uptake on the state of dendritic cell maturation is still a topic of debate. While some studies report immature dendritic cells do not undergo maturation after uptake of apoptotic cells, most reports indicate that dendritic cell maturation takes place after interaction with apoptotic cells (Gallucci et al., 1999; Albert et al., 1998a; Albert et al., 1998b; Inaba et al., 1998; Mevorach et al., 1998; Rovere et al., 1998; Wells et al., 1999). Since uptake of apoptotic cells means the dendritic cell will be taking up and processing large amounts of self-antigen, how the dendritic cell responds, and in turn, the final outcome of interaction with nai"ve T cells is of key importance. Previous hypotheses, such as those put forth by Matzinger, held that inherent danger signals exist which trigger dendritic cell maturation, and in turn a pro-inflammatory response (Matzinger, 1994; Matzinger, 1998). We now understand that dendritic cell maturation does not always lead to inflammation, and that the triggering of dendritic cell maturation by apoptotic cells is a necessary event, as the immature dendritic cell has little influence over naive T cells, and no capacity to play a role in peripheral tolerance (low MHC, costimulatory molecules, and cytokines). This mature dendritic cell can now interact with antigen-specific T cells in a manner that may induce apoptosis, anergy, or the induction regulatory T cell populations. In response to exogenous pathogens the dendritic cell will receive maturation signals such as LPS, TNF, bacterial CpGs, viral mRNA, or PGE2 (Lanzavecchia, 1999). These factors trigger the dendritic cell to progress through the previously described maturation states (Figure 2). But in this case, the number of T cells specific for these exogenous antigens is much greater than for self antigen, and the dendritic cell receives strong CD40 signals very early. This locks the dendritic cell into an IL-12 producing phenotype, where it primes Th 1 type T cell responses against these exogenous antigens (Langenkamp et al., 2000a).
PAGE 17
9 Lymphoid Dendritic Cells Lymphoid dendritic cells are distinguished from myeloid dendritic cells based on their phenotype, as lymphoid dendritic cells express high levels of the CD8a (though those in thymus do express low levels of CD8 P), low levels of CD 11 b, and high levels of DEC-205. The lymphoid dendritic cells have been shown to specifically express high levels of self-peptides within MHC complexes, further highlighting the possible role of these cells in tolerance to self-antigen (Banchereau and Steinman, 1998 ; de St Groth 1998; Steinman et al., 1997b). Increasing the number of lymphoid dendritic cells has also been shown to improve oral tolerance induction. Using Flt3L, dendritic cell numbers within peripheral lymphoid tissues can be dramatically increased (Viney et al., 1998). This increase correlates with enhanced tolerance induction to orally delivered antigen, allowing low doses of antigen to limit specific T cell proliferation as effectively as high doses. Unlike myeloid dendritic cells, it is difficult to generate lymphoid dendritic cells in vitro, thus much less is known about developmental and functional properties of this lineage of dendritic cell. Lymphoid dendritic cells were first described as residents of the thymic medulla responsible for presenting self-antigen to maturing T cells, inducing apoptosis of autoreactive T cells (Ardavin et al., 1993). Lymphoid dendritic cells develop from a T cell precursor, and reside in the thymus, as well as within T cell areas of spleen, lymph node, and Peyer's patches (Wu et al., 1996). In contrast to the migratory myeloid dendritic cell, this dendritic cell is quite sessile, residing only within the lymphoid tissues, taking up antigens contained within the lymph, as well as that carried in by other cells. Immature lymphoid dendritic cells exist on the edge of the T cell areas, moving into the central regions as they mature. As the tissue associated myeloid dendritic cell homes to the secondary lymphoid tissues, it requires certain survival signals, such as CD40L on activated T cells upon entry to the lymph node. If these signals are not received, the myeloid dendritic cell may undergo apoptosis and be taken up by the lymphoid dendritic
PAGE 18
10 cells, and its antigens presented in a tolerizing manner. Likewise, a mature myeloid dendritic cell is susceptible to Fas-mediated apoptosis if not receiving survival signals through CD40 (which upregulates bcl-2) (Bjorck et al., 1997). The lymphoid dendritic cell appears specially suited to take up these apoptotic cells within the lymph node and present the acquired peptides in high quantities on the surface with costimulatory molecules. Inaba and Steinman demonstrated that lymphoid dendritic cells were uniquely capable of capturing, processing and presenting antigen transferred from apoptotic cells captured in the lymph nodes (Inaba et al., 1998). They went on to suggest that transfer of antigen from dying myeloid dendritic cells to lymphoid dendritic cells leads to peripheral tolerance of the T cell repertoire to self. Central Tolerance: The Role of Dendritic Cells Together, lymphoid and myeloid dendritic cells work to orchestrate the complex interactions of the cell-mediated immune response, including tolerance induction and activation of specific T cells to eliminate pathogens. How might the dendritic cell system control these two diverse functions? Lymphoid dendritic cells are the sole dendritic cell type found in the thymic medulla (Banchereau and Steinman, 1998; Fairchild and Austyn, 1990). The thymic medulla plays host to migrating thymocytes that have undergone positive selection in the thymic cortex and have been sent forward in the thymic selection process toward their final CD4+ or CD8+ destination. At this stage of maturation, any strong signal through the T cell receptor will trigger apoptosis. In other words, recognition of self-antigens in the context of MHC by the TCR triggers elimination of the autoreactive T cell. It is not clear whether a lymphoid dendritic cell, thymic epithelial cells in the medulla, or thymic macrophages must provide this signal. Matzinger made the assertion that dendritic cells mediate this process by demonstrating that splenic dendritic cells were capable of triggering deletion of thymocytes (Matzinger and Guerder, 1989). However, these studies failed to address whether lymphoid or myeloid dendritic cells mediated this T cell
PAGE 19
11 deletion. The threshold required to trigger this process is lower than that required in the periphery for a nai"ve T cell to differentiate, ensuring that weakly reactive T cells will be eliminated (Dautigny et al., 1999). Peripheral Tolerance: Myeloid and Lymphoid Dendritic Cell Interactions Most potentially autoreactive T cells are eliminated in the thymus, but some do escape to the periphery. To control potentially deleterious responses, mechanisms of p eripheral tolerance have been proposed. Since the thymus is partially sequestered from pathogens (unless systemically present), it could be argued that only self-antigens present within the vascular compartment have the opportunity to reach the thymus In the periphery, the immune system functions in a manner that allows self antigen to be distinguished from the milieu of pathogenic antigen likely to be present. Though theories ha v e been proposed involving repetitive motifs on bacterial cell walls and inflammatory cytokines signaling to the dendritic cell that an antigen is indeed foreign, this system is s till not well characterized. How and why one antigen results in a pro-inflammatory response while another results in tolerance remains a central question in immunology to day. Four possible outcomes from a T cell receptor recognizing a peptide-MHC com plex in the secondary lymphoid tissues include ignorance anergy activation, or activa tio n -induced cell death. Ignorance is a complete lack of recognition of MHC-peptide complex by the T cell, or non-presentation of the antigen (sequestration). Assuming that all potentially autoreactive T cells were deleted in the thymus, this would be the response any time self-antigen was presented in the periphery. Anergy may occur if the combination of primary (TCR) signal and secondary (CD28) signals does not reach an activation threshold, best exemplified by T cells receiving Signal 1 through the TCR-CD3 complex without engagement of CD28 (Signal 2) on the T cell (Wells et al. 1997). Anergy may also be induced when CTLA-4 engagement supersedes CD28 engagement (K a randikar et al., 1996; Walunas et al., 1996). CTLA-4 shares its ligands with CD28,
PAGE 20
12 but CTLA-4 has a much higher affinity for these ligands, and effectively competes for, and binds CD80 and CD86 on the APC. CTLA-4 engagement directly inhibits phosphorylation of the TCR-CD3 complex, limiting T cell responsiveness (McAdam et al., 1998; van der Merwe et al., 1997). While unable to proliferate, anergic T cells may be capable of providing effector cytokines, and must be recognized as possible participants in autoimmune disease (McAdam et al., 1998; Mondino et al., 1996; Salojin et al., 1998; Van Parijs and Abbas, 1998) Furthermore, cytokines from anergic T cells such as IFN-y, may polarize T cell-dendritic cell interactions toward a Th1 profile Activation of a nai've T cell occurs when the TCR and CD28 signals are of appropriate strength to initiate IL-2 production and proliferation, but not so strong that the dividing population becomes fratricidal. Activation induced cell death occurs after recur rent activation via large numbers of TCR molecules. While activation-induced cell death involves an initial proliferative phase, this is followed by the Fas-mediated apoptosis of the dividing cell population. Activation-induced cell death is essential for limiting peripheral immune responses to foreign antigen, as well as to eliminate autoreactive T cells. Activation-induced cell death is dependent upon Fas-FasL interactions on CD4+ T cells, and TNF-TNFRp55 interactions on CD8+ T cells, and both humans and mice defective in either Fas or FasL suffer from multiple autoimmune disorders, due to a predominance of autoreactive T cells in the periphery (Ucker et al., 1989). Regulatory Lymphocyte Populations Rather than directly modulating effector T cells, dendritic cells may assert their effects by promoting regulatory T cell populations. Several regulatory T cell populations have been identified in the NOD, and most appear to function by steering T cell responses toward the Th2 phenotype, which is associated with protection from diabetes. One type of regulatory T cell, the NK-T cell, expresses both a TCR with invariant a and~ chains as well as NK receptors NKl .1 and Ly49. These cells appear specially suited to respond to self-peptides presented on CDld, a primitive MHC class I type
PAGE 21
13 molecule (Chiu et al., 1999). When engaged via the TCR, these cells release preformed intracellular stores of either IL-4 or IFN-y, skewing the local immune response toward Th1 or Th2 type responses. NOD mice have reduced numbers and are functionally deficient in these cells when compared to nonautoimmune strains, giving possible explanation to reduced Th2 responses and IL-4 in the NOD (Gombert et al., 1996a). Additionally, increasing the number of these cells either via cytokines or transgenic TCR prevents diabetes in the NOD (Gombert et al., 1996b; Lehuen et al., 1998). The Th3 cell appears to predominantly secrete IL-4 and TGF-~. It was shown that these cells are induced by nasal and oral administration of self-antigens, and appear to mediate resistance to autoimmune diease (Han et al., 1997; Maron et al., 1999). A third T cell population, the CD38+CD45Rb low T cell, do not secrete IL-10 or TGF-~ but have been functionally characterized as inhibiting anti-CD3-mediated T cell responses in vitro (Read et al., 1998). In addition, they are induced following CFA treatment of NOD mice, which delays the onset of diabetes (Martins and Guas, 1999). It is difficult to summarize the diverse literature on regulatory T cell populations. While the function of a regulatory T cell in prevention of autoimmune disease has been demonstrated, mechanisms accounting for ineffective generation of these cells in the NOD mouse have not been defined. If dendritic cell function is impaired in the NOD, this could result in an inability to generate effective numbers of regulatory T cells. Myeloid dendritic cells were reported to generate Th2 type responses in vitro, thus an inability to derive fully competent myeloid dendritic cells may lead to inadequate generation of regulatory T cells. Diabetes Onset and Progression in the NOD Mouse The NOD mouse exhibits lymphocytic infiltration of the peri-islet areas around 4 weeks of age Subsequently, intra-islet infiltration and cell destruction occurs in all NOD males and females, and diabetes develops in 70-80% of females, and 25-50% of males by 30 weeks of age. The infiltrates contain both CD4 and CD8 T cells expressing
PAGE 22
14 a variety of vp chains. Macrophages or dendritic cells appear to be the earliest entrants into the pancreas, as they are detected in the peri-islet areas one week after transfer to irradiated NOD females (O'Reilly et al., 1991). p cells close to the infiltrating cells in these areas express higher levels of class I, but do not express class II or CD54 (O'Reilly et al., 1991). Both CD8+ and CD4+ T cells play a role in initiation of insulitis and P cell destruction. The initial insulitis appears to be caused by CD8+ T cells, as Class I-deficient NOD mice do not develop diabetes or insulitis CD4+ T cells, in particular Th 1 cells, appear responsible for triggering actual destruction of p cells, as Class II KO mice suffer limited islet damage, but still have significant insulitis (Forster and Lieberam, 1996) Factors that lead to the eventual T cell activation and expansion in the NOD mouse can be broken down into three basic categories; underlying products within the target tissues, atypical MHC class II molecule, and deficiencies in T cell activation and apoptosis. None of these alone are sufficient to cause autoimmune disease, but together lead to the 75% occurrence of diabetes in female NOD mice. Defects in Pancreatic p Cells In juvenile NOD mice an underlying genetic defect leads to death of p cells leading to upregulation of chemokines and adhesion molecules, and infiltration of macro phages and dendritic cells (Faveeuw et al. 1994a). A similar infiltration and interaction is seen within the salivary glands which are also destroyed in NOD mice (Faveeuw et al., 1994a). These intrinsic defects in the target tissues of autoimmune disease are thought to be the factors that initiate and focus autoreactivity toward these tissues. Contributions of MHC Class II The major predisposing factor in Type I diabetes in both humans and mice is the MHC class II molecule. The NOD mouse carries the atypical I-A g7 molecule, which is necessary for the onset of diabetes in the mouse. Congenic NOD mice carrying I-Ab
PAGE 23
15 of the C57Bl/10 mouse do not develop diabetes (Wicker et al., 1992). In humans, HLA DQP 1-0201, -0302, 0101 are found in 85-95% of patients with Type I diabetes, while only possessed by 20% of the Caucasian population (McDevitt, 1998). Both the NOD and human class II alleles share a homozygous lack of Asp at position 57 in the class II p chain. Substitution of Asp at position 57 protects NOD mice from diabetes (Quartey Papafio et al., 1995). This susceptibility is most likely due to a combination of an inherent instability of the peptide-MHC complex, and an increased ability to bind and present p cell specific antigens. Defects in Activation Induced Cell Death In addition to the atypical MHC class II molecule, defects in the ability of nai"ve T cells to undergo activation and proliferation which are necessary for the elimination autoreactive T cells via activation induced cell death, contribute to expanded numbers of autoreactive T cells in the periphery of the NOD mouse. For example, production of IL-2, a cytokine critical for activation-induced cell death, is decreased in the NOD mouse (Serreze et al., 1989). In addition, Delovitch and colleagues have noted proximal defects in the intracellular signaling cascade occurring after T cell receptor ligation (Jaramillo et al., 1994). Finally, treatment of NOD mice with recombinant IL-2 or treatment of NOD mice with non-deleting anti-CD3 antibody markedly reduces diabetes incidence (Serreze et al., 1989; Vallera et al., 1992). We hypothesize that these factors alone are not solely responsible for disease. Central to all of these is the dendritic cell. It is the most potent antigen presenting cell known, the most prominent presenter of MHC class II-peptide complexes, and is found in the pancreas of very young NOD mice Transfer of dendritic cells from the pancreatic draining lymph node was sufficient to protect NOD mice from developing diabetes (Clare-Salzler, 1991 ) This study establishes that these cells are capable of inducing long lasting peripheral tolerance to self-antigens, but are unable to do so in their endogenous location. Our hypothesis suggests that dendritic cell development, maturation, and migra-
PAGE 24
16 tion could impair the ability to either eliminate autoreactive T cells or generate regulatory T cell populations, and as such contribute significantly to the autoimmune predisposition of the NOD mouse.
PAGE 25
IDENTIFICATION AND CHARACTERIZATION OF MURINE DENDRITIC CELL POPULATIONS Background Initially described in 1973 as a single population of antigen presenting cells with an extremely high potential for T cell stimulation, the dendritic cell is now recognized as the most potent of the antigen presenting cells (Knight et al. 1983; Steinman and Cohn 1973) Techniques for purification of these cells would not come until many year s later, with recognition that these were more than a single population of dendritic cells not occurring until the late I 980's (Crowley et al., 1989; Steinman et al., 1979). We now know that dendritic cells from the mouse originate from one of two distinct lineages, myeloid or lymphoid (Shortman and Maraskovsky, 1998). While cells of myeloid lineage maintain high levels of surface antigens such as CD 11 b, characteristic of all myeloid cells, lymphoid dendritic cells express the antigen CD8a, typically seen only on Class I-restricted T cells (Shortman, 2000). Whether these lineages have unique functions, or are even exclusive from one another, remains a debated topic In the late 1990s, several investigators attempted basic characterization of den dritic cell populations in murine secondary lymphoid tissues (Salomon et al. 1998; Shortman 2000; Vremec et al., 2000). This revealed that both myeloid and lymphoid dendritic cells existed in the secondary lymphoid tissues of the mouse, and that several key phenotypic differences separated the two populations. The NOD mouse develops severe insulitis and p cell destruction leading to diabetes beginning at 6 to 12 weeks of age (Bowman et al. 1994). The first cells seen infiltrating the pancreas are myeloid dendritic cells and macrophages (Shinorniya et al., 17
PAGE 26
18 2000). Studies have shown this APC infiltration may be due to defects with the cells themselves, and may be enhanced by the atypical expression of chemokines and adhesion molecules (Faveeuw et al., 1994a; Rosmalen et al., 2000a). Conversely, dendritic cells isolated from the pancreatic-draining node protect juvenile NOD mice from diabetes (Clare-Salzler et al., 1992). Why dendritic cells from the pancreatic draining node protect against disease while those within the islets themselves may promote disease was a key question that arose from these studies. To begin to address this issue we performed enumeration and characterization of dendritic cell subtypes from the uninvolved inguinal node and the pancreatic draining node, to ascertain if obvious discrepancies in dendritic cell number or activation state were pres ent. To properly perform such an analysis, we first had to describe all dendritic cell populations seen within the nodes in a manner appropriate for our future studies. Materials and Methods Mice Female C57BL/6, NOD/LtJ, Balb/c, and CBA mice between 4-16 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center. Creation of Single-Celled Suspension from Lymph Node Lymph nodes were harvested from 3 to 6 mice into ice cold Hank's Buffered Salt Solution (HBSS) without calcium or magnesium (Mediatech Herndon, VA). Nodes were then resuspended in a solution of Ix HBSS with calcium and magnesium and 100 U/rnL collagenase D (Boehringer Mannheim Indianapolis), transferred to a petri dish on ice, and needle dissected using 1.5 inch 27-gauge needles. Free cells and tissue fragments were then transferred to a 15 rnL tube and 4 rnL of 400 U/rnL collagenase D in HBSS was added. The solution was mixed, then incubated at 37 for 30 minutes. The cell suspension
PAGE 27
19 was then pipetted up and down 10 to 20 times with a disposable plastic pipette to dissociate all cell fragments and clusters. The cell suspension was immediately transferred to 40 mL of ice cold HBSS without calcium or magnesium containing 1 mM EDTA and 1 % BSA. Cells were washed twice in this solution and resuspended in 1 mL of the same solution with 4% BSA. Cells were counted on the hemacytometer and Ix 106 cells were aliquoted to each tube for staining Analysis of Surface Phenotype Surface markers were analyzed by flow cytometry. Briefly, cells were blocked in 100 L of HAB (Hanks Buffered Salt Solution without calcium or magnesium, with 1 mM EDTA, 0.01 % sodium azide, and 1 % bovine serum albumin (Calbiochem) with 4% BSA+ 1 g anti-CD 16/32 (purified clone 2.402) and stained with the appropriate surface antibodies. Antibodies used were: CDl le (clone HL3, Pharmingen), CD40 (clone 3/23, Pharmingen), CD54 (clone 3E2, Pharmingen), CD80 (clone 16-10Al, Pharmingen), CD86 (clone OL-1, Pharmingen), 1-NI-E (clone 209, Pharmingen), 1-Ak (A-k) (clone 10-3.6, Pharmingen), which reacts with I-Ag7 of the NOD mouse. Biotinylated antibodies were detected with either streptavidin-PE (Coulter-Immunotech, Miami, FL), or strepta vidin-APC (Molecular Probes, Eugene, OR). The dye 7-AAD (7-aminoactinomycin D, Molecular Probes, Eugene, OR) was added at 1 g per mL to some samples to exclude dead cells from analysis. After establishing that >90% of cells were viable in all preparations, we substituted additional fluorochrome labeled monoclonal antibodies for the viability dye. Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry, San Jose, CA).
PAGE 28
20 Results Ex Vivo characterization of Dendritic Cell Populations In order to characterize dendritic cell populations in diabetes-prone mice, we first needed to recognize all dendritic cell populations present within murine secondary lymphoid tissues. Utilizing techniques that result in the least amount of manipulation, and the best representation of the dendritic cells present in vivo, nodes were extracted and single cell suspensions created. These cells were stained and analyzed by flow cytometry to characterize and enumerate the dendritic cell populations within the lymph nodes. Five Distinct Dendritic Cell Populations in Ex Vivo Preparations By creating a suspension containing all the cells within the lymph node, we are more likely to retain all the dendritic cell populations present in vivo than if we used an ex vivo manipulation to enrich these cells The flow cytometer can easily do the work of isolating our population of interest for analysis, without the risks of losing certain cells through ex vivo manipulation. In the murine system, only dendritic cells are CDI le+. While preparations from spleen have large numbers of autofluorescent macrophages that impair the ability to discriminate other cell populations, few macrophages are found in the lymph nodes, yield ing a clear phenotypic picture for analysis. Further more, the dendritic cell populations within the spleen vary greatly from those within the lymph nodes. Utilizing antibodies against Class II MHC and CD 11 c, a two-parameter plot reveals 3 distinct CD 1 lc+ dendritic cell populations (Figure 3-1, upper panel) The first of these, labeled as population A, has very high class II expression with a wide range of CDI lc positivity The second, labeled population B, has slightly lower expression of class
PAGE 29
21 i!~ i: Ofi 0 0 200 400 600 800 1000 0 200 400 600 800 1 000 Forward Scatter FOfWllrd Scattu ~ucells 0 0 200 400 600 800 1 000 Forward Scatter Figure 3-1. Three populations of CD11c+ dendritic cells in murine lymph node preparations. After creation of a single-celled suspension as described in materials and methods, cells were stained with antibodies to class II MHC and CD I le. A minimum of I 00,000 events were acquired for each sample. Top panel represents class II MHC versus CDI le staining after gating on high forward scatter events. This allows visualiza tion of our populations using a contour plot. Lower four panels represent the scatter characteristics of those populations indicated in the upper panel. The scatter of the B cell population is given as a point of reference. II but a higher median level of CD 11 c. The third, population C, has very low expression of class II, and an intermediate level of CD 11 c staining. Adding antibodies against CD8a and CDI lb, which discriminate lymphoid and myeloid dendritic cells respectively, allows further classification of these three dendritic cell populations. Examination of population A reveals that these are very large cells, evidenced by their high forward scatter characteristics (Figure 3-1, lower panel) These cells have
PAGE 30
22 moderate amounts of CD 11 b on their surface, and no significant CD8a, indicating that these dendritic cells are of myeloid lineage (Figure 3-2). Dendritic cells within population B have lower forward scatter than those within population A, but still higher than typical lymphocytes (Figure 3-1, lower panel). Attempt ing to classify these as myeloid or lymphoid reveals a clearly bimorphic population. While a portion of these cells exhibit high CDl 1 band low CD8a expression, similar to the cells within population A, the other fraction shows lower CD 11 b expression with high expression of CD8a, indicating these cells are of lymphoid origin (Figure 3-2). Population C is quite distinct in its complete absence of CD 11 b expression While a subset of population C does express CD8a, a significant number of these cells express neither CD 11 b nor CD8a (Figure 3-2). Scatter characteristics of population C were also quite unique, with a majority of the cells falling near the low forward scatter range typically associated with larger lymphocytic cells (Figure 3-1 ). As population C had yet to be described in the literature, the possibility that these were not dendritic cells existed. Run n ing a panel of monoclonal antibodies targeting lineage specific antigens such as CD3, CD19, F4/80, and CD4, it was not possible to detect expression of any additional lineage specific molecules on these cells. Demonstration of these 5 dendritic cell populations, A, B(CD8+), B(CD8-), C(CD8+), and C(CD8-), has been repeated in our lab in at least 30 independent experiments. Having characterized lineage origins of these various dendritic cell populations, the next goal was to characterize other molecules expressed by these different popula tions. Examination of the molecules CD40, CD54 and CD86 revealed much higher levels of expression on cells within population A. This coincides with the elevated class II MHC expression on these cells relative to the myeloid dendritic cells found within population B to indicate that the myeloid dendritic cells within population A are mature myeloid dendritic cells. Population C showed little expression of CD80 or CD86, indicating the immaturity of this dendritic cell population (Figure 3-2).
PAGE 31
'
PAGE 32
z _J a.. z _J a.. A 'st o---------, A 'st o---------, 'st o--------24 'st o-------B 'st 0-.,----------, Figure 3-2 Continued C 'st o----------, 'st o----------, C 'st o---------, c.. (") () aiN c.. 0 ro co~ 8 ~~11.i..~~Plr 00 ---~~..,,..,.,~~...J 104 'st 0-r-------~
PAGE 33
.... (JQ = ; w I N I n 0 ::, .... s s:: (I) 0.. PLN CD8a PerCP 100 10 1 102 103 104 ,6c 111 I J I J1 o----------' .i,.. CD8a PerCP 10 10 1 10 2 10 3 10 4 o..J.,111111l11111J11111nl11111 o.i:.....__ _______ __, CD8a PerCP 10 10 1 10 2 10 3 10 4 ~~W""" o.i:.....__ _______ __, Inguinal CD8a PerCP 100 10 1 102 103 104 CD8a PerCP 10 10 1 10 2 10 3 10 4 o -,1,,'d"'"' = ....... 11111,d """' 0 ..._ ________ ..., .i,.. CD8a PerCP 100 10 1 102 103 104 0 ...__ ________ ..., .i,.. PLN CD8a PerCP 10 10 1 102 10 3 104 -,I llltttl II eeJ I I I I Jj )> O.i,..""-----------' CD8a PerCP 100 10 1 102 103 104 .,, I 111111 I 1111J 11111,I I 1111,41 OJ 0 ----------' .i,.. CD8a PerCP 10 10 1 10 2 10 3 10 4 -, 11 I eyrf ; 111:zl 1111 0 ----------' .i,.. Inguinal CD8a PerCP 100 10 1 102 103 104 --,I 1111111 J I I ... CD8a PerCP 100 10 1 102 103 104 0 .,, 111111111111,I 11111,l 11111 0 ----------' .i,.. CD8a PerCP 10 10 1 10 2 10 3 10 4 ....I. --I 111191b, 1111,# _;!:NJ ....... 0 ..._ ________ ..., .i,.. )> OJ n N VI
PAGE 34
26 We next stained the cells with a host of other antibodies, and found that not only did the lymphoid component of population B have significant amounts of CD38 on its surface, but population C expressed equally high amounts of this molecule suggesting a relationship between cells within population C and the CD8+ component of population B (Figure 3-3) Variations in Dendritic Cell Populations Between Anatomic Sites Although not formally addressed, the assumption amongst immunologists has been that all lymph nodes are equivalent in their cellular makeup. Since some of our original hypotheses revolved around the properties of cells within the pancreatic draining lymph node, we wished to determine whether pancreatic draining node dendritic cell composition was similar to that in other nodes. In addition, as the pancreatic draining node represents a lymph node draining an inflammatory site, effects secondary to inflam mation could occur in this compartment. The first striking difference between nodes from these different locations was the amount of MHC class II found on the surface of the dendritic cell (Figure 3-4). In the mouse, B cells constitutively express high levels of class II on their surface, and make a good reference within each node. While population A dendritic cells from peripheral nodes, such as the inguinal node, showed levels of class II which exceeded that found on B cells, population A dendritic cells from mesenteric nodes, including the pancreatic draining nodes, all have lower levels of class II. While all of the dendritic cell populations described show reduced class II expression in mesenteric nodes, the greatest differences were seen on dendritic cells within population A, the mature myeloid population (Figure 3-4). T cell costimulatory molecules such as CD86 were also decreased on dendritic cells from the pancreatic-draining node (relative to the inguinal node) (Figure 3-2). This was seen on all dendritic cell populations from lymph nodes within the mesenteric cavity. Differences in the percentage of each dendritic cell population also varied greatly between these sites. The proportion of myeloid versus lymphoid dendritic cells within
PAGE 35
27 population B was significantly different between the inguinal and mesenteric nodes, with a greater proportion of lymphoid dendritic cells in the inguinal nodes ( 1 : 1 in the inguinal node in contrast to 2: 1 myeloid versus lymphoid in the pancreatic draining node). Variations in Dendritic Cell Populations by Strain Inguinal and pancreatic draining lymph nodes were extracted from C57BL/6 and NOD mice, dissociated, and dendritic cell populations characterized. No overt signs of dendritic cell activation, such as elevated class II MHC or costimulatory molecule expres sion, were noted in NOD lymph nodes in comparison to C57BL/6 controls. Population C which appear to be the least mature dendritic cells, was significantly elevated in the non-inflamed inguinal nodes of NOD mice relative to C57BL/6 controls (Figure 3-5). A B C '
PAGE 36
.... CIC C ; w I w I n 0 = ...... 5 s:: p. CD8a PerCP _.100 10 1 102 103 104 Q =,f 1111111 I 111111 11111el IIIIP ... o~-------'----' CD8a PerCP _.100 10 1 102 103 104 Q=,I I J 111,I 11111 .. 1 ~, 0 ..._ ______ __,~..., CD8a PerCP _.10 10 1 10 2 10 3 10 4 0..,, ,,,,,,1 ,111111 ,,111,1 ,, nf, CD8a PerCP _.100 10 1 102 103 104 0 -J I 1111 I ., .... 11111,I I 11,M, Q> _. 0 ..._ ________ ....... _. o~----------' CD8a PerCP 100 10 1 102 103 104 o_"l ... ~.J:W.N ..... o~----------' CD8a PerCP _.100 10 1 102 103 104 _. 0 ------------' CD8a PerCP 100 10 1 102 103 104 o~-+-,iav~"'1 _. o~---------~ CD8a PerCP 100 10 1 102 103 104 _. ..... -+. "'ffl: lirl'i' 1_1!11,1 ) .... _. o~----------' "Tl 0 Ql a. (J) (') Ql ::i: "Tl 0 CD8a PerCP 10 10 1 10 2 10 3 10 4 0-.1 1111111 11111J 11111,I 11111111 )> _. 0 o-~--===~--~ 0 CD8a PerCP 100 10 1 102 103 104 o..J 1111111 ,,,,,,. 11111,I 11111.P, a. 1.5~ I OJ (J) (') Ql i "Tl 0 Ql a. (J) (') Ql 0 0 ._ ________ ..., 0 CD8a PerCP 100 10 1 102 103 104 0.,1 1111111 eneeenC 11111,I 1111111( 1 -n _. 0 0 ._-----==-------' 0 N 00
PAGE 37
29 ---Inguinal Node Pancreatic Draining Node Populaton A ~is ~0~1-...................................... 1,4,-,,/ly\~t __ I ....,~ /Q. 1 b0 1 b 1 1 b2 ; 103 \ 1 1 b 4 Population B CD8a+ Population B CD8aPopulation C Class II Fluorescein ,/A : 1 0 o .. w.3 ......... 4 10 10 10 10 10 Class II Fluorescein 1 ~ol )10 ...., ) CD 1+~0--........ -1~~~1..-.-.~\~b-2.,...._1_b_3~~~ .. Class II Fluorescein Figure 3-4. MHC class II variations between anatomic compartments. Cells were prepared and stained as before, then the three dendritic cell populations were gated and their class II expression displayed on a histogram for easier interpretation. The largest dif ferences occur within the myeloid dendritic cell population both within population A and population B Population C was not broken into CD8 positive and negative components due to the lack of clear separation between these populations. Additionally, population A dendritic cells made up a significantly lower proportion of the total dendritic cell milieu in pancreatic draining nodes from the NOD. We initially performed these studies segregating results based on the age of the mice sampled, which showed no differences with respect to age, and subsequent analyses were performed with all data grouped together for increased statistical power. Because of the described interactions between NK-T cells, we also examined the molecule CD Id, an MHC-like lipid binding molecule responsible for presentation of gly-
PAGE 38
35 0 20 0 0 15 Q) 0) 10 Q) 5 0 45 ui 40 c 35 UJ t 30 .... o 25 0 o 20 Q) fir 15 c 10 Q) 0.. 5 0 Strain and Population Inguinal Node .--, Strain and Population Pancreatic Draining Node Figure 3-5. Composition of dendritic cell populations between strain and anatomic location. C57BL/6 and NOD nodes were prepared as described and then analyzed by flow cytometry. The com position of the total pool of CDl lc pos itive events was broken down by each of the three distinct dendritic cell popu lations when analyzed by class II MHC versus CD 11 c. Significant differences are indicated by black overhanging bars. Five mice were pooled for each experiment. The data here represents the average of 12 independent experiments. indicates p<0.01 as determined by paired t-test. 30 colipid antigen to NK-T cells, on den dritic cells. Expression of CDld was elevated on lymphoid dendritic cells from the pancreatic draining node of NOD mice (Figure 3-6) While all den dritic cell populations in the NOD had higher basal levels of CDld than their C57BL/6 counterparts, the level found on lymphoid dendritic cells from the pancreatic draining node far exceeded that seen on other populations. Summary and Conclusions We felt a necessary first step to understanding dendritic cells in our dis ease model was to obtain a complete profile of all dendritic cell populations within murine tissues. Working with previously described protocols for tissue dissocia tion, we modified these procedures to give the greatest cellular yield and via bility with the least amount of manip ulation (Salomon et al., 1998). These techniques consistently give prepara tions of greater than 90% viability and consistent differential counts between experiments for all dendritic cell populations described.
PAGE 39
31 We observed five distinct dendritic cell populations, which coincide with various lineages of dendritic cells previously described (Shortman and Maraskovsky, 1998) Population A has high expression of class II and costimulatory molecules, lacks CD8a, and bears the molecule CD11 b, indicating this population represents mature myeloid dendritic cells Population B contains two separate dendritic cell populations, a CD8a positive component corresponding to the lymphoid dendritic cell population, and a Inguinal Pancreatic Draining Node .... .... Node C'") C'") N N C57BL/6 ~ C) a.. C)C) 0 ~,o 10 10 104 104 .... Q) a.. .... .... c::s CX) 28 88 OC'") C'") (.) N N NOD C) 27 29 c:, c:, c:, C) ~100 10 2 103 104 c:, ~100 ,02 10 3 104 CD1d APC Figure 3-6. CD1d is elevated on dendritic cells from the NOD pancreatic draining node. Cells stained and gated for population B were then analyzed for expression of CD8a and CD 1 d. Number next to each population indicates the median fluorescence intensity for that population during this experiment. Myeloid dendritic cells of population A stained in a similar manner to those shown here (CD8 negative). Population C showed no staining for CD1d. CD8a negative component which represents a more immature myeloid dendritic cell population. Studies with in vitro derived myeloid dendritic cells confirm that CD11 b
PAGE 40
32 staining decreases with maturation, supporting the lineage of the CD8a negative compo nent of population B as less mature myeloid dendritic cells (Shortman, 2000). Population C has very low surface expression of class II MHC, and is devoid of lineage markers besides CD1 lc. Several avenues suggest this cell population is composed of immature lymphoid dendritic cells. First, a recent work described isolation of a low class II dendritic cell population from spleen which, when matured, led to mature lymphoid dendritic cells (Kronin et al., 2000b ). Second, we observed very high levels of CD38 on this population. CD38 has traditionally been associated with immature cells of lymphoid origin (Malavasi et al., 1994) Within population B, the lymphoid component maintains high expression of CD38, implying that the CD8a population of cells in population B were derived from cells within population C. However population C does not uniformly express the lymphoid lineage marker CD8a. When purifying and maturing this population, Shortman and colleagues observed that all cells developed into lymphoid dendritic cells (Kronin et al., 2000b; Vremec et al., 2000). We hypothesize that even those cells within the CD8a negative component of population C will eventually upregulate CD8a and become lymphoid dendritic cells, based on these previous studies and the uniformly high expression of CD38 (Kelly et al., 2001; Vremec et al., 2000) Alternatively, population C may represent a pool of dendritic cells that develop into both mature myeloid and lymphoid dendritic cells. The characterization of dendritic cell populations in different secondary lymphoid tissues is of considerable interest, as little work has assessed whether all lymph nodes contain these dendritic cell populations (Salomon et al., 1998). Our data clearly shows lower levels of MHC class II and T cell costimulatory molecules on dendritic cells from within the peritoneal cavity. Previous literature has correlated decreased class II MHC expression with decreased dendritic cell maturity and stimulatory capacity. The epidermal Langerhan's cell, which patrols the skin and migrates to the inguinal node may have a different phenotype than those myeloid dendritic cells found in other tissues, leading to
PAGE 41
33 the increased class II MHC seen on dendritic cells from peripheral nodes. In contrast, we show that lymphoid dendritic cells from the pancreatic draining node of the NOD mouse express significantly higher levels of the glycolipid-presenting molecule CDld. The increase in CD 1 d was seen only in the pancreatic draining node and not the mesenteric lymph node suggesting that this molecule may be upregulated in lymph nodes draining inflammatory sites Finally, to follow up on previous studies demonstrating a protective effect when dendritic cells from the pancreatic draining node were transferred, but not those from the inguinal nodes, an attempt was made to phenotypically characterize dendritic cells of the pancreatic draining node and those of peripheral tissues We compared dendritic cells within secondary lymphoid tissues of the diabetes-prone NOD mouse and the control C57BL/6 strain, as well as other controls. This revealed an expanded population C within the inguinal nodes of the NOD mouse, and a decreased number of mature myeloid dendritic cells within the pancreatic-draining node. It was not clear why myeloid dendritic cells were reduced in the pancreatic draining node of NOD mice in comparison to the inguinal lymph node or the control C57BL/6 pancreatic draining node. Decreased myeloid dendritic cells within the pancre atic-draining node may relate to dendritic cells trapped within the inflamed pancreatic islets (Papaccio et al., 1999a; Shinomiya et al., 2000). Immature dendritic cells express high levels of CCR6, which chemoattracts the immature dendritic cell to areas of inflam mation (Caux et al., 2000). The islets of the NOD mouse express constitutive levels of chemokines and adhesion molecules that serve to attract immature dendritic cells to the area (Faveeuw et al., 1994a) If these myeloid dendritic cells are unable to exit the inflamed pancreas, we postulate that this would result in a decline in myeloid dendritic cells within the pancreatic draining node. Alternatively, increased levels of cell death within the dendritic cell population, or an inability to develop and proliferate from precur sors may lead to a generalized decrease in these myeloid dendritic cells as well.
PAGE 42
34 The increased number of dendritic cells within population C of the inguinal node in the NOD mouse suggests a defect in maturation of dendritic cells within both the myeloid and lymphoid lineages. This proved fortuitous for these experiments, as it allowed visualization of this population while our isolation techniques were still quite crude. If mature myeloid and lymphoid dendritic cells play a role in the uptake and presentation of antigen in a tolerizing manner, the inability to progress to a fully mature state could partially explain the predisposition of the NOD mouse toward autoimmune disease. This study provides a framework for future experiments in murine dendritic cell biology, and provides potential avenues for exploring the role of dendritic cells in the pathogenesis of not only Type I diabetes, but other diseases of immune tolerance in the mouse.
PAGE 43
DYSFUNCTIONAL DEVELOPMENT AND MATURATION OF MYELOID DENDRITIC CELLS FROM NOD MICE Background Effective control of the immune system involves the coordinated movement of both cells and molecules to defined locations at precise times. Immature myeloid dendritic cells, which exist in peripheral tissues, must move to secondary lymphoid tissues after receiving signals, whether endogenous or exogenous, to most efficiently interact with na'ive T cells, and thus initiate an immune response. Similarly, proteins within the dendritic cell must traffic to their appropriate cellular locations in order for the dendritic cell to bring about the desired outcome, whether that is immunity or tolerance (Steinman, 1991; Steinman et al., 1997 a). The NOD mouse develops insulitis by 6 weeks of age and T cell dependent autoimmune diabetes by 8-12 weeks of age. As in humans, the NOD mouse requires the presence of an atypical MHC class II molecule lacking aspartic acid at position ~57 (Corper et al., 2000; Quartey-Papafio et al., 1995). The NOD H-2g7 is necessary, but not sufficient for initiation of diabetes in this model (McDevitt, 1998; Ridgway et al., 1998; Wicker et al., 1992). Additional genetic loci contributing to defective T cell activation in the thymus and periphery, have been described in this strain (Arreaza et al., 1997; Rapoport et al., 1993b). However the MHC, IddlO, Idd3, and Idd5 appear to be most critical and can reconstitute almost full diabetes susceptibility in congenic mice (Cordell et al., 200 I). Though defects in T cell signaling could allow the survival of autoreactive cells by inhibiting AICD or generation of regulatory T cells in the periphery, a role for 35
PAGE 44
36 APC regulation of these potentially deleterious cells can be established. Early studies demonstrated that dendritic cells transferred from pre-diabetic mice were capable of limiting disease onset in the recipient. Only cells from the lymph nodes draining the pancreas were capable of this protection, suggesting that cells in this compartment either carry specific antigens, or are exposed to certain cytokines or other factors that render them tolerizing upon transfer (Clare-Salzler et al., 1992). Subsequent studies ascertained that NOD splenic dendritic cells stimulated ex vivo with IFN-yprovided similar protection as those from the pancreatic draining node (Shinomiya et al. 1999). These data suggest that endogenous NOD dendritic cells require additional stimuli to develop functions required for prevention of autoimmune disease. Additional studies demonstrated a protec tive effect after transfer of in vitro bone marrow derived myeloid dendritic cells, and again did not appear to require a diabetes related antigen to provide a protective effect (Morel et al., 1999; Papaccio et al., 2000) Dendritic cells are among the earliest leukocytes found in the pancreas of diabetes prone animals, and accumulate throughout the course of disease. These dendritic cells form networks with T cells similar to those found within secondary lymphoid tissues where they locally prime autoreactive T cells (Ludewig et al. 1998) Macrophages from the NOD mouse are limited in their capacity to differentiate in response to M-CSF and respond to IFN-y (Leiter and Serreze, 1992; Serreze et al. 1993a; Serreze et al., 1993b ). Myeloid dendritic cells develop from the same precursor as the macrophage, and may manifest similar defects to those seen in development of this myeloid lineage. To investigate this possibility, we examined lymphoid tissues from the NOD mouse and found they harbored increased numbers of immature dendritic cells (relative to the C57BL/6 control strain). We then derived myeloid dendritic cells from NOD bone marrow, and compared these with dendritic cells generated from various con trol strains for phenotype, maturation, and function. Our data demonstrates that defects within NOD hematopoeitic cells lead to generation of atypical dendritic cell populations
PAGE 45
37 which are highly resistant to typical maturation stimuli (e.g. LPS, anti-CD40). We have additionally identified a region on chromosome 3 near the ldd 10 locus that contributes approximately 50% in the defective maturation of NOD dendritic cells. We postulate that as a result of this defect, when encountering typical maturation signals and self-antigen in the pancreas, these cells are retained in the islets of Langerhans' instead of becoming activated and traveling to the draining lymph node Within the tissue these activated dendritic cells are capable of priming autoreactive T cells in a manner that leads to the destruction of cells. As a corollary, these dendritic cells, which mature and migrate to the pancreatic draining node, acquire tolerizing function and help to regulate the inflam matory process by promoting regulatory cells as previously described (Clare-Salzler et al., 1992). Materials and Methods Mice Female C57BL/6, NOD/LtJ, Balb/c, CBA, NOD.BI0-H2b, NOD.IddlO, and C57BL/6.c 17 mice 4-12 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center. Generation of Myeloid Dendritic Cells Tibias and femurs were extracted and excess tissue removed by scraping with a scalpel blade. Ends of the bones were cut, and bones flushed with 2-5 mL of 4 C RPMI-C (RPMI-1640 (Mediatech, Herndon, VA) with 10% FCS (Mediatech) and Ix Pen Strep-Neo (Life Technologies, Rockville, MD). Red cells were lysed using a standard ammonium chloride lysis buffer (0.15M NH4Cl, 1.0 mM KHC03 0.1 mM Na2EDTA, pH 7.2). Cells were resuspended at lxl06 cells per mL of RPMI-C. Murine recombinant GM-CSF (R&D Systems, Minneapolis, MN) and IL-4 (Pharmingen, San Diego, CA)) were added at 500U and lOOOU per mL, respectively. Cells were plated in 24 well culture
PAGE 46
38 dishes (Corning Inc., Corning, NY) at 1 mL per well. After 48 hours, 500 L of media was aspirated from each well, and replaced with 500 L of RPMI-C containing fresh cytokines. After 96 hours, cells were matured by adding either 1 g of LPS (Sigma Chemi cal, St. Louis), 5 g anti-CD40 (clone HM40-3, no azide, low endotoxin, Pharmingen), or a combination of LPS and 1 OOOU mouse IFN-y (R&D Systems) to each well. Cells were matured for 24-48 hours and harvested for further analysis. Analysis of Surface Phenotype Surface markers were analyzed by flow cytometry. Briefly, cells were harvested from culture and washed in HAB (calcium, magnesium, and phenol free HBSS with 1 % BSA, 0.1 % NaN 3, and 1 mM EDTA) A total of 5xl05 cells were then aliquoted to tubes for staining. Cells were blocked in 100 L of HAB with 4% BSA+ lg anti CD16/32 (purified clone 2.4G2) and stained with the appropriate surface antibodies. Antibodies used were: CDl lc (clone HL3, Pharmingen), CD40 (clone 3/23, Pharmingen), CD54 (clone 3E2, Pharmingen), CD80 (clone 16-lOAl, Pharmingen), CD86 (clone GL-1, Pharmingen), I-A/I-E (clone 2G9, Pharmingen), I-Ak (A-k) (clone 10-3.6, Pharmingen), which reacts with I-A87 of the NOD mouse. Biotinylated antibodies were detected with either streptavidin-PE (Coulter-Immunotech, Miami, FL), or streptavidin-APC (Molecu lar Probes, Eugene, OR). 7-AAD (7-aminoactinomycin D, Molecular Probes, Eugene, OR) was added at 1 g per mL to all samples to exclude dead cells from analysis. Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immunocytometry, San Jose, CA). Purification of Hematopoetic Stem Cells For all column based cell enrichments, products from StemCell Technologies (Vancouver, BC, Canada) were utilized, unless noted. Briefly, bone marrow was harvested and red cells lysed as described above, then resuspended at 5x 107 cells per mL in HBSS
PAGE 47
39 with 5% FCS (MediaTech, Herndon, VA) and 5% rat serum (Sigma, St. Louis, MO). A biotinylated lineage cocktail consisting of clones Ly-1 (CD5), B220 (CD45R), MacI (CDl lb), Gr-1 (Myeloid Differentiation Antigen), and TERl 19 (RBCs) was added to the cell suspension. An anti-biotin anti-dextran tetramer was then added, followed by dextran coated magnetic particles. The suspension was then run through the magnetic column, and effluent washed, counted, and plated as described for whole marrow cultures above Purification of CD4+ T cells For the allogeneic mixed leukocyte reaction (MLR), CD4+ T cells were purified from CBA spleens utilizing magnetic based negative selection. Enrichment was per formed exactly as that used for stem cell enrichment described above, with the following exceptions. Spleens were taken from CBA mice, smashed to release cells, and red blood cells lysed. An antibody cocktail containing clone MacI (CD 11 b ), B220 (CD45R), CDS, Gr-1 (Myeloid Differentiation Antigen), and TERl 19 (RBCs) was utilized (StemCell Technologies). Effluent was washed, counted, and verified to be >98% pure CD4+ T cells by flow cytometry Proliferation Assessment with CFSE To assess the proliferation of CD4+ T cells cocultured with various dendritic cell populations, T cell were resuspended at lxl07 cells per mL of PBS, labelled with 2 L per mL of cell suspension using a 5 mM stock of the dye CFSE (5-(and 6)-carboxyfluorescein diacetate, succinimidyl esther, Molecular Probes, Eugene, OR). Following MLR, cells were resuspended in HAB, stained with CD4 biotin followed by a streptavidin-APC secondary, and finally resuspended in a solution containg the viability dye 7-AAD. CD4+ viable cells were gated for analysis. Regions were placed around peaks of CFSE intensity, and events within those regions enumerated Number of mitotic events was computed using the techniques and formulae of Wells (Wells et al., 1997).
PAGE 48
40 Uptake of Fluorescent Labeled Proteins or Particles To assess uptake of particulate antigen, standard technique s utilizing fluorescent labeled dextran, ovalbumin, and BSA were utilized. Dendritic cells were suspended in 1 mL of RPMI C with 5 mM HEPES, and 1 mg of Rhodamine Green-Dextran, DQ Ovalbumin, or DQ-BSA (Molecular Probes Eugene, OR). Cells were incubated for 6 hours at either 4 or 37, and then analyzed for fluorescence. Both DQ-OVA and DQ-BSA are self-quenched BODIPY dyes that must be cleaved in order for fluorescence to be seen. Additionally both the Rhodamine green and BODIPY dyes are pH insensitive. Thus, no signal is lost due to the pH of the lysosomal compartment. ELISA for Cytokine in Culture Supernatants Cytokine concentrations in culture supematants were analyzed using products from Pharmingen (San Diego, CA) unless specified otherwise The following antibody pairs were utilized for cytokine detection, listed as capture and detection respectively; IL-2 (JES6-1A12 & JES6-5H4), IL-4 (BVD4-1Dl 1 & BVD6-24G2), IFN-y (R4 6A2 & XMGl.2), IL-10 (JES5-2A5 & SXC-1), IL-12p40 (Cl5.6 & Cl7.8), IL-12p70 (9A5 & C17.8), A streptavidin-HRP secondary was added, followed by addition of a TMB substrate. The reaction was stopped using IN H2S04 and read at 450 nm on a 3550-UV microplate reader (Bio-Rad, Hercules, CA). Results Identification of Protective Dendritic Cell Populations Revisiting an Old Friend In the early 1990 's, our lab performed experiments involving adoptive transfer of dendritic cell populations to juvenile NOD mice. These dendritic cells migrated from the foot pad to the popliteal lymph node where they stimulated cellular proliferation and
PAGE 49
(1) "C 0 z .c Q. E 2 Unmanipulated "o~------, 41 After metrizamide gradient "o..-----------, "o..-----------, Figure 4-1. Metrizamide gradients enrich mature myeloid dendritic cells. Single cell suspensions from lymph nodes and in vitro derived myeloid dendritic cell cultures were layered onto 14.5% metrizamide gra dients and centrifuged at 600 xg for 10 minutes. Cells at the interface were removed and either stained for flow cytometery (top right), or incubated for 2 hours with rhodamine green dextran and stained for surface expression of class II and CD 11 c. Both cases showed the majority of the cells found at the interface after cen trifugation to be mature myeloid dendritic cells. These results were repeated in three independent experiments. Data from female NOD mice is shown. protected against develop ment of diabetes At the time, we were unaware of dendritic cell subsets and the intricacies of den dritic cell maturation. To characterize the protec tive population(s), we repeated dendritic cell isolation from lymph nodes as it was initially performed, and analyzed isolated cells for phe notype and maturation state. This analysis dem onstrated that the pancre atic draining node den dritic cell, which had pro tected NOD mice from diabetes, were in fact mature myeloid dendritic cells (Figure 4-1 ). To better understand the function of the pancreatic draining node dendritic cells, we isolated dendritic cells from both the inguinal node and the pancreatic draining node and stimulated them in vitro with LPS and anti-CD40. We then assayed the supernatant for soluble IL-12p40, IL-12p70, and IL-10. Both IL-12p70 and IL-10 were undetectable with our techniques. IL-12p40 was secreted in significantly higher concentrations from
PAGE 50
42 p<0 .001 I I 2000 E 1800 --5 1600 -o 1400 Q) g 1200 1000 ... a.. 800 0 '
PAGE 51
43 (a) + GM-CSF and IL-4 Only + LPS + anti-CD40 ;; "' "' "' u u u a. a. a. C57BU6 <(N <(N <(N
PAGE 52
44 15-25% mature myeloid dendritic cells after initial culture. In addition, a third population of dendritic cells, expressing high levels of costimulatory molecules, but lacking high surface MHC class II expression, were found in NOD cultures (Figure 4-3a). We refer to this atypical third population as pseudo-mature, for reasons that will be demonstrated later NOD Dendritic Cells Do Not Respond To Maturation Stimuli To assess whether these excess immature NOD dendritic cells would respond to maturation stimuli, cultures were treated with either 1 g/mL LPS, or 5 g/mL anti-CD40 and incubated an additional 24 to 48 hours. Cells were then assessed for percent mature dendritic cells in each treatment. This showed significant differences between C57BL/6 and NOD cultures for all maturation stimuli tested, with NOD cultures undergoing significantly less maturation than their C57BL/6 counterparts (Figure 4-3). The pseudo mature population was not affected by any of these stimuli. Dendritic cells generated from additional non-autoimmune strains, Balb/c and CBA, showed a phenotype and maturation pattern identical to C57BL/6 cultures (data not shown). MHC Instability Does Not Affect NOD Dendritic Cell Phenotype Based on several reports of the inherent instability of the NOD I-Ag7 molecule, we questioned if this could lead to the maturation resistance, or the presence of the psuedo mature population seen in NOD dendritic cell cultures. Cultures were established using NOD B 10-H-2b congenic mice, which carry I-Ab from the C57BL/10. This strain showed no improvement in dendritic cell maturation, and maintained the psuedo-mature popula tion seen in stock NOD cultures (data not shown). These experiments were repeated no less than 10 times.
PAGE 53
a ... C ., ~MO ~-C h C) -2! E-., 0 "' -_g et: 0 ~ 10 0 b 6 00 12000 500 10000 4 00 i g,oo 0 18000 o 6000 200 4 000 2000 45 TQ)~ P-M ). M 101 102 10 3 104 C lass II APC ... C !!., )( 0 Cl C e ~ C) -" C E-., 0 'il -.c et: 0 ~10 0 8 000 7000 -6000 }sooo ~40 00 3000 : := 2000 101 102 10 3 1 0 4 C D86 APC Figure 4-4. Phenotypically immature dendritic cells are functionally immature. (a) After 4 days of culture, NOD dendritic cells were incubated with rhodamine green dextran for 2 hours. Cells were then stained with antibodies to Class II MHC, CDI lc, and CD86 A gate was set on CD 11 c+ events to eliminate debri s, and those cell s displayed. M are mature dendritic cells I immature and P-M are pseudomature ; (b) cytokine s present in culture s upernatant s from C57BL/6 and NOD dendritic cells treated with m a turation s timuli for 24 hours. indicates p<0.01 a s determined by by paired t-test. NOD Dendritic Cells With Immature Phenotype Take Up Particulate Antigen Though cells we described as phenotypically immature fit previously publi s hed description s of immature dendritic cells (CDl lc+, moderate s urface class II MHC, low or no costimulatory molecule expression), it was necessary to verify whether these cells functionally resembled immature dendritic cells. Cells from culture were incubated with Rhodamine Green-Dextran for 4-6 hours, and then s tained for MHC class II, CD1 lc, and CD86. Thi s demon s trated that cells cla ss ified phenotypically a s immature dendritic cell s maintained high level s of antigen upt a ke, con s istent with previou s report s of immature dendritic cell function. These s tudies also revealed th a t the atypical third dendritic cell population, which appe a rs phenotypically mature with the exception of surface cla ss II
PAGE 54
46 I ** 200
PAGE 55
47 Cytokine Dosage Does Not Affect NOD Dendritic Cell Development We noted early in our experiments that high doses of GM-CSF were capable of limiting maturation of C57BL/6 dendritic cells prior to addition of maturation stimuli (data not shown). The possibility existed that NOD stem cells could be more or less responsive to the combination of GM-CSF and IL-4 than their C57BL/6 counterparts. NOD bone marrow cultures were established with concentrations of GM-CSF and IL-4 ranging from lOOU/rnL (1/5 the standard GM-CSF dose and 1/10 the standard IL-4 dose) up to 5000U/rnL (lOx the standard GM-CSF dose and 5x the standard IL-4 dose). Cytokine concentrations were checkerboarded so multiple combinations were realized. Cultures were analyzed before and after addition of maturation stimuli, and no significant increase in maturation of NOD dendritic cells was seen with any increase or decrease in cytokine concentration (data not shown). NOD Dendritic Cells are Poor Stimulators of Allogeneic T Cells As further support, NOD dendritic cells were cultured with MHC mismatched T cells and cytokine production and proliferation assessed. Again utilizing the NOD.H-2b congenic, we cultured C57BL/6 and NOD.H-2b dendritic cells with CBA CD4+ T cells, and at 96 hours, the previously established point of greatest effector cytokine concentra tion, measured cytokines in the culture supernatants. C57BL/6 dendritic cells stimulated significantly higher levels of proliferation, IL-2, IFN-y, and IL-4 from CBA T cells than did similar NOD.H-2b dendritic cells (Figure 4-5). Investigating Causes of NOD Maturation Abnormalities Microscopy of NOD Dendritic Cell Populations To better understand the pseudo-mature dendritic cell population, we sorted and stained cells from NOD cultures and examined them by light microscopy This revealed that morphologically, the pseudo-mature dendritic cells resemble typical mature dendritic
PAGE 56
48 C57BL/6 Mature Population NOD Mature Population NOD Pseudo-Mature Population NOD Pseudo-Mature DC Figure 4-6. Microscopic analysis of dendritic cell populations. In upper panels and lower left panel cells from dendritic cell cultures were cultured stained and sorted, then cytospins were prepared from the sorted cell populations. Slides were stained with Wright's stain and viewed at lOOx. Lower right panel represents cells unique to NOD dendritic cell cultures. Cells are stained for class II MHC (green), CD86 (red), and DNA (DAPI), th e n visualized by deconvolution microscopy.
PAGE 57
(a) C57BL/6 NOD .. .. 0 0 (]) 5;, 3: Vl :8 :8 C C C ro u ... c.>-~ .... I-0 ~,ou 10 1 10' 10 10 4 1 0 4 Class II FIT C (b) +G4 + anti-CD40 0 "2 29% u ~NO V') ci ~0 -2 @ z 00 00 -,oo 10' -,oo 10' ,o' ,o' ,o Clau IIFIT C + (c) C57BL/6 NOD 'st" 2 CV) "2 5;, "2 0 u IN ~NO "'O ~-(]) u c.:: ;:: ::::, 00 00 0... -,oo ,il-' .. ,oo 10' ,., .. ClnsnFITC ClaNUFITC Figure 47. Persistent developmental defects of the NOD dendritic cell lin eage. Dendritic cell cultures were estab lished using (a) transwell systems, (b) bone marrow from NOD.SCID mice, and ( c) purified CD34+ stem cells, none of which lead to proper development of NOD dendritic cells. Number indicates percentage CD86 high dendritic cells, cal culated as the mean of three replicate experiments. 49 dendritic cells from both C57BL/6 and NOD cultures (Figure 4-6). Additionally, we adhered unsorted cells from C57BL/6 and NOD dendritic cell cultures to alcian blue coverslips, stained the cells with antibodies against class II and CD86, then visualized on the deconvolution microscope. On all slides from NOD cultures, we visualized a population of cells expressing large amounts of CD86, but little class II. These cells were never visualized on slides from C57BL/6 bone marrow cultures We concluded that these must be the psuedo-mature dendritic cell population. These cells harbor little intracellular class II, indicating the lack of surface class II expression on these cells is related to impaired transcription and/or translation of class II genes, and not an inability to bring class II peptide complexes to the cell surface (Figure 4-6). Soluble Factors Do Not Direct NOD Dendritic Cell Development To assess the role of soluble factors in dendritic cell development, NOD and C57BL/6 dendritic cell cultures were established either above or below a 0.2 mm membrane fed and matured with LPS as before, then analyzed for phenotype on day 5 of culture. NOD dendritic cells cultured either above or below C57BL/6 dendritic cells maintained a mostly immature phenotype. Additionally, the pseudo-mature pheno type was not diminished Conversely, C57BL/6 dendritic cells cultured with NOD den-
PAGE 58
i 14000 } 12000 510000 8000 6000 4000 2000 e 4000 8: 3500 ; 3000 2500 ; 2000 ; 1500 1000 500 C578U6 NOD ** ;;; o,+-i:_,..!S ,,_.~.,.~~i:_,,.!S..,._,~ V V J; C57 BU6 ... NOD....-""1400 E 1200 l1000 800 :x: 600 400 e 200 ** I C57BU6 NOD 50 dritic cells showed no inhibition of mat uration, nor induction of the pseudo mature dendritic cell population (Figure 4-7a). T, B, and NK-T Cells Do Not Affect NOD Dendritic Cell Development We established cultures using bone marrow from the NOD.SCID mouse, which lacks T, B, and NK-T Figure 4-8. NOD dendritic cells undergo a cells. Cultures derived from NOD.SCIO prolonged "maturing" phase. Spontaneously mature dendritic cells from day 4 GM-CSF mice were similar phenotypically to and IL-4 cultures were harvested and purified standard NOD cultures. These cells over 14.5 % metrizamide gradients. Mature dendritic cells were then cultured with the were slightly, but not significantly, more indicated activating stimuli. After 48 hours, supernatants were harvested and cytokines responsive to maturation with LPS. The measured by ELISA. NOD dendritic cells propseudo-mature dendritic cell population duced significantly lower levels of IL-10 and significantly higher levels of IL-12p70 than was unaffected (Figure 47b ), establishC57BL/6 controls. Data represents three replicates. indicates p<0.01, ** indicates p<0.001 ing that direct interactions with T, B, as determined by paired t test. or NK-T cells are not responsible for limiting the ability of NOD dendritic cells to fully mature, and play no role in induction of the psuedo-mature dendritic cell population. A Non-lymphocytic Cell Population Leads to Generation of Pseudo-Mature Den dritic Cells CD34+ stem cells were enriched by negatively selecting lin+ cells, and then cultured with GM-CSF and IL-4 as before. Though the ability to respond to LPS or anti-CD40 did not improve in NOD cultures, the pseudo-mature population was completely eliminated (Figure 4-7c). This indicates either that another cell population
PAGE 59
51 influences development of the pseudo-mature dendritic cell population, or the population itself is derived from a cell type eliminated during stem cell enrichment. NOD Dendritic Cells Become Arrested During Maturation Although the majority of NOD dendritic cells remained immature, a small popula tion did mature without addition of maturation stimuli. We questioned whether these "mature" dendritic cells would have similar function to mature dendritic cells from C57BL/6 cultures. Utilizing the previously described property of the metrizamide gradi ent to isolate mature myeloid dendritic cells, we ran day 4 dendritic cell cultures from NOD and C57BL/6 over gradients to isolate those dendritic cells that had acquired the mature phenotype These dendritic cells were plated and treated with the combination of LPS and anti-CD40 for 48 hours, which provides the maximual activation/maturation stimulus. Data revealed that mature dendritic cells from C57BL/6 cultures produced cyto kines consistent with the fully mature phenotype previously described, producing low levels of IL-12p70 and high levels of IL-IO (Figure 4-8). In contrast, "mature" dendritic cells from NOD cultures produced cytokines characteristic of the "maturing" dendritic cell, with very high levels of IL-12p70 and low levels of IL-10 (Figure 4-8). Genes Within lddlO Affect NOD Dendritic Cell Maturation Screening of NOD Congenic Strains We are fortunate to have a large number of NOD congenic strains available to us, allowing rapid screening of several reported diabetes susceptibility intervals in the maturation of myeloid dendritic cells. This preliminary screen revealed the IddlO interval on chromosome 3 as a candidate for further study.
PAGE 60
52 lddl O Affects Dendritic Cell Maturation We established dendritic cell cultures utilizing NOD, NOD.lddlO, and C57BL/6.c 17 (I-Ag7), giving us identical MHCs on all three strains to be tested. These cultures were assessed for phenotype, cytokine production, and stimulation of allogeneic T cells NOD.lddl O cultures had higher baseline numbers of mature dendritic cells, and responded significantly better to maturation stimuli than NOD dendritic cells (Figure 4-9). lddl O Affects Dendritic Cell Cytokine Production and T Cell Stimulatory Capacity Analysis of supernatants from these cultures revealed that NOD.lddlO dendritic cells produce significantly higher levels of ILl 2p70 and IL-IO than the NOD following addition of LPS. Culture of these dendritic cells with CD4+ CBA T cells showed that I 600 i1 i500 }1 i400 il51 ils 300 a, a, ., io 200 "t <:a :! :! 100 0 0 0 ..,. ..,. ..,. Cf) 0 ..,. Cf) 0 ..,. Cf) 0 Cf) 0 ..,. Cf) 0 Cf) 0 Cl Cl 0.. ..,. 0.. ..,. 0.. ..,. 0.. ..,. 0.. ..,. 0.. ..,. t) t) + -' Cl + -' Cl + -' Cl + -' Cl + -' Cl + -' Cl + + .... + t) Cl + t) 0 + t) .... + t) Cl + t) 0 + t) () u Cl 0 u .... + 0 Cl + 0 + u .... + 0 Cl + 0 + c:i c:i () 0 u .... z 0 Cl ,:, 0 .... z 0 Cl 0 0
PAGE 61
53 NOD.lddlO dendritic cells were significantly better at stimulating cytokine production from allogeneic T cells, and in tum produced more cytokines in response to the activated T cells (Figure 4-9). Summary and Discussion The NOD suffers a polygenic immune dysfunction, manifesting itself as a pro pensity for development of autoimmune diseases such as diabetes and Sjogrens-like syndrome. This predisposition to develop autoimmune disease has previously been attrib uted to two factors; atypical T cell activation and resistance to AICD within the T cell compartment, and the structure of the unusual Class II MHC carried by the NOD lacking aspartic acid at position ~57 (Corper et al., 2000; Jaramillo et al., 1994; Kanagawa et al., 1998; Quartey-Papafio et al., 1995; Rapoport et al., 1993a; Rapoport et al., 1993b; Zipris et al., 1991 ). In this paper, we add a third avenue, that antigen presentation in the NOD mouse is insufficient for effective peripheral tolerance due to inadequate dendritic cell development and maturation. Dendritic cells play a vital role in the initiation of immunity to pathogens in peripheral tissues. Immature dendritic cells are found in all peripheral tissues, and do not interact with large numbers of T cells, as they lack significant levels of MHC-peptide complexes, costimulatory molecules, and adhesion molecules required for these interac tions Following antigen uptake and receipt of some activating signal(s), these immature dendritic cells undergo a well-orchestrated maturation and subsequent migration to the draining lymph node. Within the node, the dendritic cell is able to interact with large numbers of na'ive T cells and rapidly initiate an appropriate clonal T cell response. In the pancreas of juvenile NOD mice, dendritic cells are found within the islets (Faveeuw et al., 1994a). Dendritic cell accumulation within the islet may attract T cells and prime them from within the pancreas itself (Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Shinomiya et al., 2000). A similar infiltration and interaction is
PAGE 62
54 seen within the salivary glands, which are also destroyed in NOD mice (Faveeuw et al., I 994a). We have also demonstrated that NOD dendritic cells are refractory to typical maturation stimuli in vitro. Myeloid dendritic cells normally produce high levels of ILI 2p70 while maturing, but down regulate production once fully mature. The fully mature dendritic cell produces large amounts of IL-10, but little or no IL-I2p70 (Langen kamp, 2000). This correlates with an initial priming of Th I type responses, followed by priming of Th2 type responses as ILI 2 production ceases and IL-10 production increases. As the dendritic cell reaches maturity, it has upregulated CCR7 and traversed the afferent lymphatics to the draining node, and has started secreting large amounts of ILI 2p70, capable of priming Th I type responses in the draining lymph nodes. After a period of time, the details of which are only now being elucidated, these dendritic cells switch to an IL-IO producing phenotype (Langenkamp, 2000). In the NOD mouse, which is predisposed to developing a Th I-mediated autoimmune pathology, this lack of transition results in impaired migration and continued IL-I2 secretion, and may prove critical in propagating islet inflammation while limiting generation of regulatory cell populations within the pancreatic draining lymph node. In juvenile NOD mice, we presume an underlying genetic defect leads to death of cells, leading to upregulation of chemokines and adhesion molecules, and infiltration of dendritic cells (Faveeuw et al., I 994a). Normally, these dendritic cells would enter the pancreas, endocytose antigen from dead and dying cells, receive maturation signals from the damaged cells, and then mature and migrate to the draining lymph node where they mediate tolerizing interactions with autoreactive T cells. This is seen to some extent by the fact that cells from the pancreatic draining lymph node do not transfer disease, and dendritic cells from that node are protective when transferred to juvenile NOD mice (Clare Salzler et al., 1992). In our model of diabetes in the NOD mouse, dendritic cells within the islet that should mature and migrate to the draining node instead enter an
PAGE 63
55 extended maturing state. They may be unable to upregulate CCR? and exit the pancreas, and continue to produce high levels of IL-12p70 and little or no ILI 0. These dendritic cells, now effectively trapped within the pancreas by defective maturation, may release chemokines, such as MIP-3~, that further attract lymphocytes and immature dendritic cells to the area (Sallusto et al., 1999). From this location, they may prime nai've autoreac tive T cells, and perpetuate~ cell destruction (Ludewig et al., 1998). Authors of previous papers have argued that dendritic cells from NOD mice are more effective at stimulating T cells than those from control strains (Feili-Hariri et al., 1999; Morel et al., 1999; Papaccio et al., 2000). As has been published, however, increasing T cell activation in NOD mice actually prevents disease (Arreaza et al., 1997; Forster and Lieberam, 1996; Serreze et al., 1989). In addition, T cells from the pancreatic draining lymph node in the NOD mouse show no signs of overt activation, the disease process takes 12-20 weeks to manifest itself, and the memory/nai've phenotypes seen are not indicative of typical ongoing immunostimulation (D' Adamio et al. 1993; Fabien et al., 1995; Forster and Lieberam, 1996). We argue that in light of these previous experiments, as well as those included in this paper, the more acceptable hypothesis is that maturation of NOD dendritic cells is inadequate, thus they fail to migrate appropriately to draining lymph nodes, and do not activate and expand antigen-specific regulatory cell populations. Our studies have clearly shown that bone marrow derived dendritic cells from NOD mice fail to mature in response to LPS and anti-CD40, and develop dendritic cell populations with atypical phenotypes in comparison to control strains. Studies by Delovitch demonstrated that treating NOD mice with anti-CD28 antibodies was sufficient to prevent disease (Arreaza et al., 1997), suggesting that antigen-presenting cells responsible for expressing CD28 activating molecules such as CD80 and CD86 are dysfunctional in NOD mice. The only cells capable of presenting MHC-peptide complexes along with high levels of CD80/86 to T cells in the absence of exogenous stimuli are dendritic cells. Salomon and
PAGE 64
56 Bluestone demonstrated that blocking CD28 signaling exacerbated diabetes in NOD mice, impairing homeostasis of CD4+CD25+ regulatory T cells. This finding suggests that some endogenous capacity to deliver signals through CD28 remains intact in the NOD mouse (Salomon et al., 2000). Work from Stockinger showed that ligation of additional TCRs was sufficient to prevent diabetes in NOD mice (Fossati et al., 1999). Since we know ligation of CD28 decreases the number of TCRs required for T cell activation, we can relate the capacity of APC to transduce sufficient signal to the T cell with the prevention of autoimmune disease (Gudmundsdottir et al., 1999; Schrum et al., 2000; Wells et al., 1997). The question of which function these dendritic cells fail to carry out is of consider able interest. It is well documented that diabetes in the NOD mouse is dominated by Th 1 type cytokines such as IFN-y (Bradley et al., 1999; Sarvetnick, 1997). Treatment of NOD mice with Th2 type cytokines, such as IL-4, prevents disease (Arreaza et al., 1997; Cameron et al., 1997b; Gombert et al., 1996b; Maron et al., 1999; Mueller et al., 1997; Rapoport et al., 1993a; Teros et al., 2000). Some treatments that prevent disease, such as the previously mentioned anti-CD28 therapy, are ineffective when anti-IL-4 antibodies are administered, indicating a substantial role for IL-4 induction in these prevention strategies (Arreaza et al., 1997). Several studies have shown that before a T cell can produce IL-4, it must undergo several cell divisions (Bird et al., 1998; Lederer et al., 1996; Schweitzer et al., 1997). This requires both the ligation of large numbers of TCRs and simultaneous costimulation. An additional requirement is to be where the action is, i.e., the T cell areas of the secondary lymphatics. As the myeloid dendritic cell fits these requirements, it then follows that the myeloid dendritic cell is the responsible ( or irresponsible) party in this scenario.
PAGE 65
a-GALACTOSYLCERAMIDEACTIVATES AND EXPANDS NK-T CELLS AND RECRUITS MYELOID DENDRITIC CELLS SPECIFICALLY WITHIN THE PANCREATIC DRAINING NODE OF NOD MICE Background NK-T cells represent a bridge between the innate and adaptive immune systems. Though they express a T cell receptor and the molecule CD3, they lack subclass specific markers such as CD4 and CD8. These cells also express a wide array of factors typically associated with NK cells, such as perforin and granzyme, and surface antigens such as NKl .1 and Ly49C (Bendelac et al., 1997) The invariant TCR of the NK-T cell is unique in that it interacts with the MHC Class I-like molecule CDld (Bendelac et al., 1995). The CDld molecule is found on many cell types, but its expression on dendritic cells is most crucial for activation of NK-T cells. First, dendritic cells appear to express higher levels of the CD 1 d molecule, and are well suited to process and present antigen in the context of this molecule. Second, the NK-T cell must receive costimulatory signals through CD28, just as a naive T cell would (Bendelac et al. 1997). The invariant TCR specifically recognizes glycolipid antigens within the context of CDld (Brossay et al. 1998a ; Brossay et al., 1998b) The synthetic glycolipid a-galactosylceramide binds to CDld and acts as a ligand for the invariant TCR, and treatment of mice with this glycolipid leads to activation and expansion of NK-T cells (Brossay et al., 1998b ). Once encountering its ligand, the NK-T cell rapidly secretes large amounts of the cytokines IFN-y and IL-4. Upon restimulation by its ligand, the NK-T cell orients to an IL-4 secretion pattern (Chen and Paul, 1997; Chiu et al., 1999; Hammond et al., 57
PAGE 66
58 1999). Many studies have shown the positive effects of IL-4 treatment on Type I diabetes (Berman et al., 1996; Cameron et al., 1997a; Fox and Danska, 1997). The NOD mouse has been shown to have an intrinsic defect which limits the number and function of NK-T cells in the periphery (Falcone et al., 1999; Gombert et al., 1996a). Increasing the number of NK-T cells has proven to protect against development of disease (Hammond et al., 1998; Lehuen et al. 1998) Previous studies support that classical T cells provide signals to the dendritic cell e.g. CD40L, which promote dendritic cell activation However, it is not known whether the NK-T cell-dendritic cell interaction leads to a similar activation of the dendritic cell As NK-T cells recognize the glycolipid-CD 1 d complex presented by dendritic cells and are activated, how might this affect the dendritic cell? A recent study has shown that dendritic cells from mice treated with a-galactosylceramide do in fact show increased IL-12 production (Kitamura et al., 1999). In previous studies, we have shown that myeloid dendritic cells from the NOD mouse do not develop and mqture properly in vitro. We also demonstrated an elevated level of the molecule CDld on dendritic cells specifically within the pancreatic draining node. Recent studies also demonstrated increased numbers of NK-T cells in this I ymph node in contrast to other lymph nodes draining non-inflamed sites (Laloux et al., 2001 ). Many studies have reported the early entrance and persistence of dendritic cells within the pancreatic islets (Dahlen et al., 1998; Faveeuw et al., 1994a; Jansen et al., 1994; Ludewig et al., 1998; Papaccio et al., 1999a; Papaccio et al., 1999b; Rosmalen et al., 2000a; Rosmalen et al. 1997; Rosmalen et al., 2000b; Shinomiya et al., 2000). Results from our collaboration with Dr. Brian Wilson (Beth Israel Deaconess Medical Center, Boston, MA) suggest that there are few NK-T cells in the islets of NOD mice in contrast to genetically similar, disease resistant strains (NOR). We hypothesized that if activated NK-T cells exited the pancreatic draining node, their entry into inflamed islets may provide maturation signals to resident dendritic cells that promote their migration to the
PAGE 67
59 pancreatic draining node. If so, this would suggest a novel role not yet described for NK-T cells, that of promoting dendritic cell emigration and resolution of inflammation Materials and Methods Mice Female NOD/LtJ mice 6-8 weeks of age were bred and housed in the Department of Pathology mouse facility at the University of Florida Health Science Center. Treatment with a-galactosylceramide Female NOD/LtJ mice were injected with a-galactosylceramide (KRN7000 Kirin Brewery Gunma, Japan) or a-mannosylceramide (AGL595, Kirin Brewery, Gunma, Japan) at a concentration of 10 g/mL in 0 5% Tween/ PBS. Two i.p. injections of 250 Leach were given the first on day O and the second on day 5. Lymph nodes were extracted from groups of 3 mice on either day 10 or day 20 Single cell suspensions were created as described previously (Chapter 3 materials and methods) a-Galactosylceramide In Vitro Recall Mice were treated, nodes harvested and single suspensions created as previously described but cells were washed in HBSS without calcium and magnesium and 1 % BSA, but 1 mM EDTA was not added to preserve complete function of the isolated cells. Cells were plated at 2 x 106 cells per mL in RPMI complete media with a 1 x concentration of a-galactosylceramide. The exact concentration of a-galactosylceramide was not known, but was titered in the lab to give optimal results, and that optimum concentration was then considered to be 1 x. After 48 hours, supernatants were harvested and assayed for cytokine concentration
PAGE 68
60 ELISA for Cytokine in Culture Supernatants Cytokine levels in culture supernatants were analyzed using products from Pharmingen (San Diego, CA) unless specified otherwise. The following antibody pairs were utilized for cytokine detection, listed as capture and detection respectively; IL-2 (JES6-IA12 & JES6-5H4), IL-4 (BVD4-IDI I & BVD6-24G2), IFN-y(R4-6A2 & XMGI.2), IL-IO (JES5-2A5 & SXC-1), IL-12p40 (Cl5.6 & Cl7. 8), IL-12p70 (9A5 & Cl 7.8), A streptavidin-HRP secondary was added, followed by addition of a TMB substrate. The reaction was stopped using IN H2S04 and read at 450 nm on a 3550-UV microplate reader (Bio-Rad, Hercules, CA). Analysis of Surface Phenotype Surface markers were analyzed by flow cytometry. Briefly, cells were washed in HAB (calcium, magnesium, and phenol free HBSS with 1 % BSA, 0.1 % NaN3, and 1 mM EDTA), 5x I 05 cells were then aliquotted to tubes for staining. Cells were blocked in 100 L of HAB with 4% BSA+lg anti-CD16/32 (purified clone 2.4G2), and then stained with the appropriate surface antibodies. Antibodies used were: CDI le (clone HL3, Pharmingen), 1-Ak (A-k) (clone 10-3.6, Pharmingen), CD8a (clone Ly-2, Pharrnin gen), CD Id (clone Ly-38, Pharmingen), CD4 (clone RM4-5, Pharmingen), TCR-b (clone H57-597, Pharmingen), CD19 (clone 1D3, Pharmingen). Biotinylated antibodies were detected with streptavidin-APC (Molecular Probes, Eugene, OR). Samples were acquired and analyzed on a 6-parameter FACSCalibur flow cytometer (Becton-Dickinson Immuno cytometry, San Jose, CA).
PAGE 69
61 Results Treatment of NOD Mice with a-Galactosylceramide Expands NK-T Cells Specifi cally Within the Pancreatic Draining Node Mice treated in vivo with a-galactosylceramide were sacrificed and inguinal and pancreatic draining nodes were extracted. Single cell suspensions were created as before, and cells were stained with antibodies against CD4, CD8, T cell receptor (TCR), and CD 19. Since NOD mice are non-reactive with the antibody against NK 1.1, the population of TCR+, CD4-CD8cells is the only means to approximate NK-T cell number in these mice. Upon opening the peritoneal cavity of the treated mice, the pancreatic draining lymph node is grossly enlarged. While the inguinal node remains unremarkable compared to untreated mice, the pancreatic draining node has almost doubled in cell number (Figure 5-1 ). No significant difference in cell number was seen between treated 20000000 -g ~15000000 !!!. tj 10000000 15000000 0 .. C 'i?, I (.) p<0.01 z ...J Q. g g (.) 3 Q. g .. Figure 5-1. Increased cellular yield from a-galactosylceramide treated pancreatic nodes. Mice were treated as described with two injections of a-galactosylceramide (a-GC), then inguinal and pancreatic drain ing nodes (PLN) were harvested, dissoci ated, and cells counted on a hemacytome ter. This data represents five separate exper iments, p value calculated by paired t-test. and untreated mesenteric lymph nodes (data not shown) Phenotypic analysis of these cells reveals significant increase in the number of TCR+, CD4-CD8events from the treated pancreatic draining node, with no detectable increase in these cells within the inguinal nodes (Figure 5-2). As stated before, this is the only technique of phenotypically assessing these cells in NOD mice, as they lack the NK 1.1 antigen found in other strains. These results have
PAGE 70
62 1 0 Days After In i t ial Tr e atment 20 Days A fter In i tial Treatment 1/) 5 1 ... .. ~ C .. I G> '8 I > z : j w ii I C CX) :i C DI i (.) .5 ,d, C o (.) .. + .. 5 1 co.. '8 I 0:: z ~ DI (.) C 3 J .... c ... f I C Q G> 2 u .., .. .. G> Q. C .. 0 D. Veh ide a m an osyl a -gat a ctosyl V eh icle a -man osyl a-ga l actosyt Onl y ceramid e ceramide Vehide a -manosyt a-galactosyl Only ce ra mi de ce ramide O n ly ceramide cer a m i d e Figure 5-2. Expansion of double negative T cells within the pancreatic draining node after a-galactosylceramide treatment. Mice were treated with two injections, as described, nodes harvested, dissociated, counted and s t a ined then the number of CD4 negative CD8 negative TCR po s itive events were enumerated. CD 19 was also included to elimi n a te B cell s from any of the s e a naly s es. Re s ults repres ent th e mean of five replicate experiments indi cate s p<0.01 a s determined by paired t-test. been confirmed by our col laborator Dr. Wil s on who ha s demonstrated increa s ed numbers of NK-T cell s by quantitative mRNA analy s i s of the invariant T cell receptor. Increased Functional Responses Ex Vivo Follow ing a-Galactosylceramide Treatment To better as s es s the effect of a-galactosyl ceramide on in vivo NK-T cell number and function, the functional characteri s tic s of NK-T cells from NOD mice treated with this lipid were assessed in vitro by recall response Following in vivo treatment with doses of a galactosylceramide or vehicle alone pancreatic draining and inguinal lymph nodes were extracted dis s ociated and single cell s u s pension s were plated with or without a-galacto s ylceramide. After 96 hours, s upernatants were harvested and s oluble cytokines mea s ured by ELISA. A 5-10 fold increase in IL 4 and IFN-y production by cells from the pancreatic draining lymph node, but not from the inguinal node, were observed in response to in v itro a-ga lactosylceramide recall (Figure 5-3). These increases were only noted in mice previously primed in vivo with a-galactosylceramide, indicating that NK-T cells were indeed activated and/or expanded within the pancreatic draining lymph node by a galactosy lceramide treatment.
PAGE 71
63 V=Vehicle only a=treated in vitro with a-galactosylceramide 7 X=l. Smg a-galactosylceramide lx=S.Omg a-galactosylceramide, 350 1600 E 300 E 1400 ....... ....... .& 1200 .& 250 VI l:'.! 200 5 1000 ::, 0 0 J: 800 J: \() \() 150 @J 600 @J 100 ,:-"'1' z 400 = 50 !:!:: 200 0 0 In vivo V lX 2X V lX 2X V lX 2X V lX 2X In vivo V lX 2X V lX 2X V lX 2X V lX 2X In vitro V V V a a a V V V a a a In vitro V V V a a a V V V a a a Inguinal PLN Inguinal PLN 600 1200 = 500 E 1000 Cl .& ....... VI 400 Cl 800 5 .& 0 VI J: 300 5 600 \() 0 J: @J \() 200 400 u.. @J Vl y 0 100 ..'.J 200 0 In vivo V lX 2X V lX 2X V lX 2X V lX 2X In vivo V lX 2X V lX 2X V lX 2X V lX 2X In vitro V V V a a a V V V a a a In vitro V V V a a a V V V a a a Inguinal PLN Inguinal PLN Figure 5-3. In vitro recall response to in vivo administered a-galactosylceramide. Mice were treated as described with vehicle or doses of a-galactosylceramide. On day 10, lymph nodes were harvested, dissociated and 2x 106 cells cultured per well in 24 well plates for 96 hours with the indicated stimulus. Supematants were harvested and cytokine concentration measured by ELISA. Data from one representative experiment of 3 replicates. In addition, we observed a dramatic increase in production of IL-10, with a dose dependent decrease in GM-CSP production after treatment with a-galactosylceramide (Figure 5-3). Secretion of these cytokines after a-galactosylceramide treatment (presum ably directly from, or induced by NK-T cells) has not been previously described in the literature Secretion of high amounts of GM-CSF is of particular interest in the context of its ability to limit dendritic cell maturation in vitro (see chapter 4).
PAGE 72
64 Increased Numbers of Myeloid Dendritic Cells in the Pancreatic Draining Node After a-galactosylceramide Treatment To determine if these expanded and primed NK-T cells would have any effect on dendritic cell emigration from the islets dendritic cell number within the pancreatic draining and inguinal nodes was assessed both with and without a-galactosylceramide treatment. We also stained sections of pancreas for CD 11 c to assess the number of dendritic cells within the inflamed islets after a-galactosylceramide treatment. Knowing that dendritic cells located within the pancreas are myeloid dendritic cells (lymphoid dendritic cells are not found in peripheral tissues), the percent and absolute number of myeloid and lymphoid dendritic cells was determined, considering all of population A and the CD8a negative component of population B as myeloid dendritic cells. This analysis showed a significant increase in the percentage of myeloid dendritic cells within the pancreatic draining node, and a greater than 2 fold increase in the 400000 350000 0 () 300000 Q) 250000 () Q) 200000 _2 150000 0 _2 100000 C:X: 50000 "O .r: a. E 3' Inguinal Nodes "O 5 .r: a. E 3' "O .r: a. E 3' Vehicle cx-ManCer cx-GalCer Pancreatic Draining Nodes "O "O "O "O "O "O 5 : 5 5 5 .r: .r: .r: a. a. a. E ::E E ::E E ::E 3' 3' 3' Vehicle cx-ManCer cx-GalCer Figure 5-4. Accumulation of myeloid dendritic cells within the pancreatic draining node following a-galactosylceramide treatment. Following treatment with vehicle, a-mannosylceramide, or a-galactosylceramide lymph nodes were harvested on day 10 and the persentage of myeloid and lymphoid dendritic cells was enumerated as previously described. Numbers here represent the absolute number of dendritic cells present within each node, and is the mean of four independent experiments. indicates p<0.01 as determined by paired t-test.
PAGE 73
700 600 500 0) C. -400 0 "
PAGE 74
66 ......-. 100 -,-;:::==================;----------, (J) 80 (l) (l) ..C 60 ro -0 "'-40 0 (l) U 20 C (l) -0 u C a-NOD/veh e-NOD/aGal ..---No D/CDKO/veh +--NOD/CD1 KO/aGal 16 20 24 Age (weeks) 28 Figure 5-6. Treatment of 4 week old NOD mice with a-galactosylceramide delays the onset of diabetes. NOD and CDld deficient NOD females were treated beginning at 4 weeks of age with weekly injections of either vehicle or a-galactosylceramide. This data and graph were generated by the laboratory of Dr. Brian Wilson, Beth Israel Deaconess Medical Center, Boston Massachusettes, and printed here with his permission. Wilson and Dr. Yuri Nuamov, and as such, some of their experimental results are critical to understanding the overall significance of these findings Treatment of 4-week-old mice with a-galactosylceramide significantly decreased the incidence of diabetes (Figure 5-6). This protection was CDld dependent, as treatment of NOD CDld KO mice with a-galactosylceramide demonstrated no protection from diabetes. As mentioned earlier, previous studies have demonstrated the persistence of den dritic cells within the inflamed pancreatic islets of the NOD mouse (Jansen et al., 1994; Ludewig et al., 1998). Our previous data have shown that dendritic cell development and maturation from NOD bone marrow is impaired in vitro, and that these cells appear to go through a prolonged maturing phase, with extended high level IL-12 production.
PAGE 75
67 That a-galactosylceramide decreases disease incidence should come as no sur prise. It's known that this glycolipid binds CDld and interacts with NK-T cells, leading to their activation and expansion. Since previous studies have demonstrated increasing NK-T cell number and function prevents diabetes, our studies are consistent with these previously published reports (Lehuen et al., 1998). Of particular interest was the signifi cant decrease in survival after treatment with a-mannosylceramide. a-mannosylceramide binds CD 1 d, but is not an activating ligand for the invariant TCR of the NK-T cell. This suggests that a-mannosylceramide b l ocks endogenous CDld-NK-T cell activation, indicating that some residual NK-T cell activity in the NOD mouse provides some protection from diabetes. Since we make the assertion that the source of increased dendritic cells in the pancreatic draining node is derived from the population within the pancreas, a logical experiment would be to look for CDl lc+ dendritic cells within the pancreas with and without treatment. Our lab has performed these experiments and the results were quite dramatic. While vehicle treated mice showed large numbers of CDl lc+ cells within the islets. These cells were almost completely lost following a-galactosylceramide treat ment. Semi-quantitative RT-PCR was also performed, and showed a loss of Il.,-12 tran scripts within the islets after a-galactosylceramide treatment. While not providing defini tive proof, the observation that dendritic cells decreased within the islet infiltrates and increased within the draining node suggests that these cells migrated to the pancreatic draining lymph node. How activated NK-T cells affect dendritic cell maturation and migration is still unknown, as is the mechanism for how these cells regulate inflammation and autoim munity We observed high levels of GM-CSF in the supernatants of pancreatic draining node cells from vehicle treated mice that were treated in vitro with a-galactosylceramide. However, when mice were treated in vivo with a-galactosylceramide a majority of this GM-CSF production was lost. Earlier cytokine titration experiments showed that high
PAGE 76
68 levels of GM-CSF limited the ability of myeloid dendritic cells to mature. This high level of GM-CSF production may not only limit the capacity of dendritic cells within the islet to mature and migrate to the draining node, but may also limit the capacity of dendritic cells resident within the pancreatic draining node itself to become functionally mature, limiting tolerance induction within this compartment. To date, no other labs have described the production of GM-CSF by NK-T cells, leaving us with little data with which to compare our results. If substantial numbers of NK-T cells are developing in the bone marrow and the liver, but are not receiving sufficient signals via TCR and CD28, these cells may not progress past the point of high GM-CSF production, and may impair myeloid dendritic cell maturation, a condition that is exaggerated within the inflamed microenvironment of the pancreatic draining node. During in vitro recall experiments, we saw that in vivo primed cells from the pancreatic-draining node produced high levels of IFN-y. This cytokine has been shown to synergize with other signals to increase dendritic cell maturation, cytokine secretion, and T cell stimulatory capacity, and has been shown to impart tolerizing capacity upon normally non-toleregenic splenic dendritic cells (Shinomiya et al., 1999). The increase in IFN-y production within the microenvironment of the pancreatic draining node and the inflamed islets may be sufficient to drive dendritic cell maturation and allow emigration from the islets. Studies examining the anti-tumor effects of NK-T cells revealed that activation with a-galactosylceramide lead to increased IL-12 production by dendritic cells, and that continued IFN-yrelease by the NK-T cell was dependent upon TCR and IL-12R-mediated signals from the dendritic cell, as well as CD40L from the NK-T cell (Kitamura et al., 1999). While these studies used CD40 blocking antibodies to establish this latter interaction, no studies have shown the presence of CD40L on the surface of an NK-T cell. These were the first experiments to demonstrate feedback from the NK-T cell to the dendritic cell. We know the myeloid dendritic cell moves from its immature state to a
PAGE 77
69 state of the "maturing" dendritic cell, where it produces high levels of IL-12, followed by its final "mature" phase, where it secretes high amounts of IL-10 (Langenkamp et al., 2000). Can the NK-T cell provide signals to the dendritic cell to move it from one phase to another? Our data suggests that after expansion and activation of NK-T cells within the pancreatic draining node, myeloid dendritic cells emigrate from the islets, and cell destruction is halted. How might this effect take place? The first possibility is that the expanded NK-T cells in the pancreatic-draining node either secrete chemokines themselves, or induce chemokine secretion from other cells in the node, leading to recruitment of those dendritic cells from within the islets to the lymphoid tissues. We know that when the dendritic cell begins to mature, the chemokine receptor CCR7 will be upregulated and this should guide the dendritic cell to the draining node (Langenkamp et al., 2000). But in the NOD, the islets themselves aberrantly express several chemokines and cell adhesion molecules, such as MIP-1 a, Lymphotoxin-a, and SLC, that may counteract this migration (Faveeuw et al., 1994a; Faveeuw et al., 1994b; Papaccio et al., 1999a; Cameron et al., 2000; Hjelmstrom et al., 2000). The activated NK-T cells may act within the draining node to increase secretion of chemokines such as MIP-3~ (ELC) and 6Ckine (SLC), driving the balance of chemokines in favor of migration to the draining node (Caux et al., 2000). Alternatively, NK-T cells may migrate out of the pancreatic draining node and home to the inflamed pancreas, where they directly interact with dendritic cells and trigger maturation and upregulation of CCR7, leading to a state capable of leaving the islet. Our collaborators did note a significant increase in invariant T cell receptor transcripts within the islets after a-galactosylceramide treatment. This hypothesis still lacks key points necessary for a full understanding of this process. First, do NK-T cells preferentially home to sites of inflammation, such as the islets or the pancreatic draining node? We see NK-T cells expanded specifically within the inflamed pancreatic draining node, but is this due to NK-T cell homing to the pancreatic draining node, or the local
PAGE 78
70 expansion stimulated by dendritic cells within that node (high CD 1 d expression)? If these cells do home to inflammatory sites, they may play a key role in resolving islet inflamma tion by promoting dendritic cell emigration. This hypothesis dramatically expands the role for NK-T cells in the immune response and autoimmune diseases where unresolved inflammation leads to tissue destruction For protection from diabetes when treating 6-week-old NOD mice with a-galactosylceramide, treatments must be repeated monthly. This indicates that once islet damage has exceeded a certain threshold, dendritic cells will simply reinfiltrate the area due to the amount of cellular damage once treatment stops. More recent experiments have shown that treating 4-week-old NOD mice with weekly injections of a-galactosylceramide gives an incidence of diabetes lower than 20%. Again, treatments must be continued for protection from disease. But is this dependance on continued treatment necessarily deleterious? First, it shows that protection of mice from diabetes is a direct effect of the a-galactosylceramide treatment, not some anomaly involved with handling the mice, a shortfall of many previous studies, or a contaminant in the lipid preparation (Atkinson and Leiter, 1999). Second it supports the previous dendritic cell transfer data. In those experiments, dendritic cells were tolerizing after transfer, but not while within their existing environment within the pancreatic draining node, seen by the development of diabetes in untreated NOD mice. Current data supports the pancreatic draining node as a site of tolerance induction, but one that becomes inadequate on its own to overcome intrinsic defects that maintain the inflammatory state within the pancreas (Kurts et al., 1997). Our data shows that even when we can reduce the islet infiltrate in 6-week-old mice, tolerance may not be regained and the islet infiltrate reaccumulates Together these studies indicate that within the environment of the pancreatic draining lymph node, dendritic cells that would normally induce tolerance to self-antigen are unable to exert their affects. This holds true even as we increase the number of dendritic cells within that compartment following treatment with a-galactosylceramide. Since
PAGE 79
71 moving dendritic cells to another location (as in the previous dendritic cell transfer studies) allows tolerance induction, this implicates the local microenvironment of that compartment (the pancreas and its draining node) in limiting tolerance induction. This work suggests that techniques to affect dendritic cell activation, maturation, and migration using intrinsic NKT cells to target these effects to inflamed tissues has great promise for manipulation of dendritic cell populations without unwanted systemic effects. Its applicability to other inflammatory diseases remains to be tested, but the presence of CD1d expressing dendritic cells and the possibility to expand NK-T cells locally should be considered as logical contributors for resolution of the inflammatory state.
PAGE 80
SUMMARY AND CONCLUSIONS This work has focused on the role of dendritic cells in the pathogenesis of Type I diabetes. Since there is little published work in this area, many gaps in the current knowledge exist. This study attempts to fill a few of those holes, creating a new framework for future understanding of dendritic cell biology in the NOD mouse, and the role of interactions with NK-T cells in autoimmune and inflammatory processes. Our first studies involved basic characterization of dendritic cell populations within the secondary lymphoid tissues. We were able to visualize dendritic cell popula tions for the first time, and generate better descriptions of others. This allowed the characterization of these populations from lymph nodes of the diabetes-prone NOD mouse The molecule CDld was significantly elevated on lymphoid dendritic cells from the pancreatic draining lymph node of the NOD, which lead to questions surrounding the role of these cells as activators of NK-T cells in this location Mature myeloid dendritic cells were decreased in number in the node that drains the pancreas, and a surplus of immature lymphoid dendritic cells existed in the peripheral inguinal lymph node. This data, coupled with the previously cited reports of dendritic cells within the pancreas, led us to question the functional aptitude of both myeloid and lymphoid dendritic cells in the NOD mouse. A key fact discovered at this point was identification of the exact dendritic cell population transferred during the protection experiments performed in the early 1990's. We found that metrizamide gradients selectively enrich mature myeloid dendritic cells, and that these isolated dendritic cells appear to be of the "exhausted" phenotype described by Langenkamp as stimulating Th2 type responses (Langenkamp et al., 2000) This 72
PAGE 81
73 means the fully mature myeloid dendritic cell in some fashion imparted the tolerance in those early studies The next study focused on generation of myeloid dendritic cells from murine bone marrow, using the autoimmune-prone NOD mouse. We found major deficiencies in the capability of NOD bone marrow-derived stem cells to differentiate to fully mature den dritic cells. No maturation stimuli tested were capable of generating the number of mature dendritic cells seen in cultures from other strains The NOD consistently maintained at least twice as many immature dendritic cells as control strains, and developed an atypical third dendritic cell population with high levels of costimulatory molecules, but low levels of class II MHC. We referred to this population as pseudo-mature, as its functional characteristics (antigen uptake) closely resemble those of typical mature dendritic cells The functional significance of these cells is unclear now, but the genetic implications are quite significant. That this population can develop in such large numbers from NOD cultures indicates the existence of serious defects in the differentiation pathways of dendritic cells in this strain. We confirmed that NOD derived dendritic cells are indeed functionally congruent with what is known about immature dendritic cells, as they take up high levels of antigen, produce low levels of cytokine, and are poor stimulators of T cells. Next we demonstrated that NOD dendritic cells appear to undergo a prolonged phase of maturation, becoming ensnared at the developmental point of high IL-I 2p70 production. This extended period of time in the IL-12 producing phase of maturation increases the chances of Th 1 responses to the antigens presented. In a mouse that develops a Th 1-mediated autoimmune disease, this implicates the myeloid dendritic cell as a major player in the process. We identified a genomic interval on chromosome 3, previously described as IddlO, that to some degree affects the maturation of NOD dendritic cells. Dendritic cells from
PAGE 82
74 NOD mice with the Iddl O interval from C57BL/6 mice were more mature phenotypically produced more cytokine, and were better stimulators of allogeneic T cells. The final phase of our study involved attempts to induce maturation or migration of the dendritic cell mass from the islets to the draining node. Knowing that dendritic cells within the islet appear to promote destruction, while those which traverse to the draining node are protective, we hypothesized that simply moving these dendritic cells from the pancreas might protect these mice from p cell destruction. While agents such as LPS, TNF, and CFA have proven effective in inducing dendritic cell maturation in vivo, these act systemically and have much greater potential for lethality than curing disease. We chose to use knowledge from our previous studies and other publications to accomplish the same goals locally. We know that CDld is abnormally elevated on dendritic cells within the pancre atic draining node. CDld is a class I like molecule that presents glycolipid antigen. The NK-T cell is a regulatory cell population which specifically recognizes this CDldglycolipid combination, and as a result becomes activated and secretes large amounts of cytokine. Most studies have focused on what the dendritic cell can do for the NK-T cell. We hypothesized that the NK-T cell could also deliver signals to the dendritic cell, possibly affecting their maturation state. We found that treatment of NOD mice with a-galactosylceramide expanded NK-T cells specifically within the pancreatic draining node, and recruited myeloid dendritic cells from the pancreas to that node. This halted p cell destruction, but was insufficient to convey long-term tolerance top cell antigens. Mice treated with a-galactosylceramide had significantly delayed onset of disease, while those treated with the CDld binding lipid a-mannosylceramide (which binds CDld but does not interact with NK-T cells) had increased disease incidence. This work establishes that dendritic cell developmental defects in the NOD mouse do exist. These defects are genetically linked and most likely contribute to the population
PAGE 83
75 of dendritic cells seen within the pancreatic islets of the NOD mouse. Myeloid dendritic cells are responsible for tolerance induction in previously performed transfer experiments and an inability to progress to the fully mature dendritic cell in vivo most likely leads to the breakdown of peripheral tolerance to cell antigens. But if we use pharmacologic agents to induce dendritic cell migration from the islets, we can stop additional cell destruction from occurring. This proof of potential has significant implications for the treatment of humans recently diagnosed with Type I diabetes Hypothesis of Dendritic Cell Polarization For the past ten years, immunologists have struggled to understand how the immune system makes decisions regarding how it will respond to which stimuli. While others have focused on delineating exogenous from endogenous antigens, and innate "danger" signals present within certain bacteria and viruses, these decisions may in fact come down to a simple numbers game. Basic mathematics Recent insights from our lab, as well as others abroad have shown that the immature and mature dendritic cell are not the only two states for this cell type In fact, the dendritic cell progresses through a "maturing" phase, where it produces high levels of the Th 1 polarizing cytokine IL-12p70. It then proceeds to its final maturation state of low IL-12p70 and high IL-10 production, which favors Th2 type responses. What this model has in its favor is redundancy, many checks and balances that must be overcome for an antigen-specific response to become pro-inflammatory in nature. What type of response will a given dendritic cell induce? Every dendritic cell, once it receives a maturation signal, will progress through the previously described stages of the maturing dendritic cell (pro-Th 1) to the mature dendritic cell (pro-Th2). At this point, the dendritic cell has constitutively high levels of both T cell costimulatory molecules such as CD86, as well as the molecule CD40, ready to receive signals back from antigen-specific T cells. When the dendritic cell receives signals via CD40 from antigen-specific T cells that exceed the predetermined threshold, it becomes locked into
PAGE 84
76 that phase of maturation and primes na"ive T cells accordingly (Langenkamp, 2000). If this occurs early, the dendritic cell will prime Th 1 type responses, if it occurs later, Th2 type responses. So what variables affect the time frame in which a dendritic cell will receive sufficient CD40 signals? First is thymic selection. The majority of T cells specific for self-antigens are eliminated in the thymus, and never make it to the periphery. Those specific for non-self antigens are released to the periphery. This significantly increases the chance that exogenous peptide will be recognized by a T cell versus those of endogenous origin. The affinity of the TCR for an exogenous peptide-MHC complex would also be of significantly greater strength, leading to more rapid upregulation of CD40L on the responding T cell. The greater the number of T cells in the periphery specific for a certain antigen, the shorter the time before the dendritic cell interacts with antigen-specific T cells, increasing the chances that the dendritic cell will be locked into a state of high IL-12 production Since a majority of autoreactive T cells are eliminated in the thymus of non autoimmune-prone mice and humans, the circulating T cell pool contains very few TCRs specific for self antigen. It is in this manner that a dendritic cell taking up self does not promote inflammatory Th 1 type responses toward the host tissues. The most important factor in this model from the standpoint of this study is the maturation time of the dendritic cell. The longer the dendritic cell takes to traverse the 'maturing' phase to arrive at the final state of exhaustion, the greater the chance of it receiving super-threshold CD40 signals while still in the IL-12 producing phase. In the NOD mouse, we see dendritic cells that spend a significantly longer period as 'maturing' dendritic cells than non-autoimmune strains. What this model offers is an explanation for polarization of immune responses without imparting a 'consciousness' upon the immune system to delineate endogenous from exogenous antigen. It can explain why tolerance to tumor antigen can be overcome
PAGE 85
77 by systemic treatment with soluble CD40L. There is no doubt active uptake of tumor antigen taking place at all times, and the ligation of CD40 on these dendritic cells locks them into a Th 1 promoting phase. In the NOD mouse, we have poor thymic tolerance leading to increased numbers of potentially autoreactive T cells in the periphery (Gerling et al., 1994; Kanagawa et al., 1998; Rapoport et al., 1993b ). When coupled with the prolonged maturation time for the myeloid dendritic cell, we start to understand how this mouse has such a high incidence of autoimmune disease. The affinity of specific antigens for MHC also play a role in how well T cells will respond, and it is well established that the atypical MHC, I-Ag7 of the NOD mouse is well suited to tightly bind certain antigenic components of the pancreatic P cell (Corper et al., 2000) The target tissues themselves likely play a role in the initiation of the disease as well, but T cell and dendritic cell defects that lead to Th 1 priming dendritic cells, coupled with the appropriate MHC haplotype, are what likely leads to total ablation of the target organ as opposed to resolution of inflammation Summation While diagnosis of Type I diabetes has steadily improved over the past 10-15 years, our ability to circumvent the inevitability of complete p cell destruction has made no progress. This study shows that if we mobilize dendritic cells from the pancreatic islets, we save intact P cells from destruction. Strategies to affect dendritic cell matura tion and migration specifically within the pancreas and its draining nodes must be explored. a-galactosylceramide is currently in clinical trial for treatment of inflammatory bowel disease, so trials utilizing it in recently diagnosed Type I diabetics would not be difficult to initiate. The use of recombinant chemokines or the introductions of viral vectors capable of delivering cytokine signals to dendritic cells within the pancreas also exist as potential avenues for future exploration. Now that we've begun to understand that resolution of the inflammatory state within the islets will spare intact beta cell mass, and that we can resolve this inflammation by targeting dendritic cells within the pancreas,
PAGE 86
78 novel strategies for targeting these cells in Type I diabetes as well as other autoimmune diseases of chronic inflammation should emerge.
PAGE 87
REFERENCES Adler, A. J., Marsh, D. W., Yochum, G. S., Guzzo, J. L., Nigam, A., Nelson, W. G., and Pardoll, D. M. (1998). CD4+ T cell tolerance to parenchymal self-antigens requires presentation by bone marrow-derived antigen-presenting cells, J Exp Med 187, 1555-64. Albert, M. L., Pearce, S. F., Francisco, L. M., Sauter, B., Roy, P., Silverstein, R. L., and Bhardwaj, N. (1998a). Immature dendritic cells phagocytose apoptotic cells via alphavbeta5 and CD36, and cross-present antigens to cytotoxic T lymphocytes, J Exp Med 188, 1359-68. Albert, M. L., Sauter, B., and Bhardwaj, N. (1998b ). Dendritic cells acquire antigen from apoptotic cells and induce class Irestricted CTLs, Nature 392, 86-9. Ardavin, C., Wu, L., Li, C. L., and Shortman K. (1993) Thymic dendritic cells and T cells develop simultaneously in the thymus from a common precursor population, Nature 362 761-3. Arreaza, G. A., Cameron, M. J., Jaramillo, A., Gill, B. M Hardy, D., Laupland, K. B., Rapoport, M. J., Zucker, P., Chakrabarti, S., Chensue, S. W., et al. (1997). Neonatal activation of CD28 signaling overcomes T cell anergy and prevents autoimmune diabetes by an IL-4-dependent mechanism, J Clin Invest 100, 2243-53. Atkinson, M. A., and Leiter, E. H. (1999). The NOD mouse model of type 1 diabetes : as good as it gets?, Nat Med 5, 601-4. Banchereau, J., and Steinman, R M. (1998) Dendritic cells and the control of immunity, Nature 392 245-52. Bendelac, A., Lantz, 0., Quimby, M. E., Yewdell, J. W., Bennink, J. R., and Brutkiewicz, R.R. (1995). CDI recognition by mouse NKl+ T lymphocytes, Science 268, 863-5. Bendelac, A. Rivera, M. N Park, H.-S., and Roark, J. H. (1997). Mouse CDl-Specific NK 1 T Cells: Development, Specificity, and Function, Annual Review of Immunol ogy 15, 535-62. Berman, M.A., Sandborg, C. I., Wang, Z., Imfeld, K. L. Zaldivar, F. J Dadufalza, V., and Buckingham, B. A. (1996). Decreased IL-4 production in new onset type 1 insulin-dependent diabetes mellitus, J Immunol 157, 4690-6. 79
PAGE 88
80 Bird, J. J., Brown, D.R., Mullen, A. C., Moskowitz, N. H., Mahowald, M.A., Sider, J. R., Gajewski, T. F., Wang, C.R., and Reiner, S. L. (1998). Helper T cell differentiation is controlled by the cell cycle, Immunity 9, 229-37. Bjorck, P., Banchereau, J., and Flores-Romo, L. (1997). CD40 ligation counteracts Fas induced apoptosis of human dendritic cells, Int Immunol 9, 365-72. Bowman, M.A., Leiter, E. H., and Atkinson, M.A. (1994) Prevention of diabetes in the NOD mouse: implications for therapeutic intervention in human disease, Immunol Today 15, 115-20 Bradley, L. M., Asensio, V. C., Schioetz, L. K., Harbertson, J., Krahl, T., Patstone, G., Woolf, N., Campbell, I. L., and Sarvetnick, N. (1999). Islet-specific Th 1, but not Th2, cells secrete multiple chemokines and promote rapid induction of autoimmune diabetes, J Immunol 162, 2511-20. Brossay, L., Chioda, M., Burdin, N., Koezuka, Y., Casorati, G., Dellabona, P., and Kro nenberg, M. (1998a). CDld-mediated recognition of an alpha-galactosylceramide by natural killer T cells is highly conserved through mammalian evolution, J Exp Med 188, 1521-8. Brossay, L., Naidenko, 0., Burdin, N., Matsuda, J., Sakai, T., and Kronenberg, M. (1998b ). Structural requirements for galactosylceramide recognition by CD Irestricted NK T cells, J Immunol 161, 5124-8. Cameron, M. J., Arreaza, G. A., Grattan, M., Meagher, C., Sharif, S., Burdick, M. D., Strieter, R. M., Cook, D. N., and Delovitch, T. L. (2000). Differential expression of CC chemokines and the CCR5 receptor in the pancreas is associated with progression to Type I diabetes, J Immunol 165, 1102-10. Cameron, M. J., Arreaza, G. A., Zucker, P., Chensue, S. W., Strieter, R. M., Chakrabarti, S., and Delovitch, T. L. (1997a). IL-4 prevents insulitis and insulin-dependent diabetes mellitus in nonobese diabetic mice by potentiation of regulatory T helper-2 cell function, J Immunol 159, 4686-92. Cameron, M. J., Arreaza, G. A., Zucker, P., Chensue, S. W., Strieter, R. M., Chakrabarti, S and Delovitch, T. L. (1997b). IL-4 prevents insulitis and insulin-dependent diabetes mellitus in nonobese diabetic mice by potentiation of regulatory T helper-2 cell function, J Immunol 159, 4686-92. Caux, C., Ait-Yahia, S., Chemin, K., de Bouteiller, 0., Dieu-Nosjean, M. C., Homey, B., Massacrier, C., Vanbervliet, B., Zlotnik, A., and Vicari, A. (2000). Dendritic cell biology and regulation of dendritic cell trafficking by chemokines, Springer Semin Immunopathol 22, 345-69.
PAGE 89
81 Cella, M., Engering, A., Pinet, V., Pieters, J and Lanzavecchia, A. (1997). Inflammatory stimuli induce accumulation of MHC class II complexes on dendritic cells [see comments], Nature 388, 7827. Chen, H., and Paul W. E (1997). Cultured NKI.1 + CD4+ T cells produce large amount s ofIL-4 and IFN-g upon activation by anti-CD3 of CD 1, J Immunol 159, 2240-9 Chiu Y. H., Jayawardena, J., Weiss, A., Lee D., Park, S H., Dautry-Varsat, A., and Bendelac, A. (1999). Distinct subsets of CDld-restricted T cells recognize self antigens loaded in different cellular compartments, J Exp Med 189, 103-10. Clare-Salzler, M. J Brooks J., Chai, A., Van Herle, K., and Anderson, C. (1992). Prevention of diabetes in nonobese diabetic mice by dendritic cell transfer, J Clin Invest 90, 741-8. Cordell, H.J., Todd, J. A., Hill, N. J., Lord, C. J., Lyons, P.A., Peterson, L.B., Wicker, L. S., and Clayton, D. G (2001). Statistical Modeling oflnterlocus Interactions in a Complex Disease Rejection of the multiplicative model of epistasis in type 1 diabetes, Genetics 158, 357-67. Corper A. L., Stratmann, T., Apostolopoulos, V., Scott, C. A., Garcia, K. C., Kang A. S., Wilson I. A., and Teyton, L. (2000). A structural framework for deciphering the link between I-Ag7 and autoimmune diabetes, Science 288 505-11 Crowley, M., Inaba, K., Witmer-Pack M., and Steinman, R. M. (1989). The cell surface of mouse dendritic cells: FACS analyses of dendritic cells from different tissues including thymus Cell Immunol 118, 108-25. D' Adamio, L., Awad, K. M., and Reinherz E. L. (1993) Thymic and peripheral apoptosis of antigen-specific T cells might cooperate in e s tablishing self tolerance, Eur J Immunol 23, 747-53. D a hlen, E. Dawe, K., Ohlsson, L., and Hedlund G. (1998). Dendritic cells and macro phages are the first and major producers of TNF-alpha in pancreatic islets in the nonobese diabetic mouse J Immunol 160 3585-93. D a utigny N Le Campion A., and Lucas, B. (1999). Timing and casting for actor s of thymic negative selection, J Immunol 162, 1294-302. d e St Groth, B. F. (1998). The evolution of self-tolerance: a new cell arises to meet the challenge of self-reactivity, Immunol Today 19 448-54. Delemarre, F. G., Simons P J., de Heer, H.J., and Drexhage, H. A. (1999). Signs of immaturity of splenic dendritic cells from the autoimmune-prone biobreeding rat: consequences for the in vitro expansion of regulator and effector T cells, J Immunol 162 1795-801.
PAGE 90
82 Fabien, N., Bergerot, I., Maguer-Satta, V., Orgiazzi, J and Thivolet, C. (1995). Pancreatic lymph nodes are early targets of T cells during adoptive transfer of diabetes in NOD mice, J Autoimmun 8, 323-34. Fairchild, P. J., and Austyn, J.M. (1990). Thymic dendritic cells: phenotype and function, Int Rev Immunol 6, 187-96. Falcone, M., Yeung, B., Tucker, L., Rodriguez, E., and Sarvetnick, N. (1999). A defect in interleukin 12-induced activation and i nterferon gamma secretion of peripheral natural killer T cells in nonobese diabetic mice suggests new pathogenic mecha nisms for insulin-dependent diabetes mellitus, J Exp Med 190, 96372. Faveeuw, C., Gagnerault, M. C., and Lepault, F. (1994a). Expression of homing and adhe sion molecules in infiltrated islets of Langerhans and salivary glands of nonobese diabetic mice, J Immunol 152, 5969-78. Faveeuw, C., Gagnerault, M. C., and Lepault, F. (1994b ). Modifications of the expression of homing and adhesion molecules in infiltrated islets of Langerhans in NOD mice, Adv Exp Med Biol 355, 137-42. Feili-Hariri, M., Dong, X., Alber, S. M., Watkins, S. C., Salter, R. D., and Morel, P.A. (1999). Immunotherapy of NOD mice with bone marrow-derived dendritic cells, Diabetes 48, 2300-8. Forster, I., and Lieberam, I. (1996). Periphera l tolerance of CD4 T cells following local activation in adolescent mice, Eur J Immunol 26, 3194-202. Fossati, G., Cooke, A., Papafio, R. Q., Haskins, K., and Stockinger, B. (1999). Triggering a second T cell receptor on diabetogenic T cells can prevent induction of diabetes, J Exp Med 190, 577-83. Fox, C. J., and Danska, J. S. (1997) IL-4 expression at the onset of islet inflammation pre dicts nondestructive insulitis in nonobese diabetic mice, J Immunol 158, 2414-24. Gallucci, S., Lolkema, M., and Matzinger, P. (1999). Natural adjuvants: endogenous activators of dendritic cells, Nat Med 5, 1249-55. Gerling, I. C., Atkinson, M.A., and Leiter, E. H. (1994). The thymus as a site for evaluat ing the potency of candidate beta cell autoantigens in NOD mice, J Autoimmun 7, 851-8. Gombert, J.M., Herbelin, A., Tancrede-Bohin, E., Dy, M., Carnaud, C., and Bach, J. F. (1996a). Early quantitative and functional deficiency of NKl +-like thymocytes in the NOD mouse, Eur J Immunol 26, 2989-98.
PAGE 91
83 Gombert, J.M., Tancrede-Bohin, E., Hameg, A., Leite-de-Moraes, M. C., Vicari, A., Bach, J. F., and Herbelin, A. (1996b). IL-7 reverses NKl+ T cell-defective IL-4 production in the non-obese diabetic mouse, Int Immunol 8 1751-8. Graser, R. T DiLorenzo, T P., Wang, F., Christianson, G. J., Chapman, H. D., Roopenian, D C., Nathenson, S. G., and Serreze, D. V. (2000). Identification of a CD8 T cell that can independently mediate autoimmune diabetes development in the complete absence of CD4 T cell helper functions, J Immunol 164, 3913-8 Gudmundsdottir, H., Wells, A. D and Turka, L.A. (1999). Dynamics and requirements of T cell clonal expansion in vivo at the single-cell level: effector function is linked to proliferative capacity, J Immunol 162, 5212-23 Hammond, K. I. L., Poulton, L. D., Palmisano, L. J., Silveira, P.A., Godfrey, D. I., and Baxter, A.G. (1998). alpha/beta-T cell receptor (TCR)+CD4-CD8(NKT) thy mocytes prevent insulin-dependent diabetes mellitus in nonobese diabetic (NOD)/Lt mice by the influence of interleukin (IL)-4 and/or IL-10., J Exp Med 187, 1047-56 Hammond, K. J., Pelikan, S. B., Crowe, N. Y., Randle-Barrett, E., Nakayama, T., Tanigu chi, M Smyth, M. J., van Driel, I. R., Scollay, R., Baxter, A. G., and Godfrey, D I ( 1999). NKT cells are phenotypically and functionally diverse, Eur J Immunol 29, 3768-81. Han, H. S., Jun, H. S., Utsugi, T., and Yoon, J. W. (1997). Molecular role of TGF-beta, secreted from a new type of CD4+ suppressor T cell, NY4 2, in the prevention of autoimmune IDDM in NOD mice, J Autoimmun 10, 299-307 Heath, W R., and Carbone, F. R. (2001). Cross-presentation, dendritic cells, tolerance and immunity, Annu Rev Immunol 19, 47-64. Hjelmstrom, P., Fjell, J., Nakagawa, T., Sacca, R., Cuff, C. A., and Ruddle, N. H. (2000) Lymphoid tissue homing chemokines are expressed in chronic inflammation, Am J Pathol 156, 1133-8. Inaba, K., Turley, S ., Yamaide, F., Iyoda, T., Mahnke, K., Inaba, M., Pack, M., Subklewe M., Sauter, B., Sheff, D., et al. (1998). Efficient presentation of phagocytosed cellular fragments on the major histocompatibility complex class II products of dendritic cells, J Exp Med 188, 2163-73. Ja n sen, A., Homo-Delarche, F., Hooijkaas, H., Leenen, P. J., Dardenne, M., and Drexhage, H. A. (1994). Immunohistochemical characterization of monocytes-macrophages and dendritic cells involved in the initiation of the insulitis and betacell destruction in NOD mice, Diabetes 43, 667-75 Ja ra millo, A., Gill, B. M., and Delovitch T. L. (1994) Insulin dependent diabetes mellitus in the non-obese diabetic mouse: a disease mediated by T cell anergy?, Life Sci 55, 1163-77.
PAGE 92
84 Kalinski, P ., Hilkens, C. M., Snijders, A., Snijdewint, F. G., and Kapsenberg, M. L. (1997). IL-12-deficient dendritic cells, generated in the presence of prostaglandin E2, promote type 2 cytokine production in maturing human naive T helper cells, J Immunol I 59, 28-35. Kanagawa, 0., Vaupel, B. A., Xu, G., Unanue, E. R. and Katz, J. D. (1998). Thymic positive selection and peripheral activation of islet antigenspecific T cells: separa tion of two diabetogenic steps by an I-A(g7) class II MHC beta-chain mutant, J Immunol I 61, 4489-92. Karandikar, N. J., Vanderlugt, C. L., Walunas, T. L., Miller, S. D., and Bluestone, J. A. (1996). CTLA-4: a negative regulator of autoimmune disease, J Exp Med 184, 783-8. Kelly, K. A., Lucas, K., Hochrein H., Metcalf, D., Wu, L., and Shortman, K. (2001). Development of dendritic cells in culture from human and murine thymic precursor cells, Cell Mol Biol (Noisy-le-grand) 47, 43-54. Kitamura, H., lwanabe, K., Yahata, T., Nishimura, S.-i., Ohta, A., Ohmi, Y., Sato, M., Takeda, K., Okumura, K., Van Kaer, L., et al. (1999) The Natural Killer (NKT) Cell Ligand alpha-Galactosylceramide Demonstrates Its Immunopotentiating Effect by Inducing Interleukin (IL)-12 Production by Dendritic Cells and IL-12 Receptor Expression on NKT Cells, J Exp Med 189, 1121-7. Knight, S. C., Mertin, J., Stackpoole, A., and Clark, J. (1983). Induction of immune responses in vivo with small numbers of veiled (dendritic) cells, Proc Natl Acad Sci US A 80, 6032-5. Kronin, V., Hochrein, H., Shortman, K., and Kelso, A. (2000a). Regulation of T cell cytokine production by dendritic cells, Immunol Cell Biol 78, 214-23. Kronin, V., Wu, L., Gong, S., Nussenzweig, M. C., and Shortman K. (2000b). DEC-205 as a marker of dendritic cells with regulatory effects on CD8 T cell responses [In Process Citation], Int Immunol 12, 731-5. Kurts, C., Kosaka, H., Carbone, F. R., Miller, J. F., and Heath, W.R. (1997). Class I-restricted cross-presentation of exogenous self-antigens leads to deletion of auto reactive CD8(+) T cells, J Exp Med 186, 239-45. Laloux, V., Beaudoin, L., Jeske, D., Camaud, C., and Lehuen, A. (2001). NK T cell induced protection against diabetes in V alpha 14-J alpha 281 transgenic nonobese diabetic mice is associated with a Th2 shift circumscribed regionally to the islets and functionally to islet autoantigen, J lmmunol 166, 3749-56. Langenkamp, A., Messi, M., Lanzavecchia, A., and Sallusto, F. (2000). Kinetics of dendritic cell activation: impact on priming of TH 1, TH2 and nonpolarized T cells, Nat Immunol 1, 311-6
PAGE 93
85 Lanzavecchia, A. (1999). Dendritic cell maturation and generation of immune responses, Haematologica 84, 23-5. Lederer, J. A., Perez, V. L., DesRoches, L., Kim, S. M., Abbas A. K., and Lichtman, A. H. (1996). Cytokine transcriptional events during helper T cell subset differentiation, J Exp Med 184, 397-406. Lehuen, A., Lantz, 0., Beaudoin, L., Laloux, V., Carnaud, C., Bendelac, A., Bach, J. F., and Monteiro, R. C. ( 1998). Overexpression of natural killer T cells protects Valpha 14Jalpha281 transgenic non obese diabetic mice against diabetes, J Exp Med 188, 1831-9 Leiter, E. H and Serreze, D. V. (1992). Antigen presenting cells and the immunogenetics of autoimmune diabetes in NOD mice, Reg Immunol 4, 263-73. Ludewig, B., Odermatt, B Landmann, S Hengartner, H and Zinkernagel, R. M. (1998). Dendritic cells induce autoimmune diabetes and maintain disease via de novo formation of local lymphoid tissue, J Exp Med 188, 1493-501. Malavasi, F., Funaro, A., Roggero, S., Horenstein, A., Calosso, L., and Mehta, K. (1994). Human CD38: a glycoprotein in search of a function, Immunol Today 15, 95-7. Maron, R., Melican, N. S., and Weiner, H. L. (1999). Regulatory Th2-type T cell lines against insulin and GAD peptides derived from orally-and nasally-treated NOD mice suppress diabetes, J Autoimmun 12, 251-8. Martins, T. C., and guas, A. P (1999). A role for CD45RBlow CD38+ T cells and costimulatory pathways of T-cell activation in protection of non-obese diabetic (NOD) mice from diabetes, Immunology 96, 600-5. Marzo, A. L., Lake, R. A., Lo, D Sherman, L., McWilliam, A., Nelson, D., Robinson, B. W., and Scott, B. (1999). Tumor antigens are constitutively presented in the draining lymph nodes [In Process Citation], J Immunol 162, 5838-45 Matzinger, P. ( 1994 ). Tolerance, danger, and the extended family, Annu Rev Immunol 12, 991-1045. Matzinger, P. (1998). An innate sense of danger, Semin Immunol 10, 399-415. Matzinger, P., and Guerder, S. ( 1989). Does T-cell tolerance require a dedicated antigen presenting cell?, Nature 338, 74-6. McAdam, A. J Schweitzer, A. N., and Sharpe, A. H. (1998). The role of B7 co stimulation in activation and differentiation of CD4+ and CD8+ T cells, Immunol Rev 165, 23 1-4 7
PAGE 94
86 McDevitt, H. 0. ( 1998). The role of MHC class II molecules in susceptibility and resistance to autoimmunity, Curr Opin Immunol 10, 677-81 Mevorach, D., Zhou, J. L., Song, X and Elkon K. B. (1998). Systemic exposure to irradiated apoptotic cells induces autoantibody production, J Exp Med 188 387-92 Mondino A., Khoruts A., and Jenkins, M. K. ( 1996). The anatomy of T-cell activation and tolerance, Proc Natl Acad Sci U SA 93, 2245-52. Morel P.A., Vasquez, A. C., and Feili-Hariri, M (1999). Immunobiology of DC in NOD mice, J Leukoc Biol 66, 276-80. Mueller, R., Bradley, L. M., Krahl, T., and Sarvetnick, N (1997). Mechanism underlying counterregulation of autoimmune diabetes by IL-4, Immunity 7, 411-8. O'Reilly, L.A., Hutchings, P. R., Crocker, P. R., Simpson, E., Lund, T., Kioussis, D., Takei, F., Baird, J., and Cooke, A. (1991). Characterization of pancreatic islet cell infiltrates in NOD mice: effect of cell transfer and transgene expression, Eur J Immunol 21, 1171-80. Papaccio, G., De Luca, A., De Luca, B., Pisanti, F. A., and Zarrilli, S (1999). Detection of dendritic cells in the non-obese diabetic (NOD) mouse islet pancreas infiltrate is correlated with Th2-cytokine production, J Cell Biochem 74, 447-57. Papaccio, G Nicoletti, F., Pisanti, F. A., Bendtzen, K., and Galdieri, M (2000) Preven tion of spontaneous autoimmune diabetes in NOD mice by transferring in vitro antigen-pulsed syngeneic dendritic cells, Endocrinology 141, 1500-5. Perez, V. L., Lederer J. A., Lichtman, A. H., and Abbas, A. K. (1995). Stability of Th 1 and Th2 populations, Int Immunol 7, 869-75. Quartey-Papafio, R., Lund, T., Chandler, P., Picard, J., Ozegbe, P., Day, S., Hutchings, P. R., O'Reilly, L., Kioussis, D., Simpson, E., and et al. (1995). Aspartate at position 57 of nonobese diabetic I-Ag7 beta-chain diminishes the spontaneous incidence of insulin-dependent diabetes mellitus, J lmmunol 154, 5567-75. Rapoport, M. J., Jaramillo, A., Zipris, D., Lazarus, A. H., Serreze, D. V., Leiter, E H Cyopick, P., Danska, J S., and Delovitch, T. L. (1993a) Interleukin 4 reverses T cell proliferative unresponsiveness and prevents the onset of diabetes in nonobese diabetic mice, J Exp Med 178 87-99 Rapoport, M. J ., Lazarus, A.H., Jaramillo, A., Speck, E., and Delovitch, T. L. (1993b). Thymic T cell anergy in autoimmune nonobese diabetic mice is mediated by deficient T cell receptor regulation of the pathway of p2 I ras activation, J Exp Med 177, 1221-6.
PAGE 95
87 Read, S ., Mauze, S., Asseman, C., Bean, A., Coffman, R., and Powrie, F. (1998). CD38+ CD45RB(low) CD4+ T cells: a population of T cells with immune regulatory activities in vitro, Eur J Immunol 28, 3435-47. Ri d gway, W. M., Ito, H., Fasso, M., Yu, C., and Fathman, C. G. (1998). Analysis of the role of variation of major histocompatibility complex class II expression on nonobese diabetic (NOD) peripheral T cell response, J Exp Med 188, 2267-75. Rosmalen, J. G., Homo-Delarche, F., Durant, S., Kap, M., Leenen, P. J., and Drexhage, H. A. (2000a). Islet abnormalities associated with an early influx of dendritic cells and macrophages in NOD and NODscid mice, Lab Invest 80, 769-77. Rosmalen, J G., Leenen P J., Katz, J. D., Voerman, J. S., and Drexhage, H. A. (1997). Dendritic cells in the autoimmune insulitis in NOD mouse models of diabetes, Adv Exp Med Biol 417, 291-4. Rosmalen, J. G., Martin, T., Dobbs C., Voerman, J. S., Drexhage, H. A., Haskins, K., and Leenen, P. J. (2000b). Subsets of macrophages and dendritic cells in nonobese diabetic mouse pancreatic inflammatory infiltrates: correlation with the development of diabetes, Lab Invest 80, 23-30. Rovere, P., Vallinoto, C., Bondanza, A., Crosti, M. C., Rescigno, M., Ricciardi-Castag noli, P., Rugarli, C., and Manfredi, A. A. (1998) Bystander apoptosis triggers dendritic cell maturation and antigenpresenting function, J Immunol 161, 446771. Sallusto, F., Lenig, D., Forster, R., Lipp, M., and Lanzavecchia, A. (1999). Two subsets of memory T lymphocytes with distinct homing potentials and effector functions [see comments], Nature 401 708-12. Salojin, K. V., Zhang, J., Madrenas, J., and Delovitch, T. L. (1998). T-cell anergy and altered T-cell receptor signaling: effects on autoimmune disease, Immunol Today 19, 468-73. Salomon, B., Cohen, J. L., Masurier, C., and Klatzmann, D (1998). Three populations of mouse lymph node dendritic cells with different origins and dynamics, J Immunol 160, 708-17. Salomon, B., Lenschow, D. J., Rhee, L., Ashourian, N., Singh, B., Sharpe, A., and Bluestone, J. A. (2000) B7/CD28 costimulation is essential for the homeostasis of the CD4+CD25+ immunoregulatory T cells that control autoimmune diabetes, Immunity 12, 431-40. Sarvetnick, N. (1997). IFN-gamma, IGIF and IDDM [editorial], J Clin Invest 99, 371-2 Schrum, A.G., Wells, A. D., and Turka, L.A. (2000). Enhanced surface TCR replenish ment mediated by CD28 leads to greater TCR engagement during primary stimula tion, Int Immunol 12, 833-42
PAGE 96
88 Schweitzer, A. N., Borriello, F., Wong, R. C., Abbas, A. K., and Sharpe, A.H. (1997). Role of costimulators in T cell differentiation: studies using antigenpresenting cells lacking expression of CD80 or CD86, J Immunol 158, 2713-22. Serreze, D. V., Gaedeke, J. W., and Leiter, E. H. (1993a). Hematopoietic stem-cell defects underlying abnormal macrophage development and maturation in NOD/Lt mice: defective regulation of cytokine receptors and protein kinase C, Proc Natl Acad Sci US A 90, 9625-9. Serreze, D. V., Gaskins, H. R., and Leiter, E. H. (1993b ). Defects in the differentiation and function of antigen presenting cells in NOD/Lt mice, J Immunol 150, 2534-43. Se rr eze, D. V., Hamaguchi, K., and Leiter, E. H. (1989). Immunostimulation circumvents diabetes in NOD/Lt mice, J Autoimmun 2, 759-76. Shinomiya, M., Fazle Akbar, S. M., Shinomiya, H., and Onji, M. (1999). Transfer of dendritic cells (DC) ex vivo stimulated with interferon-gamma (IFN-gamma) down-modulates autoimmune diabetes in non-obese diabetic (NOD) mice, Clin Exp Immunol 117, 38-43. Shinomiya, M., Nadaho, S., Shinomiya, H., and Onji, M. (2000). In situ characterization of dendritic cells occurring in the islets of nonobese diabetic mice during the development of insulitis, Pancreas 20, 290-6. Shortman, K. (2000). Burnet oration: dendritic cells: multiple subtypes, multiple origins, multiple functions, Immunol Cell Biol 78, 161-5. Shortman, K., and Maraskovsky, E (1998). Developmental options [comment], Science 282, 424-5. Steinman, R. M. (1991 ). The dendritic cell system and its role in immunogenicity, Annu Rev Immunol 9, 271-96. Steinman, R. M., and Cohn, Z. A. ( 1973). Identification of a novel cell type in peripheral lymphoid organs of mice. I. Morphology, quantitation, tissue distribution, J Exp Med 137, 1142-62. Steinman, R. M., Kaplan, G., Witmer, M. D., and Cohn, Z. A. (1979). Identification of a novel cell type in peripheral lymphoid organs of mice. V. Purification of spleen dendritic cells, new surface markers, and maintenance in vitro, J Exp Med 149, 1-16. Steinman, R. M., Pack, M., and Inaba, K. (1997a). Dendritic cell development and maturation, Adv Exp Med Biol 417, 1-6. Steinman, R. M., Pack, M., and Inaba, K. (1997b ). Dendritic cells in the T-cell areas of lymphoid organs, Immunol Rev 156, 25-37.
PAGE 97
89 Teros, T., Hakala, R., Ylinen, L., Liukas, A., Arvilommi, P., Sainio-Pollanen, S Verajan korva, E., Pollanen, P., and Simell 0 (2000). Cytokine balance and lipid antigen presentation in the NOD mouse pancreas during development of insulitis, Pancreas 20, 191-6 Uc k er, D S Ashwell, J D and Nickas, G. (1989). Activation-driven T cell death. I. Requirements for de novo transcription and translation and association with genome fragmentation, J Immunol 143, 3461-9. Vallera, D. A., Carroll, S F., Brief, S., and Blazar, B. R. (1992). Anti-CD3 immunotoxin prevents low-dose STZ/interferon-induced autoimmune diabetes in mouse, Diabetes 41, 457-64. van der Merwe, P.A., Bodian, D. L., Daenke, S., Linsley, P., and Davis, S. J. (1997). CD80 (B7-l) binds both CD28 and CTLA-4 with a low affinity and very fast kinetics, J Exp Med 185, 393-403. Van Parijs L., and Abbas, A. K. (1998). Homeostasis and self-tolerance in the immune system : turning lymphocytes off, Science 280, 243-8 Viney, J. L., Mowat, A. M., O'Malley, J.M., Williamson, E., and Fanger, N A. (1998) Expanding dendritic cells in vivo enhances the induction of oral tolerance, J Immu nol 160 5815-25 Vremec, D Pooley, J., Hochrein, H ., Wu, L., and Shortman, K. (2000). CD4 and CD8 expression by dendritic cell subtypes in mouse thymus and spleen, J Immunol 164, 2978-86. Walunas T. L. Bakker, C. Y., and Bluestone, J. A. (1996). CTLA-4 ligation blocks CD28dependent T cell activation [published erratum appears in J Exp Med 1996 Jul 1;184(1):301], J Exp Med 183, 2541-50. Wells, A. D., Gudmundsdottir, H., and Turka, L.A. (1997). Following the fate of individual T cells throughout activation and clonal expansion. Signals from T cell receptor and CD28 differentially regulate the induction and duration of a prolifera tive response, J Clin Invest 100, 3173-83. Wells, A. D., Li, X. C., Li, Y., Walsh, M. C., Zheng, X. X., Wu Z., Nunez, G Tang, A., Sayegh, M., Hancock, W.W., et al. (1999). Requirement for T-cell apoptosis in the induction of peripheral transplantation tolerance, Nat Med 5, 13037. Whittaker, D. S., Bahjat, K. S., Moldawer, L. L., and Clare-Salzler, M. J. (2000). Autoregulation of human monocyte-derived dendritic cell maturation and IL-12 production by cyclooxygenase-2-mediated prostanoid production, J Immunol 165, 4298-304.
PAGE 98
90 Wicker, L. S., Appel, M. C., Dotta, F., Pressey, A Miller, B. J., DeLarato, N. H Fischer, P A., Boltz, R. C., Jr., and Peterson, L. B. (1992). Autoimmune syndromes in major histocompatibility complex (MHC) congenic strains of nonobese diabetic (NOD) mice. The NOD MHC is dominant for insulitis and cyclophosphamide-induced diabetes, J Exp Med 176, 67-77. Wu, L., Li, C. L., and Shortman, K. (1996). Thymic dendritic cell precursors: relationship to the T lymphocyte lineage and phenotype of the dendritic cell progeny, J Exp Med 184, 903-11. Zipris, D., Lazarus, A.H., Crow, A. R., Hadzija, M., and Delovitch, T. L. (1991). Defective thymic T cell activation by concanavalin A and anti-CD3 in autoimmune nonobese diabetic mice. Evidence for thymic T cell anergy that correlates with the onset of insulitis, J Immunol 146, 3763-71.
PAGE 99
BIOGRAPHICAL SKETCH Keith S Bahjat was born and raised in the small Oklahoma town of Ponca City where his parents, Dhari and Jeanette Bahjat, still live today Following graduation from Ponca City Senior High School, he received his Bachelor's of Science degree in medical technology from Oklahoma State University which included a one year internship at Valley View Regional Hospital in Ada, Oklahoma. Keith is licensed by the American Society of Clinical Pathologists as a Medical Technologist. He was married to his lovely wife Rena on July 2"d, 1993. Following receipt of his undergraduate degree, Keith spent time in South Bend Indiana and Chicago, Illinois working as a medical technologist in the areas of clinical hematology and flow cytometry. 91
PAGE 100
I certify that I have read this study and that in m y opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation for the degree of Doctor of Philosophy .,, I J .. ( (.,/ I, Michael J. Clare-Salzler, Chair Associate Professor of Pathology, Immunology, and Laboratory Medicine I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation for the degree of Doctor of Philosophy. -V{L~ Mark Afl
PAGE 101
This dissertation was submitted to the Graduate Faculty of the College of Medicine and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August 2001 c :(l ~ ----S ~~v~--~ Dean, College of Medicine Dean Graduate S c hool
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EZ83FLW5Z_M24C2N INGEST_TIME 2017-05-24T21:01:18Z PACKAGE AA00054911_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
|
|