Strategies for Tolerance Induction and Correction of Mouse Model of Hemophilia A Using Liver Directed AAV-Mediated Gene ...


Material Information

Strategies for Tolerance Induction and Correction of Mouse Model of Hemophilia A Using Liver Directed AAV-Mediated Gene Therapy and B Cell Depletion
Physical Description:
1 online resource (97 p.)
Sack, Brandon Keith
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Medical Sciences, Immunology and Microbiology (IDP)
Committee Chair:
Herzog, Roland
Committee Members:
Srivastava, Arun
Slayton, William B
Atkinson, Mark A
Fletcher, Bradley S


Subjects / Keywords:
aav -- adeno -- associated -- b -- cd20 -- cell -- gene -- hemophilia -- immunology -- rituximab -- therapy -- virus
Immunology and Microbiology (IDP) -- Dissertations, Academic -- UF
Medical Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation


Hemophilia A is an X-linked bleeding disorder occurring inapproximately 1 in 5000 males. The disease is cause by a deficiency of clottingfactor VIII (FVIII) resulting in bleeding episodes that can be fatal to anaffected individual. The only treatment for hemophilia A is protein replacementtherapy, which is both expensive and is often complicated by formation ofantibodies against FVIII (termed “inhibitors”). Thus, therapies that addressthe burden of costly infusions as well as the immune response to FVIII arehighly desired. The use of liver-directed adeno-associated viral (AAV) genetherapy was investigated as a means of both phenotypic correction and toleranceinduction to the FVIII transgene in a hemophilia A mouse model. Treatment witha liver-specific AAV8-hFVIII vector either sensitized or tolerized BL/6-129/svor BALB/c hemophilia A mice, respectively, to an i.v. hFVIII challengeprotocol. B cell depletion concurrent with gene therapy reduced the antibodyresponse in AAV-treated BL/6-129/sv-HA mice. All of the mice had minimalcorrection of phenotype correlating to ~1% of normal FVIII activity which didnot persist past the hFVIII challenge. Codon-optimization of the transgene notonly increased expression of hFVIII to a peak of ~10% and persistent level of3-5%, but also induced tolerance in both strains. Consistent with previousstudies, this tolerance was found to be mediated by T regulatory cells. B celldepletion was also used to develop a protocol to prevent the development ofanti-AAV capsid antibodies and as an adjuvant to reverse establishedinhibitors. Reduction of ant-capsid antibodies was achieved but not to a levelthat allowed vector re-administration. Furthermore, delivery of AAV-COhFVIII tomice with pre-existing inhibitors suggests that gene transfer with or without Bcell depletion may reduce inhibitors. These results provide a significant step forward in genetherapy for hemophilia A by demonstrating that the immune responses to FVIIImay be mitigated by liver-specific expression of the transgene. Furthermore, Bcell depletion (such as with the human drug rituximab) may prove to be aessential adjuvant in gene therapy for the prevention of an immune response toboth the transgene and the viral vector.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Brandon Keith Sack.
Thesis (Ph.D.)--University of Florida, 2012.
Adviser: Herzog, Roland.
Electronic Access:

Record Information

Source Institution:
Rights Management:
Applicable rights reserved.
lcc - LD1780 2012
System ID:

This item is only available as the following downloads:

Full Text




2 2012 Brandon Keith Sack


3 To Mom


4 ACKNOWLEDGMENTS I would like to first thank my mentor, Dr. Roland Herz og, for taking me in and guiding me to this point. I thank him for his mentoring style: always giving me enough rope to hang myself but showing up to cut me down when I made the mistake to do so. He has taught me that science is more than just hypotheses a but also to whatever other endeavors I find myself in. I would also like to thank my lab mates Dr. David Markusic, Dr. Babak Moghimi, Dr Brad Hoffmann, Irene Zolothukin, Mario Cooper and Dr. Ashley Martino for enriching my graduate career by having the patience and consideration to train me, criticize me, and allow me to collaborate with them. Much gratitude is owed to my committee of Dr. Mark Atkinson, Dr. William Slayton, Dr. Bradley Fletcher, and Dr. Arun Srivastava who were essential in maintaining the quality of my research and for always offering a guiding hand. Much credit goes to my Mom, Diane Sack, for raising me and leading by ex ample. Through her I learned that hard work and ambition can go hand in hand with compassion, honesty and altruism. The unwavering support from her, my sister Amy, and the rest of my family has given me enough fulfillment in my life to attempt anything and enough. Finally, thank you to all my friends both within and outside of the IDP for providing me with the opportunity to make my near decade in Gainesville the most memo rable of my life.


5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 Hemophilia ................................ ................................ ................................ .............. 12 Treatment of Hemophilia A ................................ ................................ ............... 14 Inhibitors in Hemophilia A ................................ ................................ ................. 15 AAV Gene Therapy ................................ ................................ ................................ 17 Liver Directed AAV Gene Therapy f or Hemophilia ................................ ........... 18 2 MATERIALS AND METHODS ................................ ................................ ................ 26 Mouse S trains ................................ ................................ ................................ ......... 26 B Cell Depletion a nd Flow Cytometry ................................ ................................ ..... 26 AAV Vector Construction ................................ ................................ ........................ 27 Blood Collection ................................ ................................ ................................ ...... 28 Intraveno us hFVIII Challenge P rotocol ................................ ................................ ... 28 ELISA P rotocols ................................ ................................ ................................ ...... 28 Clotting A ssays ................................ ................................ ................................ ....... 30 Adenovirus C hallenge ................................ ................................ ............................. 31 Immunohistochemistry ................................ ................................ ............................ 32 T cell Assays and Adoptive T ransfer ................................ ................................ ....... 33 3 TRANSIENT B CELL DEPLETION IN COMBINATION WITH LIVER DIRECTED GENE THERAPY ................................ ................................ ................................ .... 36 Background ................................ ................................ ................................ ............. 36 Study Design for Transient B Cell Depletion and Gene therapy ............................. 37 B Cell Depletion with ................................ ................................ ................... 38 B Cell Depletion Concurrent with Gene Transfer Renders BL/6 129/sv HA Mice Hyporesponsive to hFVIII ................................ ................................ .................... 38 Gene Trans fer Renders BALB/c HA Mice Hyporesponsive to hFVIII R egardless of B Cell Depletion ................................ ................................ ............................... 39 T Cell Tolerance to FVIII in BALB/c HA Mice ................................ .......................... 40 hFVIII Protein Administration During Recovery o f B Cells Fails t o Achieve Lasting Tolerance ................................ ................................ ................................ 42


6 4 CODON OPTIMIZAT ION O F HUMAN F VIII TRANSGENE IMPROVES EXPRESSION AND INDUCTION O F TOLERANCE I N HEMOPHILIA A MICE REGARDLESS O F STRAIN BACKGROUND ................................ ......................... 54 Background ................................ ................................ ................................ ............. 54 Codon Optimization Improves hFVIII E xpression from AAV8 and Yields Long Term Correction of Phenotype a t Therapeutic Levels ................................ ......... 56 Immunological Tolerance to hFVIII Following AAV8 COhFVIII Gene T ransfer ....... 57 Co administration of an AAV Helper Virus Encoding a sc PP5 Transgene Under A Liver Specific Promoter Improves hFIX Expression but does n ot Effect Expression of COhFVIII ................................ ................................ ....................... 58 Co ad m inistration of scAAV8 TTR PP5 Helper Virus Does not Augment AAV8 COhFVIII Expression but Does I mprove Tolerance Induction ............................. 59 5 ALTERNATIVE STRATEGIES FOR B CELL DEPLETION COMBINED WITH GENE THER APY ................................ ................................ ................................ .... 65 Background ................................ ................................ ................................ ............. 65 Optimization of a CD20 Protocol t o Prevent Neu tralizing Antibodies Against AAV 8 Capsid ................................ ................................ ................................ ....... 66 gene Therapy to Reverse Inhibitors i n Hemophilia A Mice ................................ ................................ ................................ ..................... 69 6 DISCUSSION ................................ ................................ ................................ ......... 75 Gene Transfer May Tolerize or Sen sitize Mice Depending o n Genetic Background ................................ ................................ ................................ ......... 75 Utility of Transient B Cell Depletion f or Tolerance to hFVIII ................................ .... 77 Tolerance Induct ion with C odon optimized hFVIII ................................ .................. 79 Utility of B Cell Depletion for the Prevention o f Anti AAV Capsid Antibodies and t o Reverse Inhibitors In Hemophilia A Mice ................................ ......................... 81 Future Directions ................................ ................................ ................................ .... 83 APPENDIX: SEQUENCE OF CODON OPTIMIZED HUMAN FVIII .............................. 86 LIST OF REFERENCES ................................ ................................ ............................... 88 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 97


7 LIST OF TABLES Table page 3 1 CD3 + T cells following B cell depletion. ................................ .............................. 46 3 2 Comparison of immune response to hFVIII in BALB/c and BL/6 129/sv hemophilia A mice. ................................ ................................ ............................. 53


8 LIST OF FIGURES Figure page 1 1 The human factor VI II protein ................................ ................................ ............ 22 1 2 Simplified overview of the intrinsic clotting pathway. ................................ .......... 23 1 3 Overview of FVIII common inversion in intron 22 of the FVIII gene .................... 24 1 4 AAV genome structure ................................ ................................ ....................... 25 2 1 Cloning strategy for pAAV hAAT hFVIII mini vector plasmid ............................. 35 3 1 B cell depletion and gene therapy protocol ................................ ......................... 44 3 2 B cel l depletion following cd20 administration ................................ ................. 45 3 3 Immune response to Ad LacZ in B cell depleted mice ................................ ....... 47 3 4 B cell depletion and gene transfer in BL/6 129/sv HA mice ................................ 48 3 5 B cell depletion and gene transfer in BALB/c HA mice ................................ ....... 49 3 6 T Cell Responses in BALB/c HA mice ................................ ............................... 50 3 7 Anti CD20 treatment to prevent antibody formation in hFVIII protein replacement therapy ................................ ................................ ........................... 51 3 8 Correlation between anti FVIII IgG1 and Bethesda in BALB/c HA and BL/6 129sv HA mice ................................ ................................ ................................ ... 52 4 1 Improved expression with ................................ .... 60 4 2 Expression of codon optimized hFVIII ................................ ............................... 61 4 3 Immune responses following AAV8 COhFVIII gene transfer .............................. 6 2 4 4 scAAV8 TTR PP5 improves FIX expression at low doses ................................ 63 4 5 scAAV8 TTR PP5 fails to improve AAV8 COhFVIII expression ......................... 64 5 1 ............ 70 5 2 ........................... 71 5 3 Alternate two AAV8 antibodies ................... 72


9 5 4 Lower doses of AAV8 COhFVIII allow for prevention of antibodies without improvement after re administration ................................ ................................ .. 73 5 5 Liver directed gene therapy is able to reduce antibody titer in hFVIII challenged BALB/c HA mice ................................ ................................ .............. 74


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 STRATEGIES FOR TOLERA NCE INDUCTION AND CORRECTION OF MOUSE MODEL OF HEMOPHILIA A USING LIVER DIRECTED AAV MEDIATED GENE THERAPY AND B CELL DEPLETION By Brandon Keith Sack August 2012 Chair: Roland Herzog Major: Medical Sciences Immunology and Microbiology Hemophilia A is a n X linked bleeding disorder occurring in approximately 1 in 5000 males. The disease is cause d by a deficiency of clotting factor VIII (FVIII) resulting in bleeding episodes that can be fatal to an affected individual The only treatment for hemophilia A i s protein replacement therapy which is both expensive and is often complicated by formation of antibodies against FVIII Thus, therapies that address the burden of costly infusions as well as the immune response to FVIII are highly d esired. The use of liver directed adeno associate d viral (AAV) gene therapy was i nvestigated as a means of both phenotypic correction and tolerance induction to the FVIII transgene in a hemophilia A mouse model. T rea tment with a liver specific AAV8 hFVIII vector either sensitized or tolerized BL/6 129/sv or BALB/c hemophilia A mice, respectively, to an intravenous hFVIII challenge protocol. B cell depletion concurrent with gene therapy reduced the antibody response in AAV treated BL/6 129/sv HA mice. All of the mice had minimal correction of phenotype correlating to ~1% of normal FVIII activity which did not persist past the hFVIII challenge. Codon optimization of the transgene (COhFVIII) not only increased expression of hFVIII to a


11 peak of ~10% and persiste nt level of 3 5%, but also induced tolerance in both strains. Consistent with previous studies, this tolerance was found to be mediat e d by T regulatory cells. B cell depletion was also used to develop a protocol to prevent the development of anti AAV capsi d antibodies and as an adjuvant to reverse established inhibitors. Reduction of capsid specific antibodies was achieved but not to a level that allowed vector re administration. Furthermore, delivery of AAV COhFVIII to mice with pre existing inhibitors sug gests that gene transfer with or without B cell depletion may reduce inhibitors. These results provide a significant step forward in gene therapy for hemophilia A by demonstrating that the immune responses to FVIII may be mitigated by liver specific expre ssion of the transgene. Furthermore, B cell depletion (such as with the human drug rituximab) may prove to be a essential adjuvant in gene therapy for the prevention of an immune response to both the transgene and the viral vector.


12 CHAPTER 1 INTRODUCTIO N Hemophilia Hemophilia A is a rare X linked recessive bleeding disorder occurring in approximate ly 1 in 5000 live male births. The earliest documentation of the disease dates back to the second century when a rabbi observed that sons of carriers were susc eptible to bleeding episodes following circumcision 1 The reduced ability to form a blood clot is the most salient manifestation of the disease and is due to a deficiency in clotting factor VIII (FVIII). The FVIII protein is a large (~300kD), multi sub unit protein consisting of 2332 amino acids and is produced primarily in the liver 2 3 It exists in circulatio n in two parts: a heavy chain consisting of domains A1, A2 and B domains and a light chain with the A3, C1 and C2 domains (Fig. 1 1) Factor VIII is a part of the intrinsic coagulation cascade and is activated by thrombin whereupo n it then works with activ ated factor IX (FIX) to cleave and activate factor X (FX) (Fig. 1 2 ) 4 Activated FX cleaves pro thrombin to thrombin which drives the production of a cross linked fibrin clot that ceases the flow of blood and limits blood loss. Hemo philia B is caused by a reduction in activity or absence of FIX due to mutations in the f9 gene. Since FIX and FVIII act as cofactors at the same step in coagulation, the two diseases are clinically indistinguishable although hemophilia B is considerably m ore rare occurring in about 1 in 20000 males. The congenital form of hemophilia A is caused by mutations in the FVIII gene which limit or eliminate circulating FVIII levels or activity The FVIII gene is found on the long arm of the X chromosome at the lo cus Xq28 1 The gene itself is 186kb, of which 9kb is comprised of its 26 exons with the remainder being intervening or intronic


13 sequences 5 Given that hemophilia A is an X linked disorder, males are much more commonly affected than females and can either inherit the disease by receiving a defective allele from the mother or by a de novo mutation in the f 8 gene. The mutations that cause disease vary greatly but surprisingly, the vast majority (40 50%) are of a specific type consisting of an inversion in intron 22 resulting in less than 1% of normal FVIII levels 6 7 This is the result of homology between a region inside in t ron 22 of the f 8 II over and inversion of the f 8 gene ( Fig. 1 3 ) 6 The remaining patients with hemophilia A harbor mutations that sp an the spectrum from point mutations to large deletions. Severity of disease is directly correlated with the amount of residual FVIII activity. from a complete absence of circulating protein or minimal functional protein. This form of the disease usually results from very disruptive mutations which either prevent FVIII from being produced or sever e ly limit its function. Given the essential role FVIII plays in coagulation, patients with severe disease suffer from spontaneous bleeding episodes. These bleeds are the central complication in severe hemophiliacs who can suffer from frequent bleeding into soft tissues and joints. Bleeding into joints can cause significant swelling leading to arthropy, damage, and can eventually limit joint function to the point of necessitating arthoplasty 8 These bleeding episodes can be fatal if the bleedi ng occurs in the brain in the form of an intracranial hemorrhage and in fact is the leading cause of death among hemophiliacs not infected with HIV 9 The prognosis of hemophilia A improves with increasing amounts of FVIII activity so that patients with 5% FVIII activity) have a milder phenotype characterized by


14 infrequent spontaneous bleeds and generally are managed only during traumatic instan ( >5% FVIII activity ) only requires intervention during surgery or following injury. Treatment o f Hemophilia A The current standard of care for hemophilia is protein replacement therapy where FVIII is administer ed either prophylactically or on demand The source of FVIII has historically been derived from human plasma which carried the risk of viral infection with HIV and hepatitis C. Infection with HIV was a very serious concern amongst hemophiliacs as HIV posit ivity raised the mortality rate from <1% to >10% in severe and moderate hemophilia A patients 10 Since the institution of better screening methods following the AIDS epidemic of the early 1980s, plasma derived FVIII is considered much safer. Recently, FVIII producti on has been possible in vitro utilizing mammalian cell culture and introduction of recombinant FVIII DNA. Either product can be used to treat hemophiliacs with equal efficacy 11 Frequency of FVIII administration depends on the form of the disease. Mild and moderate hemophiliacs can be administered clotting surgical/dental work. Fo r patients with severe hemophilia A, FVIII is generally recommended prophylactically to control spontaneous bleeding events before they occ ur. Given that FVIII has a half life of approximately 8 12 hours, patients usually self administer or are given FVIII infusions every other day which prevents the majority of spontaneous bleeds and better preserves joint function 1 12 Unfortunately, production of plasma derived and recombinant FVIII is limited in the yield of FVIII. Thus, the cost of FVIII is prohibitively high. In the case of sever e hemophiliacs, prophylactic FVIII administration can cost w ell over $100,000 per year per patient 13 As a result,


15 successful treatment of hemophilia A is limited to countries that have access to expensive pharmaceuticals and widespread health insurance coverage (developed areas such as North America and Western Europe) 14 1 5 It is estimated that 70% of hemophiliacs worldwide are untreated, leaving them exposed to the almost certain fatal consequences of the disease 14 15 However, if treated with exogenous FVIII, the prognosis for patients with hemophilia A is generally good. Aside from cost, clotting factor replacement is also limited by the development of neutralizing a ntibodies against FVIII protein termed Inhibitors in Hemophilia A Inhibitors can occur in up to 30% of severe hemophilia A patients. They render treatment with FVIII ineffective and eliminate the o nly long term treatment option. activated FVII but these agents are not suitable for long term use as they are exceedingly expensive. Patients with severe hemophilia are at great est risk of developing inhibitors particularly those with large deletions, nonsense mutations and inversions of either intron 1 or 22 16 It is not entirely clear why inhibitors develop in the case of hemophilia 17 Normal lymphocyte development causes self reactive T and B cells to be removed by a process known as negative selection. T cells mature in the thymus and are presented a variety of self antigens by antigen presenting cells (APCs) in the thymus whi ch express numerous self proteins driven by the pluripotent transcription factor AIRE. If a developing T cell is able to recognize and bind a self antigen presented by these thymic APCs the result will be cell death and elimination of the self reactive T cell. B cells are


16 also similarly negatively selected in the bone marrow during development where engagement of the B cell receptor (BCR) leads to cell death. Self reactive B cells that escape negative selection can further be selected against in the periph ery where they will be rendered anergic if they encounter antigen in the absence of immunological which should have been selected against in the thymus if specific for self antigens. Th us, non affected individuals do not possess FVIII specific T and B cells as a result of this process. Individuals with severe hemophilia A, however, may lack any FVIII antigen expression at all (due to large mutations, see above) leading to the persistence of FVIII specific effector T and B cells and the most likely explanation for the higher percentage of severe hemo philiacs that develop inhibitors 18 Still, the presence of FVIII in the circulation in the absence of any immunological to such a robust humoral immune res ponse. It has been hypothesized that the role of FVIII in damaged surfaces may provide the immune stimulus to mount an immune response to the infused protein, but this has yet to be proven 19 Once established, inhibitors are very difficult to eliminate. The only established method for reversing inhibitors is a process called immune tolerance induction (ITI). This involves delivering frequent high doses of FVIII until the inhibitors are cleared which usually takes more than a year. The exact mechanism of ITI is unknown but both T cell exhaustion (as seen in chronic viral infections) and inhibition of B cell differentiation by supraphysiologic levels of FVIII have been proposed 17 20 Approximately 60 80% of pati ents will experience complete remission of inhibitors but at a high cost 17 21 22 One round of ITI can cost well over $1,000,000 in FVIII alone and


17 over $2,000,000 in total when considering the expense of bypassing agents before and during treatment 22 For those that fail, a second attempt i s usually made at which point supplementing ITI with immune suppressants such as the B cell depleting biologic drug rituximab may show some benefit 23 24 Aside from a less than perfect success rate, ITI has some considerable limitations. For one, delivering daily intravenous injections presents complications with compliance and venous access. A venous access por t can be used but at the risk of infection which tends to exacerbate the existing inhibitor response and lessens the chance of success of ITI 25 Given the high cost, number of patients that fail ITI and other complications, alternative methods to induce immunological tolerance to FVIII are desired. AAV Gene Therapy Adeno associated virus (AAV) is a replication deficient, non pathogenic parvovirus that natur ally infects humans as well as other species including non human primates 26 It contains a 4.7kb single stranded genome with two op e n reading frames: one encoding r ep genes required for viral replication and the other for cap genes which code for the three capsid genes VP1, VP2 and VP3 ( Fig. 1 4 ) In order to generate a productive, replicative infection AAV must be in the presence of a helper virus co infection such as adenovirus or herpes virus. In the absence of such viruses, AAV will establish a latent infection by integrating its genome into a locus on the human chromosome 19. Historically, AAV received little recognition and study due to their lack of pathogenicity. For this same reason however, interest in using these viruses as gene therapy vectors has increased drastically in the past 20 years. AAV vectors can be made by removing all native viral sequences except for the 145bp inverted terminal repeats (ITRs) which are required for packaging DNA into the viral capsid and for


18 initiating second strand synthesis the process of converting the single stranded viral DNA into transcriptionally active double stranded DNA. Approximately 5kb of an expression cassette coding for a therapeutic gene can be placed between these ITRs and packaged into an assembled AAV capsid. AAV vectors can be produced in vitro by transfecting HEK 293 cells with three plasmids: one encoding the vector genome, one coding for the missing viral genes rep and cap which are necessary to assemble the virus and a third providing adenoviral helper genes (E2a, VA, and E4). When these AAV vectors are delivered in vivo they efficiently target non dividing cells where the genome can exist episomally for extended periods of time (up to >8 years in dogs) 27 AAV vectors are notoriously inefficient at infecting APCs such as macrophag es an d dendritic cells which decreases the likelihood that they will be targeted by adaptive immunity 28 AAV has also been shown to be a poor stimulant of the inn ate immune system compared to other vectors such as adenovirus, further limiting the likelihood of an immune response 29 30 Initially, AAV failed to make a significant clinical breakthrough. However, recent successes in clinical trials using AAV for Leber congenital amaurosis an inherited degenerative blindness disorder and hemophilia B strongly support the app lication of AAV as a gene therapy vector 31 36 Liver Directed AAV gene therapy for hemophilia The liver produces many proteins destined for vascular circulation including FVIII, although other extra hepatic cells may contribute to circulating FVIII protein 37 In addition to being the natural site of FVIII productio n, the liver also has access to central circulation and possesses the cellular machinery to translate, process and secrete large proteins such as FVIII in abundance. The liver is also an attractive gene therapy target as it is recognized as an immune privi leged site. This was first discovered in


19 experiments where allogeneic liver grafts were readily accepted in pigs as opposed to the rejection observed in other tissues 38 This prop erty is thought to be due to the fact that the liver is directly downstream of the gut blood flow and thus receives a vast array of innocuous food and commensal bacteria antigens that do not warrant an immune response 39 40 The Herzog lab has previously demonstrated that AAV gene transfer directed at hepatocytes can not only provide sustained transgene expression but also renders immunological tolerance to the transgene. This tolerance is an active phenomenon mediated by CD4 + CD25 + FoxP3 + T regulatory cells (Treg) which are antigen specific and mediate tolerance through a variety of mechanisms including secretion of suppressive cytokines such as IL 10, TGF and by expressing inhibitory surface molecules such as CTLA 4 41 44 Once tolerance is established, the immune syste m is non responsive to antigenic challenge even in the presence of strong adjuvants 45 This has been successf ul in preclinical models of hemophilia B with FIX and has also been extended to induce tolerance in autoimmune disorders such as EAE, a mouse model of multiple sclerosis 46 Early clinical trials using AAV gene therapy in the human liver failed to show any immune response to the transgene but instead resulted in a CD8 + T cell response to the AAV capsid 47 48 However, FVIII is a much more immunogenic protein and thus there is concern of invoking an immune response to the FVIII transgene. Furthermore, if levels of FVIII are ach ieved in clinical trials that are 10%, FVIII protein therapy would still be occasionally nec essitated and tolerance to FVIII would be essential in avoiding the development of inhibitors. Therefore, liver directed AAV gene therapy for hemophilia A is an ideal protocol given the characteris tics of the FVIII protein and the immune privilege of the liver.


20 G ene therapy with AAV has seen a recent encouraging success in clinical trials where 6 individuals with severe hemophi lia B were treated with an AAV8 serotype vector carrying a genome encodi ng a self complementary (double stranded) codon optimized FIX cDNA 36 All 6 patients had sustained levels of FIX ranging from 2 12% which reduced or eliminated their need for supplementary FIX treatment and allowed for a dramatic improvement in quality of life even to the point where one patient now is a marathon runner. This is a major step forward for safety and efficacy of AAV gene therapy with all treatments groups showing an improvement in quality of life. The only adverse events were the development of a capsid specific CD8 + T cell response against transduced hepatocytes in two patients, which was easily controlled with prednisone and did not negatively affe ct treatment efficacy. W hile encouraging, extending these successes to hemophilia A presents major challenges. Preclinical studies have been limited by the large 7.3kb FVIII cDNA which exceeds the packaging capacity of AAV, the immune response to FVIII, p oor activity of human FVIII in murine models and inefficient expression and secret ion of FVIII protein Despite these challenges, preclinical progress has been made to circumvent these barriers Deletion of the non essential B domain of FVIII reduced its c DNA size to 4.5kb which is within the limitation of AAV 49 Alternatively, the FVIII cDNA has been divided into the heavy and light chain allowing co delivery o f two AAV vectors expressing each chain 50 The two chains then recombine intracellularly and comprise a functional FVIII protein capable of correcting disease alt hough concerns about imbalances in chain expression and secretion still remain 50 Since a minimum level of expression from the liver is required to induce toleran ce, poor expression of FVIII from AAV vectors often


21 feed an immune response rather than prevent it. To circumvent this, broad immune suppressi on or gene transfer to neonatal mice (when the immune system is still developing and more permissive) are often re quired 51 54


22 Figure 1 1. The human factor VIII protein. Diagram of the FVIII gene location on chromosome Xq26 along with its 26 exons in ( A) The FVIII protei n before (B) and after C) proteolytic cleavage at indicated sites to the active form of FVIII as the heavy (A1 and A2) and light chains (A3, C1 and C2). From Casta ldo, et al. 2


23 Figure 1 2 Simplified overview of the intrinsic clotting pathway.


24 Figure 1 3 Overview of FVIII common inversion in intron 22 of the FVIII gene. Copies of rows. Crossover in b between upstream A gene and A gene within intron 22 results in inversion indicated in c. From Lakich, et al. 6


25 Figure 1 4 AAV genome str ucture. (A) Layout of the AAV genome with different transcripts (from promoters p5, p19 and p40) along with their proteins to the right of each transcript. In (B) the ITR is shown which is required for DNA packaging and second strand synthesis. From Daya and Berns 26


26 CHAPTER 2 MATERIALS AND METHOD S Mouse strains All animals used were 6 10 week old mal e hemophilia A mice with a deletion in Exon 16 of the f8 gene on either BALB/c or a mixed C57BL/6 129/sv background, kindly provided by Drs. Lillicrap and Kazazian 55 These mice were generated by inserting a neomycin cassette into exon 16 of the murine FVIII gene resulting in undetectable F VIII activity and thus providing a model of severe hemophilia A. Ten week old C57BL/ 6J male mice were purchased from Jackson Laboratory (Bar Harbor, ME) and maintained at the Uni versity of Florida Colle ge of Medicine All animal experiments were performe d according to the guidelines for animal care specified by the Animal Care Services at the U niversity of Florida The Institutional Animal Care and Use Committee approved all protocols for the care and use of these mice. B cell depletion and flow cytometry Mice were depleted of B cells using 10 mg/kg of anti murine CD20 IgG2a antibody (clone 18B12, kindly provided by Biogen Idec Weston, MA ) 56 Mice were given dose s by intravenous ( i.v. ) injection To assess B cell depletion, lymphoid tissues were collected either 1 day or 7 weeks after the second dose. Peripheral blood was collected in heparinized microcapillary tubes via retro orbital plexus, centrifuged at 9600 R PM for 10 minutes at 4C to separate and remove plasma. The remaining cells were re ( BD Biosciences, San Jose, CA) and incubated at room temperature for 15 minutes before further staining for flow cytometry. Singl e cell suspensions of splenocytes and peripheral lymph node cells were made by


27 g for 10 minutes. This pellet was resuspended in PBS for counting and 10 6 added to (BD Biosciences) and incubated at room temperature for 15 minutes. Staining of all single cell suspensions was performed with fluorescently labeled antibodies for the B cell marker CD19 ( clone 1D3 V450 conjugated, BD Biosciences ) as well as CD3 (clone 145 2C11 FITC conjugated, BD Biosciences) or CD4 ( clone L3T4 FITC conjugated, BD biosciences ) to detect T cells. Lymphocyte gating was determined by forward scatter and side scatter following back gating for the area containing CD19 + cells. A B D LSRII flow cytometer (BD Biosciences, Frederick, MD, USA) and FCS Express 4 software (Denovo Software, Los Angeles, CA, USA) were used for ana lysis AAV vector construction Mice were given a tail vein injection of 10 11 vg/mouse of AAV8 containing either w ild type or codon optimized B domain deleted human hFVIII (BDD hFVIII) under a liver specific promoter. This construct was assembled by replacing the transthyretin promoter of the vector published by Lu et al. 52 w ith the ApoE/hAAT enhancer/promoter (outlined in Fig. 2 1) 44 This was accomplished by digesting the pAAV TTR hFVIII (from Lu, et al. 52 ) plasmid with SacI and XhoI and isolating the fragment containing the AAV inverted terminal repeats (ITRs) A signal. Th e pA AV hAAT hFVIII plasmid containi ng the liver specific human alpha 1 antitrypsin (hAAT) promoter and human growth hormone (hGH) poly A was also digested with MluI/SacI and the fragment containing the hAAT promoter and hFVIII cDNA was ligated to the A and backbone of the previo usly mentioned fragment to generate pAAV hAAT hFVIII mini with an approximately 5.3kb of DNA from ITR to ITR. The codon optimized hFVIII cDNA was synthesized b y GeneArt (Regensburg, Germany) and


28 cloned into the wild type construct by replacing the wild ty pe hFVIII using ClaI/ XhoI digestion and ligation into the previous plasmid (sequence in Appendix ) Viral vectors using these plasmids were produced by triple transfection of HEK 293 cells and purified by iodixonal gradient centrifugation as published prev iously 57 Viral titers were assessed by quantitative slot blot hybridization and confirmed via western blot using antibodies against viral capsid. Blood Col lection Plasma samples were collected via either the retro orbital plexus for ELISA based assays and flow cytometry or via the tail vein for aPTT and Bethesda assays. For retro orbital plexus, mice were anesthetized with isofluorane and immediately bled us ing heparinized microcapillary tubes. Tail vein bleeding was performed by heating the mice under a 125w lamp for 5 minutes and then anesthetizing with isofluorane. The tail was of blood collected in 1.5mL centrifuge tubes citrate. Tails wounds were ligated with non absorbable sutures and cauterized with sliver nitrate. Following collection via either method, plasma was collected by centrifuging samples at 9600 RPM in a table top centrifuge fo r 10 minutes at 4C and plasma removed by pipetting the supernatant. Intravenous hFVIII challenge protocol For challenges with hFVIII in mice, 1 IU BBD hFVIII (Xyntha, Pfizer, New York, i.v. once a week for 4 week s unless otherwise indicated. ELISA protocols Enzyme linked immunosorbant assay (ELISA) for hFVIII specific IgG1 was performed by coating 9 6 well plate (Corning, Tweksbury, MA) with 1ug/mL of hFVIII


29 (Xyntha, Pfizer, New York, NY) in coating buffer (0.68g Na 2 CO 3 ; 3.675g NaHCO 3 in 500mL H 2 O at pH 9.2) at 4C overnight. Standards for IgG1 were coated onto the plate in duplicate by serially diluting mouse IgG1 (Sigma Aldrich, St. Louis, MO) in coating buffer beginning at 2000ng/mL. Plates were then washed and blocked with dilution buffer (0.05% tween 20 and 6 % BSA in PBS) for 1h at room temperature. Following plate washing, plasma samples were applied in duplicate to plates at a 1:20 dilution and incubated at 37C for 2 hours. After washing, IgG1 antibodies wer e detected using a rat anti mouse IgG1 horseradish peroxidase conjugated antibody (R&D Systems, Minneapolis, MN) diluted 1:2000 in dilution buffer for 2h at 37C. Quantitation of antibody levels was performed by measuring optical density (OD) at 450nm of e ach well following room temperature incubation with SigmaFast OPD (Sigma Aldrich, St. Louis, MO) using a Bio Rad 680 microplate reader (Hercules, CA). Calculation of antibody concentrations was performed using MPM III v1.6 software. Anti AAV8 IgG2a antibod ies were measured in the same manner with some exceptions. First, an IgG2a standard was used (Sigma Aldrich, St. Louis, MO) at a starting concentration of 250ng/mL. Plates were coated with AAV8 virus at a concentration of 5x 10 1 0 vg/mL in coating buffer. Sa mples were diluted at 1:10 in dilution buffer incubated overnight at 4C and antibodies detected with a horseradish peroxidase conjugated rat anti mouse IgG2a antibody (R&D Systems, Minneapolis, MN). Plasma hFIX levels were also measured by ELISA. The sa me 96 well plates were coated with a monoclonal antibody recognizing human (and not murine) FIX (FIX 1 monoclonal anti hFIX Clone HIX 1; F2645 Sigma Aldrich, St. Louis, MO) dilute d 1:850


30 in coating buffer and incubated overnight at 4C. After washing, plat es were blocked with blocking buffer (5% non fat dry milk in PBS with 0.05% tween 20) for 2h at room temperature. Plasma samples were diluted 1:10 in blocking buffer. A standard was generated by diluting normal human plasma (TriniCHECK Level 1; Trinity Bio tech, Wicklow, Ireland) to 200ng/mL of FIX serially to 3.125ng/mL. Both samples and standards were applied to the plate in duplicate after washing and incubated overnight at 4C. Goat anti human FIX horseradish peroxidase conjugated antibody was diluted 4. performed as described above. Clotting assays Collected blood was used to determine activated par tial thromboplastin time (aPTT) as a measurement of clotting time Assay deficient plasma (Hematologic Technologies Inc., Vermont) ( TriniCLOT, Trinity Biotech, Wickow, Ireland) 25mM calcium chloride was added and recording of time to clot was immediately started using a fibrometer (Fibrosystem TM Block Scientific Inc.). Bethesda assay was performed as previously described 58 59 In detail, plasma samples were serially diluted in Imidazole buffer (MDA, Baltimore, MD) from N to Each diluted sample was mix ed w human plasma (TriniCHECK Level 1; Trinity Biotech, Wicklow, Ireland) and incubated at 37C for 2 hours. Control samples of normal human were dilut ed in Imidazole at 1:2 and 1:4 and were also incubated at 37C for 2 hours. After incub ation, all samples were transferred to ice. Standards were also made from 20% to 0% of normal human plasma


31 in Imidazole. Normal aPTT was then performed on all samples, controls and standards and Bethesda units were calculated using the controls and standar d so that 1 Bethesda unit equals a reduction of FVIII activity of 50% as described in Verbruggen, et al. 59 For hFVIII % activity, two methods were used. Fi rst, a standard curve of 0 10% was made using the same BDD hFVIII used for challenge diluted in n ave hemophilia A mouse plasma and aPTT times measured. The formula for this resulting regression was used to input the sample aPTT time and hFVIII activity de termined. Alternatively, a chromogenic assay was used ( Chromogenix Coatest SP4 Factor VIII, Diapharma, West Chester ) 20% FVIII activity in a 96 well plate. Adenovirus challenge BALB/c HA mice w weeks apart as before. Seven weeks after the second injection of CD20 mice were given 10 11 p/ms of an E1/E3 deleted human serotype 5 adenovirus expressing the lacZ gene (Ad LacZ) Two weeks after Ad LacZ challenge, mi ce were bled and plasma was collected for analysis of anti gal antibodies and neutralizing antibodies against adenovirus. To ascertain the level of anti adenoviral neutralizing antibodies, plasma samples were diluted serially from 1/16 1/512 in DMEM (Life Technologies, Carlsbad, CA) and incubated with Ad GFP in 53.5 L total volume for 1.25h at 37 C and 5%CO 2 Immediately following incubation, 14 L of the plasma/Ad GFP was added to one well of a 96 well plate containing 10 5 cells/well of HEK 293 cells in 300 L total for an MOI of 250. Each sample dilution was plated in duplicate along with uninfected wells and control wells mixed with nave mouse plasma. After 20h of incubation at 37 C, 5%CO 2 cells were collected and analyzed via flow cytometry for GFP expres sion. Percent


3 2 positive cells were counted at dilutions of 1/16 and Ad LacZ infected plasma compared with nave plasma. For anti transgene antibody levels, 96 well ELISA plates ( Easy Wash, Corning, Tweksbury, MA) were coated with gal protein at 1ng/ L in coating buffer along with serially diluted standards of mouse IgG1 (Sigma Aldrich, St. Louis, MO) from 2000ng/mL to 15.625ng/mL and incubated overnight at 4 C overnight. After washing, plates were blocked with dilution buffer for 1h at room temperature an d then washed again. Plasma samples were diluted 1:20 and added to gal coated wells and incubated for 2h at 37 C. After washing, IgG1 antibodies were detected using a rat anti mouse IgG1 horseradish peroxidase conjugated antibody (R&D Systems, Minneapolis, MN) diluted 1:2000 in dilution buffer for 2h at 37C. Quantitati on of antibody levels was performed by measuring optical density (OD) at 450nm of each well following room temperature incubation with SigmaFast OPD (Sigma Aldrich, St. Louis, MO) using a Bio Rad 680 microplate reader (Hercules, CA). Calculation of antibod y concentrations was performed using MPM III v1.6 software. Immunohistochemistry For immunohistochemical staining of hFIX producing hepatocytes, liver cryo sections from two mice per group were obtained 10 weeks post injection of AAV2 hFIX and analyzed for hFIX expression. Briefly, liver sections in O.C.T. medium (Sakura Finetek USA, Torrance, CA) were fixed in acetone at room temperature for 10 min, washed, and blocked with 5% donkey serum in PBS for 15 min at room temperature. Goat anti hFIX antibody (Aff inity Biologicals, Ancaster, ON, Canada) was applied in a 1:200 dilution in 2% donkey serum for 30 min after washing. Samples were washed and incubated with Alexa Fluor 568 conjugated donkey anti goat IgG (Molecular


33 Probes=Invitrogen, Eugene, OR) diluted 1 :200 in 2% bovine serum albumin (BSA) in PBS for 30 min at room temperature. Fluorescence microscopy was performed with a Nikon E800 microscope (Nikon, Tokyo, Japan). Images were captured with a CoolSNAP Pro camera and analyzed with Image Pro Plus softwar e (Media Cybernetics, Bethesda, MD). Percent positive cells were estimated by counting stained hepatocytes in 10 different frames per mouse and dividing by the average cell number per frame. T cell assays and adoptive transfer Single cell suspensions of s p lenocytes isolated from mice in each group were generated as described above and 10 6 cells/well were cultured in 12 well plates in RPMI mercaptoethanol, 100 mM insulin/transferrin/selenium, glutamine and penicillin /streptomycin e 37C and 5% CO 2 Cells were then removed and stained with APC efluor780 conjugated anti mouse CD3 (clone 17A2, eBiosciences, San Diego, CA), Alexa700 conjugated anti mouse CD4 (clone RM4 5, eBiosciences, San Diego, CA) and V500 conjugated anti mouse CD8 (clone 53 6.7, BD Biosciences, San Jose, CA). These cells were sorted for CD3 + CD4 + cells on an ARIA cell sorter (BD Biosciences, San Jose, CA, USA). RNA was immediately extracted from these cells using the Qiagen RNeasy isolation kit (Valencia, CA, USA) and used for quantitative RT PCR with an SA Biosciences custom RT PCR array (Qiagen, Valencia, CA) For adoptive transfer, s plenic CD4 + CD25 + cells were purified with a magnetic isolation kit from Miltenyi B iotec (Bergisch Gl adbach, Germany), which achieves ~85% purity. These cells were then pooled and adoptively transferred to nave BALB/c HA mice at 10 6 cells/mouse via tail vein injection. Recipient mice were challenged 24 hours


34 later via subcutaneous injec tion of 1 IU hFVIII in Sigma Adjuvant System (St. Louis, MO, USA).


35 Figure 2 1. Cloning strategy for pAAV hAAT hFVIII mini vector plasmid. Both pAAV TTR hFVIII mini (as in Lu, et al. 52 ) and pAAV hAAT hFVIII hGH vector containing the ApoE/hAAT promoter with MluI and XhoI restriction enzymes. The fragment from the pAAV TTR hFVIII mini containing the minimal poly A sequence and ITRs was ligated to the pAAV hAAT hFVI II hGH fragment containing the hAAT promoter and hFVIII transgene to yield the pAAV hAAT hFVIII mini transgene with approximately 5.3kb of DNA between the ITRs.


36 CHAPTER 3 TRANSIENT B CELL DEP LETION IN COMBINATIO N WITH LIVER DIRECTED GENE THERAPY Backgroun d Given the high cost, invasiveness and long time to r emission of inhibitors strategies to improve ITI or alternative means to induce tolerance to FVIII are highly desirable. One approach is through the use of the biologic drug rituximab a monoclonal ch imeric antibody directed against human CD20 originally developed to treat B cell lymphoma. Rituximab efficiently depletes CD20 expressing B cells via several mechanisms that include complement fixation antibody mediated cellular cytotoxicity and direct in duction of apoptosis 60 CD20 is expressed from the early pre B cell stage to mature B cells and short lived plasma cells but not by long lived plasma cells. Rituximab has been investigated for use in antibody mediated autoi mmune diseases such as acquired hemophilia, systemic lupus erythematosus, rheumatoid arthritis, multip le sclerosis, myasthenia gravis and others 61 Several case r eports and one national survey have revealed that rituximab can improve ITI in hemophilia A patients, especially in cases where patients have previously failed traditional ITI 24 62 63 Successful reversal of an inhibitor against factor IX (FIX) that had formed in a non human primate after gene therapy was also repo rted using rituximab combined with cyclosporine A 64 However, pre clinical studies using CD20 in hemophilic animals or in gene therapy for hemophilia are very limited. B cell depletion as a potential means of preventing (rather than reversing) inhibitor formation has also not been studied. I mportantly liver directed gene therapy with adeno associated virus (AAV) can provide both long term phenotypic correction and immune tolerance to FIX in hemophilia B animal models 44 Success in animal models has led to two clinical trials


37 for hemophilia B using liver directed AAV g ene therapy 65 Hemophilia A has been more difficult to treat with AAV gene therapy due to the increased immunogenicity of FVIII limitations in the packaging capacity of AAV and poor expression of FVIII. Typically, transient immune suppression, high vector doses, the use of canine FVIII (which has a higher specific activity in mice than hFVIII) and mice of C57BL/6 strain background (which is more promiscuous to hepatic AAV transduction), or a combination of these methods was needed to achieve lo ng term correction in hemophilia A mice 50 52 Here, we investigate liver directed AAV gene therapy in different strains of hemophilia A mice to induce tolerance t o hFVIII either alone or in combination with transient B cell depletion. Study Design for Transient B Cell Depletion and Gene therapy We sought to identify pathways toward immune tolerance to hFVIII in gene therapy and to determine the effect of transient B cell depletion on hF VIII specific immune responses using an anti muri ne CD20 monoc lonal antibody comparable to rituximab Hemophilia A mice on either a mixed BL/6 129/sv (BL/6 129/sv HA) or a BALB/c (BALB/c HA) background were divided into two treatment groups (Fig 3 prior to receiving 10 11 vg/mouse of an AAV8 vector expressing B domain deleted hFVIII under the liver specific hAAT promoter (AAV8 h F8). Two weeks following AAV8 hF8 inje hF8 vector (but not e were challenged with weekly i.v. infu sions of 1 IU hFVIII per mouse for 4 weeks (which reliably results in inhibitor formation in both strains). In case of BALB/c HA mice, another 1 month


38 challenge with weekly hFVIII injections was performed starting 22 weeks after gene transfer, and blood wa s collected within 15 min after the fourth hFVIII injection. B Cell Depletion with B cell depletion in lymphoid organs was assessed 1 day and 7 weeks following the investigated one day after the (Fig. 3 2 and data not shown for 1 d post first injection). initial hFVIII challenge), this B cell population had returned to normal levels in the spleen and lymph nodes and near normal levels in peripheral blood (Fig. 3 2 ) Absolute T cell numbers, as measured by total CD3+ cells per 10,000 total cells in each tissue were not reduced by treatment and were only significantly different from control mice in the lymph node where treated mice had higher T cell numbers ( Fig. 3 3 ). To further investigate immune competence at 7 weeks post last injection (the time of hFVIII challenge) a separate set of mice were administered 10 11 vp/m ouse of a denovirus expressing the gal transgene. Following this adenoviral challenge, neither neutr alizing antibodies against the adenovirus capsid nor antibodies against the treated mice (Fig. 3 4 ). C onsidering the data in figures 3 2 to 3 4 mice were deemed to be immune competent at time of hFVIII challenge (7 weeks po st ) in all subsequent experiments B Cell Depletion Concurrent with Gene Transfer Renders BL/6 129/sv HA Mice Hyporesponsive to hFVIII BL6 129/sv HA showed modest correction of clotting times below that of untreated mice but this correction was minim treatment (Fig. 3 4 A ). Without further manipulation, this level of correction was


39 sustained for at least 4 months. However, those mice that were challenged with supplementary hFVIII therapy lost corre ction concurrent with the development of inhib itors (Fig. 3 4 C matched mice (no gene transfer) receiving identical hFVIII challenge had an average Bethesda titer of 15444 BU, whereas mice receiving AAV8 hF8 had a higher average titer of 3 3688 ng/mL, suggesting that AAV hF8 treatment primed the mice against hFVIII. However, transient B cell depletion with hF8 gene transfer resulted in subsequent hypo responsiveness to hFVIII with a significantly lower Bethesda tit er (2211 BU). These titers were 15 times lower than in mice receiving AAV8 fold lower than mice receiving only protein challenge. These results were also reflected in total circulating anti hFVIII IgG1 as determined by ELISA (Fig. 3 4 B ). Mice in the significantly lower than both control and AAV8 hF8 mice (which had titers of 59881520 ng/mL and 7001867 ng/mL, respectively). Gene Transfer Renders BALB/c HA M ice Hyporesponsive to hFVIII regardless of B Cell Depletion The above experiments were repeated in hemophilic mice carrying the same mutation on a BALB/c background. All BALB/c HA mice receiving AAV8 hF8 had ~1% FVIII activity, which was again sustained in the absenc e of further manipulation (Fig 3 5 A). Coagulation times of mice challenged with hFVIII increased to or near the titer inhibitors (<5 BU), and undet ectable anti hFVIII IgG 1 (Fig. 3 5 B and C). This was in contrast to control mice (no gene transfer or hFVIII IgG1 and significantly higher


40 inhibitor titers (11 fold and 4 mice, respectively) (Fig. 3 5 B and (following identical time lines). These mice developed similar Bethesda titers as the other contr ol mice (Fig 3 5 B), albeit anti FVIII IgG1 titers were somewhat lower 5B ). We next wanted to determine if hypo responsiveness would be maintained after another round of challenge, and if t he supplementary exogenous hFVIII could correct the aPTT times of vector treated mice. The lack of humoral immune response was specific IgG (data not shown). In nave BALB/c HA mi ce, injection of 1 IU hFVIII resulted in correction of the aPTT to 492 se c at a 15 min time point (Fig. 3 5 of 583 sec and 604 sec, respectively. While this level of correction was not quite as good as for nave mice, the difference did not reach statistical s ignificance. No or marginal correction was achieved in control mice that had form ed high titer inhibitors (Fig. 3 5 D). In summary, we conclude that treatment of BALB/c HA mice with AAV8 hFVIII with or without B cell depletion confers a level of long term t olerance to hFVIII that protects from high titer inhibitors in subsequent protein therapy, thereby allowing for correction of coagulation with exogenous clotting factor T Cell Tolerance to FVIII in BALB/c HA Mice Inhibitor formation is dependent on T hel p and may be controlled by Treg. The BALB/c HA strain was chosen for analyses of T cell responses to FVIII because this strain responded better to tolerance induction and, representing a pure strain


41 background, could be used for adoptive transfer studies. Quantitative RT PCR on in vitro hFVIII stimulated CD4+ splenocytes was performed to test for induction of T cell tolerance. Responses in mice previously treated with AAV8 hFVIII only, or control mice (that had received protein challenge only) were measured after the last round of challenge with hFVIII pro tein had been completed. Bulk splenocytes were stimulated in vitro with hFVIII and subsequently sorted based on CD4 expression. RNA from CD4+ T cells was used in qRT PCR to analyze the transcription of 9 different genes related to T cell immunity. CD4+ cel ls from control mice, which had developed high titer inhibitors, showed increased transcription of a mix of Th1, Th2 and suppressive cytokine/marker genes including IL 2, IL 4, IL 13, IFN 4, and IL 10, an d more modestly of TGF 6 A). Mice receiving AAV8 hFVIII alone showed a substantial reduction in their IL 2 and IL 10 responses and a partial reduction in IL 4 and IL 13 gene expression, while IFN similarly up F8 + C cell unresponsiveness. Only IL 10 and CTLA 4 were marginally induced in response to FVIII. The potential role of Treg in tolerance was further investigated by adoptive transfer first challenge Adoptive transfer of cells from either group only modestly suppressed anti hFVIII formation in recipient mice (which did not reach statistical significance, Fig. 3 6 B).


42 hFVIII protein administration during reco very of B cells fails to achieve lasting tolerance Next we investigated whether B cell depletion could be combined with hFVIII protein administration to induce tolerance to hFVIII. Mice from each strain were given and 3. Starting at week 7 ( dose), when B cells were still recovering, mice were given 4 weekly IV doses of 1 IU of hFVIII. Mice were bled 2 weeks later to test for antibody formation. Initially, mice treated responsive or hyporesponsive to challenge. Only 1 of 8 BL/6 129/sv HA mice had detectable anti FVIII antibodies (513 ng/mL; 3 BU), whereas 3/7 BALB/c HA mice had detectable (albeit low titer) inhibitors (Fig. 3 7 A and B ). As titer inhibitors. The hFVIII tr eatment protocol was repeated at weeks 14 17 and 21 24. After the second challenge, only 2 of the BL/6 129/sv HA mice remained unresponsive, whereas all other mice developed an antibody response similar to controls after the first challenge. As expected, c ontrol mice showed further increases in both IgG and Bethesda titers. By the response to hFVIII similar in magnitude to the controls after two challenges (Fig. 3 7 A and B ). There fore, transient B cell depletion merely delayed the antibody response to hFVIII protein. Interestingly, when comparing the magnitude of both IgG1 titers and Bethesda titers for mice in these experiments, the two strains show several differences. First, BL /6 129/sv HA mice have higher anti hFVIII IgG1 and Bethesda titers (Figs. 3 4 B and C; 3 5 B and C). Furthermore, IgG1 and Bethesda titers after one 4 week challenge only positively correlate in BL/6 129/sv HA mice (Fig. 5 7A). BALB/c HA mice eventually


43 do form a positive correlation but only when data from multiple rounds of challenge are included again highlighting the lack of a robust Th2/antibody response of this strain to hFVIII challenge and perhaps explaining the greater permissiveness to tolerizatio n ( Table 3 2 and Discussion).


44 Figure 3 1. B cell depletion and gene therapy protocol Hemophilia A mice on either a BL/6 129/sv or BALB/c background were administered 10mg/kg anti mCD20 IgG2a 1 day prior to receiving 10 11 vg/ms AAV8 expressing hFVI II under the liver specific hAAT promoter (AAV8 hAAT FVIII) Mice were bled at indicated time points to asse s s B cell recovery and to monitor clotting times (via aPTT). weeks. Two w eeks following challenge, mice were bled again to monitor phenotypic correction as well as immune responses to hFVIII. A subset of BALB/c mice were again challenged as before starting at week 22 with the exception that mice were bled 10 15 minutes immediat ely after the last challenge to monitor clotting correction with exogenous hFVIII.


45 Figure 3 2. B cell depletion following cd20 administration Mice were i.v. injected with 10mg/kg CD20 IgG2a 3 weeks apart and sacrificed at indicated time points for flow cytometry analysis of B cell population s in the spleen, lymph node and peripheral blood. Left panel shows the gating strategy for live lymphocytes based on forward and side scatter while the middle and right pane l s show percent CD19 + lymphocytes for e ach organ at indicated time points following CD20 administration. Each plot is a representative image of three mice in each group.


46 Table 3 1. CD3 + T cells following B cell depletion. Tissue CD3 + /10,000 cells Control CD3 + /10,000 cells CD20 Spleen 2616 119 3556 878 Lymph Node 6703 39 8751 270 Peripheral Blood 2089 844 1784 312 T cells populations were determined in indicated organs 1 day following a second injection of CD20 in BALB/c HA mice. T cells were determined by CD3 + lymphocytes (lymphocytes determined by FSC/SSC gate as in Fig. 4 2) per 10,000 total cells. Means with standard deviation are shown for each subset. indicates p<0.05 vs. control as determined by t test.


47 Figure 3 3. I mmune response to Ad LacZ in B c ell depleted mice. Seven weeks following a second according to the schedule in Fig. 4 1 mice were given 10 11 vp/ms of Ad LacZ. Four weeks after challenge, mice were bled to asse s s antibodies against the gal transgene via ELISA in (A) and development of neutralizing antibodies in (B ) For ( B ) mouse plas ma from treated mice also receiving Ad LacZ nave challenged mice but were challenged with Ad LacZ Ad LacZ GFP at 37C for 1 hour prior to infection of HEK 293 cells. After 24 hours incubation, cells were analyzed for GFP expression by flow cytometry and percent transduction was determined by comparison to incubation with nave mouse plasma. Data shown is mean SD. Comparison of means of test where indicates p value <0.05. NS = not significant.


48 Figure 3 4 B cell depletion and gene transfer in BL/6 129/sv HA mice. BL/6 129 /sv HA mice were treated as described and illustrated in Fig. 4 1. (A) Correction of clotting times as measured by aPTT. Dotted lines represent expected clotting times of an untreated HA mouse, a mouse with 1% hFVIII activity and of a normal mouse. (B) FVIII specific IgG1 2 weeks following challenge with hFVIII as mea sure d by ELISA. (C) Inhibitors of hFVIII as measure d by Bethesda assay 2 weeks after challenge. indicates p value of <0.05 following ANOVA analysis with Tukey post hoc analysis. Data points in A are mean SD. Horizontal bars in B C indicate mean SD.


49 Figure 3 5 B cell depletion and gene transfer in BALB/c HA mice. BALB/c HA mice were treated as described and illustrated in Fig. 4 1. (A) Correction of clotting times as measured by aPTT. Dotted lines represent expected clotting times of an untreated HA mouse, a mouse with 1% hFVIII activity and of a normal mouse. (B) FVIII specific IgG1 2 weeks following challenge with hFVIII as measure d by ELISA. (C) Inhibitors of hFVIII as measure d by Bethesda assay 2 weeks after challenge. (D) Mice were again chal lenged on the same 4 week course. However, on the fourth and final injection, mice were bled 10 15 minutes after hFVIII infusion and this plasma was used in aPTT to determine the level of correction induced by the infused hFVIII. indicates p value of <0. 05 following ANOVA analysis with Tukey post hoc analysis. Data points in A are mean SD. Horizontal bars in B D indicate mean SD.


50 Fig ure 3 6 T Cell Responses in BALB/c HA mice. (A) Following the second round of challenge, BALB/c HA mice were sacri ficed and splenocytes stimulated in vitro with 5 g/mL of hFVIII for 48h. Following stimulation, cells were purified for CD3 + CD4 + T cells by FACS and RNA was immediately extracted for use in qRT PCR to determine fold increase in transcription of several gen es. (B) Following one round of challenge according to the schedule in Fig. 4 1, mice from indicated treatment groups were sacrificed and CD4 + CD25 + splenocytes were purified, pooled by treatment group and adoptively transferred to nave BALB/c HA mice. One day after transfer, mice were challenged with 1U hFVIII in adjuvant subcutaneously. 5 weeks following s.c. challenge plasma antibody titers were measured by ELISA. indicates p<0.05 as measured by test. Bars indicate mean SD.


51 Figure 3 7. Anti CD20 treatment to prevent antibody formation in hFVIII protein replacement therapy. BL/6 129/sv HA and BALB/c HA mice were treated with antibody as outlined in Fig. 4 1 (indicated with large arrows) followed by 4 weeks of hFVIII challenge (indi cated by small arrows) beginning at 4 twice more following the same schedu le. Antibody formation against h FVIIII was measured by Bethesda assay (A) and anti hFVIII IgG1 ELISA (B) tw o weeks after each 4 Gray triangles represent B cell recovery. Data are mean SD for n= 3 5/group.


52 Figure 3 8. Correlation between anti FVIII IgG1 and Bethesda in BALB/c HA and BL/6 129sv HA mice. Mice were challenged with standard 1U hFVIII/ms/week i.v. for 4 weeks and bled 2 weeks following the fourth challenge. (A) Correlation between hFVIII IgG1 titers as measured by ELISA and Bethesda titers after only 1 round of hFVIII challenge with regressio n line shown for each strain. (B) Correlation between hFVIII IgG1 and Bethesda titer including data from one to three challenges.


53 Table 3 2. Comparison of immune response to hFVIII in BALB/c and BL/6 129/sv hemophilia A mice. Cytokine/Marker BL/6 129/sv BALB/c IL 2 ++ + IFN + IL 4 ++ ++ IL 6 ++ IL 10 +/ ++ IL 13 ++ IL 17 TGF 1 +/ FoxP3 + Il2ra (CD25) +/ CTLA 4 ++ Data from figure 3 6A for BALB/c HA mice and previously published data for BL/6 129/sv mice from Moghimi, et al. 66 comparing cytokine and surface marker upregulation as measured by quantitative RT PCR following in vivo hFVIII challenge and in vitro restimulation of sp lenocytes.


54 CHAPTER 4 CODON OPTIMIZATION O F HUMAN FVIII TRANSG ENE IMPROVES EXPRESS ION AND INDUCTION OF TOL ERANCE IN HEMOPHILIA A MICE REGARDLESS OF STRAIN BACKGROUND Background Our previous experime nts with AAV8 expressing a wild type FVIII cDNA demonstr ated that immunological tolerance or hyporesponsiveness can be achieved depending on strain background and concurrent administration o f the B cell depleting (1% or less of normal FVIII activity) transient correction that did not withstand intravenous FVIII challenge. Any FVIII activity below 10% would necessitate on dema nd infusion with FVIII in the case of trauma or surgery. Thus, even if the levels of correction seen in our previous studies were translated to human trials the lack of tolerance seen in BL/6 129/sv HA mice would be unacceptable as the development of inhi bitors would not only render the gene therapy benefits moot but would also prevent treatment with exogenous FVIII. Prior work in our lab has shown a correlation with increased levels of antigen expression and increased tolerance induction following liver d irected gene therapy with AAV vectors 67 Wild type hFVIII has notoriously low expression which has hampered clinical trials for hemophilia A regardless of gene de livery platform 68 69 The barriers to expression seem to be at translation and secretion steps of the protein as mRNA levels do not correlate well with FVIII antigen levels in vivo 37 Sever a l attempts have been made to improve expression for gene therapy purposes inclu ding directed engineering of the hFVIII cDNA to more closely resemble porcine FVIII which secretes at a much higher rate 70 71 While these modifications have yielded drastic improvements in the expression of FVIII, the use of porcine FVIII which differs in amino acids from wild


55 type human FVIII may be less desirable due to the uncertain immunological consequenc es of delivering a partially novel antigen in the case of mild or moderate hemophiliacs. Thus a strategy to improve transgene expression which does not alter the amino acid sequence of the protein would be ideal. One such strategy is c odon optimization of the transgene which may increase expression of the transgene at the translational level The process of codon optimization is based on the premise that different species will usa ge codon optimization alters the cDNA to incorporate those codons most likely to be used in a given species thereby improving the rate at which a protein is translated 72 Furthermore, the cDNA can be further optimized to remove cryptic splice sites, reduce secondary structure of the mRNA and optimize GC conte nt all in hopes of increasing the efficiency of gene expression while preserving the amino acid sequence of the original cDNA 72 73 This strategy has been successful for multiple proteins utilized across various gene delivery platforms including a recent report where codon optimization of hFVIII improved expression from a lentiviral vector 54 73 78 An alternative approach is to increase the number of transcriptionally active vector genomes following gene transfer. Our AAV8 hFVIII constructs carry a single stranded genome, which must be converted to double stranded DNA before transcription in a this is a rate limiting step and can be overc ome by removing a DNA binding protein that binds to the single stranded genome and prevents second strand synthesis 79 Removal of this protein can be facili tated by dephosphorylation of FKBP52 via protein


56 phosphatase 5 (PP5). This PP5 protein can be provided by co AAV virus carrying a double which will be transcribed and transla ted imme diately (Fig. 4 1). This strategy has been is unknown. Thus, we investigated whether use of a helper virus containing sc PP5 transgene or a codon optimized hFV III transgene could improve expression from our AAV8 vector and if this would impact tolerance induction to the transgene. Codon optimization improves hFVIII expression from AAV8 and yields long term correction of phenotype at therapeutic levels We tested an AAV8 vector containing codon optimized hFVIII (AAV8 CO hFVIII) under the same liver specific promoter as our vector containing the wild type hFVIII transgene (AAV8 hFVIII) Doses of this vector identical to those used with AAV8 hFVIII resulted in drastic improvements of clotting times in both strains, with average aPTTs declining from 788 sec to 372 sec in the BL/6 129/sv HA mice and from 908 sec to 432 sec in BALB/ c HA mice after 4 weeks (Fig. 4 2A ). Clotting times generally rose over the following 4 weeks but then stabilized. Means of 534 sec in BL/6 129/sv HA and 616 sec for BALB/c HA mice were measure d at 18 weeks. FVIII activity was then measured with one of two functional assays. First by chromogenic assay which indicated that FVIII activity p eaked at 12.7 5.1% stabilizing to 0.8 1.7% for BL/6 129/sv HA mice and 16.6 2% stabilizing to 0.7 1.4% in BALB/c HA mice (Fig. 4 2B) Alternatively a standard curve was generated by diluting hFVIII in nave hemophilia A mouse plasma and measuring the aPTT a method which we found to be more sensitive and precise. With this method peak hFVIII activities were 10 .46 % and 81 .6 % in BL/6


57 129/sv HA and BALB/c HA mice, respectively, with a d ecline to 3.82.8 % and 2 .12.5 % (Fig. 4 2C ). Immunological tolerance to hFVIII following AAV8 COhFVIII gene transfer Factor VIII activity in this range would result in a vast improvement in quality of life for an individual with severe hemophilia A but again would necessitate occasional infusion of hFVIII. Thus it was essenti al to investigate tolerance to i.v. hFVIII following AAV8 COhFVIII gene therapy. Following challenge with hFVIII protein, 3 of 4 mice in the BL/6 129/sv showed no antibody response to hFVIII with the other having a muted IgG1 titer of 603 ng/mL compared wi th an average of 59883040 ng/mL in control animals challenged in the same manner (Fig. 4 3 A and 3 4B ). This animal also had a low Bethesda titer of 1.4 BU (Fig. 4 3 B ), resulting in a longer aPTT (64.9 sec) compared to the other mice at 18 weeks post vecto r injection. In BALB/c HA mice, no anti hFVIII antibodies were detected in 2/4 mice, while the other 2 mice had very low but detectable antibody titers after challenge (830 and 443 ng IgG1 anti hFVIII/mL, corresponding to 1.4 and 0.6 BU, Fig. 4 3 A and B ). Average anti FVIII titers of the control group were 2785693 ng/mL or 175 BU (Fig 3 5 B and C) In summary, a highly significant 9 to 10 fold reduction in antibody response to hFVIII protein was achieved with the AAV8 CO hFVIII construct including a complet e lack of response to hFVIII protein in 5/8 mice. Adoptive transfer of CD4+CD25+ Treg was also performe d as above with tolerant AAV8 COh FVIII treated BALB/c HA mice serving as donors. Recipient mice showed a significantly lower antibody response to FVIII c ompared to mice receiving CD4+CD25+ T cells from na ve mice (Fig. 3 6B ). Therefore, the use of CO hFVIII induced a more potent Treg response than wild type cDNA consistent with our previous studies that suggest a positive correlation between expression lev els and tolerance induction


58 Co administration of an AAV helper virus encoding a sc PP5 transgene under a liver specific promoter improves hFIX expression but does not effect expression of COhFVIII Given the relative insensitivities of our FVIII activity a ssays, especially near or below 1%, we first wanted to test the helper virus strategy with a hFIX vector since hFIX c an be detected in wild type C57 BL / 6 mice with a quantitative ELISA. We used an AAV2 vector carrying a single stranded hFIX transgene under the same liver specific promoter as our hFVIII vectors and delivered this i.v. at a dose of 10 10 vg/ms or 5x10 10 vg/ms with or without 10 10 vg/ms of a scAAV8 PP5 vector expressing P P5 under the liver specific transthyretin (TTR) promoter. At a dose of 10 10 vg/ ms of scAAV2 hFIX, a dministration of a helper virus was able to i ncrease expression of hFIX from sub therape utic to therapeutic levels (216 22 vs. 30 11 ng/mL 4 5% vs. <1% of normal human F IX levels) (Fig. 4 4 A) This enhancement was similar even when helper vi rus at 10 9 vg/ms was used (Fig. 4 4 A), resulting in systemic hFIX expression of 300 90 ng/mL (~6% of normal) by week 8. A 5 fold higher dose of ssAAV2 F.IX vector alone was required to achieve similar therapeutic levels. However, at this higher dose the helper vi rus only marginally increased F IX levels (Fig. 4 4 A). To investigate whether this enhancement was caused by an increase in hFIX producing hepatocytes or by increased expression from a similar number of cells, liver tissue was collected at 10 weeks post injection from mice receiving a high or low dose of ssAAV2 hAAT F.IX with or without 10 10 vg of scAAV8 TTR PP5, and immunofluorescence staining for hFIX was performed. The p ercentage of hFIX positive he patocytes i n each group is shown in Fig 4 4 B along with representative images from e ach group. The results of fluo rescence stai ning correlate with systemic hF IX expression in that the helper virus had the greatest effect in mice receiving the lower


59 dose of ssAAV2 F IX, bringing the percent posi tive hepatocytes from <1% without helper virus to 3 5% with the helper vir us (Fig. 4 4 B). The PP5 mediated increases in hF IX expressing hepatocytes were less substantial for high dose ssAAV 2 hAAT F.IX treated mice (Fig. 4 4 B). Co administration of scAAV8 TTR PP5 helper virus does not augment AAV8 COhFVIII expression but does improve tolerance induction We next wanted to determine if PP5 could be used to increase a sub therapeutic dose of AAV8 COhFVIII to therapeutic levels. To do this, we delivered 10 10 vg/ mouse of AAV8 COhFVIII with or without 10 10 scAAV8 TTR PP5 to BALB/c HA mice Clotting times were approximately 1% FVIII activity for all mice and co administration of helper virus failed to improve hFVIII expression at any time poi nt as determined by aPTT (Fig. 4 5A). All mice returned to baseline at 39 weeks similar to the low expressing AAV8 hFVIII vector treated mice which failed to show long term correction following hFVIII challenge However all mice given both AAV8 COhFVIII and scAAV8 TTR PP5 had hF VIII specific IgG1 titers that were below the dete ction limit of our ELISA two weeks following hFVIII challenge (Fig. 4 5B). On the other hand, 3/4 mice given AAV8 COhFVIII alone developed IgG1 antibodies against hFVIII following i.v. challenge ranging fro m 533 2850ng/mL (Fig. 4 5B). This is on the low end of what was observed in mice given 10 times more AAV8 hFVIII (which also had correction in the 1% range). All mice had Bethesda titers below 5 units but mice in the group treated with only AAV8 COhFVIII h ad the highest peak titer of 4.5BU. Thus, mice supplemented with the helper PP5 vector had a slightly better immune profile similar to mice treated with a log higher dose of AAV8 hFVIII but failed to show any improvement in phenotypic correction.


60 Fi gure 4 strand synthesis of a single stranded viral genome can be inhibited by binding of FKBP52 in (A) Co administration of a double stranded, self complementary vector can provide expres sion of one of two phosphatases capable of de phosphorylating FKBP52 and causing its subsequent release from the single stranded DNA in (B) Thus, second strand synthesis may occur and facilitate expression of the double stranded DNA. From Jayandharan, et al. 79


61 Figure 4 2 Expression of codon optimized hFVIII. A dose of 10 11 vg/ms of AAV8 COhFVIII was delivered i.v. to both BL/6 129/sv and BALB/c hemoph ilia A mice and plasma collected to determine phenotypic correction and FVIII activity. (A) Phenotypic correction of clotting deficiency as measured by aPTT. Dotted lines represent untreated hemophilia A mice and normal mouse clotting times while blue shad ed area indicates period of hFVIII challenge with 1U hFVIII/mouse/week i.v. (B) FVIII activity as measured by chromogenic assay following delivery of AAV8 COhFVIII. (C) FVIII activity as measured by aPTT based standard curve generated by diluting hFVIII in nave hemophilia a mouse plasma. Data points are plotted as mean SD.


62 Figure 4 3 Immune responses following AAV8 COhFVIII gene transfer. Ten weeks following i.v. delivery of 10 11 vg/mouse of AAV8 COhFVIII, mice were challenged with 1U hFVIII/mouse/week i.v. and immune responses to hFVIII measured by hFVIII specific IgG1 ELISA in (A) and by Bethesda assay in (B) Both graphs are plotted with the same y axis range as in Fig. 4 4 for comparison of scale. Bars indicate mean SD.


63 Figure 4 4 scAAV8 TT R PP5 improves FIX expression at low doses. (A) Mice on C57/BL6 background were given indicated doses (vg/ms) of AAV2 hFIX with or without indicated doses scAAV8 TTR PP5 intravenously. Plasma was monitored for hFIX concentration by ELISA and data points ar e plotted as mean SD. (B) Representative images from mouse livers at 10 weeks post injection. Livers were stained for hFIX expressing cells and p ercentage of hFIX positive cells was d etermined for multiple mice. R anges of percent positive cells are indica ted in lower right corner of each panel.


64 Figure 4 5 scAAV8 TTR PP5 fails to improve AAV8 COhFVIII expression. BALB/c HA mice were injected i.v. with 10 10 vg/ms of AAV8 COhFVIII alone or in conjunction with 10 10 vg/ms of scAAV8 TTR PP5. (A) Mice were monitored for hFVIII expression by correction of clotting time as measured by aPTT. Two weeks following 4 week challenge of 1U hFVIII/ms/week mice were bled and checked for immune responses to hFVIII by measuring IgG1 antibodies to hFVIII by ELISA in (B) a nd inhibitors by Bethesda assay in (C)


65 CHAPTER 5 ALTERNATIVE STRATEGI ES FOR B CELL DEPLET ION COMBINED WITH GE NE THERAPY Background While our data suggests a role for B cell depletion in gene therapy to correct hemophilia A and induce tolerance to hFVIII, there may be additional areas in the field of gene therapy where this specific immune suppression is helpful. For example, since AAV naturally infects humans there is a large majority of the population which possesses pre existing immunity to the virus in the form of neutralizing antibodies against the capsid. In fact, worldwide the presence of neutralizing antibodies directed against AAV2 the serotype that naturally infects humans ranges from 30 60% 80 Neutralizing antibodies to the AAV capsid can limit or pre clude treatment with AAV gene therapy vectors thus substantially reducing the pool of potential patients if measures to circumvent this problem are not found. Neutralizing antibodies against serotypes that do not normally infect humans, such as AAV5 or AAV 8, are less prevalent although there is some cross reactivity and antibodies against each are still found in a significant portion of the population 80 However, AAV8 is able to bypass pre existing immunity to AAV2 in non human primate models and has been used i n the most recent clinical trial for hemophilia B with success 36 81 Still, administration of any AAV viru s will inevitably lead to the development of neutralizing antibodies which will prevent future administration of the same vector should the patient require multiple treatments with the same vector or a vector with a different transgene but with the same AA V capsid 82 83 Reduction of these neutralizing antibodies has been demonstrated in one study using non human pr imates that had anti AAV8 antibodies due to prior exposure to i.v. administered vector 84 In this study, Mingozzi et al. used the B cell depleting drug rituximab combined with


66 cyclosporine A to reduce low titer anti AAV antibodies enough to allow for successful re administration of an AAV vector. However, this strate gy was unsuccessful in another animal which had high titer anti AAV antibodies demonstrating the limited utility of this type of post hoc immune suppression. Preventing the development of anti vector antibodies by immune suppression at the time of AAV deli very may be more feasible. One such study found immune suppression with cyclosporine A and a non depleting CD4 antibody spanning one day prior and one week after delivery of AAV8 was able to prevent the development of neutralizing antibodies to the viral c apsid and allowed for re administration with an AAV8 vector 100 days later 85 Importantly, the second injection of AAV8 allowed for transgene expression indist inguishable from mice receiving the vector for the first time. Transient immune suppression, in particular B cell depletion with rituximab, has also received attention from the hemophilia A community as an adjunct to traditional ITI. While rituximab alone fails to eradicate inhibitors in hemophilia A patients, t he literature suggests a slight advantage for using rituximab during ITI especially in patients who have previously failed ITI alone but failure rates remain fairly high 23 24 62 63 Given that up to 30% o f severe hemophilia A patients, the most likely candidates for gene therapy, have established inhibitors, it is prudent to determine if liver directed gene therapy can reverse these inhibitors and if this can be augmented with immune suppression such as an tibody mediated B cell depletion. Optimization of a CD20 protocol to prevent neutralizing antibodies against AAV8 capsid mediated B cell depletion could be used as a sole agent in a protocol to prevent the development of AAV8 antibodies in our


67 hemophilia A mice with the ultimate goal of succ essful readministration. W e have previously spleen, lymph node and peripheral blood 24h post injection There fore, we attempted to use a single dose 1 day before administering 10 11 vg/mouse of AAV8 COhFVIII i.v. into hemophilia A mice and measured plasma concentrations of anti AAV8 IgG2a antibodies 4 and 8 weeks post vector injection by ELISA (Fig. 5 1 A ). This si ngle dose protocol COhFVIII had antibody titers (453 113ng/mL) not significantly different than control mice g iven only AAV8 COhFVIII vector at 316 267ng/mL ( Fig. 5 1A and B ) Antibodies at this level completely blocked transduction of an AAV8 hFIX vector given at the same dose i.v. whereas this dose gives high (mean of 2506ng/mL) e xpression in hemophilia A mice (Fig. 5 1D). This may have failed to prevent antibody formation since B cell recovery begins at antigen to be cleared. Thus we tried a similar protocol whe re administered one day before and 2 weeks after administration o f 10 11 vg/ms of AAV8 COhFVIII (Fig. 5 2A). Again, this protocol failed to prevent antibody formation as titers were similar in both vector supplemented mice (Fig. 5 2B and C). hours prior to vector administration ma y not provide sufficient time for complete B cell depletion in all immune compartments. Theref ore we k prior to vector as well as two week s following AAV8 COhFVIII (Fig. 5 3 A). Using th is protocol, antibody titers were significantly lower at 4 weeks post


68 95ng/mL) compared to vector only mice (410 47ng/mL Fig. 5 3 B and C ). However, at 8 weeks both groups had decreased titers and were not significantly d ifferent (means of 80 21ng/mL and 89 only groups, respectively). Given the indication that this protocol could reduce anti capsid antibodies, we next wanted to determine if a lower vector dose could allow for complete eli mina tion of the antibody response and if this would allow for re administration and 3 except with only 2.5x10 10 vg/ms of AAV8 COhFVIII delivered intravenously (Fig. 5 4A). With this lower dose, anti AAV8 Ig2a antibodies were undetectable by ELISA at 4 and 8 weeks treated mice whereas vector only mice had titers of 178 21ng/mL and 330 101ng/mL at 4 and 8 weeks, respectively (Fig. 5 4B and C). Clotting correction as measured by aPTT was modest, as expected at this lower dose, with clotting times reduced to 61 8s and 54 8s for delivered another dose of 2.5x10 11 vg/ms of AAV8 COhFVIII at week 10. This d id not result in further reduction of clotting times (Fig. 5 clotting times of 64 5s and vector only mice had an average clotting time of 60 9s at 4 weeks post Several issues could have contributed to the lack of improvement in cl otting times following secondary gene transfer. First, although antibody titers were below our limit of detection, there were still some OD values above zero in treated mice even though these were comparable to pre injection levels. Therefore, perhap s even the presence of minimal antibody levels are enough to neutralize the small dose of AAV used in this experiment. Alternatively, our aPTT assay may not be given that there is little


69 distin ction between the aPTT clotting times between 1 5% FVIII activity (the range of correct ion seen at this dose). Thus, either a larger second dose or use of a transgene with a more sensitive assay (such as FIX) may be warranted. gene therapy to reverse inhibitors in hemophilia A mice Given that up to 30% of severe hemophiliacs the patient population most likely to be candidates for gene therapy will develop inhibitors, it is vital to determine the effects of gene therapy on a pre existing antibody titer and if liver directed gene transfer will be able to mitigate these inhibitors. In the presence of inhibitors, any FVIII produced by gene transfer will be neutralized and the therapy rendered null. To test this, we developed inhibitors in BALB/c HA mice with weekly infusions of 1U hFVIII/ms/week as before. One week before gene transfer, mice were boosted with a single 2U dose of hFVIII (Fig. 5 5A). Mice were then either given 10 11 vg/ms of AAV8 COhFVIII i.v. alone treated group were given another dose of antibody 2 weeks after the first. Mice were monitored weekly for anti hFVIII antibody titers via IgG1 ELISA and every 4 weeks by ELISA and Bethesda titer. In both groups, IgG1 titers decreased steadily over 4 weeks from an average of 5976 4865ng/mL to 2060 1644ng/mL in mice treated with only AAV8 COhFVIII and from 7305 3657ng/mL to 1595 794ng/mL in mice treated with both AAV8 COhFVIII and Thus, after 4 we eks liver directed gene therapy was able to reduce but no completely eliminate hFVIII antibodies in BALB/c HA mice regardless of the use of term follow up is warranted.


70 Figure 5 1. depletion does not prevent AAV8 antibodies. (A) 1 day before administration of 10 11 vg/ms AAV8 COhFVIII. Mice were bled at 4 weeks and 8 weeks post vector injection and plasma analyzed for anti A AV8 IgG2a antibodies by ELISA in (B) (C) Plasma hFIX levels as measured by ELISA 3 weeks following re administration of an AAV8 hFIX vector (10 11 vg/ms) at 10 weeks post AAV8 COhFVIII.


71 Figure 5 2. Two dose s antibodies to AAV8. (A) BALB/c i.v. injection of 10 11 vg/ms of AAV8 COhFVIII. At 4 weeks and 8 weeks mice were bled and plasma analyzed for antibodies to AAV8 capsid by ELISA. Antibody titers are shown by optical density (B)


72 Figure 5 3. Alternate two AAV8 antibodies. (A) BALB/c after i.v. injection of 10 11 vg/ms of AAV8 COhFVIII. At 4 weeks and 8 weeks post ve ctor, mice were bled and plasma analyzed for antibodies to AAV8 capsid by ELISA. Antibody titers are shown in (B) test.


73 Figure 5 4. Lower doses of AAV8 COhFVIII allow for prevention of antibodies without improvement after re adm inistration (A) BALB/c HA mice were given i.v. injection of 2.5x10 10 vg/ms of AAV8 COhFVIII. At 4 weeks and 8 weeks post vector, mice were bled and plasma analyzed for antibodies to AAV8 capsid by ELISA. At 10 weeks, mice rece ived another i.v. injection of 2.5x10 10 vg/ms AAV8 COhFVIII. Antibody titers are shown in (B) (second AAV8 COhFVIII injection). (C) Clotting times as measured by aPTT for mice treated with 2.5x10 10 vg/ms of AAV8 COhFVIII at week 0 and week 10.


74 Figure 5 5. Liver directed gene therapy is able to reduce antibody titer in hFVIII challenged BALB/c HA mice. BALB/c HA mice were challenged with 1U hFVIII/week/ms for 4 weeks and antibody titers measure 2 weeks later (time 0 on g 10 11 vg/mL AAV8 COhFVIII i.v. one day later or just the same dose of vector weeks post AAV. Plasma concentrations of h FVIII specific IgG1 were measured weekly up to 4 weeks post AAV by ELISA. Data is plotted as mean SD.


75 CHAPTER 6 DISCUSSION For many monogenetic diseases, gene therapy represents the only long term hope for a safe, non invasive cure. However, the main barr ier standing between success and failure of these treatments remains the immune system. Whether it is pre existing immunity to the viral capsid as is the case for AAV, the response to the viral vector in adenovirus or lentivirus or the response to the tran sgene in a number of disorders most prominently in hemophilia A. However, there is a large percentage of gene therapy research dedicated to overcoming these hurdles. Hemophilia A is complicated by the chance of a pre existing immu ne response to the viral vector or to the transgene (hFVIII) prior to gene therapy as well as the possibility of an immune response to the transgene after vector delivery. Thus, specific attention must be paid to learning how to manipulate the immune system to avoid these complica tions and careful choice of vector and target tissue will be critical if success is to be realized in this disease. Gene transfer may tolerize or sensitize mice depending on genetic background If gene transfer only accomplished expression at the lower end of the therapeutic range, e.g. <5% of normal, this would occasionally necessitate supplementary clotting factor administration. Hence, it is imperative that tolerance is maintained under these circumstances. Previously, sustained correction of murine hemop hilia A with AAV vectors has been shown using either cFVIII or murine FVIII, but neither study tested the effect of FVIII challenge after gene therapy 86 87 Others have combined gene transfer with immune suppression to avoid inhibitor formation 50 52 86 Our study shows that lack of inhibitor formation against FVIII after hepatic AAV gene transfer does not necessarily


76 constitute humoral immune tolerance as mice in both strains treated with a low expressing vector maintained correction in absence of but not after challenge with hFVIII ( Figs. 3 4A and 3 5A ). In fact, in the BL/6 129/sv HA strain, mice were predisposed to heightened inhibitor formation in subsequent treatment with hFVIII protein, while gene transfer had the des ired tolerogenic effect in BALB/c HA mice. Clearly, high levels of hepatic expression promote tolerance regardless of the strain background ( Fig. 4 2 and 4 2 ), which is consistent with our earlier findings on factor IX 44 However, within the range of sub optimal expression, two possibilities may explain the differing responses between the two strains. Somewhat higher expression in BALB/c HA mice may have been par tially effective in tolerance induction, while expression in the BL/6 129/sv HA strain may have been just below this threshold and thus sensitized the mice. Alternatively, similarly low levels of expression may sensitize animals with higher immune respon si veness to F VIII, while having a more tolerogenic effect in a strain with lower responsiveness. Consistent with data by others, BALB/c HA mice had lower inhibitor responses against hFVIII, and T helper responses differed between strains 88 BL/6 129/sv HA mice, after 4 weeks of hFVIII protein therapy, had 10 fold higher inhibitor titers compared to BALB/c HA. However, IgG formation against hFVIII was only 2 fold hi gher (compare Figs. 3 4A and 3 5A ). In contrast to BL/6 129/sv HA, BALB/c HA mice only showed a good correlation between BU and IgG titers after a longer course of hFVIII administration (Fig. 3 7). Previously, we documented that CD4 + T cell responses again st hFVIII in BL/6 129/sv HA mice were strongly biased to Th2, characterized by IL 4 and IL 6 production while lacking expression of suppressive/regulatory cytokines 66


77 BALB/c HA mice, however, have a more mixed response, which includes expression of Th1, Th2 and suppressive molecules ( Fig. 3 6A and Table 3 2 ). This difference in the T helper response likely explains differences in inhibitor responses and in ability to induce tolerance in the 2 strains. Despite evidence for Th1 activation as part of the CD4 + T cell response in BALB/c HA, we found that mice of both strains forme d IgG2a against hFVIII only rarely and inconsistently (data not shown) suggesting that antibody formation is primarily driven by the Th2 component, thus consistently resulting in high titer IgG1. Utility of transient B cell depletion for tolerance to hFV III Rituximab has been used to treat patients with inhibitors with varying degrees of success 24 62 63 In acquired hemophilia, a rare form of the dise ase where inhibitors appear in hemostatically normal individual s treatment with rituximab alone can eradicate inhibitors 89 This response is long lasting and has been observed to last up to 5.5 years 89 In contrast, rituximab use in X linked hemophilia A is usually concurrent with traditional ITI. In this setting, data is limited largely to case reports and well controlled studies are lacking. However, a consecutive national cohort analyzed in the UK investigating the use of rituximab as an adjunct to traditional ITI suggests a modest benefit of B cell depletion for eradicating in hibitors in patients who have previously failed ITI 63 Still, a number of these individuals (6/10) relapsed and required additional immune therapy 63 The data in this thesis al While B cell depletion effectively prevented anti hFVIII formation in protein therapy, such a protocol failed to induce lasting hyporesponsiveness, suggesting that protein administration fails to tolerize the new ly emerging B cells. This conclusion is supported by a recent clinical study where immune responses to neoantigens following Rituximab treatment were muted during B cell depletion but returned to near normal once the


78 effects of the drug waned 90 Furthermore, our experiments using B cell depleting low efficacy gene therapy priming the immune system to hFVIII. This indicates a role of B cells in the generation o f a primary immune response to hFVIII. This was not due to long control animals to form antibodies once B cell frequencies had recovered. To further confirm this point, we ad ministered an adenoviral vector expressing galactosidase 7 against the virus and antibodies against galactosidase at titers indistinguishable from control mice (Fig. 3 3A and B). Low levels of hepatic expression resulted only in a partial reduction of the T cell response to FVIII in BALB/c HA mice, despite the protective effect against inhibitor formation. The reduction of the Th2 response, in conjunction with preserving Treg responses, was likely instrumental in achieving this effect. Nonetheless, gene transfer up regulation of Treg markers was also abrogated. This was somewhat reflected in adoptive transfer experiments, where CD4+CD25+ cells from AAV8 F8 only treated F8 + statistical significance. In contrast, high expression fro m the codon optimized construct increased the suppressive response in the presence of B cells, which raises the question of how the combination of these two approaches may shape the response. B cells play a dual role in the immune response as antigen prese nting cells and as producers of antibodies. Zhang et al. demonstrated


79 protein challenge in hFVIII primed hemophilia A mice 91 Interesti ngly, these authors found that IgG1 instead of IgG2a marginal zone B cells and caused accumulation of Treg after further exposure to hFVIII. Furthermore, a recent study by Xiang et. al demonstrated an effect of B cell depletion on T cell funct ion and phenotype 92 In their model, Treg were initially reduced following B cell depletion but actually increased when B cells had recovered, and there was als o evidence for emergence of suppressive B cells. Thus, a more differential method for B cell targeting may further improve tolerance induction by this approach. Tolerance induction with codon optimized hFVIII It is encouraging that a more effective codon optimized hFVIII expression cassette induced humoral tolerance, resulting in sustained correction of coagulation after challenge with hFVIII protein, independent of the strain background. This is significant in light of the fact that hFVIII has been notori ously difficult to express in mice at levels high enough for phenotypic correction. Often, alternative species FVIII have been used such as canine FVIII or porcine FVIII because of either higher expression or superior secretion 50 87 93 94 Here, we were able t o achieve tolerance to hFVIII and long term phenotypic correction with a single AAV vector expressing codon optimized hFVIII in the absence of immune suppression. Codon optimization is a process whereby the cDNA to be expressed is analyzed and altered to improve expression based on a variety of factors including but not limited to: usage of species specific codo n s, increased GC content, removal of cryptic slice sites, and stabilizing RNA structure 73 This process has been used in the context of gene therapy to improve expression of several transgenes from different vectors 54 74 75 77 78 95 Only one study used this approach for hFVIII


80 expression. Investigators observed up to a 44 fold increase in FVIII expression following delivery of a codon optimized hFVIII via lentivirus to neonatal hemophilia A mice 54 However, the immunological consequences of increased expression in the context of a mature immune system were not investigated. Here, we demonstrate that a CO hFVIII construct improves not only efficacy but also tolerance induction in adult hemophilia A mice of two different strains. In adult animals, Treg induction is a critical component of tolerance induction by hepatic gene transfer. Our adoptive transfer experiments showed that use of CO hFVIII facilitated Treg induction, thereby conferring better suppression of antibody formation against hFVIII. Although the 4 5% long term expression we observed in mice is therapeutic, the likelihood that these levels would translate directly to human a pplication of this strategy is unlikely. Thus, we attempted to further increase our hFVIII transgene expression by co PP5 vector. This vector has been able to increase second strand synthesis and therefore expression of si ngle stranded vector s including GFP and in this thesis, hFIX 79 96 While this helper virus increase d the expression of low dose ssAAV2 hFIX vector by approximately 5 fold (Fig. 4 3A), this strategy did not have similar effects on a higher dose of the same vector or on our ssAAV8 COhFVIII vector (Figs. 4 3A and 4 4A). This could be for several reasons. First, the murine liver is much more permissive to AAV8 where a dose at this level should transduce >90% of hepatocytes 67 In addition, AAV8 genomes are delivered more efficiently to the nucleus 97 Therefore, the number of transcriptionally active vector genomes may not be a limiting factor in our AAV8 COhFVIII transgene. Also, given that our transgene was codon optimized, it is likely that there is efficiency of translating the available transcripts


81 is high thus overcoming the known translational barrier for hFVIII production 37 What this may reveal is that for hFVIII, a significant barrier may be post translational. Indeed, studies with porcine FVIII suggest that targeted modifications in the amino acid sequence of hFVIII c a n greatly increase its expression and that this increase is due to enhanced secretion of the protein as opposed to greater transcription or translation 71 93 Thus further studies that aim to increase hFVIII expression should incorporate considerations for enhanced protein secretion. Utility of B cell depletion for the prevention of anti AAV capsid antibodies a nd to reverse inhibitors in hemophilia A mice Even in the most successful clinical trials with hemophilia B, transgene expression levels were high enough to remove the need for supplemental clotting factor in 2 of 6 patients 36 While gene therapy did reduce the need for clotting factor in the other patients the main advantage of AAV gene transfer would be to have a one time or extremely infrequent treatment that is a replacement for protein therapy. Therefore, one strategy might be to give repeated administrations of the same vector for increased protein expression and decreased dependence on protein therapy. Unfortunately, upon delivery of AAV, neutralizing anti bodies will inevitably be formed against the vector capsid and will preclude addition treatment with the same vector. In this thesis, we attempted to develop a protocol that would prevent the formation of capsid antibodies using only a single drug agent an alogous to the drug rituximab a clinically safe biologic with wide use and minimal side effects. Ideally, one would use as few doses of immune suppressant as possible to prevent capsid antibodies and thus we tried a single dose of given one day before gene transfer. This did not reduce the anti capsid antibody and the levels seen at 8 weeks (211 325ng/mL) prevented detectable expression from


82 an AAV8 hFIX vector delivered at 10 weeks after the initial injection (Fig. 5 1B D) B cells are expected to begin recovering by 3 weeks and as they recover they may allow for antibodies to form against capsid antigen that may still be present. However, even an addition al dose given 2 weeks after vector administration removing B cells for approximately 5 weeks after AAV injection failed to prevent or reduce antibody development (Fig. 5 2 B and C). Alternatively, giving one week rather than one day before initial AAV delivery did reduce antibody titers at 4 weeks post injection bu t they were still present and were not significantly different at 8 weeks from control mice given only AAV8 COhFVIII (Fig. 5 3B and C). This suggests that B cell depletion may not be robust at 24h post injection and suggests that a dose should be giv en at least one week prior to vector. Finally, the only protocol that eliminated AAV8 capsid antibodies below our limit of detection was 2 doses of given one week prior and two weeks following a reduced dose of AAV8 COhFVIII at 2.5x10 11 vg/ms (Fig. 5 4 B and C). However, administration of another dose of the same vector and dose did not further decrease the clotting times of mice given and AAV8 COhFVIII. This may be due to a lack of sensitivity of the aPTT to detect small changes in FVIII activit y, or there may be still be a small neutralizing antibody titer that prevented this small amount of vector from transducing mouse hepatocytes. In conclusion, it appears that AAV capsid antigen is present for greater than 5 weeks post injection at a moderat e, 10 11 vg/ms, dose of virus. Furthermore, complete depletion of B cells must be accomplished at the time of vector injection. Both of these observations argue for more aggressive dosing of to prevent antibodies and perhaps earlier re administration of secondary vector while B cells are still depleted.


83 Thus far, patients enrolled in clinical trials for hemophilia A and B have been inhibitor negative given the uncertain effect that gene therapy will have on an individual with an existing immune response to the clotting factor. However, up to 30% of severe hemophilia A patients have inhibitors and it is these patients who have the poorest prognosis and are most burdened by heavy factor us e. It is also these patients that would benefit the most from a protocol that could not only correct disease but reverse their inhibitors as well. Our attempts to model this in mice demonstrated that antibody titers could be reduced in the short term (Fig. 5 5) but not completely eliminated. Similar experimen t in dogs revealed that most inhibitors were gone in most dogs by 5 weeks inhibitors in one dog were not reversed for 80 weeks after gene transfer 98 Thus, prolonged observation of these mice is warranted and underway. Future Directions The work presented in this thesis highlights a number of challenges left to overcome in the field of gene therapy for hemophi lia A which still include the limited expression and inherent immunogenicity of FVIII. We were able to achieve sustained expression of 3 5% hFVIII activity in a murine model of hemophilia A with a scalable vector dose Furthermore, this expression maintain ed tolerance to hFVIII throughout the length of this experiment. If this result were exactly recapitulated in humans it would represent a drastic improvement in the quality of life of a person with sever e hemophilia A and would greatly reduce their depende nce on expensive and invasive clotting factor as seen in the recent human trials for hemophilia B 36 However, it is unclear if these results will translate dire ctly and therefore further improvement of expression at the same or lower doses is highly desired. Codon optimization seemed to provide a n enhancement of expression at the transcription al and translation al level but confirmation


84 of which of these will be i mportant for identifying what, if any, improvements can be made to the cDNA to improve expression at these steps. Furthermore, it is known that secretion of hFVIII is a limiting step and therefore use of FVIII cDNA carrying modifications to the amino acid sequence and not just cDNA sequence that enhance secretion may prove fruitful in this vector system. Replicating modifications such as those by Doering et al. in our codon optimized hFVIII protein could further improve the phenotypic correction beyond what was seen in the experiments presented here 70 Expression could also be improved by use of alternative viral capsids including those modified to decrease ubiqui tin meidated degradation by removing phosphorylation sites on the viral capsid surface 99 Another important aspect will be determining the effect of B cell de pletion on tolerance induction following liver directed AAV gene therapy. While we found that our optimized vector induced tolerance to hFVIII in both high and low responding strains of mice in the absence of any immune suppression, there were still a few mice with low level antibody responses to hFVIII challenge. Given that B cell depletion has mixed effects on T cell responses especially Treg and that rituximab has a history of use in hemophilia A as well as AAV gene therapy, it will be prudent to evaluat e the use of concomitant use of B cell depletion and delivery of a tolerizing gene therapy vector 100 Finally, up to 30% of severe hemophilia A patients have pre existing antibod ies against hFVIII. To date, clinical trials evaluating gene therapy for hemophilia have involved inhibitor free patients. However, patients with inhibitors represent the highest risk patients with the fewest treatment options. It will be interesting to al so investigate whether or not our optimized vector can reverse inhibitors in mice of either strain and if


85 B cell depletion will have any role in the efficacy of this treatment. There has been success in a canine model of hemophilia A where investigators we re able to reverse inhibitors in 4 hemphilia A dogs after giving liver directed gene therapy 98 In this study there was indication that long established inhibitors may be more difficult to eradicate than recently developed inhibitors as the time to reverse inhibitors in the former was 80 weeks opposed to 4 5 weeks for the latter. However, more complete characterization of this model in mice is necessary as is an und erstanding of the mechanism behind eliminating hFVIII specific B cells. Our codon optimized vector and mouse strains should allow for better observation of this phenomenon in the future. In conclusion, the immune response remains a major hurdle for gene th erapy and particularly for hemophilia A gene therapy. Obtaining high level, liver specific expression of the hFVIII transgene appears to be crucial to eliciting both tolerance and long term correction. Choice of vector, route of administration, transgene m odifications and transient immune suppression all must be considered for effective and safe delivery of future vectors. Several hurdles remain such as pre existing immunity to viral vectors and the poor expression of hFVIII transgenes and our findings here should help clear these hurdles as the possibility for treating hemophilia A with AAV gene therapy becomes a reality.






88 LIST OF REFERENCES 1 Marshall A. Lichtman, T. J. K., Uri Seligsohn, Kenneth Kaushansky, Josef T. Prchal in Williams Hematology Vol. 8 Ch. Hemostasis and Thrombosis, (McGraw Hill Med ical, 2010). 2 Castaldo, G. et al. Haemophilia A: molecular insights. Clinical chemistry and laboratory medicine : CCLM / FESCC 45 450 461, doi:10.1515/CCLM.2007.093 (2007). 3 Vehar, G. A. et al. Structure of human factor VIII. Nature 312 337 342 (1984). 4 van Dieijen, G., Tans, G., Rosing, J. & Hemker, H. C. The role of phospholipid and factor VIIIa in the activation of bovine factor X. The Journal of biological chemistry 256 3433 3442 (1981). 5 High, K. A., & Roberts, H. R. Molecular basis of thrombosi s and hemostasis 6th edn, (M. Dekker, 1995). 6 Lakich, D., Kazazian, H. H., Jr., Antonarakis, S. E. & Gitschier, J. Inversions disrupting the factor VIII gene are a common cause of severe haemophilia A. Nature genetics 5 236 241, doi:10.1038/ng1193 236 (1993). 7 Higuchi, M. et al. Molecular characterization of severe hemophilia A suggests that about half the mutations are not within the coding regions and splice junctions of the factor VIII gene. Proc Natl Acad Sci U S A 88 7405 7409 (1991). 8 Luck, J. V., Jr. et al. Hemophilic arthropathy. The Journal of the American Academy of Orthopaedic Surgeons 12 234 245 (2004). 9 Darby, S. C. et al. Mortality rates, life expectancy, and causes of death in people with hemophilia A or B in the United Kingdom who we re not infected with HIV. Blood 110 815 825, doi:10.1182/blood 2006 10 050435 (2007). 10 Darby, S. C. et al. The impact of HIV on mortality rates in the complete UK haemophilia population. AIDS 18 525 533 (2004). 11 Seremetis, S. et al. Human recombinant DNA derived antihaemophilic factor (factor VIII) in the treatment of haemophilia A: conclusions of a 5 year study of home therapy. The KOGENATE Study Group. Haemophilia : the official journal of the World Federation of Hemophilia 5 9 16 (1999). 12 Manco Johnson, M. J. et al. Prophylaxis versus episodic treatment to prevent joint disease in boys with severe hemophilia. The New England journal of medicine 357 535 544, doi:10.1056/NEJMoa067659 (2007).


89 13 Globe, D. R., Curtis, R. G. & Koerper, M. A. Utilizat ion of care in haemophilia: a resource based method for cost analysis from the Haemophilia Utilization Group Study (HUGS). Haemophilia : the official journal of the World Federation of Hemophilia 10 Suppl 1 63 70 (2004). 14 O'Mahony, B. & Black, C. Expand ing hemophilia care in developing countries. Seminars in thrombosis and hemostasis 31 561 568, doi:10.1055/s 2005 922228 (2005). 15 Stonebraker, J. S., Brooker, M., Amand, R. E., Farrugia, A. & Srivastava, A. A study of reported factor VIII use around the world. Haemophilia : the official journal of the World Federation of Hemophilia 16 33 46, doi:10.1111/j.1365 2516.2009.02131.x (2010). 16 Gouw, S. C. et al. F8 gene mutation type and inhibitor development in patients with severe hemophilia A: systematic review and meta analysis. Blood doi:10.1182/blood 2011 09 379453 (2012). 17 Waters, B. & Lillicrap, D. The molecular mechanisms of immunomodulation and tolerance induction to factor VIII. Journal of thrombosis and haemostasis : JTH 7 1446 1456, doi:10.11 11/j.1538 7836.2009.03538.x (2009). 18 van Haren, S. D., Wroblewska, A., Fischer, K., Voorberg, J. & Herczenik, E. Requirements for immune recognition and processing of factor VIII by antigen presenting cells. Blood reviews 26 43 49, doi:10.1016/j.blre.20 11.10.001 (2012). 19 Lollar, P. Pathogenic antibodies to coagulation factors. Part one: factor VIII and factor IX. Journal of thrombosis and haemostasis : JTH 2 1082 1095, doi:10.1111/j.1538 7836.2004.00802.x (2004). 20 Hausl, C. et al. High dose factor V III inhibits factor VIII specific memory B cells in hemophilia A with factor VIII inhibitors. Blood 106 3415 3422, doi:10.1182/blood 2005 03 1182 (2005). 21 Di Michele, D. M. Immune tolerance induction in haemophilia: evidence and the way forward. Journal of thrombosis and haemostasis : JTH 9 Suppl 1 216 225, doi:10.1111/j.1538 7836.2011.04349.x (2011). 22 Odeyemi, I. A. & Dano, A. M. Optimising immune tolerance induction strategies in the management of haemophilia patients with inhibitors: a cost minimis ation analysis. Current medical research and opinion 25 239 250, doi:10.1185/03007990802633863 (2009). 23 Franchini, M. et al. Immune tolerance with rituximab in congenital haemophilia with inhibitors: a systematic literature review based on individual pa tients' analysis. Haemophilia : the official journal of the World Federation of Hemophilia 14 903 912, doi:10.1111/j.1365 2516.2008.01839.x (2008).


90 24 Streif, W. et al. Inhibitor treatment by rituximab in congenital haemophilia A Two case reports. Hamos taseologie 29 151 154 (2009). 25 Santagostino, E. & Mancuso, M. E. Barriers to primary prophylaxis in haemophilic children: the issue of the venous access. Blood transfusion = Trasfusione del sangue 6 Suppl 2 s12 16 (2008). 26 Daya, S. & Berns, K. I. Gen e therapy using adeno associated virus vectors. Clinical microbiology reviews 21 583 593, doi:10.1128/CMR.00008 08 (2008). 27 Niemeyer, G. P. et al. Long term correction of inhibitor prone hemophilia B dogs treated with liver directed AAV2 mediated factor IX gene therapy. Blood 113 797 806, doi:10.1182/blood 2008 10 181479 (2009). 28 Jooss, K., Yang, Y., Fisher, K. J. & Wilson, J. M. Transduction of dendritic cells by DNA viral vectors directs the immune response to transgene products in muscle fibers. Jo urnal of virology 72 4212 4223 (1998). 29 McCaffrey, A. P. et al. The host response to adenovirus, helper dependent adenovirus, and adeno associated virus in mouse liver. Molecular therapy : the journal of the American Society of Gene Therapy 16 931 941, doi:mt200837 [pii] 10.1038/mt.2008.37 (2008). 30 Zaiss, A. K. et al. Differential activation of innate immune responses by adenovirus and adeno associated virus vectors. Journal of virology 76 4580 4590 (2002). 31 Ashtari, M. et al. The human visual cort ex responds to gene therapy mediated recovery of retinal function. J Clin Invest 121 2160 2168, doi:10.1172/JCI57377 (2011). 32 Bennett, J. et al. AAV2 gene therapy readministration in three adults with congenital blindness. Science translational medicine 4 120ra115, doi:10.1126/scitranslmed.3002865 (2012). 33 Cideciyan, A. V. et al. Human RPE65 gene therapy for Leber congenital amaurosis: persistence of early visual improvements and safety at 1 year. Human gene therapy 20 999 1004, doi:10.1089/hum.2009. 086 (2009). 34 Maguire, A. M. et al. Age dependent effects of RPE65 gene therapy for Leber's congenital amaurosis: a phase 1 dose escalation trial. Lancet 374 1597 1605, doi:10.1016/S0140 6736(09)61836 5 (2009). 35 Simonelli, F. et al. Gene therapy for Le ber's congenital amaurosis is safe and effective through 1.5 years after vector administration. Molecular therapy : the journal of the American Society of Gene Therapy 18 643 650, doi:10.1038/mt.2009.277 (2010).


91 36 Nathwani, A. C. et al. Adenovirus Associ ated Virus Vector Mediated Gene Transfer in Hemophilia B. The New England journal of medicine doi:10.1056/NEJMoa1108046 (2011). 37 Hollestelle, M. J. et al. Tissue distribution of factor VIII gene expression in vivo -a closer look. Thrombosis and haemosta sis 86 855 861 (2001). 38 Calne, R. Y. et al. Induction of immunological tolerance by porcine liver allografts. Nature 223 472 476 (1969). 39 Crispe, I. N. et al. Cellular and molecular mechanisms of liver tolerance. Immunological reviews 213 101 118 (2 006). 40 Mackay, I. R. Hepatoimmunology: a perspective. Immunol Cell Biol 80 36 44, doi:1063 [pii] (2002). 41 Dobrzynski, E. et al. Induction of antigen specific CD4+ T cell anergy and deletion by in vivo viral gene transfer. Blood 104 969 977, doi:10.11 82/blood 2004 03 0847 2004 03 0847 [pii] (2004). 42 Hoffman, B. E. et al. Nonredundant roles of IL 10 and TGF beta in suppression of immune responses to hepatic AAV factor IX gene transfer. Molecular therapy : the journal of the American Society of Gene Th erapy 19 1263 1272, doi:10.1038/mt.2011.33 (2011). 43 LoDuca, P. A., Hoffman, B. E. & Herzog, R. W. Hepatic gene transfer as a means of tolerance induction to transgene products. Curr Gene Ther 9 104 114 (2009). 44 Mingozzi, F. et al. Induction of immune tolerance to coagulation factor IX antigen by in vivo hepatic gene transfer. The Journal of clinical investigation 111 1347 1356, doi:10.1172/JCI16887 (2003). 45 Cao, O. et al. Induction and role of regulatory CD4+CD25+ T cells in tolerance to the transg ene product following hepatic in vivo gene transfer. Blood 110 1132 1140 (2007). 46 Luth, S. et al. Ectopic expression of neural autoantigen in mouse liver suppresses experimental autoimmune neuroinflammation by inducing antigen specific Tregs. J Clin Inv est 118 3403 3410 (2008). 47 Manno, C. S. et al. Successful transduction of liver in hemophilia by AAV Factor IX and limitations imposed by the host immune response. Nat Med 12 342 347, doi:nm1358 [pii] 10.1038/nm1358 (2006). 48 Mingozzi, F. et al. CD8(+ ) T cell responses to adeno associated virus capsid in humans. Nat Med 13 419 422, doi:nm1549 [pii] 10.1038/nm1549 (2007).


92 49 Gnatenko, D. V. et al. Human factor VIII can be packaged and functionally expressed in an adeno associated virus background: appl icability to haemophilia A gene therapy. British journal of haematology 104 27 36 (1999). 50 Sarkar, R. et al. Total correction of hemophilia A mice with canine FVIII using an AAV 8 serotype. Blood 103 1253 1260, doi:10.1182/blood 2003 08 2954 (2004). 51 Jiang, H. et al. Multiyear therapeutic benefit of AAV serotypes 2, 6, and 8 delivering factor VIII to hemophilia A mice and dogs. Blood 108 107 115, doi:10.1182/blood 2005 12 5115 (2006). 52 Lu, H. et al. Complete correction of hemophilia A with adeno as sociated viral vectors containing a full size expression cassette. Human gene therapy 19 648 654, doi:10.1089/hum.2007.0182 (2008). 53 Hu, C. & Lipshutz, G. S. AAV based neonatal gene therapy for hemophilia A: long term correction and avoidance of immune responses in mice. Gene therapy doi:10.1038/gt.2011.200 (2012). 54 Ward, N. J. et al. Codon optimization of human factor VIII cDNAs leads to high level expression. Blood 117 798 807, doi:10.1182/blood 2010 05 282707 (2011). 55 Bi, L. et al. Targeted disr uption of the mouse factor VIII gene produces a model of haemophilia A. Nat Genet 10 119 121, doi:10.1038/ng0595 119 (1995). 56 Uchida, J. et al. The innate mononuclear phagocyte network depletes B lymphocytes through Fc receptor dependent mechanisms duri ng anti CD20 antibody immunotherapy. J Exp Med 199 1659 1669, doi:10.1084/jem.20040119 (2004). 57 Zolotukhin, S. et al. Recombinant adeno associated virus purification using novel methods improves infectious titer and yield. Gene therapy 6 973 985, doi:1 0.1038/ (1999). 58 Cao, O. et al. Impact of the underlying mutation and the route of vector administration on immune responses to factor IX in gene therapy for hemophilia B. Mol Ther 17 1733 1742, doi:10.1038/mt.2009.159 (2009). 59 Verbruggen B. et al. The Nijmegen modification of the Bethesda assay for factor VIII:C inhibitors: improved specificity and reliability. Thrombosis and haemostasis 73 247 251 (1995). 60 Verweij, C. L. & Vosslamber, S. New insight in the mechanism of action of ritu ximab: the interferon signature towards personalized medicine. Discov Med 12 229 236 (2011). 61 Weiner, G. J. Rituximab: mechanism of action. Semin Hematol 47 115 123, doi:10.1053/j.seminhematol.2010.01.011 (2010).


93 62 Aleem, A. et al. Rituximab as a sing le agent in the management of adult patients with haemophilia A and inhibitors: marked reduction in inhibitor level and clinical improvement in bleeding but failure to eradicate the inhibitor. Haemophilia 15 210 216, doi:10.1111/j.1365 2516.2008.01865.x ( 2009). 63 Collins, P. W. et al. Rituximab and immune tolerance in severe hemophilia A: a consecutive national cohort. J Thromb Haemost 7 787 794, doi:10.1111/j.1538 7836.2009.03332.x (2009). 64 Nathwani, A. C. et al. Self complementary adeno associated vi rus vectors containing a novel liver specific human factor IX expression cassette enable highly efficient transduction of murine and nonhuman primate liver. Blood 107 2653 2661, doi:10.1182/blood 2005 10 4035 (2006). 65 High, K. A. Gene therapy for haemop hilia: a long and winding road. J Thromb Haemost 9 Suppl 1 2 11, doi:10.1111/j.1538 7836.2011.04369.x (2011). 66 Moghimi, B. et al. Induction of tolerance to factor VIII by transient co administration with rapamycin. Journal of thrombosis and haemostasis : JTH 9 1524 1533 (2011). 67 Cooper, M. et al. Improved induction of immune tolerance to factor IX by hepatic AAV 8 gene transfer. Human gene therapy 20 767 776, doi:10.1089/hum.2008.161 (2009). 68 Powell, J. S. et al. Phase 1 trial of FVIII gene transfe r for severe hemophilia A using a retroviral construct administered by peripheral intravenous infusion. Blood 102 2038 2045, doi:10.1182/blood 2003 01 0167 (2003). 69 Roth, D. A., Tawa, N. E., Jr., O'Brien, J. M., Treco, D. A. & Selden, R. F. Nonviral tra nsfer of the gene encoding coagulation factor VIII in patients with severe hemophilia A. The New England journal of medicine 344 1735 1742, doi:10.1056/NEJM200106073442301 (2001). 70 Doering, C. B. et al. Directed engineering of a high expression chimeric transgene as a strategy for gene therapy of hemophilia A. Molecular therapy : the journal of the American Society of Gene Therapy 17 1145 1154, doi:10.1038/mt.2009.35 (2009). 71 Doering, C. B., Healey, J. F., Parker, E. T., Barrow, R. T. & Lollar, P. Ide ntification of porcine coagulation factor VIII domains responsible for high level expression via enhanced secretion. The Journal of biological chemistry 279 6546 6552, doi:10.1074/jbc.M312451200 (2004). 72 Gustafsson, C., Govindarajan, S. & Minshull, J. C odon bias and heterologous protein expression. Trends in biotechnology 22 346 353, doi:10.1016/j.tibtech.2004.04.006 (2004).


94 73 Fath, S. et al. Multiparameter RNA and codon optimization: a standardized tool to assess and enhance autologous mammalian gene expression. PloS one 6 e17596, doi:10.1371/journal.pone.0017596 (2011). 74 Foster, H. et al. Codon and mRNA sequence optimization of microdystrophin transgenes improves expression and physiological outcome in dystrophic mdx mice following AAV2/8 gene tran sfer. Molecular therapy : the journal of the American Society of Gene Therapy 16 1825 1832, doi:10.1038/mt.2008.186 (2008). 75 Li, C., Xiao, P., Gray, S. J., Weinberg, M. S. & Samulski, R. J. Combination therapy utilizing shRNA knockdown and an optimized resistant transgene for rescue of diseases caused by misfolded proteins. Proc Natl Acad Sci U S A 108 14258 14263, doi:10.1073/pnas.1109522108 (2011). 76 Ng, Y. Y. et al. Correction of B cell development in Btk deficient mice using lentiviral vectors with codon optimized human BTK. Leukemia : official journal of the Leukemia Society of America, Leukemia Research Fund, U.K 24 1617 1630, doi:10.1038/leu.2010.140 (2010). 77 Padegimas, L. et al. Optimization of hCFTR Lung Expression in Mice Using DNA Nanopart icles. Mol Ther 20 63 72, doi:10.1038/mt.2011.196 (2012). 78 Wang, L. et al. Preclinical evaluation of a clinical candidate AAV8 vector for ornithine transcarbamylase (OTC) deficiency reveals functional enzyme from each persisting vector genome. Mol Genet Metab 105 203 211, doi:10.1016/j.ymgme.2011.10.020 (2012). 79 Jayandharan, G. R., Zhong, L., Li, B., Kachniarz, B. & Srivastava, A. Strategies for improving the transduction efficiency of single stranded adeno associated virus vectors in vitro and in viv o. Gene therapy 15 1287 1293, doi:10.1038/gt.2008.89 (2008). 80 Calcedo, R., Vandenberghe, L. H., Gao, G., Lin, J. & Wilson, J. M. Worldwide epidemiology of neutralizing antibodies to adeno associated viruses. J Infect Dis 199 381 390, doi:10.1086/595830 (2009). 81 Davidoff, A. M. et al. Comparison of the ability of adeno associated viral vectors pseudotyped with serotype 2, 5, and 8 capsid proteins to mediate efficient transduction of the liver in murine and nonhuman primate models. Molecular therapy : t he journal of the American Society of Gene Therapy 11 875 888, doi:10.1016/j.ymthe.2004.12.022 (2005). 82 Wang, L. et al. AAV8 mediated hepatic gene transfer in infant rhesus monkeys (Macaca mulatta). Molecular therapy : the journal of the American Societ y of Gene Therapy 19 2012 2020, doi:10.1038/mt.2011.151 (2011).


95 83 Wang, L. et al. Impact of pre existing immunity on gene transfer to nonhuman primate liver with adeno associated virus 8 vectors. Human gene therapy 22 1389 1401, doi:10.1089/hum.2011.031 (2011). 84 Mingozzi, F. et al. Pharmacological Modulation of Humoral Immunity in a Nonhuman Primate Model of AAV Gene Transfer for Hemophilia B. Molecular therapy : the journal of the American Society of Gene Therapy doi:10.1038/mt.2012.84 (2012). 85 McI ntosh, J. H. et al. Successful attenuation of humoral immunity to viral capsid and transgenic protein following AAV mediated gene transfer with a non depleting CD4 antibody and cyclosporine. Gene therapy 19 78 85, doi:10.1038/gt.2011.64 (2012). 86 Ishiwat a, A. et al. Liver restricted expression of the canine factor VIII gene facilitates prevention of inhibitor formation in factor VIII deficient mice. J Gene Med 11 1020 1029, doi:10.1002/jgm.1391 (2009). 87 Sarkar, R., Xiao, W. & Kazazian, H. H., Jr. A sin gle adeno associated virus (AAV) murine factor VIII vector partially corrects the hemophilia A phenotype. J Thromb Haemost 1 220 226 (2003). 88 Qadura, M. et al. Immunoglobulin isotypes and functional anti FVIII antibodies in response to FVIII treatment i n Balb/c and C57BL/6 haemophilia A mice. Haemophilia 17 288 295, doi:10.1111/j.1365 2516.2010.02397.x (2011). 89 Singh, A. G., Hamarneh, I. S., Karwal, M. W. & Lentz, S. R. Durable responses to rituximab in acquired factor VIII deficiency. Thromb Haemost 106 172 174, doi:10.1160/TH11 02 0097 (2011). 90 Pescovitz, M. D. et al. Effect of rituximab on human in vivo antibody immune responses. J Allergy Clin Immunol doi:10.1016/j.jaci.2011.08.008 (2011). 91 Zhang, A. H., Skupsky, J. & Scott, D. W. Effect of B cell depletion using anti CD20 therapy on inhibitory antibody formation to human FVIII in hemophilia A mice. Blood 117 2223 2226, doi:10.1182/blood 2010 06 293324 (2011). 92 Xiang, Y. et al. The Dual Effects of B Cell Depletion on Antigen Specific T Cell s in BDC2.5NOD Mice. J Immunol doi:10.4049/jimmunol.1103055 (2012). 93 Doering, C. B., Healey, J. F., Parker, E. T., Barrow, R. T. & Lollar, P. High level expression of recombinant porcine coagulation factor VIII. J Biol Chem 277 38345 38349, doi:10.1074 /jbc.M206959200 (2002). 94 Gangadharan, B., Parker, E. T., Ide, L. M., Spencer, H. T. & Doering, C. B. High level expression of porcine factor VIII from genetically modified bone marrow derived stem cells. Blood 107 3859 3864, doi:10.1182/blood 2005 12 49 61 (2006).


96 95 Sun, B. et al. Immunomodulatory gene therapy prevents antibody formation and lethal hypersensitivity reactions in murine pompe disease. Molecular therapy : the journal of the American Society of Gene Therapy 18 353 360, doi:10.1038/mt.2009.1 95 (2010). 96 Jayandharan, G. R. et al. Optimized adeno associated virus (AAV) protein phosphatase 5 helper viruses for efficient liver transduction by single stranded AAV vectors: therapeutic expression of factor IX at reduced vector doses. Human gene the rapy 21 271 283, doi:10.1089/hum.2009.100 (2010). 97 Thomas, C. E., Storm, T. A., Huang, Z. & Kay, M. A. Rapid uncoating of vector genomes is the key to efficient liver transduction with pseudotyped adeno associated virus vectors. Journal of virology 78 3110 3122 (2004). 98 Finn, J. D. et al. Eradication of neutralizing antibodies to factor VIII in canine hemophilia A after liver gene therapy. Blood 116 5842 5848, doi:10.1182/blood 2010 06 288001 (2010). 99 Markusic, D. M. et al. High efficiency transduc tion and correction of murine hemophilia B using AAV2 vectors devoid of multiple surface exposed tyrosines. Molecular therapy : the journal of the American Society of Gene Therapy 18 2048 2056, doi:10.1038/mt.2010.172 (2010). 100 Xiang, Y. et al. The dual effects of B cell depletion on antigen specific T cells in BDC2.5NOD mice. J Immunol 188 4747 4758, doi:10.4049/jimmunol.1103055 (2012).


97 BIOGRAPHICAL SKETCH Brandon Sack was born in Longwood, Florida to Diane and Jerome Sack. He grew up in and aroun d central Florida always enjoying the outdoors and in particular the water. He was first introduced to the world of science by his mother who provided him with a microscope and a pack of slides on which to explore everything from pond scum to dog hair. He carried this interest into the University of Florida fr om which he received a Bachelor of Science degree in biology and a minor in classical s tudies. He joined the University of Florida IDP in August of 2007, was awarded a Grinter fellowship and a Universi ty of Florida Alumni Fellowship to support his research. He eventually joined the lab of Dr. Roland Herzog following rotations with Dr. Kenneth Warrington and Dr. Dietmar Siemann The Herzog lab focused on gene therapy for hemophilia and the immunological consequences thereof. He received recognition for his work by being awarded an NIH Infection Disease and Microbiology Training Grant in 2009 and the Medical Guild silver medal in 2012. During his tenure he published one first author research paper, one co author research paper, a first author review article and one first author book chapter. Brandon accepted a post doctoral research position at the Seattle Biomedical Research institute where he will provide immunological insight into the development of a m alaria vaccine. He hopes to continue working in academic research inv olving global health challenges while also teaching and preparing future generations of scientists through teaching and mentoring.