PAGE 1
ANALYSIS OF THE RIBOZYME MEDIATED CLEAVAGE OF THE RNA ASSOCIATED WITH RETINAL DISEASE By PATRICK WHALEN 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 2000
PAGE 2
Copyright 2000 by Patrick 0. Whalen
PAGE 3
AKNOWLEDGMENTS I would like to thank my mentor, Dr. Alfred Lewin, for his support and advice. I appreciate the freedom he gave me in carrying out my project. I would also like to thank all my colleagues in the lab for their assistance. I would like to give special thanks to my family for all of their support through the years. lll
PAGE 4
TABLE OF CONTENTS ACKNOWLEDGMENTS ................................................................. ................... ...... ID LIST OF FIGURES ...................................... .... .... ....... ................................................... VI ABSTRACT ..................................................................................................................... VII CHAPTERS 1 BACKGROUND AND SIGNIFICANCE ...... ......... ..... ..... ........................................ 1 Gene Therapy ............................................. . . . ..... ............. ..... ...................... 1 Ribozymes . .... . . ..... .... . ................ ........................................................ 2 Selfsplicing Intrans ........................................ ..... ...... .. .. .. .. ...... .. .. .. .. .. ...... 6 RNase P RNA ............. .......................................... ...... ...... ............. ... ..... 12 Self-cleaving Ribozymes .. .. .. .... ............. .. ..... .... .................. ..................... 15 Therapeutic Ribozymes ............................................... ..... ...... .. .. .. .. ........... 22 The Rod Photoreceptor and the Visual Cascade ........ ... .... . ..... ......... ......... 24 Retinitis Pigmentosa ......................................................................................... 30 Genetics of retinitis pigmentosa ....................................................................... 32 Animal Models of RP ...... ............................... ............................... ................ 35 Exogenous methods of gene and nucleic acid delivery to the cells ..... ....... 38 Viral mediated gene delivery .... ............................ ...... . . ..... . .............. 40 Adenoviral Vectors .................................................................................... 43 Retroviral vectors ................................................................... ....... .. .......... 45 Herpes Simplex Virus ................................................................................ 50 Adena associated virus ....................................................................... ...... 53 Comparison of viral vectors and delivery of genes to the retina .............. 58 Promoter control ............... .... .................. ... ..... ........................................ ....... 59 Experimental Aim ......................................................................... ............ ....... 60 2 MATERIALS AND METHODS ................ ................... ........ . . ..... . . .... .... 68 The cloning of ribozyme constructs ... .................................. . ........ .... ...... .... 68 In vitro transcription for the generation of ribozyme RNA .......... ............ ....... 72 Preparation of target oligonucleotide ....... ........... ... ............ ......... ... ... ....... 74 In Vitro cleavage reactions ........................... . .... .... .... ................... ......... 75 Initial characterization of ribozyme activity by an in vitro cleavage reaction. 76 Time course of the ribozyme cleavage reaction ..... ... ... .... ......... .... ..... ....... 76 Multiple turnover kinetic analysis .......... .................... ........ .............. . .... 77 IV
PAGE 5
V Dependence of ribozyme cleavage on magnesium concentration .................... 78 3 RES UL TS ........ ................... .... .......................... ............................. ............................ 87 Ribozymes Design ......... ............... .............. ...................................................... 87 The initial characterization of the catalytic activity of ribozymes .................. 94 Time course of the cleavage carried out under substrate excess ...................... 99 Kinetic analysis ... ........ ......................... ...... ..................................................... 115 The dependence of the ribozyme cleavage reaction on magnesium concentration. .. .. ..... ........ .. ................................... .. ...... .... ....... ................ 122 Cleavage of an oligonucleotide with the mutant mRNA sequence in the presence of increasing amounts of oligonucleotide with the wild type sequence ......... .. ........ .. ........ .. ...... ........................ .......... ....................... 130 Cleavage of retinal RNA by ribozyme ........................................................... 136 4 DISCUSSION .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 140 Autosomal Dominant genetic diseases are amenable to ribozyme gene therapy ........................................ ........................................................................... 140 Design of ribozymes for the treatment of ADRP ........................................... 142 Analysis of the catalytic efficiency of the ribozymes in vitro .................. ...... 144 Examination of the effectiveness of ribozyme therapy in vivo ....................... 154 REFERENCES ........... .......... ........... ........................ ........ ... ... ......... ......................... 160 BIOGRAPHICAL SKETCH .......................................................................................... 17 4
PAGE 6
LIST OF FIGURES Figure page 1. Trans-acting ribozymes can be designed to specifically cleave an mRNA substrate ......................................................................................................... 4 2. The secondary structure of self-splicing introns ................................................... 7 3. The design of a trans-splicing group I intron for gene therapy applications .......... 9 4. The splicing mechanism of the group II intron .............. ...... .... ..... ...................... 11 5. Cleavage of the tRNA 5' leader sequence by RNase P ....................................... 13 6. An RNA guide sequence allows the cleavage of a specific mRNA sequence by RNase P ....................................................................................................... 14 7. Self-cleaving ribozymes resolve concatemers formed by rolling-circular replication into individual genomic molecules .............................................. 17 8. The structure of the trans-cleaving HDV ribozyme ............................................ 18 9. The structure of the trans-cleaving hammerhead ribozyme .................... ............ 19 10. The structure of the trans-cleaving hairpin ribozyme ......................................... 21 11. Structure of the rod photoreceptor cell ............................................................... 26 12. Rod cells are intimately associated with RPE cells ............................................. 27 13. Structure of the rod opsin protein ......... .... ......... ..... ........... ........ ......... ..... ........... 28 14. The visual transduction cascade ......................................................................... 29 15. Retinitis Pigmentosa results in a gradual loss of peripheral vision ...................... 31 16. The G90D mutation disrupts a salt bridge required for the normal structure of the rod opsin protein .......................................................................................... 34 17. The nucleotide and amino acid sequence of the transgenes used to create the mouse models for this project.. ................... ................................... ......................... ... 37 vi
PAGE 7
VII 18. Delivery of genes to the nucleus of the cell by a novel liposome system ............. 39 19. Recombinant viruses can be used as vectors to deliver genes to the nucleus of cells .................................................................................................................. 41 20. Infectious cycle of adenovirus .............................................................................. .42 2 1 The adenoviral genome ....................................................................... .................. 44 22. The infectious cycle of Retrovirus ........................................................................ 46 23. The retroviral genome ... ............................. ..... . .... ...... . ..... ..................... ..... 47 24. The lentiviral genome ....................................................................................... .... 49 25. Herpes Simplex Virus .................. .......... ................ ............................................... 51 26. AA V infectious cycle ............................................................................................ 54 27. AAV genome and gene products .... .... .... ..... ....... . . ............ ....................... 55 28. The AAV gene therapy vector and packaging strategy ......................... ........... .... 57 29. Ribozymes able to cleave the mRNA of mutant rhodopsin genes may lead to a delay or prevention in the onset of ADRP ........... ..... .................................. .... 61 30 An RT-PCR method for quantitation of the cleavage of mRNA by a ribozyme .. 82 31. Recombinant AA V ribozyme construct. .................. ......... .................................. 84 32. The pTRUF vector and cloning strategy ......... ........ ...... .... .................................. 85 33. The sequence and secondary structure of the G90Dl hammerhead ribozyme ..... 90 3 4 The sequence and structure of the G90D2 Ribozyme ....... ..... ............................... 91 35. The sequence and structure of the G90D3 hammerhead ribozyme ................ ..... 93 36. The sequence and structure of the HHl hammerhead ribozyme .......... ..... ..... ..... 95 37. The sequence and structure of the HPl hairpin ribozyme ..... ......... ..... ................ 96 38. The sequence and structure of the HP2 hairpin ribozyme ................................ .... 97 39. The sequence and structure of the HP3 hairpin ribozyme .................................... 98
PAGE 8
viii 40 Cleavage of an RNA oligonucleotide by a ribozyme . ....... ....... ...... .... ...... . 100 41. Cleavage rate of the G90Dl hammerhead ribozyme .......................................... 102 42. The G90Dl hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence ........ .............. ....... ........ ...... 103 43. Non-specific degradation of the G90D target ............. ..... .......... ............... ....... 104 44 Cleavage rate of the G90D2 hammerhead ribozyme .......................................... 105 45. The G90D2 hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant and the wild-type sequence ................ .......... ....................... 106 46. Cleavage rate of the G90D3 hammerhead ribozyme ................................... ...... 108 47. The G90D3 hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence .... ........................................ 109 48. Non-specific degradation of the G90D target ..................................................... 110 49. Cleavage rate of the HHl hammerhead ribozyme .............. ............................... 111 50. The HHl hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence ......................................... ..... ...... 112 51. Cleavage of an RNA oligonucleotide by three different VPP hairpin ribozymes ................... ...... ............... . ......... . ............... ............. ....... . ............ ... 113 52 Cleavage rate of the HPl hairpin ribozyme ................................................ ....... 114 53. Cleavage rate of the HP2 hairpin ribozyme ........................... ........ ......... ........... 116 54. Cleavage rate of the HP3 hairpin ribozyme ........... .... ......................................... 117 55. Non-specific degradation of the VPP target. .......... .... ........... ............................. 118 56. Non-specific degradation of the VPP target.. ...................................................... 119 57. G90Dl hammerhead ribozyme multiple turnover cleavage reaction ......... ........ 121 58. G90D3 hammerhead ribozyme multiple turnover cleavage reaction .... .............. 123 59. HP2 hairpin ribozyme multiple turnover cleavage reaction ......... ...................... 124
PAGE 9
ix 60. HP3 hairpin ribozyme multiple turnover cleavage reaction ........ .... .......... ........ .. 125 61. The magnesium dependence of the G90Dl ribozyme .......................... .... .......... 127 62. The magne s ium dependence of the G90D3 ribozyme ........................................ 128 63. The magnesium dependence of the HP2 hairpin ribozyme ........ .... ........ ............ 129 64. The magne s ium dependence of the HP3 hairpin ribozyme ................................. 131 65. The G90D 1 cleavage reaction i s not effected by the presence of an RNA oligonucleotide with the wild-type sequence ................................................ 132 66. The cleavage reaction catalyzed by the G90D3 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence .. ........ ..... 133 67. The cleavage reaction catalyzed by the HP2 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence ........... 134 68. The cleavage reaction catalyzed by the HP3 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence ............... 135 69. Cleavage of the G90D mRNA by the G90Dl hammerhead ribozyme ... ............ 137 70. Cleavage of the VPP transgene mRNA by the HP3 hairpin ribozyme .......... ..... 139 71. A comparison of the cleavage reaction of the different G90D ribozymes ........ .. 147 72. A comparison of the cleavage by VPP ribozymes .............................................. 149
PAGE 10
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 AN ANALYSIS OF RIBOZYME-MEDIATED CLEAVAGE OF THE RNA ASSOCIATED WITH RETINAL DISEASE By Patrick Whalen May 2001 Chairperson: Alfred Lewin Major Department: Molecular Genetics and Microbiology The discovery of RNA sequences called ribozymes that could act in trans led to the idea that they could be used as a gene therapy. Ribozymes have been examined primarily as a treatment for cancer and AIDS. Autosomal dominant diseases may be particularly amenable to treatment by ribozymes as the expression of a single mutant gene results in disease Inhibition of the expression of this mutant gene might prevent disease. Genetic forms of blindness are easier to treat than other types of genetic diseases since the eye is easily accessible and is an immune privileged site. This facilitates the delivery and expression of genes in the eye. The goal of this project was to design ribozymes for the treatment of two different retinal diseases. Both of these retinal diseases are caused by mutations in the rod opsin gene. The first retinal disease is caused by a glycine to aspartic acid conversion in amino acid 90, the G90D mutation. The G90D mutation results in night blindness. This mutation disrupts a salt bridge in the rhodopsin X
PAGE 11
XI molecule that results in a rhodopsin protein that is locked into an active confirmation A proline to histidine conversion at amino acid 23, the P23H mutation, results in autosomal dominant retinitis pigmentosa. This mutation is reported to prevent the incorporation of the rod opsin into the proper cellular compartment of rod cells at the same rate at which it is lost. The success of this project depended on the completion of three specific aims The first aim was to design ribozymes able to degrade specifically RNA sequences corresponding to the mutant rhodopsin sequences responsible for disease. The second aim was to characterize these ribozymes in vitro. The purpose of the second aim was to identify those ribozymes that had the greatest chance for success as a gene therapy The third aim of this project was to create an rAA V vector capable of expressing the selected ribozymes in the eye. The mRNA of two different rhodopsin mutations were targeted for degradation by ribozymes. Three different hammerhead ribozymes were designed against the first rhodopsin mutation, the G90D mutation. Only two of these ribozymes were found to cleave specifically the mutant rhodopsin RNA. In vitro analysis determined that both these ribozymes had catalytic coefficients sufficient for in vivo activity. A hammerhead ribozyme and three different hairpin ribozymes were designed to cleave a second rhodopsin mutation the VPP mutation All of these ribozymes were able to cleave correctly the VPP RNA. Only two modified hairpin ribozymes cleaved their target RNA efficiently however. The two G90D ribozymes and the two VPP hairpin ribozymes were cloned into rAA V vectors. These ribozymes are being examined for their therapeutic effectiveness in animal models.
PAGE 12
CHAPTER 1 BACKGROUND AND SIGNIFICANCE Gene Therapy The advent of molecular biology allowed a wealth of information to be gathered on the biology of the cell leading to a greater understanding of the mechanisms responsible for many diseases. The idea that diseases could be treated at a genetic level developed as an increasing number of gene products and their cellular interactions were identified. In 1971, Rogers and colleagues attempted an entirely new approach to the treatment of disease by injecting two young girls with wild-type Shope Papilloma Virus. These girls suffered from a disease caused by elevated serum levels of the amino acid arginine. It was thought that the viral genome contained an arginase gene that could reduce the girls' serum arginine to normal levels and permanently correct the disease. Unfortunately, the Shope Papilloma Virus did not encode an arginase gene and was incapable of infecting human cells. Consequently, the experiment failed and little useful information was obtained at the time(]). The idea that foreign DNA could be used as a drug, however, led to the concept of gene therapy Initially, gene therapy was conceived as a way to compensate for a genetic defect by expressing a missing protein in the cells of patients. In 1990, the first approved clinical trial for this type of gene therapy was carried out as a treatment for Severe Combined Immunodeficiency Disease (SCID). This disease is caused by an adenosine deaminase (ADA) deficiency that leads to an accumulation of dATP and deoxyadenosine in the cells of affected individuals. The accumulation of these
PAGE 13
2 products is toxic to antibody-producing B cells, and to T cells that are responsible for the regulation of the immune system (2) In this first clinical trial, mononuclear cells were removed from a four-year-old patient and grown under conditions that stimulated their differentiation into T cells. The ADA gene was delivered to the exogenous T cells using a retroviral vector. When the patient was infused with the transformed T cells, there was an immediate improvement in the immune function and clinical condition of the patient (3) Although the improvement in immune function was temporary, it demonstrated the possibilities of gene therapy. In 1999, there were over 300 gene therapy clinical trials underway worldwide (4). Gene therapy is increasingly being examined as a treatment for cancer, AIDS, cardiovascular disease, arthritis, and a number of infectious and monogenic diseases. Advances have expanded the potential of gene therapy, so that it is no longer solely defined as the long-term restoration of missing or defective genes. Currently most gene therapy protocols focus on the killing of disease related cells through the transfer of toxic genes or genes that activate the immune system (5) Other approaches seek to inhibit the expression of genes responsible for pathogenesis (6). Ribozymes The idea that oligonucleotides could be used for gene therapy arose from the discovery of naturally occurring RNA molecules that were able to regulate gene expression (7;8). Antisense RNA regulation was first discovered in the bacterial ColEl plasmid. This plasmid expresses an RNA molecule that is complimentary to the RNA primer required for DNA replication (9;10) Hybridization of the antisense RNA to the RNA primer controls the copy number of the plasmid. A number of naturally occurring cases of antisense RNA control have now been documented (11-13). These discoveries
PAGE 14
3 led to the idea that oligonucleotides could be designed to down-regulate the expression of genes responsible for pathogenesis At about the same time, the study of RNA processing led to the discovery of naturally occurring RNA sequences that possessed catalytic properties (14) These RNA molecules were called ribozymes, an abbreviation for ribonucleotide enzyme, stressing the ability of these RNA molecules to act as endonucleases Ribozymes form base pair specific complexes with their RNA substrates that are able to catalyze the hydrolysis or phosphoryl exchange at a phosphodiester bond. Naturally occurring ribozymes are involved in RNA processing (15-18).The discovery that ribozymes could be designed to act in trans led to the development of a second generation of antisense technology (Figure I). A trans-acting ribozyme's ability to inactivate or modify multiple mRNA molecules gives it an advantage over an antisense oligonucleotide that acts stociometrically (19). Furthermore, ribozymes do not rely on the host cellular machinery to degrade their substrate like antisense oligonucleotides. Thus, trans-acting ribozymes can be expected to inactivate their target RNA more efficiently than antisense oligonucleotides In fact, in a direct comparison of antisense oligonucleotides and ribozymes, ribozymes were found to be 2to 10-fold more efficient (20) RNA and protein enzymes have several similarities. First, the activity of ribozymes and enzymes both depend on complex secondary and tertiary structures (21-23). Second, ribozymes and proteins use some of the same mechanisms of acid base catalysis to carry out reactions (24). Third, ribozymes and enzymes both stabilize the transition-state between substrate and product formation (25). Finally, catalysis in each case is driven by
PAGE 15
4 Hammerhead Rlbozyme / c~~vagEt / __ ,/ -Figure 1. Trans-acting ribozymes can be designed to cleave an mRNA substrate.
PAGE 16
5 the intrinsic binding energy resulting from interactions between the enzyme and its substrate at sites distant from the catalytic core(25). Ribozymes, though, offer some advantages over proteins as therapeutic agents. First, they recognize their RNA substrate through base pairing. As a result specificity is easily manipulated and very stringent. Second, ribozymes are less likely to evoke an immune response than foreign or altered proteins (26). In contrast, a number of animal studies have shown that genetically transduced cells are eliminated in vivo due to the expression of foreign or transgenic proteins (27;28). Third, RNA is a natural component of cells and can have a relatively short half-life. As a result, unintended toxic effects due to the expression of ribozymes have not been observed (29-35). However, the phosphorothioate bonds used to increase the stability of chemically synthesized ribozymes and antisense oligonucleotides have been shown to be toxic in mammalian cells (36;37). Finally, ribozymes are small making them inexpensive and easy to produce. Their size also facilitates the insertion of genes encoding the ribozymes into viral vectors used for their delivery to cells in vivo. Naturally occurring RNA molecules possess a number of features not found in naturally occurring DNA molecules. These properties favored the evolution of RNA rather than DNA as catalysts. The most important difference between the two polymers is the 2' hydroxyl group associated with the sugar moiety of RNA. This 2'hydroxyl group can participate directly in a chemical reaction, and is able to enhance the reactivity of the adjacent 3'-hydroxyl group. In addition, the 2'hydroxyl group of RNA is a hydrogen bond donor and acceptor that gives RNA molecules more versatility than DNA in the formation of tertiary structure (38). Another important property of RNA is that it is
PAGE 17
6 usually folded from a single strand, whereas DNA is found predominantly as a base paired duplex of complementary strands. RNA molecules are able to form extensive secondary structure due to pairing of imperfect complementary sequences in the RNA strand. Single stranded regions punctuate this secondary structure leading to a variety of RNA conformations Finally, RNA has uridine rather than thymidine as a nucleotide base. Uridine allows the formation of unique secondary structures such as the uridine tum seen in transfer RNA and hammerhead ribozymes (39;40) This structure results in an abrupt tum in the direction of the polynucleotide backbone that allows the formation of complex tertiary interactions. This uridine tum is stabilized by hydrogen bonding and Van Der Waals interactions between uridine and surrounding nucleotides (38;41). Ribozymes can be classified into different groups based on their catalytic activities. The three main groups of ribozymes are the self-splicing introns, RNase P, and the small self-cleaving ribozymes. Currently each type of ribozyme is being examined for its possible use as a genetic therapy. Self-splicing Introns The first type of ribozyme to be discovered was a self-splicing intron (Figure 2). Two different classes of self-splicing introns have been identified based on their conserved secondary structure and splicing mechanisms; the group I and group II introns (42;43) The group I intron is found in a wide number of species including eubacteria, bacteriophages, fungal mitochondria, plant chloroplasts, as well as the rRNA of lower eukaryotes (44;45). Splicing proceeds via two consecutive transphosphoesterfication reactions (Figure 3). The first step in this reaction is a nucleophilic attack by an extramolecular guanosine at the 5' splice site. The guanosine becomes covalently linked
PAGE 18
7 A. B DOMA .. t Tetrahymena tflermophi/a LSU rRNA GenBenlc# ..LJ7 235 EUCi1,Yll. Protoctisf.i. Ci/iophora (/Cl) J.m09, 7994 a I' Figure 2 The secondary structure of self-splicing introns A) The group I intron ( 46) B) The group II intron ( 4 7).
PAGE 19
8 by a 3'-5' phosphodiester bond to the first nucleotide of the intron. The 5' exon is left with a 3' hydroxyl group that then carries out a nucleophilic attack on the 3' splice site. Following this action the intron is released and the exons are ligated (l 5;46). Although the self-splicing reaction is normally intramolecular, the group I intron can be modified to allow the reaction to occur in trans (14; 15). Currently, several groups are examining the ability of trans-splicing ribozymes to repair the mRNA of genes containing a pathogenic mutation (Figure 3). A trans-splicing ribozyme has several advantages over other types of ribozymes for the treatment of diseases like cystic fibrosis. Different mutations in the Cystic Fibrosis Transmembrane Conductance Regulator gene in different patients have been shown to cause the same disease (48). In theory, a single trans-splicing ribozyme could be used to repair the many different pathogenic mRNAs found in different patients. Furthermore, the minimum sequence requirement for the trans-splicing reaction is a uridine in the substrate. Thus, the chance of finding a target site in any one mRNA is very high. Finally, the complex secondary and tertiary structure of the group I intron may increase the stability of these ribozymes to nucleases There are only a few reports that document the ability of a group I intron to repair mRNA. Sullenger and Cech showed that a trans-splicing ribozyme could be used to create translatable ~-galactosidase mRNA in vivo (49). This ribozyme was found to be active in E. coli and mammalian cells (50;51). Another ribozyme was recently designed to repair a ~-globin mRNA carrying the mutation responsible for sickle cell anemia (52). Similarly a group I intron was used to repair the mRNA of the gene responsible for myotonic dystrophy (53). In each case, the ribozyme was reported to correctly repair a
PAGE 20
A B 9 ---Gc:====::::::i guanosine cofactor binds, attacks phosphate at 5' JunctJon ---G'----~ 5' junction cleavage, ligation of G to lntron 5'-end '------::::nm~~4-~ structural rearrangement puts 3'-lntron end (G) in G-{)inding site, 3'-exon forms P1 o 3~hydroxyl of 5'-exon attacks 3' junction phosphate, lntron removed Uulant lranlcript Rlbozyme wtth wfld-type 3' exon cc:..:====::::i,~~~S:S:\~'-~"i.'S\:S:'-S:\S~s~s.SJS1 Correc:100 transcrilJf + Ribozyme Mutant 3' e,ccn Figure 3. The splicing mechanism of the group I intron and its use as a gene therapy applications ( 49).
PAGE 21
IO mutant mRNA message However, the reaction was shown to lack specificity in studies designed to look for unintended splice products. The trans-splicing reaction was also shown to be relatively inefficient (50). The infidelity of the trans splicing reaction was due in part to the fact that the ribozymes only had a six-nucleotide recognition sequence. A specific six-nucleotide sequence would be expected to be found every 4096 nucleotides Furthermore the ribozymes used in the previous studies were modified group I intrans that lacked some of the natural RNA structures shown to increase the fidelity of the splicing reaction (54). The loss of fidelity could lead to cellular toxicity due to the elimination of mRNAs required for viability. Two groups are examining modifications of the group I intron intended to increase the specificity and catalytic efficiency of trans-splicing reaction (55;56). The second self-splicing intron is the group II intron found primarily in the organelles of fungi and plants (47) The group II intron follows the same splicing pathway described for the processing of nuclear pre-mRNA (Figure 4). As with the group I intron, the splicing occurs via two consecutive transphosphoesterfication reactions. The main difference between the splicing mechanism of the two different intrans is the nature of the hydroxyl group that initiates the initial transphosphoesterfication reaction. With the group I intron, the reaction is initiated by the 3 '-hydroxyl group of an exogenous guanosine The splicing reaction of the group II intron is initiated by the 2' -hydroxyl group of an internal adenosine. This 2' -hydroxyl group attacks and cleaves the 5' splice site by forming a 2' -5' phosphodiester bond with the first nucleotide of the intron. This results in a lariat intron structure also found in the processing of pre-mRNA (57) Next,
PAGE 22
11 -OH-+ 2'-hydroxyl group of an internal adenosine attacks the phosphate at the 5' splice junction. The 3'-hydroxyl of the first exon attacks the splice site of the second exon. The first and second exon are ligated and the intron lariat is removed. Figure 4. The splicing mechanism of the group II intron (57).
PAGE 23
12 the 3 '-hydroxyl group of the upstream exon attacks the 3' -splice site resulting in the removal of the intron lariat and the ligation of the two exons ( 47). The gene therapy potential of the group II intron has not been examined. In fact, there is only a single report describing the use of a trans-splicing group II intron. In this report, a modified group II intron was used as a tool for creating chimeric proteins (58). This ribozyme was able to correctly create a recombinant mRNA. However, the efficiency and specificity of the trans-splicing reaction were not examined in detail. RNasePRNA The second group of naturally occurring ribozymes is RNase P RNA. RNase Pis an endoribonuclease that removes the 5' leader sequence from precursor tRNA (Figure 5). RNase P has an essential RNA and protein subunit. The RNA is the catalytic component of the complex as demonstrated by its ability to cleave precursor tRNA in vitro in the absence of the protein subunit. The protein component increases the turnover rate of the reaction is required for cleavage of precursor tRNA in vivo. In some cases, the protein subunit can also influence the affinity of an RNase P complex for its substrate (59;60). One of the more interesting aspects of RNase Pis its ability to recognize and cleave 60 different precursor tRNA substrates even though there is no sequence homology around the cleavage site (61). The RNase P complex recognizes the structure of tRNA. Only a minimal tRNA structure is required for the creation of RNase P cleavage site. The Altman laboratory took advantage of this discovery by designing an external guide sequence (EGS) that hybridized to a specific mRNA sequence of~ galactosidase. After hybridizing to the target mRNA, the EGS forms a structure similar to
PAGE 24
13 3' ~.: .c .c ; Figure 5. Cleavage of the tRNA 5' leader sequence by RNase P (59).
PAGE 25
14 5 Substrate mRNA "---NNNNNNNNN -f N Cleavage site N N N N N l) 3' A C C Guide Sequence A N N N N N N cG U NCUUCC G GAAGG C GC U U A AA N N N N AA Cc '-s Figure 6. An RNA guide sequence allows the cleavage of a specific rnRNA sequence by RNase P.
PAGE 26
15 the acceptor stem of precursor tRNA (Figure 6). This guide sequence facilitated cleavage of p-galactosidase mRNA molecule by RNase P with kinetics similar to those of a tRNA substrate (62;63). The cleavage reaction was tested in animal cells by creating a plasmid encoding a guide sequence against chloramphenicol acetyltransferase (CAT) mRNA The polill promoter was used to drive expression of the EGS, and a polill termination signal was placed downstream. When this plasmid was co-transfected with a plasmid containing the CAT gene in human lung cells, expression of CAT was decreased by 60% in the first 36 hours following transfection (62). In another study, the therapeutic potential of RNase P was examined by creating C 127 mouse cells that expressed guide sequences targeting two different influenza genes. When these cells were subsequently infected with influenza, assays indicated that virus production was inhibited by up to 80% in some C127 clones (64) RNase P has several advantages over other types of ribozymes as a therapeutic agent. First, it is found naturally and gene therapy strategies employing RNase P exploit the abundance of RNase Pin the human cells. Second, RNase Pis the only ribozyme that evolved to perform multiple turnover cleavage reactions in trans. Third, RNase P can catalyze the cleavage of RNA in a wide variety of cells (65) Finally, because only the secondary RNA structure created by the EGS is needed to create the cleavage site, a wide variety of RN As are expected to be recognized by the RNase P complex (62). However, since RNase Pis only found in the nucleus of the cell the effectiveness of this ribozyme may be decreased when targeting mRNAs that localize in the cytoplasm. Self-cleaving Ribozyrnes The final group of natural ribozymes is the small nucleolytic RN As associated
PAGE 27
16 with viruses and satellite RNA. These ribozymes can catalyze an RNA cleavage reaction in the absence of protein. Magnesium facilitates the reaction which results in products with a 2'-3' cyclic phosphate and a 5' hydroxyl terminus (66). These ribozymes evolved to play an essential role in the life cycle of the RNA molecules in which they are found (Figure 7). During the replication of these RNA molecules, circular forms of the RN As act as templates for the production of long transcripts that contain multiple copies of the RNA genome. Self-cleaving ribozymes resolve the concatameric transcript to create unit length RNA molecules (67). Although there are several different types of small ribozymes, the most extensively examined are the hepatitis deltal virus (HOV) ribozyme, and the hammerhead and hairpin ribozymes derived from the tobacco ring spot virus satellite RNA. HDV is a short single-stranded RNA virus found in some patients infected with human hepatitis B. The circular RNA genome of HDV encodes a ribozyme in both orientations. Like the other self-cleaving ribozymes, HDV replicates through a rolling circular mechanism. The ribozyme resolves genomic multimers into monomers in a cis cleavage reaction (68). The self-cleaving domain was identified and modified to cleave RNA substrate in trans with multiple turnover (Figure 8). The substrate is recognized via an eight-nucleotide sequence. Cleavage occurs at a GGX sequence where Xis C, U, or A. The efficiency of the trans-cleavage reaction is much lower than the cis-cleavage reaction (69). The hammerhead and hairpin ribozymes, though, have received the most attention for potential clinical applications. A comparison of the self-cleaving RNA sequence of a number of different plant viroid infectious RNA molecules led to the discovery of the
PAGE 28
17 Figure 7. Self-cleaving ribozymes resolve concatemers formed by rolling-circular replication into individual genomic molecules (67)
PAGE 29
18 cu G U 3 RCl GC UA C-G II C C-G G C I A-U G C C-GI A I C G 111 GC A U A G-C G G C C U C G C G C G G U C uC~ Cleavage ~C G sate U !G U G --C-G s U A AU rv G -C c-a G-C A U C A U A UAA t spite J4'ftlon Figure 8. The structure of the trans-cleaving HOV ribozyme (69). .'
PAGE 30
19 Helix 3 Helix 1 NNNNUXNNNN I I I I I I I I I NNNNA NNNN A CuG A A G AGN N-N N-N Helix2 N-N N-N A G AA Target Catalytic Core Figure 9. The structure of the trans-cleaving hammerhead ribozyme.
PAGE 31
20 catalytic domain of the hammerhead ribozyme (Figure 9). A relatively small ribozyme of thirty-five nucleotides was discovered that was able to cleave an RNA target in trans. The hammerhead ribozyme binds its substrate to form a structure that consists of three stem loops and a catalytic core with a conserved nine-nucleotide sequence. A mutation in any of these conserved nucleotides prevents RNA cleavage (17) The catalytic core of the hammerhead ribozyme has two functions. First, it destabilizes the substrate strand andtwists it into a cleavable conformation. Second, it binds the metal cofactor needed for catalysis (70). The hammerhead ribozyme recognizes the substrate sequence on either side of an NUX cleavage site by means of two flanking arms that hybridize to form helices ill and I. Cleavage occurs 3' of the NUX cleavage site, where N = any nucleotide, and X = any nucleotide except G. Cleavage sites do not all show the same efficiency. In general, GUC is the most efficient cleavage site, followed by CUC, UUC, and AUC. The rest of the cleavage sites are cleaved at least ten-fold more slowly than the GUC site (71). The hairpin ribozyme is derived from the cis-cleaving sequence in the negative strand of the Tobacco Ringspot Virus satellite RNA. A 50-nucleotide RNA sequence is responsible for the generation of the 359-nucleotide satellite RNA following replication. The hairpin ribozyme binds its substrate to form a structure with four helices and two loops (Figure 10). The arms of the hairpin ribozyme hybridize to the substrate molecule to form a six base pair helix 1 and a four base pair helix 2. The sequence of these arms can be manipulated to allow binding to a variety of RNA substrates. Loop A contains the BNGUC target sequence required for cleavage where B = G, C, or U and N is any nucleotide (72) There are no conserved nucleotides in any of the helixes, although a Gin
PAGE 32
21 Cleavage sit\ Helix 1 Helix 2 5' NN N BNGUCN N N N NN Helix 3 1111 A 111111 N N N NN NN N N N N N N-N AG N-N N-N N-N N-N AC N U N U A A A N A u B G A N-N A C H I 4 N-N e IX N-N U G u Target 5' Figure 10. The structure of the trans-cleaving hairpin ribozyrne.
PAGE 33
22 the substrate portion of helix 2 adjacent to loop A leads to an increased rate of cleavage (73). There are a number of conserved nucleotides found in loop A and loop B, which fold on top of one another to form the catalytic core (74). Loop B of the hairpin ribozyme shows a similarity to several other RNA molecules including loop E of eukaryotic 5S rRNA, and the a-saracin loop of 26s rRNA (75). The structure of the hairpin ribozyme loop B may represent an RNA structural element utilized by a number of functionally important RNA molecules. The hammerhead and hairpin ribozymes are being examined as gene therapies for a number of different diseases. There are a number of reasons for this. As mentioned they are small and can be easily cloned and packaged into many of the existing viral vectors for delivery to target cells. One advantage of the hammerhead ribozyme is that it can recognize a greater number of cleavage sites than HDV or hairpin ribozymes. An advantage of the hairpin ribozyme is its complex secondary structure that may increase stability to nucleases. In addition, it is able to carry out cleavage reactions in the presence of magnesium concentrations that are closer to physiological levels than other small ribozymes (76). Finally the hairpin ribozyme has the highest catalytic rate of any naturally occurring ribozyme (77). Therapeutic Ribozymes There are a number of reports of the ability of hammerhead and hairpin ribozymes to control the expression of specific genes in cell culture. For example, a hammerhead ribozyme designed to cleave the mRNA encoding the c-Ha-Ras mutation inhibited formation of the foci of transformed cells by 50%(78) Hammerhead and hairpin ribozymes were used to reduce the steady state levels of ~-amyloid peptide precursor
PAGE 34
23 mRNA in cos7 cells by 67%-80% (79). In addition, hammerhead ribozymes targeting the mRNA of a BCR/ABL fusion protein responsible for Chronic Myelogenous Leukemia, reportedly eliminated expression of this protein in K562 cells (80). A number of groups have also designed ribozymes as potential HIV therapies. The most successful of these was a hairpin ribozyme targeting the 5' transcribed leader sequence of the virus (81) Proviral DNA synthesis was reduced by 50% to 100% in human T cells stabily transfected with the ribozyme. Viral RNA and protein production was reduced by 95%. In another study, expression of this hairpin ribozyme in stem cells taken from the fetal cord showed a 90% reduction in viral RNA following infection (82). Finally, a hammerhead ribozyme expressed in Huh7 cells reduced hepatitis B viral production by 83% (83). The effectiveness of hammerhead and hairpin ribozymes has also been demonstrated in animal models. In one of these experiments, transgenic mice carrying the bovine lactalbumin gene were crossed with mice that expressed a ribozyme designed to cleave the mRNA of this gene. The resulting offspring showed a 50% to 78% reduction in bovine lactalbumin mRNA levels (84). In a second study, the adenoviral-mediated transfer of a ribozyme to a transgenic mouse expressing human growth hormone resulted in a transient 96% reduction in hepatic growth hormone mRNA (85). Finally, a recombinant AAV vector expressing a ribozyme directed against the mRNA of a mutant form of rhodopsin was constructed. This construct was delivered to the retina of a rat model of retinitis pigmentosa by a subretinal injection. Expression of this ribozyme in the animal model led to significant degree of rescue from retinal degeneration (86).
PAGE 35
24 Currently, most of the therapeutic applications of ribozymes are for the control of cancer and AIDS, but these diseases may be the least amenable to treatment by ribozymes. For example, one of the hallmark s of cancer is genomic instability. This can lead to an increase in the copy number of certain genes as well as the formation of sequence variants in the target substrate called escape mutants, which can not be cleaved by a single ribozyme The formation of escape mutants poses a similar problem for the treatment of viral infections. In addition, metastatic cancer cells and cells infected with HIV can be widely distributed throughout the body. This makes it difficult to ensure that ribozymes are delivered to all of the affected cells. Because of these problems associated with the treatment of cancer and HN, it will be difficult to determine the potential of ribozyme therapy in the treatment of these diseases. Genetic disorders inherited in an autosomal dominant fashion may be particularly susceptible to ribozyme therapy These diseases are often caused by the product of a single mutant allele that interferes with the function of the protein encoded by the wild type allele. A reduction in the expression of the mutant allele may ameliorate the disease. There are a number of features that make the autosomal dominant form of retinitis pigmentosa, ADRP, a prime candidate for ribozyme therapy An understanding of the visual cascade and the cell biology of the photoreceptors involved in the visual process is necessary to understand why ADRP is such a good candidate for treatment by ribozymes The Rod Photoreceptor and the Visual Cascade Photoreceptor cells found in the outer retina mediate vision. In humans, rod cells represent 97% of all photoreceptor cells and are responsible for vision in dim light. Rod cells are concentrated in the peripheral region of the retina. Cone cells are responsible for
PAGE 36
25 detailed and color vision. Their density increases toward the center of the retina in an area known as the macula. A rod cell is divided into an outer segment (ROS), an inner segment (RIS), and a synaptic terminus (Figure 11). The apical portion of the ROS is embedded in the retinal pigment epithelium (RPE) The synaptic terminus links the photoreceptor cells to the neural cells which carry the visual signal to the brain for processing The ROS is composed of a series of disks that are stacked one on top of another. Approximately 10% of the disks are shed from the top of the ROS on a daily basis. The shed disks are phagocytosed by the RPE and degraded (Figure 12). Corresponding numbers of disks are replaced at the base of the ROS during the course of the day (87) Rhodopsin is the major protein component of the ROS and constitutes up to 85% of the total protein Rhodopsin is composed of rod-specific opsin, a seven transmembrane domain spanning protein, which binds the light-sensitive chromophore 11-cis-retinal (Figure 13) The visual cascade begins with the absorption of a single photon of light by 11-cis-retinal leading to a conformational change in the chromophore to all-trans retinal (Figure 14). This conformational change allows rhodopsin to activate the G protein transducin A single rhodopsin molecule is able to activate hundreds of transducin molecules, resulting in a significant amplification of the initial signal. Transducin then stimulates cyclic guanosine monophosphate phosphodiesterase to hydrolyze cyclic guanosine monophosphate, ( cGMP). The resulting decrease in cGMP concentration closes the cGMP gated cation channels in the photoreceptor plasma membrane. Closing these channels in response to light prevents the influx of cations that normally occurs in the dark, but the efflux of calcium continues. This closure leads to hyperpolarization of
PAGE 37
Outer Segment Inner Segmet Nucleus Fiber Synaptic Ending I L = c:::=:J C :I C -. C -1..;;;; ,= c:: c:::: c::::=. "=-c::= a= c=-c::=. -I I \ ) ) (';. I I N 26 --. .. Rhodopsin Molect,dlit .. Rhodopsin Disk lnterdiscal Surface Figure 11. Structure of the rod photoreceptor cell (88). C
PAGE 38
27 Figure 12. Rod cells are intimately associated with RPE cells (87).
PAGE 39
28 ,-COOH l NH2 Figure 13. Structure of the rod opsin protein (88).
PAGE 40
29 ( meta II R di k osm Figure 14. The visual transduction cascade (89).
PAGE 41
30 the photoreceptor cell membrane. Deactivation of the phototransduction pathway begins with the phosphorylation of rhodopsin by rhodopsin kinase. Next, rhodopsin is bound by the protein arrestin which prevents further activation of transducin and causes the release of all-trans-retinal. Arrestin binding results in an increase in the intracellular levels of cGMP and then of calcium. Finally, transducin is deactivated and rhodopsin is regenerated with the binding of a new molecule of 11-cis-retinal (90). While the major role of rhodopsin is to initiate the visual phototransduction cascade, recent findings indicate that rhodopsin also plays an important role in the structure of the ROS. This was demonstrated by the generation of homozygous and heterozygous rhodopsin knockout mice (91). In the homozygous knockout mouse, the rod outer segment was not detected In addition, rod photoreceptor cells had all completely degenerated by the third month following birth. In the heterozygous animals, while there was no degeneration of the rhodopsin cells, there was some disorganization in the ROS. As the mice aged the disorganization increased and the ROS became increasingly shorter. Because of the important role of rhodopsin in the visual cascade and in maintaining the structure of the ROS, it is not surprising that rhodopsin would also play a role in retinal disease. In fact, studies in the last decade have identified a number of rhodopsin mutations that can lead to retinal disease (92;93) Retinitis Pigmentosa Retinitis Pigmentosa (RP) is a general term for a heterogeneous group of retinal dystrophies. This term is only used to refer to those dystrophies that begin in the peripheral retina with the loss of rod photoreceptor cells. One of the first symptoms of the disease is a loss of the mid-peripheral visual field and impaired night vision. This
PAGE 42
31 Figure 15. Retinitis Pigmentosa results in a gradual loss of peripheral vision.
PAGE 43
32 symptom is followed by night blindness As the disease progresses, there is a: loss of the far peripheral visual field (Figure 15). Eventually, the cone cells also begin to degenerate, leading to a loss of the central visual field and blindness. Retinitis pigmentosa can be detected early in life by measuring the light evoked electrical response from the retina with an electroretinogram (ERG). Abnormal ERGs have been detected in asymptotic children in some cases a decade before any diagnostic change can be detected by an ocular examination. In fact, children six and older with a family history of RP and normal ERGs have not been observed to develop RP themselves (93;94). Genetics of Retinitis Pigmentosa RP is the most common form of inherited blindness, affecting one in every 3000 to 4000 individuals (95). Retinitis Pigmentosa can be inherited by an autosomal dominant, autosomal recessive, X-linked, or digenic mode (96). While the symptoms and progression of the disease are characteristic of RP, the age of onset of the symptoms of RP and blindness are variable and are partially dependent on the mode of inheritance. All forms of RP are similar, though, in that there is currently no treatment or cure. Pedigree analysis indicates that the autosomal dominant form of RP is less prevalent than the autosomal recessive form of the disease. According to one study ADRP accounts for approximately 30% of all cases of RP (97) A number of different genes have been identified as being responsible for ADRP. The first mutation found to be responsible for ADRP was in the rhodopsin gene (93). Various mutations in rhodopsin were initially found to account for 25% of all cases of ADRP (92). A more recent and comprehensive study has implicated rhodopsin mutation as the cause of 50% of all cases of ADRP (98). Both studies agree that mutations in rhodopsin represent the most
PAGE 44
33 common cause of retinitis pigmentosa, representing 10% of all cases of RP in the United States. Some of these mutations have been found to affect the phototransduction cascade while others interfere with rhodopsin's role in maintaining the structure of the ROS. For example, the replacement of glycine at the amino acid position 90 of rhodopsin with aspartic acid leads to a very mild form of retinitis pigmentosa in humans. The most serious symptom associated with this mutation is night blindness. This night blindness is reported to be caused by the disruption of a salt bridge formed by glutamine 113 and lysine 296 by aspartic acid 90 (Figure 16). This disruption leads to a rhodopsin molecule that has a conformation that is active in the absence of light. As mentioned previously, in order to depolarize a typical photoreceptor rod cell only 200 of the 108 rhodopsin molecules need to be in the active conformation. The consequence of the G90D mutation is a desensitization of the rod cells leading to a higher background signal and stationary night blindness. Stationary night blindness is characterized by an inability to see in dim light, but there is no or little retinal degeneration (99). The replacement of proline with histidine at amino acid 23 of rod opsin accounts for 15% of all cases of ADRP in the United States (93;94;100) The first symptom of ADRP associated with the P23H mutation, impaired night vision, appears between the ages of 10 and 30. The majority of these individuals are legally blind by age 60. Recent findings suggest that the P23H mutation impedes incorporation of rhodopsin into the rod outer segment at the rate required to replace that which is lost as part of the natural shedding process (101;102). This reduction results in a ROS that becomes progressively shorter over time until, ultimately, the rod cell undergoes apoptosis (103).
PAGE 45
34 Gly{Asp)90 Glu113 IV Figure 16. The G90D mutation disrupts a salt bridge required for the normal structure of the rod opsin protein.
PAGE 46
35 Many different factors make diseases caused by rhodopsin mutations good candidates for gene therapy. First, ERG can detect retinal disease prior to the clinical presentation and can be used to follow the course of the disease. Second, the retinal diseases caused by rhodopsin mutations progress very slowly Together with early detection, the gradual nature of these diseases allows ample time for treatment. Third, ocular disorders are particularly susceptible to gene therapy Since the eye is easily accessible, delivery of genes to the eye is less difficult than in other tissues. The eye is also an immune privileged site, which means that neither B cells nor T cells have access to the eye (104). Thus, the viral vector and infected cells are less likely to be eliminated as the result of an immune response Finally, animal models are available for a number of rhodopsin point mutations, including the G90D and P23H mutation Animal Models of RP The study of animal models of human retinal dystrophies offers many obvious benefits. In fact, the use of animal models has already led to a greater understanding of these diseases in humans. For example, the identification of a number of genes responsible for human retinal dystrophy were first identified in mouse models (105-107). In addition, detailed histopathological, molecular, and electrophysical studies of animals which are impossible in human patients have increased our understanding of the pathology associated with various mutations The use of mouse models is particularly attractive due to their cost effectiveness, short generation interval, and well-characterized genetics. Furthermore, mice bearing a disruption in one rhodopsin gene are currently available. Over-expression of rhodopsin has been found to be toxic to cells and has been shown to be responsible for the retinal degeneration seen in some of the animal models
PAGE 47
36 for this disease (108). This toxicity obscures the role the mutant protein plays in the disease process and the actual effect gene therapy directed against these mutations will have. Transgenic mouse models can be crossed with the rhodopsin knockout mice to produce an animal with the normal copy number of the mutant and wild-type rhodopsin. This eliminate s the experimental ambiguity due to the toxicity of rhodopsin. There are two different mou s e models for the G90D mutation and a number of different models for the P23H mutation The first mouse model for the G90D mutation was created by Dr. Muna Naash (Naash et al. unpublished data). This model expresses a transgene with a sequence that differs from the wild-type sequence at six nucleotide positions (Figure 17). A GA to AT nucleotide mutation in codon 90 creates the G90D mutation A T to C nucleotide mutation in codon 100 creates a restriction length polymorphism (RFLP) by deleting a Nco I site in the transgene Finally, mutations in the last nucleotide of codons 91, 92, and 93 create sequence changes that allow the differentiation of the normal gene and the transgene as well as the differentiation of the mRNA of these genes. A second mouse model for the G90D mutation was created by Dr. Paul Sieving (Sieving et al. unpublished data). In this model, a rhodopsin transgene was created that only had the two nucleotides conversions in codon 90 responsible for the G90D mutation. This transgene is closer to the sequence of the human G90D sequence than the transgene used by Dr Naash. Dr Naash and colleagues also created a transgenic mouse that expresses the P23H mutant rhodopsin protein (109). The mutant rhodopsin transgene used to create these transgenic mice has two restriction fragment length polymorphisms (RFLP) and three amino acid substitutions not found in the wild type rhodopsin protein (109). The RFLPs
PAGE 48
A WT A.A. Seq WT Seq. G90D Seq. G900 A.A. Seq B WT A.A. Seq WT Seq. G90D Seq. G900 A.A. Seq C WT A.A. Seq WT Seq. VPP Seq. VPPA.A. Seq 3 7 V F G G F T T GTC TTC GGA GGA TTC ACC ACC AT T G G D V F G G F T T GTC TTC GGA GGA TTC ACC ACC AT D V R s p F E Q p GGT CGG AGC CCC TTC GAG CAG CCG G T A C T G H L Figure 17. The nucleotide and amino acid sequence of the transgenes used to create the mouse models for this project. A) The G90D transgene in the Naash mouse model. B) The G90D transgene in the Sieving mouse model. C) The VPP transgene in the Naash mouse model.
PAGE 49
38 can be used to distinguish the mutant mRNA and DNA from that of the normal rhodopsin gene Likewise, a polyclonal antibody has been produced which is capable of preferentially recognizing the unique epitope in the P23H transgene rhodopsin (110). These models offer an opportunity to test the effectiveness of ribozyme based gene therapy. In order to prevent rod cell death, though, the ribozyme must by delivered to the photoreceptors. The ribozyme must then be expressed at a high level for an extended period of time Two major methods can be used to deliver ribozymes to the retina. A ribozymes can be delivered to its target cell as an exogenous RNA oligonucleotide, or alternatively a gene encoding the ribozyme can be delivered to a cell using a recombinant viral vector. Each method has its own advantages regarding efficiency, safety, and ease of production. Exogenous Methods of Gene and Nucleic Acid Delivery to the Cells Pre-synthesized ribozymes can be delivered to the target cells exogenously, but this leaves them exposed to extracellular as well as intracellular nucleases. Modifying the 2' hydroxyl groups can increase the stability of these ribozymes in serum or nuclear suspensions (111-113). The uptake of nucleic acids by mammalian cells, though, is very inefficient. In fact, naked DNA and RNA are not taken up by the photoreceptor cells or RPE cells in the retina (Maguire, Sun, Zack, Bennett, unpublished data). Various delivery systems can be used to enhance the uptake of nucleic acid by cells. Liposomes were one of the first systems developed for the exogenous delivery of genes to cells. Liposomes are microscopic vesicles composed of unilamellar or multilamellar lipid bilayers surrounding aqueous compartments. Nucleic acids can be incorporated in liposomes composed of neutral or anionic lipids to protect them from
PAGE 50
0 0 0 0 () 0 PIHmid DNA Nuclear Protein (HMG1) 39 20-C _. 60 min e .... .,, e e ONAHMG1 Complex Lipids Vortex I Ultrasonlcallon 't ._ Llposome 30 min HVJ Llposome Inactivated HVJ Figure 18. Delivery of genes to the nucleus of the cell by a novel lipo s ome system ( 117).
PAGE 51
40 nuclease degradation (114). Liposomes however do not allow the targeted delivery of genes or oligonucleotides to cells Furthermore, liposomes have also been shown to be toxic in a dose dependent manner (115). Finally while liposomes might allow the delivery of ribozyme s or plasmid s to a cell, they do not ensure entry into the cytoplasm or nucleus of the cell (116) Hangai and colleagues developed a novel liposome system that delivered the Lacz reporter gene to photoreceptor cells (Figure 18). In this system, the reporter gene was complexed with the nonhistone nuclear protein high mobility group I (HMGI). This protein facilitates trafficking of the reporter gene to the nucleus, the site of gene expression This complex is then encapsulated in a liposome fused with inactive Hemagglutinating Virus of Japan (HJV), a Sendai virus HJV helps to mediate liposome cell membrane fusion to introduce the reporter gene directly into the cytoplasm of the cell (117). Delivery of this reporter gene to the cytoplasm of the cell was 100 to 10,000 times more efficient than delivery by simple liposomes alone. Viral mediated gene delivery An alternative to the exogenous delivery of oligonucleotides to cells is the use of viral vectors. Viruses are obligate intracellular parasites that have evolved to infect cells, often with a great degree of specificity. Replacing the genes involved with replication of the virus with a transgene creates a recombinant virus capable of delivering genes to the cells the virus would normally infect (Figure 19). Currently, replication-deficient viral vectors are the most effective way to deliver genes to the cytoplasm or nucleus of a cell. The ideal viral vector for use in gene therapy has a number of requirements. First, it must be produced in high titers Second, it should integrate safely into a specific site in the host genome, or be maintained as a stable episome. Ideally, it should target specific cell types
PAGE 52
41 Gene replacement -"--"""IC...._ -,,. -Gene inhibition Figure 19. Recombinant viruses can be used as vectors to deliver genes to the nucleus of cells.
PAGE 53
42 Matue Virion Attachment DNA released Hexo:os Figure 20 Adenovirus infection (123 ).
PAGE 54
43 Finally, it must be nonimmunogenic and nonpathogenic (118). Though a number of viral vectors have been developed, the four that have received the most attention are adenoviruses, retroviruses, herpes viruses, and the adeno-associated viruses. Adenoviral Vectors The adenoviruses are a family of non-enveloped linear double stranded DNA viruses They are capable of infecting both dividing and nondividing cells (Figure 20). There are over 40 different serotypes of adenovirus, most of which cause benign respiratory infections in humans. Subgroup C serotypes 2 and 5 are most often used as vectors for gene therapy applications. The adenoviral genome exists as an episomal element in the nucleus of the host cell (119). The wild type genome is 35 kb, and up to 30kb can be replaced by the transgene (Figure 22) (120;121). The first replication deficient adenoviral vector was created by replacing the El gene, which is required for replication, with a gene of interest. The recombinant vector was replicated in a cell line that expresses the El gene. High titers of virus were produced using this cell line (122) Investigators examining this recombinant adenoviral vector in animal models encountered a number of difficulties. Although cells infected with recombinant adenovirus were able to express high levels of the transgene, the expression of the transgene was transient and lasted only 5-10 days. This transient expression was partially due to the strong cell mediated immune response elicited by the recombinant adenovirus. This immune response led to the elimination of infected cells (124; 125). The duration of expression was increased by introducing the virus into immunosuppressed mice and those lacking an immune system (124). An improved version of the virus, in which all of the viral genes were removed, allowed expression of a transgene for 84 days (126).
PAGE 55
ITR I I I E1a E1b Promoter 44 Major Late Transcript -1 t cDNA pA Figure 21. The adenoviral genome. E3 ITR .. E2 E4
PAGE 56
45 Nevertheless, a majority of the human population has been exposed to adenovirus due to a natural infection. Reinfection by recombinant adenovirus may be limited by a humoral response to the coat proteins of the recombinant adenoviral vector (126; 127). In addition, the recent report of a death following injection of a recombinant adenovirus raises questions about the safety of adenoviral vectors (128) Retroviral Vectors Retroviruses are a class of enveloped viruses with a single-stranded RNA genome. Following infection (Figure 23), the retrovirus genome undergoes a reverse transcription reaction that converts the viral RNA genome to double stranded DNA. The DNA proviral transcript then integrates randomly into the genome of the host cell (119). The viral genome of the retrovirus is approximately 10kb in size and contains a gag pol, and env gene that are required for replication (Figure 23). The retroviruses also posses a number of DNA elements required for replication. These elements include the long terminal repeat (LTR) that flanks the retroviral genes. The LTR contains the enhancer and promoter elements that control expression of the retroviral genes, as well as the DNA integration sites Two other important sequence elements are the packaging signals required for inclusion of the viral RNA genome into the virion particles, and the binding sites required for the reverse transcription reaction For gene therapy purposes, only these DNA sequences are included in the recombinant retroviral vector (129). The other viral genes are replaced with a therapeutic transgene of up to 7.5kb in size to create a replication-deficient recombinant viral vector The recombinant vector is introduced into a packaging cell line to create a virion capable of infection, reverse transcription, and integration This packaging cell line expresses the genes required for replication of the
PAGE 57
Adsorpllon to specific receptor 46 Retra-~'s;-1io -n -/-@~ =~ provirus 1 f Retrovirus. s eye e o 22 The infeetiou Figure Budding
PAGE 58
47 LTR gag pol env LTR Figure 23. The retroviral genome.
PAGE 59
48 virus, but none of these genes have a packaging signal, so only the recombinant viral genome is incorporated into the virion (Figure 24 ) A disadvantage of these vectors, most of which are based on the mouse mammary tumor virus, is that they are only able to infect nondividing cells (130 ; 131). Lentiviruses are a subclass of the retrovirus family, that are able to infect both dividing and nondividing cells (132). The best known lentivirus is the human immunodeficiency virus (HIV-1 ). Initial results using HIV-1 vectors showed prolonged expression of a marker gene in muscle, liver, and neuronal cells (133; 134).The genome of the lentivirus is more complicated than that of simple retroviruses and contains six additional genes (Figure 24 ). Whether any of these six genes are essential for infection remains to be determined. An HIV vector with a deletion in two of these six genes was able to infect neurons, but not muscle or liver cells (134). There are problems that need to be overcome before retroviruses can be widely used as vectors for gene therapy. First, only low viral titers in the range of 104 7 Colony Forming Units (CFU)/ml are obtained from current strategies using packaging cell lines (135-138). A second problem with retroviral vectors is transcriptional silencing. The retroviral genome contains a number of elements that lead to silencing of the endogenous promoter (139; 140). Cameron et. al. have identified a number of regions in the genome of retroviral vectors that are able to prevent the expression of a transgene from mammalian promoters (141). The silencing can be ameliorated by the deletion of these regions, but high titers of these vectors cannot be produced. The main problem associated with these vectors however, is the concern about their safety. The generation of replication competent retrovirus from cell lines is the primary concern regarding the
PAGE 60
LTA gag 49 pol Figure 24. The lentiviral genome.
PAGE 61
50 safety of retroviral vectors Even improved packaging cell lines produce replication competent virus, though the frequency is low. The replication competent virus is the result of recombination between the viral vector and the viral genes used to make the packaging cell line (142;143). The production ofreplication competent virus is very rare, and can be further reduced by the elimination of homologous overlap between the vector and viral DNA sequences in the cell line. Insertional mutagenesis is a second area of concern. Retroviruses have been shown to cause oncogenesis, either by disruption of a tumor suppressor gene, or by activation of an oncogene (144; 145). The risk of cellular transformation increases with an increase in the rate of insertional mutagenesis. Powell and colleagues examined the ability of replication deficient retroviral vectors to transform cells in vitro (146). They estimated that insertional mutagenesis leading to transformation occurred at a rate of less than one for every 1.1 x 105 insertions. This was less than the observed rate of spontaneous cellular transformation. In addition, there have been no reports of cancer associated with the use of retroviral vectors in clinical trials It is difficult to draw conclusions from these trials, however, since only a small number of individuals were involved and the patients in these clinical trials have a shortened life span. The possibility of cellular transformation due to insertion of a retrovirus can not be ruled out. Herpes Simplex Virus Herpes Simplex Virus (HSV) is a linear double stranded virus of approximately 150-kb, encoding over 80 viral proteins. HSV has two glycoproteins that it uses to bind to epidermal cells. The virus enters cells by fusion, and the linear viral genome circularizes. The lytic replication cycle is regulated by a temporally-coordinated sequence of gene
PAGE 62
51 A UL Us TR TR TR' TR' II I II 11 II DH I ~5; i t~ t lE~~s~E'& !,! i B /ICP22 rr,cP~+ + IE E Viral DNA L \"'-3-J!__yfsvnthesl s +. VP16 ICPO -::5' ICP27 C UL U5 TR TR TR TR Ill I II II 11 lllll I i E' 11 i1~x !CPO alCPO alCP22 Figure 25. Herpes Simplex Virus (147). A) The HSV-1 genome B) The HSV-1 genetic expression regulatory cascade. C) The HSV-1 gene therapy vector.
PAGE 63
52 expression (Figure 25). The viral protein VP16 that is packaged in the HSV particles along with the DNA genome initiates viral expression in the cell. This protein stimulates the expression of a set of immediate early genes. The immediate early genes are transactivating factors that allow the expression of a second set of genes, the early genes, which are required for viral DNA synthesis. These early genes, in turn, activate the expression of a series of late genes required for viral packaging. The result of this regulated gene expression is the synthesis of viral particles that are released from infected cells by lysis. The entire cycle takes less than ten hours and results in the death of the host cell (119). Next, the virus released by the lysis of the epidermal cells migrates to neuronal cells Following infection of the neuron, HSV either proceeds to another lytic cycle or persists as an intranuclear episome in a latent state. Two different latency associated transcripts (LATs) regulate the expression of genes during latency. LATs appear to play a role in both the establishment of latency and reactivation from the latent state (148;149). One of the properties that make HSV a promising vector is its tropism to neuronal cells that may allow the delivery of genes specifically to neuronal cells In addition, the ability to achieve long term persistence as an episomal element in the nucleus of the cell may allow long term expression of a transgene. Furthermore, about half of the HSV genes are not essential for infection, and can be deleted allowing the insertion of 40-50kb of foreign DNA (150-152). Deleting one or more of the early protein genes allows the construction of a replication-deficient virus that can be produced in a complementing cell line (Figure 25). There are still a number of problems with the use of HSV vectors, though, including toxicity. This toxicity can be decreased by the deletion of one of the
PAGE 64
53 intermediate early genes (150). In addition, a large segment of the population possesses antibodies to HSY that may prevent infection with the recombinant vector. In fact, strong inflammatory responses have been observed in animal models at both the primary and secondary sites of infection. Up to 20% of the animals used in this experiment died (153). Another problem with HSY vectors is that transgene expression is shutdown within one week after infection in human cells A number of groups are currently searching for promoters that do not undergo transcriptional silencing during latency (154). If these problems can be eliminated, HSY vectors show promise in a number of gene therapy applications. Adeno Associated Virus Currently, the virus that comes closest to fulfilling the requirements for the ideal vector for gene delivery is recombinant adeno-associated virus (AA Y). AA Y is a non pathogenic human parvovirus that is dependent on a helper virus for proliferation. AA Y requires host cell proteins for transcription and for replication. These helper viruses activate the expression of host proteins allowing a productive AA Y infection. The usual AA Y helper viruses are adenovirus, herpesvirus and vaccinia (155-157). In the absence of a helper, AA Y is unable to replicate and instead establishes a latent infection (Figure 26). During this latent period the virus integrates preferentially into the q arm of chromosome 19 (158) There is no pathology associated with AA Y integration. The provirus resulting from this integration appears to be stable. However, if the cell is later infected with one of the AA Y helper viruses, the AA Y genome is excised and undergoes the normal replicative cycle (159; 160). This integration appears to be a mechanism for ensuring the survival of AA Y in the absence of helper virus.
PAGE 65
54 Lytle + Adenovlrus Figure 26. AA V infi ectlous cycle.
PAGE 66
55 TR rep cap TR oe e e to pS p19 p40 Rep78 4.2kb n Rep68 3.9 kb V Nn Rep52 3.6 kb n Rep40 3.3 kb V t'. n VP-1 2.3 kb 87 kDa V t-An VP-2 2.3 kb 73 kDa t-An VP-3 2.3 kb 62 kDa rAn Figure 27. AA V genome and gene products ( 14 7).
PAGE 67
56 AAV has a single strand DNA genome of 4 5kb that encodes the rep gene and the cap gene (Figure 27). The rep gene contains three promoters and one intron. Transcription from the rep promoters, and alternative splicing, results in the synthesis of at least four different proteins. These rep proteins are responsible for several steps in the replication of AA Vand are involved in the integration of AA V into the host cell genome. The cap gene codes for capsid structural proteins. At both ends of the genome are 145 nucleotide inverted terminal repeats (ITR). The ITR contains the packaging sequences needed for incorporation into a virion particle. In fact, the terminal repeat is the only sequence required in cis for packaging into the AA V virion. In the recombinant viral vector, rep and cap are replaced with the transgene of interest. The length of the transgene should not exceed 4.5kb in order to ensure efficient packaging into the virion. The recombinant virus is produced by cotransfection of cultured cells, usually HEK 293 cells, with a second helper plasmid (PDG). PDG expresses the AA V rep and cap genes the adenovirus El a, El b, E2a, and E4, genes required for replication (Figure 28) (161). This method allows the production of high titers of virus. Unfortunately, this recombinant virus is not able to integrate into the host genome like the wild type virus The recombinant AAV has been used to infect a large variety of cell types (162164). Furthermore, AAV is able to infect quiescent cells (153). One question concerning AA V vectors is whether the natural immune response to the virus will diminish it efficiency as a vector for use in gene therapy. This is an important consideration since a majority of the population has been exposed to wild type AAV. A recent study of cystic fibrosis patients found that while virtually all had antibodies to AAV, only 32% of the
PAGE 68
57 t5 2NCela I\ e e e Figure 28 The AA V gene therapy vector and packaging strategy ( 14 7).
PAGE 69
58 patients had neutralizing antibodies (J 65). Furthermore, only 5% had peripheral blood lymphocytes that proliferated in response to AAV. A second study examined the immune response of rhesus macaques to infection with wild type AA V alone and in the presence of adenovirus. These studies showed that a cell mediated response was only seen in the presence of adenovirus. A humoral response to AA V was reported in both types of infection (166). Bennett and colleagues have found that circulating antibodies do not prevent retinal reinfection in rhesus macaques (J 67). One of the main limits of the AA V vector is that this vector can only accommodate 4.5 to 4.8 KB of foreign DNA. A recombinant AA V that is 10% larger than the normal AA V can still be packaged. While this limited capacity may exclude the use of AA V to deliver larger genes, it will have no impact on the delivery of ribozymes by recombinant AA V vectors. Comparison of Viral Vectors and Delivery of Genes to the Retina Studies comparing the merits of all four viral vectors in a single experiment have not been performed. This makes the choice of vectors for the delivery of genes to the retina difficult. There are a number of reports of the delivery of a reporter gene to the retina by adenovirus (168-170). Expression of the trans gene, however, was transient. Furthermore, there is at least one report of a significant immune response due to the subretinal injection of adenovirus (J 69). A retroviral construct was used to infect mouse retinal progenitor cells that are still dividing in mice at birth. These progenitor cells stop dividing during the first trimester of gestation in humans The use of retroviral vectors for the delivery of genes to the retina is limited in mice and not practical for use in humans (171). A lentiviral vector however was found to efficiently deliver the GFP reporter gene to both rod and RPE cells. Furthermore, GFP was found to be expressed for at least 12
PAGE 70
59 weeks (133) In a more recent report, a lentiviral vector was able to deliver the cGMP phosphodiesterase P-subunit (PDE-P) gene to rod photoreceptor cells. The expression of the PDE-P transgene seemed to delay the retinal degeneration resulting from its absence (J 72) There are no reports of the use of HSV vectors to deliver genes to the rod cells. There is, however, one report of the delivery of the P-galactosidase reporter gene to various ocular cells using HSV (173). A significant inflammatory response was observed following the injection of this vector, though. Furthermore, the duration of the expression of the reporter gene was not reported. The most promising vector for the delivery of genes to the retina, currently, is AAV. A number of groups have already used AAV vectors to deliver different genes to the retina (162;174-176) Furthermore, persistent expression of a reporter gene in rod photoreceptor cell for at least one year has been demonstrated (177). Promoter Control Once gene delivery is obtained, transgene expression can be controlled at the level of transcription by a promoter. The ideal promoter is tissue specific, regulatable, and strong. Experiments in transgenic mouse lines indicated that only the mouse opsin proximal promoter elements, -222 to 170, are needed to achieve cell type specific expression of P-galactosidase in the retina (178) The proximal mouse opsin promoter (MOPS), was then examined for its activity in rod photoreceptor cells following infection with rAAV (162) In this experiment, the MOPS promoter was used to drive expression of a Lacz or green fluorescent protein (GFP) reporter gene These constructs were packaged into an AAV vector and delivered to the retina by a subretinal injection. The results showed the reporter gene was expressed in 100% of the photoreceptor cells in the
PAGE 71
60 area of the injection. In addition, the reporter gene was only expressed in rod photoreceptor cells. Although the MOPS promoter is not regulatable, its ability to give high levels of expression specifically in rod cells makes it an ideal promoter for the expression of a transgene in the rod cell photoreceptor. In the rAA V construct the MOPS promoter is separated from the ribozyme by the SV40 intron. Studies have shown that transcripts from genes without introns are rapidly degraded in the nucleus leading to a reduction in expression (179). Inclusion of an intron has been reported to enhance the expression of a gene up to 100 fold in cell culture (180). Experimental Aim The goal of this project was to examine the effectiveness of ribozymes in the treatment of autosomal dominant disease. More precisely, ribozymes were designed to specifically cleave the mRNA of rhodopsin genes encoding mutations responsible for ADRP. The hypothesis was that cleavage of this mRNA would lead to a decrease in the expression of the mutant rhodopsin protein (Figure 30). A second hypothesis was that this decrease in the mutant protein would slow or prevent the progression of ADRP. In previous experiments targeting cancer and HIV, several ribozymes were designed to target different sites in the same mRNA substrate. The most active ribozymes were then examined for their ability to prevent disease. Unlike these ribozyme approaches, the selection of the target site for this project was limited by the fact that the ribozymes had to specifically target a short region of mRNA containing the point mutation responsible for ADRP. Since the target site could not be changed, the only way to increase the activity of the ribozymes used in this project was to change the ribozyme. The first step of this project was to design ribozymes able to cleave the sequence in the
PAGE 72
61 11111 RNA processing l mRNA ....... ._ ______ __ mRNA ....... ._ ________ __ WT rhodopsin mRNA Mutant rhodopsin mRNA l l mRNA _,,,A..__ mRNA Translation Mutant mRNA associates l ~ ,.. ,.. i .. ..... I \ I \ R, R, I! 11 ,.; .. ft: \( et' ,.. i= =t i ,. ...... \ I .. ..' ,I ... Wt rhodopsin protein Mutant mRNA is cut and then degraded by cellular nucleases. Figure 29. Ribozymes able to cleave the mRNA of mutant rhodopsin genes may lead to a delay or prevention in the onset of ADRP.
PAGE 73
62 mutant but not the wild type rhodopsin mRNA. Two different hammerhead ribozymes, G90D 1 and G90D2, were created targeting different regions of the mRNA of the transgene in the G90D mouse model created by Dr. Muna Naash (Naash et al. unpublished data) A third hammerhead ribozyme, G90D3, was designed to target the mRNA in the G90D mouse model created in the laboratory of Dr. Paul Sieving (Seiving et al. unpublished data) In addition, a hammerhead ribozyme and three different hairpin ribozymes were designed to target the transgene mRNA in Dr. Naash's P23H animal model for ADRP (109). Helix 4 of the first hairpin ribozyme, HPl, had the sequence found in the naturally occurring hairpin ribozyme. The second hairpin ribozyme, HP2, had an extended helix 4 and a tetraloop that had been shown in previous experiments to increase the stability of the secondary structure of the ribozyme (181). The third hairpin ribozyme, HP3, was similar to the second hairpin ribozyme except that it also had a hairpin loop between helix two and helix three. This structure was expected to increase the ability of the hairpin ribozyme to fold into the catalytically active form of the ribozyme (182). The second step of this project was to develop in vitro assays to examine ribozyme activity. These assays were used to determine the ribozymes most likely to be effective for the treatment of ADRP in animal models. Prior to the start of this project, there were no standard methods or conditions used to determine ribozyme activity. This made it difficult to compare the results of different studies. In fact, in early experiment different groups using the same ribozyme often reported different cleavage efficiencies. For example, early experiments examining the hairpin ribozyme derived from the negative strand of the Tobacco Ringspot Virus reported cleavage rates of 0.06/min to
PAGE 74
63 7 1/min (72;77;183-186). The difference in these rates may be due to different temperatures and buffers used in the different experiments. The differences in the rates, however, are greater than would be expected from experiments that have examined the effect of temperature, pH and divalent ion concentration on cleavage by the hairpin ribozyme (76; 77; 187). The results may instead be due to experimental design. Reactions carried out under different conditions might have different rate-limiting steps. At low concentrations of ribozyme and substrate, for instance substrate-binding will be rate determining. Similarly, in reactions with substrate in excess are limited in many cases by the rate of product dissociation (188; 189) The Cech laboratory was the first to characterize ribozymes employing experiments traditionally used to examine protein enzymes (46). These experiments measure the catalytic rate of the ribozyme, the affinity of the ribozyme for its substrate, and the overall efficiency of the ribozyme reaction (190) Determination of these kinetic coefficients gives a basis for the comparison of ribozymes used in different studies thereby allowing the requirements for ribozyme activity in vivo to be defined. This also allows the affect of various ribozyme modifications on the cleavage reaction to be established. For this project, ribozymes were chosen for their ability to catalyze more than one cleavage reaction, referred to as the ability to carry out a multiple turnover reaction. In order to measure the turnover of a ribozyme, cleavage reactions were set up with a similar ribozyme concentration and increasing concentrations of substrate. The concentration of substrate for each reaction was at least five times the concentration of ribozyme. Cleavage reactions in which substrate was in excess were performed to determine the kinetic parameters of KM and kcat. Under these conditions, KM is the concentration of ribozyme at which the reaction is
PAGE 75
64 at half of its maximum velocity. Likewise, kcat is the rate of the overall reaction and is a measure of the ability to cleave and turnover under these conditions. The overall efficiency of a reaction is determined by the ration of kcailKM (190). Once the kinetic parameters for each ribozyme were measured, other factors affecting the ribozyme reaction were examined. The first of these factors to be studied, was the dependence of the ribozyme cleavage reaction on the concentration of magnesium A number of experiments have demonstrated an increase in ribozyme catalysis in the presence of divalent metal ions. Experiments suggest that the mechanism of action may be different for the hammerhead and the hairpin ribozyme. With the hammerhead ribozyme Mg+2 Mn+2 Ca+2 Co+2 and to a lesser extent Sr+2, and Ba+2 are able to act as cofactors. Experiments suggest that the divalent cation plays two important roles in the catalytic mechanism of the hammerhead ribozyme. First, it promotes the folding of the ribozyme into the active conformation. One of the first experiments to demonstrate the importance of divalent cations in promoting the folding of a ribozyme showed that Zn +2 and Cd+2 both allowed the hammerhead ribozyme to catalyze a cleavage reaction in the presence of spermine (191) Later Bassi et al. used electrophoresis to show that the hammerhead ribozyme folds in response to the types and concentrations of ions present (41). Menger et al. used fluorescent residues and Orita et al. used NMR to examine the conformational change in the hammerhead ribozyme in response to increasing magnesium concentrations (192;193). Both studies showed the important role divalent cations play in the promotion of the folding of ribozyme and substrate into a catalytically active complex. The structure of the hammerhead ribozyme in the presence of magnesium was determined by x-ray crystallography (70; 194). For the
PAGE 76
65 hammerhead ribozyme, however, divalent cation also acts as a catalytic cofactor (191). Dahm and colleagues found that the reaction rates were dependent on the pH of the reaction and were correlated with the pKa of the cations present (24) This suggested that the metal ion formed a metal hydroxide in solution that acted as a base in the cleavage reaction. As with the hammerhead ribozyme, divalent metal cations act as a cofactor in the reaction catalyzed by the hairpin ribozyme. Chowrira and colleagues showed that Mg+2 Sr+2 and ca+2 are all capable of acting as cofactors for the hairpin ribozyme (195). Of the three cations, though Mg +z appears to be the most important cofactor. They also showed that spermidine could induce a slow rate of cleavage in the absence of divalent cations. In addition, the concentration of divalent cations required for the cleavage reaction was decreased in the presence of spermidine. This result suggested that divalent cations played an important role in promoting the folding of the ribozyme substrate complex into an active conformation. Butcher and Burke (196) showed that the hairpin ribozyme did not need a cofactor to cleave its substrate. Formation of the tertiary structure required for catalysis, however, required a divalent cation, in this case Mg+2 Young and colleagues carried out a series of experiments which cast doubt on the ability of magnesium to act as an electron donor or as an electron acceptor required for activation of an attacking or leaving group (197). A second group was able to demonstrate catalysis in the absence of divalent cations (198). In their experiments, ~leavage was carried out in the presence of an aminoglycoside antibiotic or polyamine spermine. Cleavage reactions carried out in the presence of optimum concentrations of these alternative agents proceeded at rates comparable to reactions carried out in the presence of magnesium.
PAGE 77
66 For reactions in vivo, however, divalent cations are expected to play an important role in the cleavage reaction of hammerhead and hairpin ribozymes. Of these divalent cations, Mg2+ is probably the main ionic contributor to ribozyme activity under physiological conditions. Most groups carry out ribozyme reactions in vitro using concentrations of Mg2+ that are much higher than the free intracellular concentration of Mg2+ which ranges for 0.5mM to lmM. Thus, cleavage reactions were carried out in the presence of different concentrations of Mg2+ to determine the ability of a ribozyme to cleave its substrate in the presence of physiological concentrations of Mg2+. In addition, changes in the Mg2+ dependence of a ribozyme reflect the ability of a ribozyme to fold into its active conformation. Mg2+ dependence, then, can be used to compare the ability of different modifications to aid in the folding of a ribozyme into its active conformation. Furthermore, since the aim of this project is to specifically degrade a mutant rhodopsin rnRNA, experiments were carried out to determine whether the wild type rnRNA interfered with cleavage of the mutant mRNA. The ability of a ribozyme to inhibit the expression of a mutant gene would be diminished if the wild-type mRNA inhibited cleavage of the mutant mRNA by the ribozyme. It is important to design a ribozyme that is unaffected by the presence of the wild-type mRNA. Therefore, a constant amount of ribozyme and mutant RNA substrate were incubated with increasing amounts of the wild-type RNA. Finally, the ability of each ribozyme to cleave the mutant mRNA was examined in vitro using a RT-PCR assay. A number of factors can prevent cleavage of a mRNA substrate by a ribozyme. In order for a ribozyme to cleave its substrate, the RNA sequence must be accessible for hybridization by the ribozyme. An initial test of a
PAGE 78
67 ribozyme's ability to cleave its mRNA substrate was carried out in vitro. Cleavage of the mRNA was quantitated using a RT-PCR assay This RT-PCR assay can be used to detect cleavage of both the G90D and P23H rhodopsin mRNA. In addition, this assay can be used to determine the cleavage of the mutant mRNA by ribozymes in experiments carried out in vivo The most efficient ribozymes were cloned into a recombinant AA V construct for analysis in a mouse model of ADRP. Expression of the ribozyme was controlled by the MOPS promoter. The ribozyme AA V construct was then packaged into virions. Some of these ribozymes have already been delivered to the retina of the appropriate mouse model by subretinal injection for analysis in vivo
PAGE 79
CHAPTER2 MATERIALS AND METHODS The Cloning of Ribozyme Constructs Two partially complementary DNA oligonucleotides, overlapping by nine to eleven nucleotides were heated to 65C for 2 minutes and then allowed to cool for 10 minutes at room temperature. The ends were filled out with the large fragment of the DNA polymerase I (New England Biolabs; Beverly MA) for 1 hour at 37 c in the presence of 10 mM deoxynucleoside triphosphates (Amersham Pharmacia; Pistcataway, NJ) and polymerase buffer (10 mM Tris-HCl, pH 7.5, 5 mM MgCh. 7 5 mM dithiothreitol) This reaction generated a double stranded DNA fragment coding for the ribozyme flanked by Not 1 and Mlu 1 restriction sites. These DNA fragments were then digested with the Eag I and Mlu I according to manufacturer s protocols (New England Biolabs; Beverly MA). Next, the product of this reaction was ligated into the pHC expression plasmid that had a cis acting hairpin ribozyme immediately downstream of an Mlu I restriction site (199). This downstream ribozyme performs a cleavage reaction that results in a precise 3' end following an intracellular transcription reaction. T4 polynucleotide ligase was used according to the manufacturer's specifications (New England Biolabs; Beverly MA). Ribozyme plasmids were maintained in transformed E. coli DH5-alpha cells. Clones were screened by colony hybridization following a protocol based on the method originally developed by Grunstein and Hogness (200). Briefly, a nitrocellulose membrane was placed on an agar plate with the bacterial colonies. This 68
PAGE 80
69 membrane was then soaked in 0 5 M NaOH to lyse the colonies transferred to the membrane. The membrane was then rinsed with chloroform to remo v e bacterial membrane lipids and the DNA was crosslinked to the nitrocellulose membrane by ultraviolet light (Stratagene UV Stratalinker Model 1800). Next the membrane is incubated for 30 minutes at 50 C in 10 mL of pre-hybridization fluid. The sense oligonucleotides used in the cloning of the ribozymes were also used to probe the membranes. The target oligonucleotide to be used in cleavage reactions was radioactively labeled at the 5 end using T4 polynucleotide kinase (New England Biochemicals ; Beverly MA). These reactions were set up as follows: 6 L sterile water 1 L (lOpmoles) ofRNA oligonucleotide 1 Ly[32P] ATP (167 uCi ICN) 1 L kinase buffer (0.7 M Tris-HCl pH 7.6 0.1 M MgClz 50 mM dithiothreitol) 1 L T4 polynucleotide kinase ( New England Biochemicals 10 units) Following the kinase reaction, the reaction was diluted to 100 Land the free y[32P] ATP was removed using a sephadex column. The activity of the probe is calculated by pipeting 1 L of the kinase reaction onto a glass fiber filter (Gelman Instrument Company ; Ann Arbor MI). The filter was allowed to dry and then it was added to a vial along with 5 mL of scintillation fluid (ICN; Irvine, CA). Radioactive decay was measured in a liquid scintillation counter (Beckman; Palo Alto CA). Finally the membrane was incubated overnight at 50C in 10 mL of hybridization fluid to which five million counts per minute of probe was added. The membrane is then washed for ten minutes three times with wash solution at 3 7 C. The membrane is then
PAGE 81
70 exposed to film overnight. The film is developed and the colonies labeled by the probe are isolated and grown in a 3 mL culture. Plasmid DNA from each culture was prepared by the alkaline-SOS method (201). Finally, the clones were verified using the T7 DNA polymerase sequencing kit (USB; Cleveland, Ohio). The following deoxynucleotides were used to clone the ribozymes this project. Sense Strand Oligonucleotide which has the same sequence as the ribozyme mRNA. Antisense Oligonucleotide which has a sequence complimentary to ribozyme mRNA G90Dl Hammerhead Ribozyme Active Ribozyme Sense Strand-5'CCG GGA TCC GTC GTA ACT GAT GAG CCG CTT CGGC Antisense Strand-5'GCC ACG CGT CGG AGA ITT CGC CGC CGA AGCGG Inactive Ribozyme Sense Strand-5'CCG GGA TCC GTC GTA ACT GCT GAC CCG CTT CGGC Antisense Strand-5'GCC ACG CGT CGG AGA TIT CGC CGC CGA AGCGG G90D2 Hammerhead Ribozyme Active Ribozyme Sense Strand-5'CGG GAA TTC ATC TCC CTG ATG ACG GCG AAA GCC GGA AAA GAC CAC GCG TCG G Antisense Strand-5'CCG ACG CGT GGT CIT ITC CGG CIT TCG CCGTCA TCAGGGAGA TGAATTCCCG
PAGE 82
G90D3 Hammerhead Ribozyme Active Ribozyme 71 Sense Strand -5'CAG GCG GCC GCG GTG GTC TGA TGA GCC CG Antisense Strand-5'GTC ACG CGT AGA CTT TTC GCC CTI TCG GGC TCA TCA GAC Inactive Ribozyme Sense Strand-5'CAG GCG GCC GCG GTG GTC TGA TGA GCC CG Antisense Strand-5'GTC ACG CGT AGA CTT TGT GCC CTT TCG GGC TCA TCA GAC C HHl Hammerhead Ribozyme Active Ribozyme Sense Strand-5'CTC CGG CCG AAG TCT G Antisense Strand-5'GAG CAG CGT CGG AGT TTC GCG CTT TCG CGC TCA TCA GAC TTC GGC CGG AG HPl Hairpin Ribozyme Active Ribozyme Sense Strand-5'GCA GAA TIC AGC GGC CGC ACG AAG TAG AAC CGA ACC AGA GAA ACA CAC G Antisense Strand-5'GCC ACG CGT ACC AGG TAA TAT ACC ACA ACG TGT GTI TCT CTG G HP2 Hairpin Ribozyme Active Ribozyme Sense Strand-5'GCA GAA TIC AGC GGC CGC ACG AAG TAG AAC CGA ACC AGA GAA ACA CAC G Antisense Strand-5'GCC ACG CGT TAC CAG GTA ATG TAC CAC GAC TT A CGT CGT GTG TTT CTC TGG
PAGE 83
72 Inactive Ribozyme Sense Strand-5'ACG GGC CGC ACG AAG TAT TAC CGA CAC CAO AGA AAC ACA CG Antisense Strand-5'GCC ACG COT TAC CAO OTA ATG TAC CAC GAC TTA COT COT GTG TTT CTC TOG HP3 Hairpin Ribozyme Active Ribozyme Sense Strand-5' AOC GGC CGC ACG AAG TAG AAC CGA CCC GAC GAC OTA AGT COT CCC CAC CAO AGA AAC ACA CG Antisense Strand-5'GCC ACG COT TAC CAO OTA ATG TAC CAC GAC TTA COT COT GTG TTT CTC TOG Inactive Ribozyme Sense Strand-5' AOC GGC CGC ACG AAG TAT TAC CGA CCC GAC GAC OTA AGT COT CCC CAC CAO AGA AAC ACA CG Antisense Strand-5'GCC ACG COT TAC CAO OTA ATG TAC CAC GAC TTA COT COT GTG TIT CTC TOG The nucleotides in bold were sequence alterations used to create the inactive ribozyme (202;203). In Vitro Transcription for the Generation of Ribozyme RNA Plasmids containing the ribozyme constructs were linearized with the Mlu I restriction enzyme, which cuts downstream of the ribozyme. An internally labeled ribozyme was generated by a T7 RNA polymerase reaction carried out in the presence of a-[32P] uridine triphosphate (ICN, Costa Mesa, Ca). The transcription reaction was prepared in a 100 L volume using the standard method as follows: 83 L of water deionized and purified by a filtration system (Barnstead; Dubuque, IA) and then autoclaved.
PAGE 84
73 10 L of buffer (400 mM Tris-HCl [pH 7.9), 6 mM MgCh, 2 mM spermidine, 10 mMDTT) 4 L 20mM NTP (Amersham Pharmacia; Pistcataway, NJ) 1 L Placental RNase inhibitor (RNAsin, Promega, Madison WI.) 1 L [a-32P] UTP (10 uCi/uL)(ICN; Irvine, Ca) 1 L T7 RNA polymerase (5 units) This reaction was incubated for 1 hour at 37 c. The product of the transcription reaction was purified on an 8% denaturing polyacrylamide urea gel. The RNA was eluted overnight in an elution buffer containing 1 % SDS, 50 mM Tris-HCl pH 7.4, and 0.2 M ammonium acetate The supernatant was ethanol precipitated twice and resuspended in 50 L of water. The specific activity was then determined for each ribozyme. First, 1 L of the ribozyme solution was spotted onto a glass fiber filter (Gelman Instrument Company; Ann Arbor, Ml). The filter was allowed to dry and then it was added to a vial along with 5 mL of scintillation fluid (ICN; Irvine, CA). Radioactive decay was measured in a liquid scintillation counter (Beckman; Palo Alto, CA). To calculate the specific activity the concentration of cold UTP was divided by the radioactivity, in microcuries (Ci), of the a-[32P] UTP added after taking into account the radioactive decay of the a-[32P] UTP. The value of this calculation was divided by both the number of uridine residues in the transcript and the conversion factor 2.22 x 106 Ci/cpm. The result of the previous calculation was multiplied by the cpm of the sample determined by the scintillation counter assuming 100% counting efficiency. This calculation gives a value of the concentration of the ribozyme in nM/L.
PAGE 85
74 Preparation of Target Oligonucleotide Approximately 400 g of oligonucleotide was deprotected overnight by constant mixing of the oligonucleotide in 400 mL of trihydrofluoride (Sigma Chemical Company; St. Louis, Ml) at room temperature. The oligonucleotide was precipitated with absolute butanol (Fischer Scientific; Fair Lawn, NJ) and centrifuged at maximum speed for 30 minuets in an eppendorff microcentrifuge. The pellet was washed with absolute ethanol, dried and resuspended in 20 L of sterile RNase free water. The oligonucleotide was run on a 20% denaturing polyacrylamide urea gel and excised using an UV shadowing technique to visualize the band The RNA oligonucleotide was eluted overnight as in the previous method for the ribozymes. The supernatant was ethanol precipitated twice using dextran blue (Sigma Chemical Company; St. Louis, MO) as a carrier and resuspended in 20 L of water. Then the oligonucleotide was quantitated by measuring the absorbance at 260 nM in an UV spectrophotometer (Hewlett Packard; Palo Alto CA). The volume of the RNA oligonucleotide stock solution was then diluted to attain a concentration of 10 pMoles/L. The target oligonucleotide to be used in cleavage reactions was radioactively labeled at the 5' end using T4 polynucleotide kinase (New England Biochemicals; Beverly, MA). These reactions were set up as follows: 6 L sterile water 1 L (lOpmoles) of RNA oligonucleotide 1 Ly-[32P] ATP ( 167 uCi, ICN) 1 L kinase buffer (0.7 M Tris-HCl, pH 7.6, 0.1 M MgCli. 50 mM dithiothreitol) 1 L T4 polynucleotide kinase ( New England Biochemicals, 10 units) The following RNA oligonucleotides were used in this research:
PAGE 86
75 Mutant G90Dl: 5'GGA GAU UUU ACG AC Wild Type G90Dl: 5'GGA GGA UUC ACC AC Mutant G90D2: 5'UGG UCU UCG GAG AUU Wild Type G90D2 : 5'UGG UCU UCG GAG GAU Mutant G90D3 : 5' AGA CUU CAC CAC C Wild Type G90D3: 5' AGG AUU CAC CAC C Mutant VPP target: 5'UCG GAG UCA CUU CG Wild Type VPP: 5'UUC GGA GUC ACU UCG In Vitro Cleavage Reactions A standard protocol was followed for in vitro ribozyme cleavage reactions. First, separate dilutions of ribozyme and target were heated at 85C for 2 minutes. The ribozyme was removed and allowed to cool at room temperature for 10 minutes. Magnesium buffer (400 mM Tris-HCl, pH 7.5, 200 mM MgCh) was added, and the ribozyme was incubated at 37C for 30 minutes prior to starting the reaction. The target was quick-cooled on ice following heating. Adding an equivalent volume of the target dilution and the ribozyme solution started the cleavage reaction. The cleavage reaction was incubated at 37C and stopped by adding 3 volumes of a gel loading dye containing 8 M urea (BioRad; Hercules, CA), 100 mM EDTA (Sigma; St. Louis, MO), 0.1 % xylene cylanol (Sigma; St. Louis, MO), and 0.1 % bromophenol blue (Sigma; St. Louis, MO) to 1 volume of the cleavage reaction. Samples were then heated to 85C for 2 minutes and quick cooled on ice. Next, the samples were separated on a 15% polyacrylamide, 8 M urea gel. The gel was then fixed for 20 minutes in a solution of 10% methanol 10% acetic acid volume to volume, dried under vacuum, and quantitated on a Phosphor Imager
PAGE 87
76 (Molecular Dynamics, Costa Mesa, CA). Conditions such as the substrate concentration, magnesium concentration, and reaction time were varied in individual experiments (see below). Initial Characterization of Ribozyme Activity by an In Vitro Cleavage Reaction An initial cleavage reaction was carried out in order to determine the ability of a ribozyme to catalyze the cleavage of an RNA oligonucleotide. In this reaction, 1000 nanomoles of ribozyme and 1000 nanomoles of an RNA oligonucleotide substrate were diluted to a volume of 9 L. This reaction was heated at 85C for 2 minutes and allowed to cool at room temperature for 10 minutes. The reaction was started by adding 1 L of a concentrated ( 1 OX) magnesium cleavage buffer. The reaction was allowed to incubate at 37C for 16 hours. The reaction was stopped by adding 15 L of the urea gel loading dye. The samples were then heated at 85C for 2 minutes, quick cooled on ice, and separated on a 15% polyacrylamide, 8 M urea gel. The gel was fixed, dried, and quantitated on a Phosphor Imager (Molecular Dynamics, Costa Mesa, CA). Time Course of the Ribozyme Cleavage Reaction For this experiment, the ribozyme was diluted with sterile, purified water to a concentration of 20 nM and a volume of 40 L. At the same time, the RNA oligonucleotide substrate was diluted to a concentration of 400nM and a volume of 50 L in a separate tube. Both tubes were heated for 2 minutes at 85C. The target RNA oligonucleotide was quick cooled on ice. The ribozyme was allowed to slow cool at room temperature for 10 minutes. Then 10 L of the magnesium buffer was added to the ribozyme solution, which was then incubated at 37C for 30 minutes Next, the target dilution of either mutant or wild type oligonucleotide was added to the ribozyme solution.
PAGE 88
77 A reaction containing only the target RNA oligonucleotide with the magnesium buffer was also prepared. Aliquots of 5 L were taken at various intervals and added to 15 L of a gel loading dye For the G90Dl, HHl and HPl ribozymes, aliquots were taken at 0, 1.5, 3.5, 6,13, 24, 52, and 100 hours. For the G90D2 ribozyme, aliquots were taken at 0, 0.37, 0 75, 1.5, 3, 6,12, 24, and 48 hours. For the G90D3 ribozyme aliquots were taken at 0, 0.75, 1.5, 3, 6, and 12 hours. For HP2 and HP3 ribozymes, aliquots were taken at 0, 0.75, 1.5, 3, 6,12, 24, and 48 hours. Each reaction was repeated in triplicate Samples were run on a 15% urea gel. The samples from the time course of a single reaction were loaded in sequential order with increasing intervals going from left to the right side of the gel. Each of the three different reactions was run side by side on a single gel to allow a comparison. The gel was quantitated on a Phosphor Imager. The percentage of substrate cleaved at each time point was determined by dividing the radioactivity in the product by the sum of the radioactivity of the product and target. In addition, the total radioactivity at each interval was measured in order to determine the amount of non-specific degradation that occurred. Multiple Turnover Kinetic Analysis In order to determine kinetic parameters, a series of standard cleavage reactions were performed. For these experiments, the final concentration of ribozyme was kept constant at 10 nM. The final concentration of target RNA oligonucleotide used in reactions with the G90Dl, G90D3, and HP2 ribozymes was 40 nM, 60 nM, 80 nM, 100 nM, 120 nM, 140 nM, and 160 nM. The concentration of target RNA oligonucleotide used in reactions with the HP3 ribozyme was 80 nM, 100 nM, 120 nM, 140 nM, 160 nM, and 180 nM. For these reactions a two-fold (2x) concentration of ribozyme in a 4 L
PAGE 89
78 volume was heated at 85C for 2 minutes and allowed to cool at room temperature for 10 minutes. Next, 1 L of thelOx cleavage buffer was added to the ribozyme and the 5 L mixture was incubated at 37C for 30 minutes A 2x concentration of the substrate in a 5 L volume was heated at 85C for 2 minutes and quick cooled on ice. Two different reactions were carried out. The first reaction was begun by adding the 5 L ribozyme mixture to the 5 L substrate dilution. The second reaction was begun by adding 4 L of sterile water and 1 L of !Ox-cleavage buffer to the 5 L substrate dilution. The samples were then incubated at 37C for a pre-determined period The reaction time for each ribozyme kinetic reaction was the time at which 10% of the substrate had been cleaved in the previous time course experiment. The reaction time for the G90D 1 ribozyme was 12 hours The reaction time for the G90D3 ribozyme was 2 hours The reaction times for HP2 and HP3 were 5 and 3 hours. The reactions were stopped by adding 15 L of the urea gel loading dye. The samples were then heated at 85C for 2 minutes, quick cooled on ice, and separated on a 15% polyacrylamide, 8 M urea gel. The gel was then fixed, dried, and quantitated on a Phosphor Imager. The velocity of the cleavage reaction was determined by dividing the moles of the product of the cleavage reaction by the reaction time A double reciprocal plot of velocity versus velocity divided by substrate concentration was used to determine the values of V max and KM for each ribozyme Dependence of Ribozyme Cleavage on Magnesium Concentration In this experiment, reactions were prepared with a final concentration of 200 nM of target RNA oligonucleotide and 10 nM of each ribozyme. These reactions were incubated with increasing amounts of magnesium in a 10 L total reaction volume.
PAGE 90
79 Ribozyme and target were mixed together in a 9 L reaction volume, heated at 85C for 2 minutes, and allowed to cool at room temperature for 10 minutes. The reaction was started by adding 1 L of cleavage buffer, which gave a final magnesium concentration of 1.0 mM, 5 0 mM, 10 mM, 20 mM, 40 mM, 80 mM, or 160 mM. The ribozyme reactions were allowed to incubate the same amount of time as the reactions in the previous kinetic experiments. Samples were then mixed with a loading dye and run on a 15% urea gel. The gel was quantitated on a Phosphor Imager as previously described. Test of Competition by a RNA Oligonucleotide with the Wild Type Opsin Sequence In this experiment 100 nM of target RNA oligonucleotide and 10 nM of each ribozyme were incubated in a 10 L total reaction volume. In addition, RNA oligonucleotide with the wild type opsin sequence were added to a series of reactions to a final concentration of 50 nM, 100 nM, 150 nM, 200 nM, 300 nM and 350 nM. Ribozyme and target were mixed together, heated at 85C for 2 minutes, and then allowed to cool at room temperature for 10 minutes The reaction was started by adding the standard 1 Oxcleavage buffer. The reactions were incubated for the intervals used in the kinetic reaction. Samples were then mixed with the loading dye and run on a 15% urea gel. The gel was quantitated on a Phosphor Imager. Total Retinal RNA Extraction and Cleavage Mice were euthanized in the laboratory of Dr. Muna Naash and the retina were removed and snap frozen in liquid nitrogen, stored at 70 C, and sent to our laboratory. Up to six retinas were added to 1 mL of Trizol (Life Technologies; Grandview, NY) and homogenized by repeated passage through an 18-gauge syringe needle. The retinas were then incubated in Trizol for 5 to 15 minutes. Next, 200 L of chloroform were added, the
PAGE 91
80 solution was vortexed, and allowed to incubate at room temperature for 2 minutes Then the solution was centrifuged at top speed in an eppendorff microcentrafuge for three minutes, and the aqueous layer containing the retinal RNA was removed. The RNA was precipitated by adding an equal volume of isopropanol (Fischer Scientific; Fair Lawn, NJ) to the aqueous phase, incubating five minutes at room temperature, and spinning at maximum speed in an eppendorff microcentrafuge. The RNA was then washed two times with cold 70% ethanol, dried, and resuspended in 20 L of sterile water. Approximately 6 g of total RNA extract was incubated with 600 nM ribozyme and placental RNase inhibitor (Promega Madison, WI) to prevent degradation by residual cellular RNases in cleavage buffer at 37 C. Aliquots were removed from the reaction at 0 6 12,24,and 48 hours. Absolute ethanol was added to the aliquots to stop the cleavage reaction and precipitate the RNA. RT-PCR Reaction The aliquots were ethanol precipitated, washed twice with cold 70% ethanol, and dried. A reverse transcription reaction was started by resuspending the cleavage reaction in 11 L of sterile water and a 1 L solution containing 2 pM of an antisense ~ actin primer and 1 pM of an antisense mouse opsin primer. This mixture was heated at 70 C for 10 minutes and quick chilled on ice. Next 4 L of 5x buffer (250 mM Tris-HCl pH 8 3, 2 L 0 1 M DTT, 1 L lOmM dNTP) was added to each reaction. Following a 5 minute incubation at 42 C, 1 L of Superscript II (Life Technologies; Grandview, NY) was added and the reaction was incubated for 50 minutes at 42 C. The product of the reverse transcription reaction was amplified by a 25 cycle polymerase chain reaction (PCR) reaction in the presence of 60 pM of ~-actin primer and 30 pM of opsin specific
PAGE 92
81 primer. The reactions were heated at 95 C for 30 seconds, annealed at 53 C for 30 seconds, and extended at 74 C for 45 seconds. When this PCR reaction was finished, 10 Ci of [ a32P] ATP was added and an additional cycle was carried out. The PCR product was then digested with Nco I (Roche Molecular Biochemicals; Indianapolis, IN). Two Nco I restriction sites are found in the normal opsin PCR product but only the upstream restriction site was found in the transgene PCR product (Figure 30). Digestion yielded a 279 base pair product from the mutant transcript, a 197 and an 80 base pair product from the normal transcript, and a 157 base pair band common to both transcripts. A comparison of the ratio of wild type product or mutant product relative to the ~-actin product in reactions with and without ribozyme allowed for a determination of the fraction of mutant transcript cleaved by the ribozyme. The mouse opsin 3' antisense primer used for the reverse transcription reaction and the following PCR reaction was a 17 nucleotide DNA oligonucleotide with the sequence 5'CCA CAG GGC GAT TTC AC. This primer hybridized to nucleotides 3266 to 3283 of mouse opsin CDNA. This oligonucleotide overlapped the 3' end of exon 1 and the 5' end of exon 2. The 5' primer for this reaction was a 17 nucleotide DNA oligonucleotide with the sequence 5' AAG CAG CCT TGG TCT CT that was complementary to nucleotides 1376 to 1393 of mouse opsin cDNA. The mouse ~-actin 3' antisense primer used for the reverse transcription reaction and the following PCR reaction was a 20 nucleotide DNA oligonucleotide with the sequence 5' ACT CCT GCT TGC TGA TCC CT. This primer hybridized to nucleotides 1146 to 1167 of mouse ~-actin cDNA. This oligonucleotide overlapped the 3' end of exon 1 and the 5' end of exon 2. The 5' primer for this reaction was a 21
PAGE 93
82 Perform RTPCR Reaction Wild-Type Rhodopsin RT-PCR Fragment NCOI NCOI 103bp Mutant Rhodopsin RT-PCR Fragment NCOI VPP Mutation G90D Mutation Digest with NCOI l 197 bp 103 bp Actin Standard 649bp 300 bp 168 bp Figure 30. An RT-PCR method for quantitation of the cleavage of mRNA by a ribozyme.
PAGE 94
83 nucleotide DNA oligonucleotide with the sequence 5'GTT TGA GAC CTI CAA CAC CCC that was complementary to nucleotides 449 to 470 of mouse opsin cDNA. Creation of recombinant AA V ribozyme constructs. Ribozymes to be tested in vivo were cloned into a recombinant AA V construct based on the pTR-UF2 vector (204) The ribozymes were expressed from a 472 base pair murine opsin promoter (mops500) (205). Each ribozyme was followed by an internally cleaving hairpin ribozyme derived from the pHC plasmid (199). The ribozyme cassette was preceded by an SV40 intron and followed by an SV40 polyadenylation signal to promote nuclear export of the ribozyme. Mops 500 was also used to drive expression of the GFP reporter gene (Figure 31). This allowed the efficiency of transfection with the rAA V construct to be determined. DNA constructs were packaged into AA V particles by standard procedures in the lab of Dr.William Hauswirth. Ribozyme was delivered to the retinal via subretinal injection. An rAA V suspension was delivered to the subretinal space in the posterior retina. The opposite eye was injected with a control rAA V containing an inactive ribozyme or no ribozyme (86). The first step in cloning a ribozyme into the pTR-MOPS-UF2 plasmid was to carry out a PCR reaction to add Not I restriction sites to both ends of the ribozyme construct. The total volume of this reaction was 50 L. This reaction was carried out for 25 cycles in the presence of 0.05 g of the pHC plasmid containing the ribozyme construct, 45 pM of the 5' sense primer, and 45 pM of the 3' antisense primer. The 138bp product of this reaction was gel purified on a 2% agarose gel. This product was then
PAGE 95
TR 84 MOP ss RZ PA MOP GFP Figure 32 The AAV cassette used to create rAAV as a vector for a ribozyme: TR= AAV terminal repeat, BOP= bovine opsinpromoter SS = splice donor acceptor site, Rz = ribozyme cloning site PA = polyandenylation sequence, MOP= mouse opsin promoter GFP = green fluorescent protein TR
PAGE 96
ori 85 TR MOPS500 promoter ----, o,1(683) pMOPS500GFP 7229 bp 'GFPh --Natl ( 1 4 15) AmpRes;st~ SV40 poly(A) Sa/1( 1622) PYF441 enhancer HSV-tk neoR Sa/I (2721 ) bGH poly(A) TR F igure 32. The plasmid used to create the rAAV ribozyme construct.
PAGE 97
86 digested with Eag I to generate ends compatible with those generated by Not I. This fragment was then cloned into the pTR-MOPS-UF2 plasmid and amplified in DH5a cells (Figure 32) This plasmid was extracted using the alkaline lysis method. A second PCR reaction was carried out to add Sal I sites to each end of the ribozyme ca s sette. This reaction was carried out for 25 cycles in the presence of O Olg or ribozyme containing pTR-MOPS-UF2, 45 pM of 5' sense primer, and 45 pM of the 3' antisen s e primer. The product of thi s reaction was gel purified and dige s ted with Sal I This fragment was then ligated into the Sal I site of the pTR-MOPS-UF2 plasmid. This procedure creates a recombinant AAV vector containing a ribozyme and GFP The expression of both is driven by the MOPs 500 promoter. This plasmid was amplified in DH5a cells, and extracted using the alkaline lysis method The sequence of the ribozyme was verified and 0.1 ug of plasmid was digested with Smal to ensure that the AA V construct has both terminal repeats intact. Once these two requirements have been met, the plasmid is maintained in sure cells.
PAGE 98
Chapter 3 Results Ribozymes Design In ADRP expression of the mutant protein results in retinal dystrophy (93). The hypothesis underlying this research project was that a decrease in the expression of the mutant protein would delay or prevent the disease. In this study, several different ribozymes were designed to specifically cleave one of two different mutations, G90D and P23H, in the rhodopsin mRNA. Each mutation causes retinal degeneration in a mouse model of the corresponding human disease. The success of ribozyme therapy depends on the identification of an appropriate RNA target site. The hammerhead ribozyme is able to cleave an NUX target sequence as previously described, but the efficiency of the cleavage reaction can vary by as much as 100-fold depending on the triplet. The most efficient cleavage triplet is a GUC sequence followed by CUC, which is five-fold less efficient, and UUC, which is 10 times less efficient. All other sites are cleaved at least 50 times less efficiently (7 J). The hairpin ribozyme has slightly different target site requirements than the hammerhead ribozyme The target cleavage site of the hairpin is less flexible. Mutational analysis demonstrated that only the GUC triplet is cleaved efficiently (72;206). The length of the hybridizing arms of the ribozyme can also affect the cleavage reaction. An increase in the length of the hybridizing arms of a ribozyme decreases the rate of the dissociation of cleaved product from the ribozyme. Therefore, the turnover and 87
PAGE 99
88 overall rate of the cleavage reaction decreases (189). An increase in the length of the hybridizing arms of the ribozymes can also lead to a decrease in the specificity of the cleavage reaction This occurs when mismatched substrates bind to the ribozyme and are cleaved before the effect of the mismatch is manifested. Conversely, if the hybridizing arms are too short, the ribozyme may not stay bound to the substrate long enough to ensure cleavage (207). Furthermore, it is estimated that a minimum of 11 to 15 nucleotides are needed to distinguish unique RNA sequences (208;209). As a rule, for hammerhead ribozymes with hybridizing arms greater than six nucleotides, the dissociation of ribozyme and product is the rate-limiting step in a bi-molecular cleavage reaction. For hammerhead ribozymes with hybridizing arms of 5 nucleotides or less, turnover is enhanced and cleavage of the substrate by the ribozyme becomes the rate limiting step in a bi-molecular reaction (189 ; 210). In general hammerhead ribozymes with hybridizing arms of six nucleotides have the highest catalytic efficiencies in vitro and in vivo (189 ; 211-213) For the hairpin ribozyme, the length the hybridizing arm formed by helix 2 must be four nucleotides (Figure 10). The other hybridizing arm, helix 1, is usually kept at six nucleotides in length (184). Two strategies were used to ensure that a ribozyme was able to specifically cleave the mRNA of a mutant but not of a wild-type gene. The first approach was to identify a mutation that creates a cleavage triplet target in the mRNA of the mutant gene. The second strategy was to identify a mutation in close proximity to the cleavage triplet in the substrate. These mutations allow cleavage of the mutant mRNA but not the wild type mRNA because the mismatched base pairs disrupt the formation of the catalytic core of the ribozyme (186;206) Werner and Uhlenbeck (214) have shown that a mismatch in
PAGE 100
89 any of the first four innermost nucleotides of helix ill of a hammerhead ribozyme prevented cleavage (Figure 9). Only a mismatch in the innermost nucleotide of helix I prevented cleavage. Mismatches in the more distal regions of either arm had no effect on cleavage. Another study showed that a mismatch four nucleotides from the catalytic core would allow specific cleavage of a target substrate by a hammerhead ribozyme. Shortening the length of the arm with the mutation and increasing the length of the other arm prevented cleavage of the mismatched substrate (215). Similarly, Berzal-Herranz and colleagues (206) showed that mismatches in the first three nucleotides adjacent to loop A of the hairpin ribozyme in either helix 1 or helix 2 abolished cleavage (Figure 10). Mismatches in the more distal nucleotides of helix 1 or 2 resulted in a reduction of cleavage by the ribozyme We have designed ribozymes to cleave the mutant rod opsin mRNA found in transgenic mouse models of retinal dystrophy The first rhodopsin mutation targeted was the G90D mutation that results in congenital stationary night blindness in humans (90). In the G90D trans gene present in the N aash mouse model of the disease, a GA to AT mutation creates an AUU target site in the transgene (Figure 18) (Naash et al. unpublished). A hammerhead ribozyme, G90D 1, was designed using this target sequence (Figure 33). A second hammerhead ribozyme, G90D2, was designed that recognized a UUC cleavage triplet found in both the wild-type and mutant mRNA (figure 34). The UUC target site is cleaved 22 times more efficiently than the AUU cleavage site. Specificity in this case depends on a targeting arm that can hybridize to the mRNA of the transgene containing a GA to AT mutation. The A and T mutations are the most distal nucleotides in helix I of G90D2. A third ribozyme, G90D3, was designed to cleave a
PAGE 101
90 Cleavage site GA t C C 51 GGAGAUUUUACGA I 1111 111111 Wt Sequence Mut sequence CCUCUA AAUGCU 51 A Cu ... C A GA... U G AGU C-G... C C-G G-C G-C A G A A Figure 33. The sequence and secondary structure of the G90Dl hammerhead ribozyme. The specificity of this reaction is the result of a GA to AT mutation that creates a unique AUU cleavage triplet in the mRNA sequence of the mutant rhodopsin gene. This triplet is highlighted in yellow. The blue letters are the nucleotides found in the inactive ribozyme.
PAGE 102
91 Cleavage site y GA Wt Sequence 5'GGUCUU CGGAGAU Mut sequence 111111 111111 CCAGAA CCUCUA 5' A Cu C A GA~ U G AGU C-G C C-G G-C G-C A G AA Figure 34. The sequence and structure of the 09002 Ribozyme. This ribozyme was designed to target a UUC cleavage triplet (highlighted in yellow) found in the wild-type and mutant mRNA of the rhodopsin gene. The specificity of the cleavage reaction depended on the ability of the ribozyme to hybridize to the mutant but not the wild-type mRNA due to a GA to UA mutation. The blue letters are the nucleotides found in the inactive ribozyme.
PAGE 103
92 UUC cleavage triplet found in the transgene used by Sieving to create a second mouse model for this form of retinal dystrophy (Figures 17 ,35) (Sieving et al. unpublished) The sequence of the Sieving trans gene is slightly different from the sequence of the Naash transgene targeted by the previous two ribozymes and is closer to the human G90D opsin sequence The specificity of this ribozyme is the result of a GA to AC mutation adjacent to the UUC cleavage triplet of the ribozyme. This mutation creates the G90D mutation responsible for stationary night blindness. The G90D3 ribozyme recognizes the AC mutation, and the GA mismatch in the wild-type substrate was expected to prevent cleavage (214). In the transgene used to create the VPP mouse model for ADRP there is a C to A mutation in codon 23 that leads to the P23H mutation resulting in ADRP (Figure 17) (109). There is also a silent C to T mutation in codon 22 that creates a GUC cleavage site in the transgene mRNA. This is the optimal cleavage triplet for both the hammerhead and hairpin ribozyme. Previous experiments have shown that the ability of hammerhead and hairpin ribozymes to cleave an RNA substrate can vary depending on the sequence of the target substrate (188;189 ; 216). The hairpin ribozyme has been shown to achieve the highest catalytic rates of any naturally occurring ribozyme in vitro under standard reaction conditions (pH 7 .5, 10 mM Mg2+ 25C) (188). Studies in cell culture, however, have shown that in some cases the hammerhead ribozyme is more efficient at cleaving a specific RNA sequence (217). Currently, there is no way to predict which ribozyme will cleave a particular target most efficiently so both types of ribozymes were examined A hammerhead ribozyme and three different hairpin ribozymes were constructed using the cleavage site present in codon 22 of the VPP transgene. Helix I and helix ID of
PAGE 104
93 Cleavage site GA 5' AGACWUCACCACC II I I I 1111 UCUGAA UGGUGG Wt sequence Mut sequence 5' A CuG A A~~C u G AGU C-G~~C-G G-C G-C A G AA C Figure 35. The sequence and structure of the G90D3 hammerhead ribozyme. This ribozyme targeted a UUC target triplet (highlighted) found in both the wild-type and mutant mRNA of rhodopsin. Specificity of cleavage by this ribozyme depends on GA to AC mutation adjacent to the UUC cleavage triplet. A mismatch at this position should prevent formation of the catalytic core of the hammerhead ribozyme. The blue letters indicate the nucleotides found in the inactive ribozyme.
PAGE 105
94 the hammerhead ribozyme, HH 1, were both six nucleotides in length (Figure 36). The first hairpin ribozyme, HPl, had the helix 4 and base stem loop found in the wild type hairpin ribozyme (Figure 37) (72) In the second hairpin, HP2, helix 4 contained an additional 3 base pairs and a stable tetra-loop was introduced at the base of the ribozyme (Figure 38). This modification has been reported to enhance the stability of the secondary structure of the hairpin ribozyme increasing the cleavage efficiency of the ribozyme (218). The third hairpin ribozyme HP3, added a hairpin loop in the hinge region of the hairpin ribozyme separating helix two and helix three (Figure 39). A similar modification has been reported to increase the percentage of correctly folded ribozymes (74). The authors of this report, however, tested this modification using a bi-molecular ribozyme that lacked the tetraloop Finally, inactive ribozymes corresponding to G90Dl, G90D3, HP2, and HP3 active ribozymes were created to determine whether the results of in vitro and in vivo experiments were due to an antisense effect or mRNA cleavage. These ribozymes contained mutations in conserved nucleotides required for catalysis by the hammerhead and hairpin ribozymes (Figure 9,10) (202;203). The initial characterization of the catalytic activity of ribozymes. Trans acting ribozymes were cloned into the pHC plasmid, which contains a downstream hairpin ribozyme (199). Ribozymes were generated by a transcription reaction using T7 RNA polymerase. The pHC plasmid containing the cloned ribozyme was used as a template for this reaction following digestion with Miu! to create a linear molecule. RNA oligonucleotides were radiolabled at the 5' end by a T4 polynucleotide kinase reaction (See Materials and Methods). In an initial reaction, a 10 L sample
PAGE 106
95 Cleavage site C l e 5'CGGAGUCA CUUCG 111111 111111 GCCU C A UGAAGC5' A CuG A A .... ~----U G AG C-G C C-G G-C C-G A G AA WT Mut C u Figure 36. The sequence and structure of the HHI hammerhead ribozyme The C to U conversion in the substrate creates a GUC cleavage triplet (highlighted) that allows the specificity of the cleavage reaction. The nucleotide in the blue box creates the P23H mutation. The red letters above the substrate indicate the nucleotides found in the wild type sequence that differ from the mutant sequence. The blue letters are the nucleotides found in the inactive ribozyme.
PAGE 107
96 Cleavage site \ C G A GU c c 5'UCGG A CUUCG 1111 111111 WT Mut AGCC UGAAGC 5' AU-A A A G-C AG U G-C \U-A C-G C A U G u A C A u A A A C G A o ~G-C U-A G-C U G u Figure 37. The sequence and structure of the HPl hairpin ribozyme. The red letters indicate the nucleotides found in the wild-type rhodopsin sequence. The letters in yellow indicate the nucleotide substitutions used to create the inactive ribozyme. The P23H mutation is caused by the nucleotide in the blue box. The cleavage site is indicated by the arrow.
PAGE 108
97 Cleavage site "'C G AG U C C 5'UCGG ACUUCG 1111 111111 WT Mut AGCC UGAAGC 5' AU-A A A G-C AG G-C U-A u c-G "'-c A u u A C A u G G A A A C A C ..,._G-C U-A G-C C-G U-A G-C A G A U Figure 38. The sequence and structure of the HP2 hairpin ribozyme The red letters indicate the nucleotides found in the wild-type rhodopsin sequence. The letters in yellow indicate the nucleotide substitutions used to create the inactive ribozyme. The P23H mutation is caused by the nucleotide in the blue box. The cleavage site is indicated by the arrow.
PAGE 109
98 Cleavage site \ C c, G AG U C C WT t' v 5' UCGG ACUUCG Mut t0//"~ 11 I I I I 11 11 5 v~/'/'~~vvvAGCCA AUGAAGC c,~ AG C, C, C, AU-A G-C G-C U U-A \;c-GA U G U A A C A A A U C G A C-+G-C U-A G-C C-G U-A G-C A G A U Figure 39. The sequence and structure of the HP3 hairpin ribozyme. This ribozyme had an extra hairpin between Helix 2 and Helix 3 to aid the folding of the ribozyme into an active conformation. The red letters indicate the nucleotides found in the wild-type rhodopsin sequence. The letters in yellow indicate the nucleotide substitutions used to create the inactive ribozyme. The P23H mutation is caused by the nucleotide in the blue box. The cleavage site is indicated by the arrow.
PAGE 110
99 mixture containing final ribozyme and substrate concentrations of 100 nM each was incubated for sixteen hours at 37 C. The purpose of this reaction was to determine if a ribozyme was able to cleave its substrate, and to ensure that the cleavage product was of the correct size. An example of this reaction is shown in Figure 40 All three hammerhead ribozymes targeting the G90D mutation and the hammerhead ribozyme targeting the VPP mutation were able to correctly cleave their RNA substrates. The three different hairpin ribozymes targeting the VPP mutation were tested, and in each case, the RNA substrate was correctly cleaved Time Course of Ribozyme Cleavage Carried Out Under Substrate Excess Once the ribozymes were shown to correctly cleave their RNA substrates, the rates of three different cleavage reactions were compared for each ribozyme Each reaction was repeated at least three times in order ensure that the results were reproducible and statistically significant. In the first reaction, the ribozyme was incubated with the mutant RNA oligonucleotide In the second reaction, the ribozyme was incubated with the wild-type RNA oligonucleotide. In the third reaction, an RNA oligonucleotide was incubated without the ribozyme to determine the uncatalyzed rate of oligonucleotide cleavage. These reactions were carried out to determine the specificity of the cleavage reaction and to measure the efficiency of the ribozyme cleavage reaction. Specificity of the cleavage reaction was demonstrated by the cleavage of the mutant but not the wild-type RNA oligonucleotide. The efficiency of the cleavage reactions was calculated by comparing the rate of cleavage by the ribozyme with the uncatalyzed rate of RNA oligonucleotide degradation Different ribozymes were compared by examining the time it took each ribozyme to cleave approximately 10% of its substrate. The rate of
PAGE 111
Ribozyme Substrate 100 -+ GOOD G900 WT Figure 40. Cleavage of an RNA oligonucleotide by a ribozyme. An initial cleavage react i on containing 100 nM of ribozyme and 32P-labled RNA oligonucleotide is incubated for 16 hours at 37 Cat standard reaction conditions (40 mM Tris-HCl, pH 7 5, 20 mM MgCh). The reaction was stopped by adding 30 L of RNA gel loading dye to 10 L of reaction. The sample is run on a 15% polyacrylamide 8 M urea gel. Below is a typical autoradiogram of this reaction.
PAGE 112
101 cleavage was quantitated by following the formation of the cleavage product. Cleavage of the substrate by a hammerhead ribozyme resulted in a ?-nucleotide 5' end-labeled product. Cleavage of the oligonucleotide by the hairpin ribozyme led to the formation of a 5-nucleotide 5' end-labeled product. The total amount of product and substrate was also measured at each time point to determine the amount of nonspecific degradation and to ensure that the reaction was quantitative The first ribozyme to be examined was the G90Dl hammerhead ribozyme. G90Dl cleaved an RNA oligonucleotide of the mutant sequence. The amount of product increased linearly for the duration of the experiment, 100 hours, with 10% of the substrate cleaved by 12 hours (Figure 41) G90Dl was unable to cleave an RNA oligonucleotide of the wild-type sequence, showing specificity (Figure 42). Similarly, no cleavage of the mutant substrate was detected in the absence of ribozyme. There also appeared to be little non-specific degradation of the substrate. At 12 hours, there was only 9% nonspecific degradation and only 14% nonspecific degradation over the 100 hours of the experiment (Figure 43). The second ribozyme to be examined was the G90D2 hammerhead ribozyme. This ribozyme was also able to cleave a mutant RNA substrate. The cleavage reaction was linear for the first six hours of the reaction, after which the reaction began to plateau (Figure 44) The activity of this reaction was much higher with approximately 10% of the product cleaved by 4 hours. Substrate continued to be cleaved for the 100-hour duration of the experiment. G90D2, however, also cleaved the RNA oligonucleotide of the wild type sequence (Figure 45) and there was no significant difference in the cleavage of the wild-type and mutant oligonucleotide Again, there was no cleavage of the
PAGE 113
102 100 80 G) 60 a, cu > cu Cl) u 1: 40 Cl) u ... Cl) 20 __._ Mutant Substrate a. -oWild-Type Substrate 0 0 20 40 60 80 100 120 Time(Hours) Figure 41. Cleavage Rate of the G90D 1 hammerhead ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown. Each reaction was repeated a minimum of three times.
PAGE 114
103 Mutant RNA Substrate WIid Type RNA Substrate Tlme(hours} o 1 5 3 5 6 13 24 52100 O 1.5 3.5 6 13 24 52 100 Substrate Product Figure 42 The G90Dl hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence. This figure is a scan of an autoradiogram of a 15% acrylamide 8 M urea gel. The substrate is 13 nucleotides and the product is 7 nucleotides.
PAGE 115
104 110 100 90 C: 80 0 i 70 1J RI 60 ... CJ Q) 0 60 1: Q) 40 u ... Q) a.. 30 20 10 0 0 20 40 60 80 100 120 Time(Hours) Figure 43. Non specific degradation of the G90D target. The sum of the substrate and product at each time point allows the determination of the non-specific degradation during a time-course of the cleavage of substrate by G90D 1.
PAGE 116
105 100~----------------------~ 80 60 40 _.o 20 _._ Mutant Substrate ... o -.. Wild-Type Substrate 0 0 20 40 60 80 100 120 Time(Hours) Figure 44. Cleavage Rate of the G90D2 hammerhead ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown. Each reaction was repeated a minimum of three times.
PAGE 117
Wt Substraie Time (oouns,) 0 .37 75 L:5 3 6 12 24 48 106 Mutant substrate O ;37 .75 L5 3 6 l2 24 48 RNA Substrate Cleavage Pro:lucl Figure 45. The G90D2 hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant and the wild-type sequence. This figure is a scan of an autoradiogram of a 15% acrylamide 8 M urea gel. The substrate is 13 nucleotides and the product is 7 nucleotides.
PAGE 118
107 oligonucleotide in the absence of ribozyme The G90D2 ribozyme was not characterized any further because of its lack of selectivity. The third and final ribozyme targeting the G90D mutation was the G90D3 hammerhead ribozyme. Incubation of the ribozyme with the mutant RNA oligonucleotide resulted in an increase in the formation of product during the 12 hour course of the reaction (Figure 46) This was the fastest of the G90D ribozymes, with 10% of the substrate cleaved in 2 hours. The cleavage reaction was specific and there was no detectable cleavage of the wild-type oligonucleotide (Figure 47). Similarly, there was no degradation of the oligonucleotide in the absence of ribozyme. At 2 hours, there was no detectable nonspecific degradation and only 3% nonspecific degradation over the 12 hours of the experiment (Figure 48) The fourth ribozyme to be examined, HH 1, was a hammerhead ribozyme targeting the VPP transgene. This ribozyme exhibited the lowest catalytic activity of all the ribozymes examined, with 10% of the substrate cleaved by 24 hours (Figure 49). The cleavage reaction appeared to be linear throughout the 52 hour duration of the experiment. This ribozyme was unable to cleave the wild-type oligonucleotide and no degradation of the oligonucleotide was seen in the absence of ribozyme (Figure 50). This ribozyme was not characterized any further due to its low activity. The last ribozymes to be analyzed were the three hairpin ribozymes that targeted the VPP mutation All of these ribozymes specifically cleaved the mutant RNA oligonucleotide without cleaving the wild-type oligonucleotide (Figure 51) There was no degradation of oligonucleotide in the absence of ribozyme. HPl was the slowest of the hairpin ribozymes with 10% of the substrate cleaved in 12 hours (Figure 52). Since, this
PAGE 119
30 25 G> 20 C) ftl > ftl 15 G> u 1: 10 G> u ... G> Q. 5 0 0 2 4 108 _._ Mutant Substrate -oWild-Type Substrate 6 Time(Hours) 8 10 12 14 Figure 46. Cleavage Rate of the G90D3 hammerhead ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are s hown. Each reaction was repeated a minimum of three time s
PAGE 120
Time( hours} Substrate Product 109 Mutant RNA Substrate 0 .75 1.5 3 6 12 WIid Type RNA Substrate 0 .75 1 5 3 6 12 Figure 47. The G90D3 hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence This figure is a scan of an autoradiogram of a 15% acrylamide 8 M urea gel. The substrate is 13 nucleotides and the product is 7 nucleotides.
PAGE 121
110 110 100 90 C 80 0 i; 70 "O (U 60 ... C, CD 0 50 .... C CD 40 u ... CD 0... 30 20 10 0 0 2 4 6 8 10 12 14 Time(Hours) Figure 48. Non specific degradation of the G90D target. The measurement of the sum of the substrate and product at each time point allows the determination of the non-specific degradation during a time-course of the cleavage of substrate by G90D3. )
PAGE 122
111 25 20 Q) 15 Cl Ill > Ill Q) u 10 1: Q) u __._ Mutant Substrate ... Q) 6 -0Wild-Type Substrate Q. 0 0 10 20 30 40 50 60 Time(Hours) Figure 49. Cleavage Rate of the HH 1 hammerhead ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown. Each reaction was repeated a minimum of three times.
PAGE 123
112 Mutant RNA Substrate WIid Type RNA Substrate Time( hours } 0 1 5 3 5 6 13 24 52 100 0 1 .6 3 5 6 13 24 52 100 Substrate Product Figure 50. The HHl hammerhead ribozyme cleaves a 32P-labled RNA oligonucleotide with the mutant but not the wild-type sequence. This figure is a scan of an autoradiogram of a 15% acrylamide 8 M urea gel. The substrate is 13 nucleotides and the product is 7 nucleotides.
PAGE 124
A B C 113 WIid Type RNA Substrate Mutant RNA Substrate Time(hours) O 1.5 3.5 6 13 24 52 100 0 1.5 3.5 6 13 24 52 100 Substrate Product Tim e(hours) Substrate Product Tim e(hours) SubstratG Product 4 --..... Mutant RNA Substrate WIid Type RNA Substrate /1 7r:, 1 <; .1 R 1 ;, ;,4 4A (l 7<; 1 C, :~ R 1;, ;,4 4A -" -----...... --Mutant RNA Substrate WIid Type RNA Substrate 0 .75 1.5 J 6 1 2 24 48 0 .75 1.5 J 6 12 24 48 ...... ------, ----Figure 51. Cleavage of the VPP oligonucleotide by three different VPP hairpin ribozymes. The hairpin ribozymes specifically cleaves the mutant VPP but not the wild type oligonucleotide A) Cleavage time course of HP 1. B) Cleavage time course ofHP2. C) Cleavage time course ofHP3. These figures are scans of autoradiograms of a 15% acrylamide 8 M urea gels. The substrate is 13 and the product is 6 nucleotides.
PAGE 125
114 35 30 25 G> c:, 20 tU > tU G> u 15 1: G> u 10 .... _._ Mutant Substrate G> Q. --0Wild-Type Substrate 5 0 .s 0 10 20 30 40 50 60 Time(Hours) Figure 52. Cleavage Rate of the HP 1 hairpin ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown Each reaction was repeated a minimum of three times.
PAGE 126
115 ribozyme had a very low catalytic activity it was not tested further. HP2 cleaved 10% substrate in 5 hours (Figure 53). The reaction appeared to be linear for the first 10 hours of the experiment. Substrate continued to be cleaved for the entire 52 hours of the experiment. The final ribozyme, HP3, cleaved 10% of its substrate in 3 hours and was linear for the first 6 hours of the reaction (Figure 54 ). Product formation occurred during the entire 48 hours of the cleavage reaction. HP2 and HP3 were chosen for further characterization There was no detectable nonspecific degradation at 3 hours for reactions using the HP2 or HP3 ribozyme and no more than 9% nonspecific degradation overall (Figures 55, 56) Kinetic Analysis The time course reaction was also used to determine the interval needed for the determination of the kinetic parameters of the ribozyme. In order to ensure cleavage of RNA substrate in vivo, it is important to design ribozymes with the highest possible catalytic activity. This requires the determination of the turnover number (kcat) and Michaelis constant (KM) of the ribozymes. These parameters are determined by experimental methods develop by Michaelis and Menten (190). These methods rely on a number of assumptions The first is that the initial rate of the reaction is measured Measuring the initial rate of the reaction ensures that changes in the formation of product and depletion of substrate do not affect the rate of the reaction. By convention, kinetic measurements are made when no more than 15% of the substrate has been converted to product. For this project, the reaction time chosen for the multiple turnover kinetic reactions was the time point at which approximately 10% of the substrate had been cleaved. A second premise of the Michaelis-Menten analysis is that the concentration of
PAGE 127
G> c:, 60 40 tU G> u 1: t 20 ... G> 0... 0 116 __._ Mutant Substrate -oWild-Type Substrate -20 ...___--~----.-------.----..-----.----r------1 0 10 20 30 40 50 60 Time(Hours) Figure 53. Cleavage Rate of the HP2 hairpin ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown. Each reaction was repeated a minimum of three times. ; :
PAGE 128
117 100 80 G) 60 Cl as > m G) 0 40 1: -----Mutant Substrate G) u ---oWIid-Type substrate ... G) 20 Q.. 0 -20 0 10 20 30 40 50 60 Time(Hours) Figure 54. Cleavage Rate of the HP3 hairpin ribozyme. The time courses of the cleavage reaction for both the mutant and wild type oligonucleotide are shown. Each reaction was repeated a minimum of three times.
PAGE 129
118 110 100 .........____. 90 8 80 j 70 1111 bi 60 50 ... i 40 u .... CD a.. 30 20 10 0 0 2 4 6 8 10 12 14 Time(Hours) Figure 55. Non specific degradation of the VPP target. The measurement of the sum of the substrate and product at each time point allows the determination of the non-specific degradation during a time-course of the cleavage of substrate by HP2.
PAGE 130
119 110 100 90 C 80 0 i 70 ,, IQ 60 ... OJ 4) 0 60 1: 4) 40 u ... 4) Q. 30 20 10 0 0 10 20 30 40 50 60 Time(Hours) Figure 56 Non specific degradation of the VPP target. The measurement of the sum of the substrate and product at each time point allows the determination of the non-specific degradation during a time-course of the cleavage of substrate by HP3.
PAGE 131
120 the ribozyme is lower than the concentration of substrate. A low ratio of ribozyme to target ensures that the velocity of the reaction is proportional to the concentration of the ribozyme This allows the turnover rate of the reaction to be determined Multiple turnover kinetic experiments were performed by preparing reactions with 10 nM of transcribed ribozyme and increasing concentrations of the substrate The catalytic parameters were determined using at least eight reactions at different substrate concentrations. In addition, each experiment was performed in triplicate The initial reaction rate (V 0) was calculated by dividing the concentration of reaction product by reaction time Eadiee-Hofstee analysis was used to calculate VMax and KM by performing linear regression analysis of 1/ VO versus VO /[S]. Eadiee-Hofstee analysis was used instead of other linear regression methods because it places less weight on the reactions containing lower ratios of substrate to ribozyme, which are less accurate than reactions with higher ratios of substrate to ribozyme. In Eadiee-Hofstee analysis V Max is equal to the y intercept, and the slope of the line gives the value of KM. KM is the concentration of substrate at which the reaction rate is at half of its maximum value. V Max can be divided by the concentration of ribozyme to give kcat, the turnover number for the reaction The catalytic efficiency of a ribozyme can be determined dividing kcat by KM (190) For the G90D 1 ribozyme, the reaction rate was measured 12 hour after the reaction was started. Product was first detected with the incubation of ribozyme and 80 nM substrate (Figure 57). The ribozyme was saturated at a substrate concentration of approximately 160 nM. The G90D 1 ribozyme was found to have a KM of 24 nM and a kcat of 1.33 x 10-3/min. The efficiency of this ribozyme, kcat1 KM, was 5 5 x 104 min-1 M-1
PAGE 132
121 A No Rlbozyme Co ntro l Rl bozym e R e ac tio n S ubstrate concentration 40 60 so 100 2 0 40 1so 1 80 40 60 80 oo 1 20 1 40 1 60 ,so (n M) S u b s trate Ban d P ro duct Band B 0.0110 0.0105 0.0100 c D.CXS6 i O.OCIIO :I: .s 0.0086 0.0080 0.0075 0.0070 +----~--~--~-----~--___, 0.00012 0.00014 0.00016 0.00018 0.00020 0.00022 0.00024 Vol[S](mln"1 ) Figure 57. G90Dl hammerhead ribozyme multiple turnover cleavage reaction. A) An image of the autoradiogram of the G90Dl multiple turnover cleavage reaction. B) A graph of the Eadiee-Hofstee plot used to calculate the kinetic coefficients of the G90Dl ribozyme. These experiments were performed in triplicate.
PAGE 133
122 For the G90D3 ribozyme, a 2 hour reaction time was used to determine the kinetic constants of the ribozyme With this ribozyme product was first detected with a reaction at 80 nM of s ubstrate, and saturation was achieved at 160 nM (Figure 58). Analysis from the Eadiee-Hofstee plot indicated that the ribozyme had a KM of 18 nM a kca1of 2.5 x 10-2 /min. The efficiency of this ribozyme, kcau KM, was 1.4 x 106 min-1 M-1 For the HP2 ribozyme, the kinetic reactions were analyzed 5 hours after starting the reaction Product was first detected when the ribozyme wa s incubated with 80 nM of substrate (Figure 59) Saturation occurred at a substrate concentration of 140 nM. The HP2 ribozyme has a KM of 45 nM and a kca t of 1.8 x 10-2/min Similar values were obtained at 4 and 5 hours indicating that the reaction was in the linear velocity range. The efficiency of this ribozyme, kca u KM, was 4 x 105 min-1 M-1 The final ribozyme, HP3, was analyzed 3 hours after the start of the reaction. Product was first detected with incubation of ribozyme and 80 nM of substrate (Figure 60). Analysis indicated that the ribozyme had a KM of 300 nM a kcat of 0.29/min. The efficiency of this ribozyme, kcau KM, was 9 x 106 min-1 M-1 The catalytic parameters of HP3 are similar to the catalytic parameters of the hairpin ribozyme derived from the Tobacco Ring Spot virus, which has a KM of 34 nM and a kcat of 0.3/min. The efficiency of that ribozyme kcau KM, was 9 x 107 min-1 M-1 in a bi-molecular reaction (219). The Dependence of the Ribozyme Cleavage Reaction on Magnesium Concentration Experiments were carried out to determine the dependence of each ribozyme on the concentration of Mg2+. There were two objectives for these experiments. First, Mg2+ is the main ionic contributor to ribozyme activity within cells (195) The free intracellular
PAGE 134
A B S u bstrate conce ntration (nM) S ubstrate Band Prod u c t B a n d 0.18 0.16 0.14 c E :E 0.12 C -1; 0.10 0.08 123 No Rlbozy m e Cont r o l Rlb ozyme Reaction 40 60 80 100 2 0 40 160 40 60 80 00 1 .20 140 6 0 0.06 -i-----.----.------.------.-----.----,---------1 0.0004 0.0005 0.0006 0.0007 0.0008 0.0009 0.0010 0.0011 VoJ[S](min1 ) Figure 58. G90D3 hammerhead ribozyme multiple turnover cleavage reaction. A) An image of the autoradiogram of the G90D3 multiple turnover cleavage reaction. B) A graph of the Eadiee-Hofstee plot used to calculate the kinetic coefficients of the G90D3 ribozyme. These experiments were performed in triplicate.
PAGE 135
124 A No Rlbozy m e Contro l Rlbo zym e Reac tio n S u bstr a te concentratio n 40 60 80 100 2 0 40 160 180 40 60 80 100120140 so 8 0 (nM) S u bstrat e Ba n d Prod uct Ba n d B 0.056 0.060 0.045 c 0.040 I en G> 0.035 15 ::E C J 0.0~ 0.025 0.0~ 0.015 0.00040 0.00045 0.00050 0.00055 0.00060 0.00065 0.00070 Vol[S] Figure 59. HP2 hairpin ribozyme multiple turnover cleavage reaction. A) An image of the autoradiogram of the HP2 multiple turnover cleavage reaction. B) A graph of the Eadiee Hofstee plot used to calculate the kinetic coefficients of the HP2 ribozyme. These e x periments were performed in triplicate
PAGE 136
A B S u bstrate Concentratio n (nM) S u bstrate Ban d P roduct B and 0.7 0.6 0.5 C 0.4 E 7n 0.3 t) c5 :E 0.2 C 0 1 0.0 .Q.1 -0.2 0.0014 125 No RH:rnzyme Control Rlbozyme ReacUon 80 100 120 140 HID 180 80 100 120 1-40 160 180 0.0016 0.0018 VoJ(S] 0.0020 0.0022 Figure 60. HP3 hairpin ribozyme multiple turnover cleavage reaction. A) An image of the autoradiogram of the HP3 multiple turno ver cleavage reaction. B) A graph of the Eadiee Hofstee plot used to calculate the kinetic coefficients of the HP3 ribozyme. These experiments were performed in triplicate.
PAGE 137
126 concentration of Mg2+, though, is only 0.5mM-lmM (220). The first goal of these experiments was to determine whether a given ribozyme would be able to cleave substrate RNA at low concentrations of Mg2+ The second goal was to examine ribozyme modifications that decrease the magnesium dependence in order to determine those changes that may increase the ability of the ribozyme to fold into an active conformation. The experiments to determine the magnesium dependence of a ribozyme were carried out in a manner very similar to those used to determine the kinetic parameters for the ribozyme. The same reaction times used for the kinetic reactions were used for these experiments. The concentration of magnesium required to achieve half of the maximal velocity of the reaction is referred to as KMg. The determination of KMg allowed for the comparison of the magnesium dependence of different ribozymes. The value of KMg was calculated using nonlinear regression of the plot of the initial rate as a function of ion concentration Experiments to determine the Mg2+ dependence of the G90D 1 ribozyme were allowed to react for 12 hours. The formation of cleavage product was detected at all concentrations of Mg2+, including 1 mM Mg2+ at which 0.18% of the substrate was cleaved (Figure 61) The KMg for the G90Dl ribozyme was 25 mM Mg2+ Reactions using the G90D3 ribozyme were allowed to incubate for 12 hours. At a concentration of 1 mM Mg2+ 0 25% of the substrate was cleaved (Figure 62). The KMg for the G90D3 ribozyme was 23 mM Mg2+. However, maximum velocity was not achieved even at 160 mM Mg2+ Experiments using HP2 were allowed to react for 5 hours. At 1 mM Mg2+ 0.06% of the substrate was cleaved (Figure 63). The KMg for the HP2 ribozyme was 28 mM Mg.
PAGE 138
127 A Mg Coricentratlo n (11M) 1 5 1 0 2 0 4 0 80 160 S u b stra t e Ba n d 1 P ro d uct B an d B 60 60 G> 40 ICD lU 30 G> u 20 G> KMg= 25 mM u ... G> Q.. 10 0 0 20 40 60 80 100 120 1) 160 180 Concentration of Mg2+ Figure 61. The magnesium dependence of the G90Dl ribozyme. A) An autoradiogram showing the effects of increasing concentrations of magnesium on the cleavage of an RNA oligonucleotide b y the G90Dl ribozyme. B) A graph used to determine the KMg of the G90D 1 ribozyme by nonlinear regression.
PAGE 139
A B G> Cl G> u 1: G> u ... G> 0.. 128 M g Co ncen tr a ti o n ( n M) 1 5 1 0 2 0 40 8 0 160 S u b stra t e Ba n d P r o d uct B a n d 40-.------------------------, 30 20 10 0 KMg= 29mM 0 20 40 60 80 100 120 140 1a) 1EK> Concertratlon of Mrt'+ (mM) Figure 62. The magnesimn dependence of the G90D3 ribozyme. A) An autoradiogram showing the effects of increasing concentrations of magnesimn on the clea v age of an RNA oligonucleotide by the G90D3 ribozyme. B) A graph used to determine the KMg of the G90D3 ribozyme by nonlinear regression
PAGE 140
A B 35 30 25 Cl) CJ al 20 ::,, al Cl) u 15 c Cl) u 10 ... Cl) Q.. 5 0 129 Mg Concent ration ( n M) o.s 1 5 1 o 2 0 40 a o 160 suos tr a te Band Product Band KMg= 26 mM 0 20 40 60 80 100 120 1.l 160 1EK> Concentration of Mg2+ (mM) Figure 63. The magnesium dependence of the HP2 hairpin ribozyme. A) An autoradiogram showing the effects of increa sing concentrations of magnesium on the cleavage of the VPP RNA oligonucleotide by the HP2 ribozyme. B) A graph used to determine the KMg of the HP2 ribozyme by nonlinear regression.
PAGE 141
130 The final ribozyme tested, HP3, was allowed to react for 3 hours. At a lmM concentration of Mg2+ 1.0% of the substrate was cleaved (Figure 64) The maximum velocity of the cleavage reaction catalyzed by HP3 is higher than the maximum velocity of the other ribozymes. The increased reaction velocity is reflected in the KMg value of 17 mM. HP3 has the lowest KMg value of the four ribozymes tested Cleavage of an Oligonucleotide with the Mutant mRNA Sequence in the Presence of Increasing Amounts of Oligonucleotide with the Wild Type Sequence A series of six cleavage reactions were prepared such that the concentration of ribozyme and mutant substrate were kept constant, while increasing amounts of an oligonucleotide with the wild type rhodopsin sequence were added. In these experiments, 10 nM of ribozyme was incubated with 50 nM of mutant RNA oligonucleotide The six reactions had a wild-type RNA oligonucleotide concentration of O nM, 50 nM, 100 nM, 150 nM, 200 nM, and 250nM. The reactions were incubated the same length of time to allow approximately 10% of the substrate to be cleaved. G90Dl was incubated for 12 hours, G90D3 was incubated 2 hours, HP2 was incubated 5 hours, and HP3 was incubated 3 hours. The reactions were quantitated by finding the percentage of mutant oligonucleotide cleaved for each reaction. For each of the ribozymes, although there was some variation in the reactions, approximately the same amount of mutant oligonucleotide was cleaved in all of the reactions (Figures 65, 66, 67, 68). Thus, there was no inhibition of the cleavage of the mutant oligonucleotide by the wild type oligonucleotide even at a 1 :5 ratio of mutant to wild-type RNA.
PAGE 142
A B 0.7 0.6 0.6 t) Cl 0.4 :> ., u 0.3 1: t) u 02 ... t) 0... 0.1 0.0 131 Mg Concentration (nM) 0 5 1 5 1 o 20 40 80 160 Substrate Band Product Band ~g=17mM 0 20 40 60 80 100 120 140 100 1~ Concertration of Mg 2+ (mM) Figure 64. The magnesium dependence of the HP3 hairpin ribozyme. A) An autoradiogram showing the effects of increasing concentrations of magnesium on the cleavage of the VPP RNA oligonucleotide by the HP3 ribozyme. B) A graph used to determine the KMg of the HP3 ribozyme by nonlinear regression.
PAGE 143
A B Ratio of .\lulant lo \\,.ild T, Substrate Substrate Hand Proch1<'t Rand 16 14 12 & 10 Ill > Ill u 8 "t: u 6 ... 11.. 4 2 0 -60 0 132 1:0 1:2 I :.J I: I I:.\ I:~ 60 100 160 200 260 300 [Wt SUbstrate](nM) Figure 65. The G90Dl cleavage reaction is not effected by the presence of an RNA oligonucleotide with the wild-type sequence. A) An autoradiogram of the G90Dl cleavage reaction. B) A graph showing the effect of increasing wild-type RNA oligonucleotide on the cleavage of a mutant RNA oligonucleotide by the G90Dl ribozyme.
PAGE 144
A B GI DI Ill :> GI 0 GI u ... GI a.. Ratio of ~ lutanl lo \\,.ild TJpc SuhslrJle Suhst rate Uaml Product Band 16 14 12 10 8 6 4 2 0 0 60 133 I :O I' I :.J I: I I:. \ I:, 100 160 200 260 300 [Wt Substra.te](nM) Figure 66. The cleavage reaction catalyzed by the G90D3 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence. A) An autoradiogram of the G90D3 cleavage reaction. B) A graph showing the effect of increasing wild-type RNA oligonucleotide on the cleavage of a mutant RNA oligonucleotide by the G90D3 ribozyme.
PAGE 145
A B Ratio of \.folant lo Wild T) pc Substratt' Sub-;t rate Rand Product Band 134 1:1.1 1:2 1 : 4 I :I 1: l 1 :i-5~------------------------, 3 2 1 0+----'-...---,.__-'-...---L--...,__...---....__.__ .......... .___.__ .......... .___.__,.........._----i -60 a 60 100 160 200 260 300 [Wt Substrate](nM) Figure 67 The cleavage reaction catalyzed by the HP2 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence A) An autoradiogram of the HP2 cleavage reaction B) A graph showing the effect of increasing wild type RNA oligonucleotide on the cleavage of a mutant RNA oligonucleotide by the HP2 ribozyme
PAGE 146
A B 26 CD 2Q Ill C> u 16 1: C> u ... 3?,. 10 5 Ratio of .\ Jutanl to \\,.ild 1) pc SubslrJk Substrate Hanel Produd Hand 135 I :0 1 -1' I :4 1.6 I :I I:.\ l:S -~ ... 0 +--........ --'----'---r--'--'--......_..__,r---"-___. _.__,_-r'-......_...__ ......... _--t 0 100 200 300 400 [Wild-fype](nM) Figure 68 The cleavage reaction catalyzed by the HP3 ribozyme is not effected by the presence of an RNA oligonucleotide with the wild type sequence A) An autoradiogram of the HP3 cleavage reaction B) A graph showing the effect of increasing wild-type RNA oligonucleotide on the cleavage of a mutant RNA oligonucleotide by the HP3 ribozyme
PAGE 147
136 Cleavage of Retinal RNA by a Ribozyme Many factors can affect the cleavage of substrate RNA by a ribozyme in vivo The binding of mRNA by an antisense oligonucleotide is affected by secondary and tertiary structure of the mRNA (221). In addition, the mRNA may be bound by proteins. The ability of a ribozyme to cleave its substrate is determined by the accessibility of the region of the mRNA to which the ribozyme hybridizes. Substrate target sites that are obscured by secondary structure are poorly cleaved (203;222) In addition, it is not known if the presence of cellular RNA interferes with the ability of a ribozyme to cleave its mRNA substrate. For this reason, cleavage reactions were carried out with each ribozyme and mRNA extracted from the retina of the appropriate mouse model. To examine the G90Dl ribozyme RNA was extracted from the retina of the 6086 G90D mouse line created by Dr. Muna Naash. This mouse line has a 1: 1 ratio of the mutant transgene to wild-type rhodopsin Reactions were prepared in triplicate containing 4 g of retinal extract and 600 nM ribozyme in a 40 L reaction volume. The reactions were incubated at 37C in the presence of 20 mM Mg2+. Aliquots of 10 L were removed at 0, 12, 24, and 48 hours. The cleavage of mRNA was quantitated using an RTPCR assay. The presence of an Ncol site in the wild-type but not the mutant RTPCR fragment allowed quantitation of the two forms of rhodopsin mRNA. Incubation of the G90D 1 ribozyme appeared to result in a significant loss of rhodopsin transgene mRNA. The amount of wild-type rhodopsin mRNA remained essentially constant (Figure 69) Over the 48 hour duration of the experiment approximately 66% of the mutant transgene was specifically degraded. Interestingly,
PAGE 148
137 A lime 0 6 12 24 48 ONA (hours) standard ~. .. i 8-Adin w I ... -.G'100 ...... ... Wil d-type w ecommon tf Wild-type B 0 7 0 6 0 5 2 iii 0:: !: 0 4 "' a. 0 "O 0 3 0 .c. 0:: 0 2 0.1 0 10 20 30 40 50 60 Time(Hours) Figure 69. Cleavage of the G90D mRNA by the G90Dl hammerhead ribozyme. A) An autoradiogram of an 6% non-denaturing polyacrylamide gel of an RT-PCR assay of the cleavage ofG90D by the G90Dl ribozyme. B) A graph comparing the amount ofG90D and wild-type rhodopsin mRNA after incubation with ribozyme for 0, 6 12, 24, 48 hours The time point at 12 hours had a high background and could not be quantitated.
PAGE 149
138 there was no degradation of the G90D transgene in the first six hours of the experiment. This may reflect the time it takes the ribozyme and G90D substrate to form a catalytically active complex. RNA was also extracted from the retina of the VPP mouse created by Dr. Muna Naash. This mouse line has a 1: 1 ratio of the mutant VPP transgene to wild-type rhodopsin Reactions were prepared in triplicate containing 6 g of retinal extract and 600 nM of HP2 or HP3 in a 60 L reaction volume. The reactions were incubated at 37C in the presence of 20 mM Mg2+ Aliquots of 10 L were removed at 0,12, 24, and 48 hours. The cleavage of mRNA was quantitated as previously described. As with the G90D ribozyme, incubation of HP3 ribozyme with the RNA retinal extract from the VPP mouse again resulted in no significant decrease in the transgene mRNA levels. Similarly, the amount of the wild-type rhodopsin mRNA remained constant throughout the course of the experiment (Figure 70).
PAGE 150
139 A Time 0 12 24 48 DNA (hours) standard --- -...... t B-Adin --~- VPP --~Wild-type ........ Common I t .... ....... Wild-type : '"' ; j #. t:r i B 2 01 --------------~----1 8 I 1 6 I 1 4 1.2 0 1 0 ... C iii 0. 0 8 0 "8 0 6 .s:::. 0::: 0.4 0 2 0 0 0 10 20 30 40 50 60 Time(Hour) Figure 70. Cleavage of the P23H transgene mRNA by the HP3 hairpin ribozyme. A) An autoradiogram of an 6% non-denaturing polyacrylarnide gel of an RT-PCR assay of the cleavage of P23H by the HP3 ribozyme. B) A graph comparing the amount of P23H and wild-type rhodopsin mRNA after incubation with ribozyme for 0, 12 24, 48 hours.
PAGE 151
CHAPTER4 DISCUSSION The overall goal of this experiment was to examine the therapeutic potential of ribozymes. The hypothesis underlying this project was that ribozymes could inhibit the expression of a pathogenic gene and delay or prevent disease. Autosomal dominant genetic diseases are a good target for ribozyme therapy since these disorders are caused by the expression of a single mutant gene. For this project several ribozymes were designed to target the mRNA of mutant rhodopsin protiens responsible for Autosomal Dominant Retinitis Pigmentosa. In vitro assays were developed to compare the efficiency of each ribozyme. The most efficient of these ribozymes were cloned into a recombinant AA V vector for in vivo analysis in animal models of retinal disease. The results of these animal experiments will determine the therapeutic potential of ribozymes for the treatment of ADRP. The animal models used in this project are also ideal for future experiments that examine ways to improve the activity of ribozymes in vivo. Autosomal Dominant Genetic Diseases Are Amenable to Ribozyme Gene Therapy As previously mentioned, a majority of ribozymes have been designed to target the mRNA of genes associated with cancer or AIDS. There are over 160 anticancer ribozymes and over 600 antiviral ribozymes. Yet, after a decade of research, there are no reports of the successful use of ribozymes to treat cancer or AIDS in an animal model. Autosomal dominant genetic disorders are more amenable to treatment using ribozymes. These diseases are often caused by the product of a single mutant allele that interferes 140
PAGE 152
141 with the function of the protein encoded by the wild-type allele. A reduction in the expression of the mutant allele would be expected to prevent or ameliorate the disease. In fact, one of the first ribozymes designed to treat an autosomal dominant genetic disease was also the first ribozyme shown effective in the treatment of a disease in an animal model (86) This ribozyme was designed for the treatment of ADRP caused by the P23H mutation However, these studies were carried out in a transgenic rat model of the disease. One of the problems with many of the animal models for ADRP caused by rhodopsin mutations, is that rhodopsin is often over-expressed. The over-expression of rhodopsin is toxic to photoreceptor cells (108). This makes it difficult to determine the etiology for the degeneration that results from a mutation in rhodopsin. A rhodopsin knockout mouse was created, and crossing a transgenic mouse with knockout mouse can produce a transgenic mouse with an exact copy number of the normal and transgenic rhodopsin gene (90). The mouse model of ADRP, therefore, is better than the rat and other models of ADRP since retinal disease is due entirely to a mutant gene instead of the over-expression of rhodopsin. As described in the experimental aims, the overall goal of this project was to examine the ability of ribozymes to treat autosomal dominant genetic diseases. More specifically, ribozymes were designed to cleave the mRNA of two different transgenes with mutations that lead to retinal dysfunction in mouse models of ADRP Then, in vitro assays were developed to identify the ribozymes that had the greatest chance for success in vivo. The ribozymes found to be the most efficient using these assays were then cloned into a recombinant AA V vector so they could be studied in animal models
PAGE 153
142 Design of Ribozymes for the Treatment of ADRP For this project, ribozymes were created to cleave the mRNA of a transgene used to create mouse models for retinal disease due to rhodopsin G90D and P23H mutations. The hammerhead and hairpin ribozymes were used for these experiments. These ribozymes have the ability to catalyze an RNA cleavage reaction in trans (17;72). The hammerhead and hairpin ribozyme are also the ribozymes that have been the most thoroughly studied for use in gene therapy. Two different trans genes were used to create mouse models of the disease caused by the rhodopsin G90D mutation The first transgene was created by Dr.Muna Naash, and differs from the wild-type gene by six nucleotides (Figure 18) (Naash et al. unpublished data). The GA to AT mutation in codon 90 creates the G90D mutation The additional mutations were created to allow the differentiation of the transgene and wild-type rhodopsin gene The hammerhead ribozymes, G90D 1 and G90D2, were designed to specifically cleave the Naash transgene but not the wild-type RNA sequence. The specificity of the G90D 1 ribozyme is due to the GA to AT mutation that creates an A UU cleavage triplet in the transgene mRNA (Figure 33) The G90D2 ribozyme recognizes a UUC cleavage triplet that is found in mRNA of both the transgene and the wild-type gene (Figure 34). The UUC triplet is cleaved 22 times more efficiently than the AUU triplet, and it was thought that the G90D ribozyme would have a higher catalytic activity. Specificity in the case of this ribozyme, again depended on the GA to AT mutation. The A and T mutation are the most distal nucleotides in helix I of G90D2. It was thought that the GA mismatch in helix I due to the hybridization of the G90D2 ribozyme and the mRNA of the wild-type gene might prevent cleavage
PAGE 154
143 The second G90D transgene was created by Dr. Paul Sieving (Figure 17) (Sieving et al. unpublished data). The sequence of the Sieving trans gene is closer to the human G90D sequence than the Naash transgene. The sieving transgene has a GA to AC mutation in codon 90 that creates the G90D mutation. This mutation is adjacent to a UUC cleavage triplet. A hammerhead ribozyme, G90D3, was designed to cleave this mutant transgene (Figure 35). The specificity of the cleavage reaction depended on the mismatch in helix ill formed when the ribozyme binds to the wild-type rhodopsin RNA. Dr. Naash also created a transgenic mouse model for the form of ADRP caused by the P23H mutation (Figure 17) (109) This transgene, termed the VPP transgene, differs from the wild-type rhodopsin gene by six nucleotides. A C to A conversion in codon 23 was responsible for the P23H mutation. A C to T conversion in codon 22 created the ideal GUC cleavage triplet for the hammerhead and hairpin ribozymes. Single nucleotide mutations in codon 20 and 27 result in amino acid substitutions. The VPP transgene also has an RFLP to allow differentiation of the transgene and wild-type rhodopsin gene. A hammerhead, HHl, and three different hairpin ribozymes, HPl, HP2, and HP3 were designed to target this transgene (Figures 36, 37, 38, 39). The specificity of the cleavage reaction catalyzed by these ribozymes was due primarily to the GUC cleavage triplet found in the VPP transgene but not the wild-type rhodopsin gene. The hammerhead and hairpin ribozymes were both designed against this transgene since the ribozymes cleave different substrates with different catalytic efficiencies There are reports that some substrates are cleaved more efficiently by hammerhead ribozymes, and similar reports of substrates that are cleaved better by hairpin ribozymes (217) The three
PAGE 155
144 different hairpin ribozymes were all slightly different. HPl is the hairpin ribozyme found in the negative strand of the Tobacco Ringspot Virus (72) HP2 had a helix 4 with three additional base pairs and a stable stem loop. These modifications stabilize the secondary structure of the hairpin ribozyme and increase its catalytic activity The third hairpin ribozyme, HP3, also had the modifications found in HP2. HP3 also had a stem loop between helix 2 and helix 3 This stem loop was reported to increase the ability of the hairpin ribozyme to fold into its proper conformation (74). Analysis of the Catalytic Efficiency of the Ribozymes In Vitro Perhaps the most important result of this research was the development of the methodology needed to determine the efficiency of a particular ribozyme. There are a number of factors that can prevent or inhibit the cleavage of a mRNA substrate by a ribozyme. The sequence of the RNA substrate recognized by the hybridizing arms of the ribozyme can effect the rate of cleavage of the substrate by the ribozyme (188; 189;216) In addition, the catalytic activity of a ribozyme might be effected by the secondary or tertiary structure of the RNA substrate, or by the binding of a protein to the substrate (203;222). The possibility that an mRNA molecule will be cleaved by a ribozyme can be increased by designing a number of different ribozymes to cleave a mRNA, or by modifying a ribozyme to increase its efficiency. A methodology was needed, therefore, in order to compare the efficiencies of different ribozyme in vitro to determine the ribozymes that had the best chance of success as a gene therapy. Furthermore, a comparison of the effectiveness of a ribozyme in vitro and in vivo can be used to determine the requirements for catalytic activity of the ribozyme in cells
PAGE 156
145 The first step in designing the in vitro ribozyme assays was to determine the conditions to be used for the cleavage reactions. There were two important criteria to consider. First since the ribozymes were being designed for use in vivo it was important that the cleavage reaction simulate in vivo conditions. Second, in order to compare the ribozymes used in this s tudy with ribozymes examined in previous experiments, the conditions should kept similar to the conditions used by a majority of the researchers in the field. For the ribozymes in this project, the standard reaction was carried out in 40 mM Tris-HCl, pH 7.6, 20 mM MgCli In addition, all ribozyme reactions were carried out at 37C In initial cleavage reaction 100 nM of ribozyme and substrate were allowed to incubate for 16 hours under standard conditions All of the ribozymes were able to correctly cleave their substrate. In the next set of experiments, a time course of the cleavage reaction was performed using ribozyme and an RNA oligonucleotide with the transgene or wild-type sequence. The purpose of these time courses was to determine the specificity of the ribozyme cleavage reaction, and to get an estimation of the efficiency of the ribozymes needed for the next set of experiments. The G90D 1 and G90D3 hammerhead ribozymes was able to specifically cleave a RNA oligonucleotide containing the G90D mutation. The G90D2 ribozyme, though, cleaved the transgene and wild-type sequence with the same efficiency. These results demonstrate that two strategies for ensuring the specificity of the hammerhead cleavage reaction can be effective. One strategy that can be used to ensure the cleavage of the mutant but not the wild-type RNA substrate is to design a ribozyme that targets a unique cleavage triplet found only in the mRNA of a mutant gene A second strategy that can be used to ensure cleavage of the mutant but not the wild-type
PAGE 157
146 RNA substrate is to design a ribozyme that targets a unique sequence found only in the mRNA of the mutant gene Hybridization of the ribozyme with the wild-type mRNA will create mismatches in the arms of the hammerhead ribozyme As shown in this project, however, a hammerhead ribozyme can still cleave an RNA substrate with mismatches in the terminal nucleotides of the hybridizing arms. Detailed analysis showed that mismatches in the terminal nucleotides of the hybridizing arms of the hammerhead ribozyme could not ensure the specificity of the cleavage reaction (223;224) The more distal mismatches are tolerated since these nucleotides contribute only to the binding of the ribozyme and substrate (225). The base pairs adjacent to the catalytic core of the hammerhead ribozyme contribute to the chemical activity of the ribozyme. These base pairs are required for the formation of the active conformation of the ribozyme. To prevent cleavage mismatched base pairs need to be within four nucleotides of the catalytic core. As previously mentioned, the G90Dl hammerhead ribozyme recognizes an AUU cleavage triplet. The G90D2 and G90D3 ribozymes targeted a RNA substrate with a UUC cleavage triplet. The UUC triplet is cleaved 22 times more efficiently than the AUU triplet recognized by the G90D 1 ribozyme. A comparison of the time it took each ribozyme to cleave 10% of its substrate under similar conditions, showed that the G90D2 ribozyme was only three times more efficient than the G90D 1 ribozyme (Figure 71 ). Similarly, G90D3 was only four times more efficient than the G90Dl ribozyme. The fact that G90D2 and G90D3 were not 22 times more efficient than G90D 1 shows the impact
PAGE 158
Q) O> cu > cu Q) 0 C Q) 0 .... Q) a.. 100 80 60 40 20 0 0 ___.,.._ G9001 O G9002 ----Y-G9003 147 0 20 40 60 80 100 120 Time(Hours) Figure 71. A comparison of the cleavage reaction of the different G90D ribozymes.
PAGE 159
148 the sequence of the hybridizing arms have on the cleavage reaction. G90D 1 and G90D3 were chosen for further characterization because of there specificity. All of the ribozymes targeting the VPP transgene were able to specifically cleave the mutant oligonucleotide. An approximation of the effectiveness of each ribozyme was gathered by comparing the times required to cleave 10% of their substrates under similar conditions (Figure 72) HHl cleaved 10% of its substrate in 24 hours and HPl cleaved 10% of its substrate in 12 hours showing that HPl was twice as efficient as HHl. Similarly, a comparison of the cleavage of substrate by the naturally occurring hammerhead and hairpin ribozymes derived from the Tobacco Ringspot Virus showed that the hairpin ribozyme was twice as efficient as the hammerhead ribozyme (189). HPl was inefficient compared to HP2 and HP3 which were 3 and 4 times more active. Therefore, only the modified hairpin ribozymes were chosen along with G90Dl, G90D3, for further characterization. Following the time course reactions, the ribozymes were incubated with increasing concentrations of the RNA substrate. The lowest concentration of substrate was at least five times the concentration of ribozyme. These multiple turnover reactions were used to determine the overall rate and concentration dependence of the ribozyme cleavage reaction Additional reactions in vitro determined that these ribozymes possessed the catalytic coefficients needed for activity in vivo (Table 1). The determination of these coefficients allows the comparison of the ribozymes used in this project with those in the literature. This allows the most efficient ribozymes to be selected for in vivo analysis. Therapeutic ribozymes have catalytic efficiencies reported to be between 10-5-10-7 min-1 M -1 Of the four ribozymes characterized in this project,
PAGE 160
80 60 Q) C) ro > 40 ro Q) 0 C: Q) u 20 .... Q) a.. 0 0 ---HH1 o HP1 ----THP2 ----'v HP3 / I / I / .. J' /1 .1/ 10 149 ---v -v----~ / -----/ 0 20 30 40 50 Time(Hours) Figure 72. A comparison of the cleavage by VPP ribozymes. 60
PAGE 161
150 Table 1. The catalytic coefficients of the G90D and VPP ribozymes. Ribozyme kcat(min-1 ) KM(nM) kcai!KM KMg (mM) (nM-1 min-I) G90Dl 1.33 X 10-3 24 5.5 X 104 25 G90D3 2.5 X 10-2 18 1.4 X 106 23 HP2 1.8 X 10-2 45 4 X 105 28 HP3 0.29 300 9 X 106 17
PAGE 162
151 G90D3, HP2, and HP3 possessed catalytic efficiencies within these parameters (Table 1). G90Dl had a catalytic efficiency of 5 5 x 10-5 min -1 M -1 The low efficiency of this ribozyme is due to the low turnover rate of G90D 1 In contrast, G90D 1 had a KM of 24 nM and should hybridize to its substrate in vivo. Although cleavage of its substrate will be low, the ribozyme would still be expected to have an effect in vivo. While the catalytic efficiencies for G90D3, HP2, and HP3 were within normal parameters the rate of cleavage of these ribozymes was lower than expected. There are several possible reasons for these low cleavage rates. First, the G90D and VPP sequence recognized by the ribozymes may interfere with the ability of these ribozymes to form catalytic cores There are a number of reports that have shown that different RNA sequences are cleaved with different efficiencies Second, hairpin ribozymes can catalyze product ligation as well as substrate cleavage. Ligation of the cleavage product would lead to an experimentally derived cleavage rate that is lower than the actual cleavage rate. At optimal conditions, the ligation reaction is favored over cleavage by 10-fold. A saturating concentration of cleavage product is required for this optimal ligation reaction. In cleavage reactions and in vivo the concentration of cleavage substrate is much greater than the concentration of cleavage product. There should be little if any hairpin ribozyme mediated ligation in a bimolecular reaction under these conditions A number of recent studies have shown, however, that a slow dissociation of bound product and the stability of the ribozyme complex can led to a decrease in the observed cleavage rate (l 88;226228). A decrease in the product dissociation rate of a ribozyme increases the chance of product ligation. In addition, an increase in the stability of the ribozyme structure leads to an increase in the rate of ligation. In order to achieve maximal cleavage rates, the
PAGE 163
152 ribozyme must be stable enough to assemble into an active confirmation but not so stable that cleavage is reversed by ligation. Another reason for the low observed cleavage rate may be a systemic error associated with the kinetic experiments One possible source of error is the quantitation of the ribozyme. In this project, ribozymes were quantitated by measuring the amount of radiolabled UTP incorporated by the ribozyme transcript. This method of quantitation requires the removal of unincorporated free label and knowledge of the exact amount of cold UTP added to the transcription reaction. The presence of unincorporated label and degradation of cold nucleotide could lead to an over estimation of the amount of ribozyme RNA. These are common problems associated with transcribed RNA. In addition, RNA transcripts that incorporate labeled nucleotides degrade much more rapidly than unlabeled RNA. These factors could lead to the use of less ribozyme in a cleavage reaction than expected. This in turn would result in the calculation of the turnover rate that is too low, without effecting the calculation of KM. Following the determination of the kinetic efficiency of each ribozyme, experiments were carried out to further examine the potential in vivo efficiency of G90Dl, G90D3, HP2, and HP3 ribozymes. The first experiment measured the magnesium dependence of the ribozyme cleavage reaction. The free intracellular concentration of magnesium is 0.5mm-1.5mM (220). For this project the standard cleavage reaction contained 20mM Mg2+. Therefore, it is important to ensure that the ribozymes can cleave an RNA substrate at physiological concentrations of Mg2+. In these experiments, ribozyme reactions were carried out in the presence of increasing concentrations of Mg2+. This experiment determined the ability of the ribozymes to
PAGE 164
153 cleave their substrate at low physiological Mg2+ concentrations. In addition, this experiment enabled the quantitation of the magnesium dependence of a ribozyme (Table 1). All four ribozymes were able to cleave the appropriate substrate in the presence of lmM Mg2+. The magnesium dependence of the ribozymes in the cleavage reactions, though, was approximately 25 times higher than the physiological concentration of Mg2+. This may seem alarming at first, but magnesium has been shown to play an important structural role in the folding of the hammerhead and hairpin ribozymes into an active conformation (192; 193; 196; 197). Other divalent cations, monovalent cations, and polyamines found intracellularly can facilitate the folding of ribozymes into an active conformation (76;191;198;229). For the purpose of this experiment, the magnesium coefficient was used mainly to predict a ribozyme's ability to fold into its active conformation. One method of facilitating the folding of the hairpin ribozyme into its active conformation is to include a structure between helix 2 and helix 3 (74). HP3 was designed with a hairpin loop between helix 2 and helix 3. The KMg for HP3 is 17 mM and the KMg for HP2 is 28 mM. From these experiments, the extra hairpin loop in HP3 does seem to enhance the ability of this hairpin ribozyme to fold into an active conformation. Furthermore, in order to inhibit the expression of a mutant gene, a ribozyme must be designed such that cleavage of the mutant mRNA is not inhibited by the wild-type mRNA. Otherwise, inhibition of the cleavage reaction will increase with a decrease in the mutant mRNA due to ribozyme cleavage. In this project, the wild-type rhodopsin mRNA did not appear to interfere with the cleavage of the mutant substrate by the G90D 1, G90D3, HP2, or HP3 ribozymes.
PAGE 165
154 Finally, in order to examine the ability of a ribozyme to cleave the full length mRNA, the RNA was extracted from the retina of the G90D and VPP mouse. The retinal RNA was incubated with the appropriate ribozyme, and RT-PCR assay was used to quantitate the cleavage of the mutant transgene mRNA. Neither G90D 1 nor HP3, however, was able to cleave the full-length transgenic mRNA substrate in these reactions. There a several possible reasons that cleavage was not observed. First, the ribozyme target sequence in the transgene mRNA may not be accessible to the ribozyme because of the secondary or tertiary structure of the mRNA. Alternatively, the mRNA used in these experiments was extracted from the retina of six month old mice. Retinal degeneration in these mice begins at one month and by six months is a significant loss of rod cells would be expected. In the RTPCR experiments 1 g of RNA extracted from the retina is used to determine the ability of a ribozyme to cleave full length mRNA. At six months the ratio of rod specific mRNA in this retinal extract may be greatly diminished. This may prevent the detection of cleavage of the transgenic mRNA by the ribozyme. Examination of the Effectiveness of Ribozyme Therapy In Vivo G90Dl, G90D3, HP2, and the HP3 ribozyme have been cloned and packaged into recombinant AA V vectors. AA V was chosen as the vector for these ribozymes because it is non-pathogenic, it does not elicit a strong immune response by itself, and it can infect quiescent cells (153;165;230) In addition, a number of groups have already used AAV to deliver genes to the retina (162;174-176). The ribozymes are delivered to the retina of transgenic mice via subretinal injection. The transgenic mouse models used for these experiments have been crossed with the rhodopsin knockout mouse. This ensures that the phenotype of the retinal disease
PAGE 166
155 is due to the presence of a transgenic rhodopsin gene and not over-expression of rhodopsin. The retinal disease caused by the G90D is expected to be more difficult to treat than the disease caused by the VPP mutation. This is due to the fact that the VPP mutation interferes with the incorporation of rhodopsin in the ROS (101;102). A decrease in the mutant rhodopsin should increase the incorporation of rhodopsin in the ROS. The G90D mutation, though, disrupts a salt bridge and leads to a rhodopsin protein that is frozen in the active conformation. Since only 200 of the 108 rhodopsin molecules need to be in the active state to depolarize a rod cell, a decrease in the expression of the mutant rhodopsin protein may not prevent the disease (99). A number of factors make the G90D and VPP mouse models ideal for the determining the effectiveness ribozymes in vivo. First, the eye is easily accessible making gene delivery easier than other most other target sites. Second, the eye is an immune privileged site (104). Therefore, a strong immune response to the recombinant vector would not be expected. Third, the retinal degeneration in these animals is slow, which gives adequate time for the ribozyme to have an effect. Fourth, the results of animals experiments are typically subject to large variations. This can make it difficult to interpret the data. There is less variation in the results of animal experiment when the results are obtained from a single animal. Consequently, injecting the right eye of a mouse with the ribozyme construct and the left eye with a construct that acts as a control should allow an accurate determination of the activity of a ribozyme in vivo Finally, a change in the ERG is a hallmark of the retinal degeneration associated with these forms of ADRP (93;94). Thus, ERG can be used to measure the efficiency of a ribozyme without having to sacrifice mice that have been injected with the ribozyme. Thus, the effects of a ribozyme
PAGE 167
156 can be followed over time and the in vitro and in vivo activity of a ribozyme can be compared. This in turn will aid in the design ribozyme with increased therapeutic activity This will be especially important for ribozymes designed to treat autosomal dominant genetic diseases For most gene therapy applications, a number of ribozymes are designed to target different sites on the RNA of interest. The most active ribozyme is then analyzed for activity in vivo This is done because differences in the sequence of the hybridizing arms of the ribozyme can have a dramatic affect on the ability of a ribozyme to achieve an active conformation. This can be a problem for ribozymes designed to target the mRNA sequence of a mutant gene. In this case, all the ribozymes are designed to recognize the same target sequence. The sequence of the hybridizing arms of the ribozyme required to recognize the mutant substrate might not allow cleavage with the efficiency required for therapy in vivo (188;189;216). Furthermore, the mRNA site may not be accessible in the cell due to RNA structure (203;222). When the choice of the target site is constrained, the efficiency of the cleavage reaction can only be increased by modifying the ribozyme Two strategies can be used to design new, more efficient ribozymes. The first method is rational design. With this method, the ribozyme is modified based upon experimental observations. One example of a modification made based upon the rational design of a ribozyme is the length of the hybridizing arms of the ribozyme. As previously mentioned, the length of the hybridizing arms effect several of the kinetic parameters of the RNA cleavage reaction (184;189;210). In addition, long hybridizing arms aid in the cleavage of mRNA with cleavage site blocked by secondary structure. Long hybridizing arms are thought to hybridize to single stranded portions of the mRNA. The ribozyme is
PAGE 168
157 then displaces the mRNA secondary structure by hybridization with the mRNA in a zippering effect (231). Other modifications that have been made based on rational design are an increase in helix 4 of the hairpin ribozyme and inclusion of a stable tetraloop. These modifications stabilize the secondary structure of the hairpin ribozyme. Another modification based on rational design is the inclusion of a hairpin loop between helix 2 and helix 3 of the hairpin ribozyme. This hairpin loop has been reported in some cases to increase the ability of a ribozyme to fold into an active conformation (74) Rational design can also be used to increase the intracellular activity of a ribozyme. First, ribozyme transcripts can be modified to increase their intracellular stability to RNase degradation. Second, the ribozyme transcript can be modified to traffic the ribozyme to the same cellular compartment as the substrate mRNA. Third, protein binding sites can be inserted into the ribozyme, or into upstream or downstream sequences. The protein binding sites can be chosen for their ability to fold the ribozyme into an active conformation, increase the stability of the ribozyme, or for their ability to increase the turnover of the ribozyme reaction. Sullenger and Cech designed a hammerhead ribozyme that contained a retroviral packaging signal (232). This hammerhead ribozyme was able to reduce the expression of a reporter gene found in a recombinant retroviral vector, but had no effect on the translation of the same reporter gene when transcribed as a cellular mRNA. Burke and colleagues designed a hairpin ribozyme that contained the binding site for the Rl 7 bacteriophage coat protein. The activity of this ribozyme was increased two fold in the presence of the protein. This protein increases the folding of the ribozyme into its active conformation (233). Several groups have shown that the HIV nucleocapsid protein (NC) and hnRNP A 1 protein are
PAGE 169
158 also able to increase the activity of a ribozyme. The interaction of these proteins with a ribozyme is nonspecific. They act by facilitating the annealing of the ribozyme to its substrate and by increasing the rate of dissociation of the ribozyme from its cleavage products. Like the R 17 protein, NC and A 1 also help the ribozyme fold into its active conformation (234;235). Sioud has also reported that endogenous proteins could bind to a hammerhead ribozyme and prevent its degradation by cellular ribonucleases (236). Irrational design is a second method that can be use to improve ribozymes. This method uses in vitro or in vivo techniques to select for new or improved ribozyme activities (237-239). This approach relies on the probability that a given pool of random oligonucleotides will include those able to catalyze the desired reaction. The problem is to identify and isolate these rare molecules with the desired properties. Once a method of separating a cleavage site molecule based on its unique properties is developed, the oligonucleotide is then selectively amplified using standard RNA and DNA replication procedures. This procedure has been termed in vitro selection. Improvements in the desired activity are achieved by randomly introducing mutations during the amplification process. Increasing the stringency of the selection process selects for those molecules with increased activity. This process has been termed in vitro evolution. These in vitro techniques have been used to create a number of ribozymes with unique catalytic activities. Gerald Joyce used an in vitro evolution procedure to modify a group I intron so that it could efficiently cleave DNA (240-242). Lehman and Joyce also used in vitro evolution techniques to create a group I intron able to cleave RNA in the presence of Ca2+ instead of the normal cofactors Mg2+ or Mn2+ (243;244). Ribozymes have also been created to catalyze various chemical reactions such as phosphoanhydride formation, RNA
PAGE 170
159 phosphorylation, and reactions with ester and amide bonds (237;245-247) Of more relevance for gene therapy, in vitro selection was used to construct a hammerhead like ribozyme able to recognize an AUG rather than the natural NUX cleavage triplet (248). This ribozyme expands the range of RNA substrates able to be cleaved by a ribozyme Similarly, Tang and Breaker used in vitro selection to identify twelve novel self-cleaving ribozymes (249). Of the twelve novel ribozymes, eleven had rate constants below the range displayed by natural ribozymes It is not certain that the twelve novel ribozymes offer any advantage over the natural ribozymes and the therapeutic potential of the novel ribozymes remains to be determined Nevertheless, this study suggests the potential of in vitro selection for the design of therapeutic ribozymes.
PAGE 171
160 References 1. Rogers, S., Lowenthal A., Terheggen, H. G., and Columbo, J.P. (1973) J. Exp Med 137, 1091-1096. 2 Culver, K. W., Ander s on, W. F., and Blaese R. M (1991) Hum Gene Ther 2, 107109 3. Blaese, R. M., Culver, K. W Chang, L., Anderson, W. F., Mullen, C., Nienhuis, A., Carter, C., Dunbar, C., Leitman, S and Berger, M (1993) Hum. Gene Ther. 4, 521-527 4. (1999) Hum. Gene Ther. 10, 1043-1092 5. Mountain, A. (2000) Trends Biotechnol. 18, 119-128. 6. Rosenfeld M. E., Wang, M., Siegal, G. P., Alvarez, R. D., Mikheeva, G., Krasnykh, V., and Curiel, D. T. (1996) J Mol. Med. 74, 455-462. 7. Zamecnik, P C. and Stephenson, M. L. (1978) Proc. Natl. Acad. Sci. USA 75, 280284. 8 Blake, K. R., Murakami A., and Miller, P. S. (1985) Biochemistry 24, 6132-6138. 9. ltoh, T. and Tomizawa, J. (1980) Proc. Natl. Acad. Sci. USA 77, 2450-2454 10. Lacatena, R. M. and Cesareni, G. (1981) Nature 294, 623-626 11. Stougaard, P., Molin, S., and Nordstrom, K. (1981) Proc. Natl. Acad. Sci USA 78, 6008-6012. 12. Simons, R. W. and Kleckner, N (1988) Annu. Rev. Genet. 22, 567-600. 13. Mizuno, T., Chou, M. Y., and Inouye, M. (1984) Proc. Natl. Acad. Sci USA 81, 1966-1970. 14. Inoue, T., Sullivan, F. X., and Cech, T. R. (1985) Cell 43, 431-437. 15. Zaug, A. J., Been M. D., and Cech, T. R. (1986) Nature 324, 429-433. 16. Couture, S., Ellington, A. D., Gerber, A. S., Cherry, J.M., Doudna, J. A., Green, R., Hanna, M., Pace, U., Rajagopal, J and Szostak, J. W. (1990) J. Mol. Biol. 215, 345-358.
PAGE 172
161 17. Haseloff, J. and Gerlach, W. L. (1992) Biotechnology 24, 264-269. 18. Bartkiewicz, M., Gold, H., and Altman, S. (1989) Genes Dev. 3, 488-499 19. Cech, T. R. (1988) JAMA 260, 3030-3034 20. Hormes, R., Homann, M., Oelze, I., Marschall, P., Tabler, M Eckstein, F., and Sczakiel, G. (1997) Nucleic Acids Res. 25, 769-775. 21. Pley H W., Flaherty, K. M., and McKay D. B. (1994) Nature 372, 111-113 22. Ferre-D'Amare, A. R., Zhou, K., and Doudna, J A. (1998) Nature 395, 567-574. 23 Walter, N. G and Burke, J.M. (1998) Curr. Opin. Chem. Biol. 2, 24-30. 24 Dahm, S C., Derrick, W. B., and Uhlenbeck, 0. C. (1993) Biochemistry 32, 1304013045 25. Narlikar G J. and Herschlag, D. (1997) Annu Rev. Biochem. 66, 19-59. 26. Schur, P.H., Stollar, B. D., Steinberg, A. D., and Talal, N (1971) Arthritis Rheum 14, 342-347. 27. Boulanger S. C., Faix, P.H., Yang, H., Zhuo, J., Franzen, J. S Peebles, C. L., and Perlman, P. S. (1996) Mal. Cell Biol. 16, 5896-5904. 28. Zimmerly, S., Guo H., Eskes, R., Yang, J., Perlman, P. S., and Lambowitz, A. M. (1995) Cell 83, 529-538. 29. Cameron, F H. and Jennings, P.A. (1989) Proc. Natl. Acad. Sci. USA 86 91399143. 30. Sarver N., Cantin E. M., Chang, P. S., Zaia, J. A., Ladne, P.A., Stephens, D. A., and Rossi, J. J. (1990) Science 247, 1222-1225. 31. Rossi, J. J., Elkins, D., Taylor N., Zaia, J., Sullivan, S and Deshler, J. 0. (1991) Antisense Res. Dev. 1, 285-288. 32. Sioud, M. and Drlica, K. (1991) Proc. Natl. Acad. Sci. USA 88, 7303-7307. 33. Kashani-Sabet, M., Funato, T., Tone, T., Jiao, L., Wang, W., Yoshida, E Kashfinn, B. I., Shitara, T., Wu, A. M., and Moreno, J. G. (1992) Antisense Res. Dev. 2, 3-15. 34. Sioud M., Natvig, J.B., and Forre, 0. (1992) J. Mal. Biol. 223, 831-835. 35. Inokuchi, Y., Yuyama, N., Hirashima, A., Nishikawa, S., Ohkawa, J., and Taira, K. (1994) J. Biol. Chem. 269, 11361-11366.
PAGE 173
162 36. Agrawal, S., Zhao, Q., Jiang, Z., Oliver, C., Giles, H., Heath, J., and Serota, D. (1997) Antisense Nucleic Acid Drug Dev. 7, 575-584. 37. Monteith, D. K., Horner, M. J., Gillett, N. A., Butler, M., Geary, R., Burckin, T., Ushiro-Watanabe, T and Levin, A. A. (1999) Toxicol Pathol. 27, 307-317. 38. Quigley, G. J. and Rich, A. (1976) Science 194, 796-806. 39. Moras, D., Comarmond, M. B., Fischer, J., Weiss, R., Thierry, J.C., Ebel, J.P., and Giege, R. (1980) Nature 288, 669-674. 40. McKay, D. B. (1996) RNA 2, 395-403. 41. Bassi, G S., Murchie, A. I ., and Lilley, D. M (1996) RNA 2, 756-768. 42. Michel, F. and Westhof, E. (1990) J. Mol. Biol. 216, 585-610. 43. Burke, J.M. (1988) Gene 73, 273-294. 44. Michel, F., Ellington, A. D., Couture, S., and Szostak, J. W. (1990) Nature 347, 578-580. 45. Heuer, T S Chandry, P. S Belfort, M., Celander, D. W., and Cech, T. R. (1991) Proc Natl. Acad. Sci. USA 88, 11105-11109. 46. Herschlag, D. and Cech, T. R. (1990) Biochemistry 29, 10172-10180. 47. Michel, F., Hanna, M., Green, R., Bartel, D. P., and Szostak, J. W. (1989) Nature 342, 391-395. 48. Tsui, L. C. (1992) Trends Genet. 8, 392-398. 49. Sullenger, B. A. and Cech, T R. (1994) Nature 371, 619-622. 50. Jones, J. T., Lee, S. W., and Sullenger, B. A. (1996) Nat. Med. 2, 643-648. 51. Jones, J. T and Sullenger, B. A. (1997) Nat Biotechnol. 15, 902-905. 52. Phylactou, L.A., Darrah, C., and Wood, M J (1998) Nat. Genet. 18, 378-381. 53. Lan, N., Howrey, R. P., Lee, S. W., Smith, C. A., and Sullenger, B. A. (1998) Science 280, 1593-1596. 54. Ayre, B. G., Kohler, U., Goodman, H. M., and Haseloff, J. (1999) Proc. Natl. Acad. Sci. USA 96, 3507-3512. 55. Zarrinkar, P. P. and Sullenger, B. A. (1999) Biochemistry 38, 3426-3432.
PAGE 174
163 56. Kohler, U., Ayre, B. G., Goodman, H. M., and Haseloff, J. (1999) J. Mol. Biol. 285, 1935-1950. 57. Peebles, C. L., Perlman, P. S., Mecklenburg, K. L., Petrillo, M. L., Tabor, J. H., Jarrell, K. A., and Cheng, H L. (1986) Cell 44, 213-223. 58. Mikheeva, S. and Jarrell, K. A. (1996) Proc. Natl. Acad. Sci USA 93, 7486-7490. 59. Kole, R. and Altman, S. (1981) Biochemistry 20, 1902-1906 60. Darr, S C Zito, K., Smith, D and Pace, N. R. (1992) Biochemistry 31, 328-333. 61. Bothwell, A L., Stark, B. C., and Altman, S (1976) Proc. Natl. Acad. Sci. USA 73, 1912-1916. 62. Yuan, Y., Hwang, E S and Altman, S. (1992) Proc. Natl. Acad. Sci. USA 89, 8006-8010. 63. Talbot, S. J and Altman, S. (1994) Biochemistry 33, 1406-1411. 64. Plehn-Dujowich, D. and Altman, S. (1998) Proc. Natl. Acad. Sci. USA 95, 73277332. 65. Altman, S., Baer, M. F., Bartkiewicz, M., Gold, H., Guerrier-Takada, C., Kirsebom, L. A., Lumelsky, N., and Peck, K. (1989) Gene 82, 63-64. 66. Cedergren, R. (1990) Biochem Cell Biol 68, 903-906. 67. Feldstein, P .A. and Bruening, G. (1993) Nucleic Acids Res. 21, 1991-1998. 68. Perrotta, A. T. and Been, M. D. (1991) Nature 350, 434-436. 69 Perrotta, A. T and Been, M. D. (1992) Biochemistry 31, 16-21. 70. Pley, H. W., Flaherty, K. M., and McKay, D. B. (1994) Nature 372, 68-74. 71. Shimayama, T., Nishikawa, S., and Taira, K. (1995) Biochemistry 34, 3649-3654. 72. Hampel, A., Tritz, R., Hicks, M., and Cruz, P. (1990) Nucleic Acids Res. 18, 299304. 73. Joseph, S. and Burke, J.M. (1993) J. Biol. Chem. 268, 24515-24518. 74. Esteban, J. A., Walter, N. G., Kotzorek, G., Heckman, J.E., and Burke, J.M. (1998) Proc. Natl. Acad. Sci. USA 95, 6091-6096. 75. Butcher, S. E. and Burke, J.M. (1994) Biochemistry 33, 992-999.
PAGE 175
164 76. Chowrira, B. M., Berzal-Herranz, A., and Burke, J.M. (1993) Biochemistry 32, 1088-1095. 77 Hampel, A. and Tritz, R. (1989) Biochemistry 28, 4929-4933. 78. Kashani-Sabet, M., Funato, T., Florenes, V. A., Fodstad, 0., and Scanlon, K. J. (1994) Cancer Res. 54, 900-902. 79 Denman, R. B., Smedman, M., Ju, W., Rubenstein, R., Potempska, A., and Miller, D. L. (1994) Nucleic Acids Res. 22, 2375-2382. 80. Lange, W., Daskalakis, M., Finke, J., and Dolken, G. (1994) FEBS Lett. 338, 175178. 81. Yamada, 0., Kraus, G Leavitt, M. C., Yu, M., and Wong-Staal, F. (1994) Virology 205, 121-126. 82. Li, X., Gervaix, A., Kang, D., Law, P., Spector, S. A., Ho, A. D., and Wong-Staal, F. (1998) Gene Ther. 5, 233-239. 83. Beck, J. and Nassal, M. (1995) Nucleic Acids Res. 23, 4954-4962 84 L'Huillier, P. J., Soulier, S., Stinnakre, M. G Lepourry, L., Davis, S. R., Mercier, J. C., and Vilotte, J. L. (1996) Proc. Natl. Acad. Sci. USA 93, 6698-6703. 85. Lieber, A. and Kay, M.A. (1996) J. Viral. 70, 3153-3158. 86. Lewin, A. S., Drenser, K. A., Hauswirth, W.W., Nishikawa, S., Yasumura, D., Flannery, J. G., and La Vail, M. M. (1998) Nat. Med. 4, 967-971. 87. Matsumoto, B., Defoe, D. M., and Besharse, J.C. (1987) Proc. R. Soc. Lond B Biol. Sci. 230, 339-354. 88. Hargrave P A. and McDowell J. H. (1992) FASEB J 6, 2323-2331. 89. Polans A., Baehr W., and Palczewski K (1996) Trends Neurosci 19, 547-554. 90. van Soest, S., Westerveld, A., de Jong, P. T., Bleeker-Wagemakers, E. M., and Bergen, A. A. (1999) Surv. Ophthalmol. 43, 321-334. 91. Humphries, M. M., Rancourt, D., Farrar, G. J., Kenna, P., Hazel, M., Bush, R. A., Sieving, P.A., Sheils, D. M., McNally, N., Creighton, P., Erven, A., Boros, A., Gulya, K., Capecchi, M. R., and Humphries, P. (1997) Nat. Genet. 15, 216-219. 92. Inglehearn, C. F., Bashir, R., Lester, D. H., Jay, M., Bird, A. C., and Bhattacharya, S.S. (1991) Am. J. Hum. Genet. 48, 26-30. 93. Dryja, T. P., McGee, T. L., Reichel, E., Hahn, L.B., Cowley, G. S Yandell, D. W., Sandberg, M.A., and Berson, E. L. (1990) Nature 343, 364-366.
PAGE 176
165 94. Berson, E. L., Rosner, B., Sandberg, M A., and Dryja, T. P (1991) Arch. Ophthalmol. 109, 92-101. 95. Boughman, J. A., Conneally, P. M., and Nance, W. E. (1980) Am. J. Hum. Genet. 32, 223-235. 96 Kajiwara, K., Berson, E L., and Dryja, T. P. (1994) Science 264, 1604-1608 97. Bunker, C.H. Berson, E. L., Bromley, W. C., Hayes, R. P., and Roderick, T H. (1984) Am. J. Ophthalmol 97, 357-365 98 Inglehearn, C. F., Tarttelin, E. E., Plant, C., Peacock, R. E., Al Maghtheh, M., Vithana, E Bird, A. C., and Bhattacharya, S.S. (1998) J. Med. Genet. 35 1-5. 99 Rao, V R., Cohen, G. B., and Oprian, D. D. (1994) Nature 367, 639-642. 100. Sung, C.H., Davenport, C. M., Hennessey, J.C., Maumenee, I. H., Jacobson, S G Heckenlively, J. R., Nowakowski, R., Fishman, G., Gouras, P and Nathans, J. (1991) Proc. Natl. Acad. Sci. USA 88, 6481-6485. 101. Wang, M Lam, T T Tso, M 0 and Naash, M. I. (1997) Vis. Neurosci. 14, 5562. 102. Wu, T. H., Ting, T. D., Okajima, T. I., Pepperberg, D. R., Ho, Y K., Ripps, H and Naash, M. I. (1998) Neuroscience 87, 709717. 103. Naash, M. L., Peachey, N. S., Li, Z. Y., Gryczan, C. C., Goto, Y Blanks, J., Milam, A.H., and Ripps, H. (1996) Invest Ophthalmol. Vis. Sci. 37, 775-782. 104. Tompsett, E., Abi-Hanna, D., and Wakefield D. (1990) Curr Eye Res 9, 11411145. 105. Bowes C., Danciger, M., Kozak, C. A., and Farber, D. B. (1989) Proc. Natl. Acad. Sci. USA 86, 9722-9726. 106. McLaughlin, M. E., Sandberg, M.A., Berson, E. L., and Dryja, T. P. (1993) Nat. Genet. 4, 130-134. 107. Travis, G. H. Groshan, K. R., Lloyd, M., and Bok, D. (1992) Neuron 9 113-119. 108. Olsson, J.E. Gordon, J. W., Pawlyk, B. S Roof, D., Hayes, A., Molday, R. S., Mukai, S Cowley, G. S., Berson, E. L., and Dryja, T. P. (1992) Neuron 9, 815-830. 109. Naash, M. I., Hollyfield, J. G Al Ubaidi, M. R., and Baehr, W. (1993) Proc. Natl. Acad. Sci USA 90, 5499-5503. 110 Liu, X., Wu, T H. Stowe, S., Matsushita, A., Arikawa, K., Naash, M. I., and Williams, D S. (1997) J. Cell Sci. 110 ( Pt 20) 2589-2597.
PAGE 177
166 111. Beigelman, L., McSwiggen, J. A., Draper, K. G., Gonzalez, C., Jensen, K., Karpeisky, A. M., Modak, A. S., Matulic-Adamic, J., DiRenzo, A. B., and Haeberli, P. (1995) J Biol. Chem. 270, 25702-25708. 112. Heidenreich, 0., Benseler, F., Fahrenholz, A., and Eckstein, F. (1994) J. Biol. Chem. 269, 2131-2138. 113. Heidenreich, 0., Xu, X., Swiderski, P., Rossi, J. J., and Nerenberg, M (1996) Antisense Nucleic Acid Drug Dev 6, 111-118. 114 Ellens, H., Doxsey, S., Glenn, J. S and White, J.M. (1989) Methods Cell Biol. 31, 155-178. 115. San, H., Yang, Z. Y Pompili, V. J., Jaffe, M. L., Plautz, G E Xu L., Feigner, J H Wheeler, C. J., Feigner, P. L., and Gao, X. (1993) Hum. Gene Ther. 4, 781-788. 116. Dowty, M. E., Williams, P., Zhang, G., Hagstrom, J.E., and Wolff, J. A. (1995) Proc Natl. Acad. Sci. USA 92, 4572-4576. 117. Hangai, M., Kaneda, Y., Tanihara, H., and Honda, Y. (1996) Invest Ophthalmol. Vis. Sci. 37, 2678-2685. 118. Verma, I. M. and Somia, N. (1997) Nature 389, 239-242. 119. Fields Virology (1990) Raven Press, New York. 120. Smith, L. C., Eisensmith, R. C., and Woo, S. L. (1995) Adv. Exp. Med. Biol. 369, 79-88. 121. Verma, I. M. and Somia, N. (1997) Nature 389, 239-242. 122. Yeh, P. and Perricaudet, M. ( 1997) F ASEB J. 11, 615-623. 123. All the Virology on the WWW. 2000. Ref Type: Electronic Citation 124. Dai, Y., Schwarz, E. M., Gu, D., Zhang, W.W., Sarvetnick, N., and Verma, I. M. (1995) Proc. Natl. Acad. Sci USA 92, 1401-1405. 125. Yang, Y., Nunes, F. A., Berencsi, K., Furth, E. E., Gonczol, E., and Wilson, J.M. (1994) Proc. Natl. Acad. Sci. USA 91, 4407-4411. 126. Chen, H. H., Mack, L. M., Kelly, R., Ontell, M., Kochanek, S., and Clemens, P.R. (1997) Proc. Natl. Acad. Sci. USA 94, 1645-1650. 127. Mack, C. A., Song, W.R., Carpenter, H., Wickham, T J., Kovesdi, I., Harvey, B G., Magovern, C. J., Isom, 0. W., Rosengart, T., Falck-Pedersen, E., Hackett N. R., Crystal, R. G., and Mastrangeli, A. (1997) Hum. Gene Ther. 8, 99-109. 128. Teichler Zallen D (2000) Trends Genet 16, 272-275.
PAGE 178
167 129 Kim, S H., Yu, S.S. Park, J. S Robbins, P. D., An, C. S., and Kim, S (1998) J V i ral. 72, 994-1004. 130 Miller, A. D., Miller D G., Garcia, J. V and Lynch, C. M. (1993) Methods Enz y mol. 217, 581-599. 131. Robbin s, P D. Tahara, H., Mueller, G Hung, G Bahnson A., Zitvogel, L., Galea-Lauri, J Ohas hi, T., Patrene K. and Boggs, S S. (1994) Ann. N Y Acad. Sci. 716, 72-88 132 Lewi s P Hensel M and Emerman, M. (1992) EMBO J 11, 3053-3058. 133 Miyo s hi, H., Takahashi, M., Gage, F. H., and Verma, I. M. (1997) Proc. Natl. Acad. Sci. USA 94 10319-10323. 134 Kafri T., Blomer, U., Peterson, D. A., Gage, F. H., and Verma, I. M. (1997) Nat. Genet. 17, 314-317. 135 Buchschacher, G L., Jr. and Wong-Staal, F. (2000) Blood 95, 2499 2504. 136. Fekete D. M. and Cepko, C. L. (1993) Mal. Cell Biol. 13, 2604-2613. 137. Paul, R. W., Morris, D., Hess, B. W Dunn, J., and Overell, R. W. (1993) Hum Gene Ther 4, 609-615. 138. Kotani, H., Newton, P B ., ill, Zhang, S Chiang Y L., Otto, E Weaver, L., Blaese R. M. Anderson, W. F and McGarrity, G J. (1994) Hum. Gene Ther. 5, 19-28. 139. Weiher, H., Barkli s E Ostertag, W and Jaenisch, R. (1987) J. Viral. 61, 27422746 140 Colicelli, J. and Goff, S. P. (1986) J. Viral. 57, 674-677. 141. Osborne CS, Pasceri P Singal R, Sukonnik T, Ginder GD, and Ellis J (1999) J. Viral. 73, 5940-5946 142 Otto, E Jones-Trower, A., Vanin, E. F., Stambaugh, K., Mueller, S. N Anderson, W. F and McGarrity, G. J. (1994) Hum. Gene Ther. 5 567-575 143 Chong H. and Vile, R. G. (1996) Gene Ther 3, 624-629. 144 Kung H.J., Boerkoel, C., and Carter T. H (1991) Curr. Top. Microbial. Immunol. 171 1-25. 145. Bellacosa, A., Testa, J R., Staal, S P., and Tsichlis, P. N. (1991) Science 254 274277.
PAGE 179
168 146. Powell, S. K., Kaloss, M., Burimski, I., Weaver, L., Long, Z., Lyons, R., McGarrity, G. J., and Otto, E. (1999) Hum. Gene Ther. JO, 2123-2132. 147. Kotin, R. M. (1994) Hum Gene Ther. 5, 793-801. 148 Steiner, I., Spivack, J. G., Lirette, R. P., Brown, S. M., MacLean, A. R., Subak Sharpe, J. H., and Fraser, N. W. (1989) EMBO J. 8, 505-511. 149. Sawtell, N. M. and Thompson, R. L. (1992) J Viral. 66, 2157-2169. 150 Marconi P Krisky, D., Oligino, T., Poliani, P. L., Ramakrishnan, R., Goins, W. F., Fink, D. J., and Glorioso, J C. (1996) Proc. Natl. Acad. Sci USA 93, 11319-11320. 151. Wu, N., Watkins, S. C., Schaffer, P.A., and DeLuca, N A. (1996) J. Viral. 70, 6358-6369. 152. Zhu, Z., DeLuca, N. A., and Schaffer, P.A. (1996) J. Viral. 70, 5346-5356. 153 Wood, M J Byrnes, A. P., Kaplitt, M. G., Pfaff D W., Rabkin, S. D., and Charlton H. M (1994) Exp. Neural. 130, 127-140. 154 Lokensgard, J. R., Bloom, D. C., Dobson, A. T., and Feldman, L. T (1994) J. Viral. 68 7148-7158. 155 Blacklow, N. R., Dolin, R., and Hoggan, M. D. (1971) J. Gen. Viral. JO, 29-36 156. Hoggan, M D., Shatkin, A. J Blacklow, N. R., Koczot, F and Rose, J A. (1968) J. Viral. 2, 850-851. 157. Buller, R. M., Straus, S. E., and Rose, J A. (1979) J. Gen Viral. 43, 663-672. 158. Kotin, R. M., Siniscalco, M ., Samulski, R. J., Zhu, X. D., Hunter, L., Laughlin, C. A., McLaughlin, S., Muzyczka, N Rocchi, M., and Berns, K. I. (1990) Proc. Natl. Acad. Sci. USA 87, 2211-2215. 159. Berns, K. I., Pinkerton, T. C., Thomas, G. F., and Hoggan, M. D. (1975) Virology 68, 556-560. 160 Cheung, A. K., Hoggan, M. D., Hauswirth, W.W., and Berns, K. I. (1980) J. Viral. 33, 739-748. 161. Grimm, D ., Kern, A., Rittner, K., and Kleinschmidt, J. A. (1998) Hum. Gene Ther 9, 2745-2760. 162. Flannery, J. G., Zolotukhin, S., Vaquero, M. I., LaVail, M. M., Muzyczka, N., and Hauswirth, W.W. (1997) Proc. Natl. Acad. Sci. USA 94, 6916-6921.
PAGE 180
169 163. Flotte, T. R., Afione, S. A., Conrad, C., McGrath, S. A., Solow, R., Oka, H., Zeitlin, P. L., Guggino, W. B., and Carter, B J. (1993) Proc. Natl. Acad. Sci. USA 90, 10613-10617. 164. Kessler, P. D., Podsakoff, G. M., Chen, X McQuiston, S. A., Colosi, P. C., Matelis, L.A., Kurtzman, G. J., and Byrne, B. J. (1996) Proc Natl. Acad. Sci USA 93, 14082-14087. 165. Chirmule, N., Propert, K., Magosin, S., Qian, Y., Qian, R., and Wilson, J (1999) Gene Ther. 6, 1574-1583 166. Hernandez, Y. J., Wang, J., Kearns, W. G., Loiler, S., Poirier, A., and Flotte, T. R. (1999) J. Viral. 73, 8549-8558. 167. Bennett, J., Maguire, A. M., Cideciyan, A V., Schnell, M., Glover, E., Anand, V., Aleman, T. S., Chirmule, N., Gupta, A. R., Huang, Y., Gao, G. P., Nyberg, W. C., Tazelaar, J., Hughes, J., Wilson, J.M., and Jacobson, S G (1999) Proc. Natl. Acad. Sci USA 96, 9920-9925. 168 Bennett, J., Zeng, Y., Bajwa, R., Klatt, L., Li, Y., and Maguire, A. M (1998) Gene Ther. 5, 1156-1164. 169 Ali, R.R., Reichel, M B., Byrnes, A. P., Stephens, C J., Thrasher, A. J., Baker, D., Hunt, D M and Bhattacharya, S.S. (1998) Gene Ther 5, 1561-1565. 170. Cayouette, M., Behn, D., Sendtner, M., Lachapelle, P., and Gravel, C. (1998) J. Neurosci. 18, 9282-9293. 171. Kido, M., Rich, K. A., Yang, G., Barron, E., Kohn, D. B., Al Ubaidi, M. R., Blanks, J.C., and Lang, G. (1996) Curr. Eye Res. 15, 833-844. 172. Takahashi, M., Miyoshi, H., Verma, I. M., and Gage, F. H. (1999) J. Viral. 73, 7812-7816. 173. Liu, X., Brandt, C. R., Gabelt, B. T., Bryar, P. J., Smith, M. E., and Kaufman, P. L. (1999) Exp. Eye Res. 69, 385-395. 174. Ali, R.R., Reichel, M. B., Thrasher, A. J., Levinsky, R. J., Kinnon, C., Kanuga, N., Hunt, D. M., and Bhattacharya, S.S. (1996) Hum. Mol. Genet. 5, 591-594. 175. Jomary, C., Vincent, K. A., Grist, J., Neal, M. J., and Jones, S. E. (1997) Gene Ther. 4, 683-690. 176. Grant, C. A., Ponnazhagan, S., Wang, X. S., Srivastava, A., and Li, T. (1997) Curr. Eye Res. 16, 949-956. 177. Guy, J., Qi, X., Muzyczka, N., and Hauswirth, W.W. (1999) Arch. Ophthalmol. 117, 929-937.
PAGE 181
170 178. Zack, D. J., Bennett, J., Wang, Y Davenport, C., Klaunberg, B., Gearhart, J., and Nathans, J. (1991) Neuron 6, 187-199. 179. Ryu, W. S. and Mertz, J.E. (1989) J. Viral. 63, 4386-4394. 180. Huang, M T and Gorman, C. M. (1990) Nucleic Acids Res. 18, 937-947. 181. Yu, M., Poeschla, E., Yamada, 0., DeGrandis, P., Leavitt, M C., Heusch, M ., Yees, J K., Wong-Staal, F., and Hampel, A. (1995) Virology 206, 381-386. 182. Walter N. G Burke, J.M., and Millar, D. P (1999) Nat. Struct. Biol. 6, 544-549 183 Feldstein, P A., Buzayan, J M ., van Tol, H., deBear, J., Gough, G. R., Gilham, P T., and Bruening, G. (1990) Proc. Natl. Acad. Sci. USA 87, 2623-2627. 184. Chowrira, B. M and Burke, J.M. (1991) Biochemistry 30, 8518-8522. 185. Komatsu, Y., Koizumi, M Sekiguchi, A., and Ohtsuka, E. (1993) Nucleic Acids Res. 21, 185-190. 186. Grasby, J. A., Mersmann, K., Singh M., and Gait, M. J. (1995) Biochemistry 34, 4068-4076. 187. Nesbitt, S M., Erlacher, H A., and Fedor, M. J. (1999) J Mol. Biol 286, 10091024. 188. Hegg, L.A. and Fedor, M J. (1995) Biochemistry 34, 15813-15828. 189. Fedor, M. J. and Uhlenbeck, 0. C. (1992) Biochemistry 31, 12042-12054. 190 Westley J. ( 1969) Enzymatic catalysis Harper and Row, New York 191. Dahm, S. C. and Uhlenbeck, 0. C. (1991) Biochemistry 30, 9464-9469. 192. Menger, M ., Tuschl, T., Eckstein, F., and Porschke, D. (1996) Biochemistry 35, 14710-14716. 193. Orita, M ., Vinayak R., Andrus, A., Warashina, M., Chiba, A., Kaniwa, H., Nishikawa, F., Nishikawa, S., and Taira, K. (1996) ]. Biol. Chem. 271, 9447-9454. 194. Pley, H. W., Lindes, D.S., DeLuca-Flaherty, C., and McKay, D. B. (1993) J. Biol. Chem. 268, 19656-19658 195. Chowrira, B. M., Berzal-Herranz, A., and Burke, J.M. (1993) Biochemistry 32, 1088-1095. 196. Butcher, S. E. and Burke, J.M. (1994) J. Mol. Biol. 244, 52-63 197. Young, K. J., Gill, F., and Grasby, J. A. (1997) Nucleic Acids Res. 25, 3760-3766.
PAGE 182
171 198. Earnshaw, D. J. and Gait, M. J. (1998) Nucleic Acids Res. 26, 5551-5561. 199. Altschuler, M., Tritz, R., and Hampel, A. (1992) Gene 122, 85-90. 200. Grunstein, M. and Hogness, D.S. (1975) Proc. Natl. Acad. Sci. USA 72, 39613965 201. Kado, C. I. and Liu, S. T. (1981) J Bacterial. 145, 1365-1373. 202 Berzal-Herranz, A., Joseph, S., and Burke, J.M. (1992) Genes Dev. 6, 129-134. 203. Ruffner, D. E., Stormo, G.D., and Uhlenbeck, 0. C. (1990) Biochemistry 29, 10695-10702. 204. Gnatenko, D., Arnold, T. E., Zolotukhin, S., Nuovo, G. J., Muzyczka, N., and Bahou, W. F. (1997) J. Investig. Med. 45, 87-98. 205 Sioud, M. (1999) Int. J. Mol. Med. 3, 381-384. 206. Joseph, S., Berzal-Herranz, A., Chowrira, B. M., Butcher, S. E., and Burke, J.M. (1993) Genes Dev. 7, 130-138. 207. Herschlag, D. (1991) Proc. Natl. Acad. Sci. USA 88, 6921-6925. 208. Helene, C. and Toulme, J. J. (1990) Biochim. Biophys. Acta 1049, 99-125. 209 Miller, P. S. and Ts'o, P. 0. (1987) Anticancer Drug Des 2, 117-128. 210. Hendry, P. and McCall, M. (1996) Nucleic Acids Res. 24, 2679-2684. 211. Hertel, K. J., Herschlag, D., and Uhlenbeck, 0. C. (1994) Biochemistry 33, 33743385. 212 Heidenreich, 0 and Eckstein, F. (1992) J. Biol. Chem. 267, 1904-1909. 213. Goodchild, J. and Kohli, V. (1991) Arch. Biochem. Biophys. 284, 386-391. 214. Werner, M. and Uhlenbeck, 0. C. (1995) Nucleic Acids Res. 23, 2092-2096 215. Phylactou, L.A., Tsipouras, P., and Kilpatrick, M. W. (1998) Biochem. Biophys. Res. Commun. 249, 804-810. 216 Chowrira, B. M., Pavco, P.A., and McSwiggen, J. A. (1994) J. Biol. Chem. 269, 25856-25864 217. Tang, X B., Hobom, G., and Luo, D. (1994) J. Med Viral. 42, 385-395. 218. Lian, Y., Young, M. D., Siwkowski, A., Hampel, A., and Rappaport, J. (1999) Gene Ther. 6, 1114-1119.
PAGE 183
172 219. Hegg, L.A. and Fedor, M J. (1995) Biochemistry 34, 15813-15828. 220 Alvarez-Leefmans, F J., Giraldez, F., and Gamino, S. M. (1987) Can J Physiol. Pharmacol. 65, 915-925. 221. Lima, W. F., Monia, B P Ecker, D. J., and Freier, S. M. (1992) Biochemistry 31, 12055-12061. 222 Taylor, N. R. and Rossi, J. J. (1991) Antisense Res Dev 1, 173-186 223. Werner, M and Uhlenbeck, 0. C. (1995) Nucleic Acids Res 23, 2092-2096. 224. Hertel, K. J. Herschlag, D., and Uhlenbeck, 0. C. (1996) EMBO J. 15, 3751-3757 225. Hertel, K. J., Peracchi, A., Uhlenbeck, 0. C., and Herschlag, D. (1997) Proc. Natl. Acad. Sci. USA 94, 8497-8502. 226. Fedor, M. J. (1999) Biochemistry 38, 11040-11050. 227. Donahue, C. P. and Fedor, M. J. (1997) RNA. 3, 961-973. 228. Hertel, K. J., Pardi, A., Uhlenbeck, 0 C. Koizumi, M., Ohtsuka, E., Uesugi, S., Cedergren, R., Eckstein, F Gerlach, W. L., and Hodgson, R. (1992) Nucleic Acids Res. 20, 3252. 229. Murray, J.B., Seyhan, A. A., Walter, N. G Burke, J M and Scott, W. G. (1998) Chem. Biol. 5, 587-595. 230 Hernandez, Y. J., Wang, J., Kearns, W. G., Loiler, S., Poirier, A., and Flotte, T R. (1999) J. Viral. 73, 8549-8558 231. Denman, R. B (1996) FEES Lett. 382, 116-120 232. Sullenger, B. A. and Cech, T R. (1993) Science 262, 1566-1569. 233. Sargueil, B., Pecchia, D. B., and Burke J.M. (1995) Biochemistry 34, 7739-7748 234. Tsuchihashi, Z., Khosla, M., and Herschlag, D. (1993) Science 262, 99-102. 235 Herschlag, D., Khosla, M., Tsuchihashi, Z., and Karpel, R. L. (1994) EMBO J 13, 2913-2924. 236. Sioud, M., Opstad, A., Zhao, J. Q., Levitz, R., Benham, C., and Drlica, K. (1994) Nucleic Acids Res 22, 5571-5575 237. Bartel, D. P. and Szostak, J. W. (1993) Science 261, 1411-1418. 238. van der Vlugt, R. A., Prins, M., and Goldbach, R. (1993) Virus Res. 27, 185-200.
PAGE 184
173 239 Breaker, R.R. and Joyce, G F. (1994) Chem. Biol. 1, 223-229 240. Robertson, D L. and Joyce, G. F. (1990) Nature 344, 467-468. 241. Beaudry, A. A. and Joyce, G. F. (1992) Science 257, 635-641. 242. Tsang, J. and Joyce, G. F. (1994) Biochemistry 33, 5966-5973. 243. Lehman, N. and Joyce G. F. (1993) Nature 361 182-185 244 Wright, L., Wilson, S. B., Milliken, S., Biggs, J., and Kearney, P. (1993) Exp. Hematol. 21, 1714-1718. 245. Wilson, C. and Szostak, J W (1995) Nature 374, 777-782. 246. Illangasekare, M., Sanchez, G., Nickles, T and Yarus, M (1995) Science 267, 643647. 247. Lohse, P.A. and Szostak, J. W. (1996) Nature 381, 442-444. 248. Yaish, N. K., Heaton, P.A., Fedorova, 0., and Eckstein, F (1998) Proc. Natl. Acad. Sci USA 95, 2158-2162. 249. Tang, J. and Breaker, R.R. (2000) Proc. Natl. Acad. Sci. USA 97, 5784-5789.
PAGE 185
BIOGRAPHICAL SKETCH Patrick Whalen obtained his undergraduate degree in chemistry at the University of Florida in 1989 He entered graduate school in 1992 and joined the Department of Molecular Genetics and Microbiology In 1994 joined the laboratory of Dr. Alfred Lewin for his graduate research 174
PAGE 186
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. I certify that I have read this study and that in my opinion it conforms to accep!able st~dards of scholarly presentation is fully adequaf in scope and quality, as a dissertation for the degree of Doctor of Philosophy l ,., J q i L1Ul4J1~ William H. Hau~ Professor of Molecular Genetics and Microbiology 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 ~o.u~ es B. Flaneg :fessor of Molecular Genetics and Microbiology 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. Professor of Pathology, Immunology, and Laboratory Medicine
PAGE 187
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. May 2001 Dean College of Medicine Dean, Graduate School
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 E9T75Q3ZN_09F3E0 INGEST_TIME 2017-05-24T21:18:51Z PACKAGE AA00054909_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
|