951 novel AAV FVIII single chain vectors that produce high levels of FVIII activity in hemophilia a mice

951 Novel AAV FVIII Single Chain Vectors That Produce High Levels of FVIII Activity in Hemophilia A Mice 949 Design of Gene Correction Devices for Improved Recognition by Protein Involved in Homologou[.]

949 Design of Gene Correction Devices for Improved Recognition by Protein Involved in Homologous Recombination Hiroyuki Kamiya,'? Hiroyuki Tsuchiya,'? Masayuki Uchiyama,' Kazuhiro Hara," Hideo Inoue.' Hideyoshi Harashima.F [Faculty 0/Pharmaceutical Sciences lJokkaido University, Sapporo, Japan; lCRES T, Japan Science and Technology, Kawaguchi, Japan; JGraduate School 0/Engineering, Osaka City University, Osaka, Japan [Background] Personalized medicine based on an individual's genetic information, as gene therapy, is expected to start in the ncar future.An attractivestrategy is gene correction, by which a mutated gene isconvertedto one withthe normal(or desired)sequence However, the current gene correction methodyielded the low correction efficiency, and thus,the methodmustbe improved.Asingle-stranded (ss) DNA fragment, prepared from ss phagemid DNA, corrects a targetgene with more than 1O-fold efficiencyas compared to a PCR fragment, which is the conventional type of DNAfragment used in gene correction [I] This gene correction possibly involveshomologous recombination, since target DNAbecomesradioactive when a radioisotope-labeled ss DNAfragmentis used [2] In this study,new gene correction devices were designed to improve recognition by protein(s) involved in homologous recombination.FMethods] The 606-nt ss fragmentwas prepared by restriction enzyme digestion of the ss phagemid containing a desired sequence [1] The new gene correction devices were constructed by mixing thc ss DNAwith an oligonucleotide The gene correction efficiencies of these devices were tested with an inactivated episomal hygromycin-resistance (Hyg) and enhancedgreen fluorescence protein(EGFP) fusiongene witha basesubstitution(TCA:Serto TGA: Stop; S34X)mutationas a model target The new devices and the target plasmid DNA were cotransfectedwith cationic lipids into CHO-KI cclls.Ikesults] The mixing of the ss DNA with an oligonucleotide increased gene correction efficiency in CHO-Kl cells Position, length, and backbone structureof'the oligonucleotidesaffectedgene correction efficiency, In the best cases, the newly designed DNA fragments provided a -2-fold increase in the gene correction activity, as compared to the 606-nt ss fragment alone Nearly 4% gene correction frequencies were observed when a 400-fold molar excess of these new devices were cointroduced with the target plasmid DNA [Concluslcns] The newly designed DNA fragments increased gene correction efficiency, These results indicate that gene correction could be improvedby efficientrecognitionby protein(s)involvedin homologous recombination [I] Tsuchiya et aI., J Gene Med., 7,486-493 (2005) [2] Tsuchiya et aI., Biochem Biophys, Res Commun 336, I 194-1200(2005) INBORN ERRORS OF METABOLISM II - HEMOPHILIA PLUS OTHERS 950 Intraarticular Expression of Factor IX from Adeno-Associated Virus (AAV) Vectors Decreases Development of Hemophilic Joint Pathology Junjiang Sun; Narine Hakobyan.' LeonardA Valentino.l Paul E Monahan.I [Gene Therapy Center; University 0/North Carolina Chapel Hill, NC; 2!JemophilialThrombophilia Center; RUSH University, Chicago, IL Objective: In the treatment of severe hemophilia B, maintaining circulatingplasma factor IXactivityadequateto preventhemophilic arthropathy (HA) is challenging and costly using protein replacement, and has not been achieved with systemic gene therapy in clinical trials Given that the major morbidity of hemophiliaoccurs primarily in six joints, we investigateddirect transduction of joints S362 with AAV expressing clotting factor to protect joints from bloodinduced degeneration Methods: Serotype AAVl, AAV2, AAV5 and AAV8 vectors were used to transduce (I) human joint tissues in vitro with green fluorescent protein or (2) mouse knee joints in vivo via intraarticular (LA.) injection of vectors expressing firel1y luciferase followed by serialbioluminescence imagingover5 weeks Toestablishwhetherconcentrated humanfactorIX(hFIX)withinthe joint space (rather than circulating in plasma) provides hemostasis, the hind kneejoints offaetor IX knockout(FIXKO) mice were subjected to a needlepuncture injury Simultaneouswithjoint puncture, either normal saline or hFIX doses ranging from the equivalentof IUlkg to 20 IU/kg were instilled intraarterially (LA.) Comparison groups received the same injury and intravenous (LV.) hFIX 25 IUlkg to 100 IU/kg.Two weeks after injury,joints were harvested Joint deterioration was graded (0-10) with a standardized mouse hemophilic synovitis histopathology scale Finally, FIXKO mice (n=o4/group) were injected in the left knee with either a lower dose AAV2 or AAV5 (2.5 x 109 vector genomes) or higher dose AAVI, AAV2, AAV5, AAV8 (I x 1OIU vg) hFIX vectors under hemostatic coverage with LV FIX concentrate; right knees were injected with Saline as control.After weeks, bilateral knee puncture injury was induced.Twoweeks later,joints were harvestedand pathologygraded Results: The marker gene transduction studies determined that differentAAV serotypesdisplaygreater than 1OO-fold differencesin expressionand varyingpatternsof tissuetropismafterdeliveryto the joint; these include differencesin isolated local expression (AAV2, AAV5) versus extraarticular spread (AAV8) and in efficiency of transduction of synoviocytes versus chondrocytes Examining the injury model in FIXKO mice, 100%of29 untreated control knees showed subacute joint damage (mean pathology score 4, range 27) FIXKO mice receiving protein doses as low as 10 IUlkg hFIX showed significant protection from synovitis when compared to untreated control and to all LV treatment groups Following gene therapy, AAV1.hFIX-treated joints were not consistently protected from subacute damage Mean pathology scores in all other AAVhFIX treated groups were at least 50% lower than injured controls Higher dose AAV2-or AAV5-treatedjoints showed minimal or no signs of synovitis, with most joints scoring 0-1 Median synovitis scores for I x I0 111 vg AAV2-, AAV5-hFIX and control knees were 0,0.67, and 4.9, respectively Conclusion: Intraarticularhemostasis and joint-directed gene therapy may ameliorate the events that lead to hemophilicjoint destruction 951 Novel AAV-FVIII Single Chain Vectors That Produce High Levels of FVIII Activity in Hemophilia A Mice Denise E Sabatino.P Amy M Lange,' Samuel Vidal,2 Rita Sarkar,' ValderR Arruda,' Haig H Kazazian, Jr.2 [Dept 0/ Pediatrics, Children 50 Hospital 0/ Philadelphia, Philadelphia, PA; lDept ofGenetics, School ofMedicine, University 0/ Pennsylvania, Philadelphia, PA Gene therapy is a promising approach for the treatment ofhemophiliaA (HA)which requiresonly ~ I% ofnonnal factorVIII(FVIII) activity to alter the clinical phenotype from severe to moderate disease The limited capacity of the adeno-associatedviral (AAV) vector makes it challenging to accommodate the 4.5 kb B-domain deleted FVIII eDNA and its regulatory elements We have utilized a two-chain delivery approach to administer one AAV vector expressing the FVIII heavy chain and one expressing the FVlII light chain This approach results in therapeutic levels ofFVIII in mice and dogs, however, there may be safety concerns associated with using high vector doses Thus, we have developed an approach based on use of single chain FVIIL Previously, we demonstrated that HAmice administered an AAV8-cFVIII (5.6 kb) vectorat a dose of3x IOlUgc/mouse initiallyexpressed 80% of normal FVIIIactivity Molecular Therapy Volume 15 Supplement I• \ b)' 2007 Co pyright ~ The Americm Society o f Gen e The rapy that plateaued at 50% activity (Sarkar et al., 2004) However, the inability to produce high titer AAV vector limits application ofsuch a large transgene construct in large animals We hypothesized that a :0;;5.2 kbAAV construct would package more efficiently and result in a high titer vector without sacrificing high levels ofFV111 expression Nine AAV-cFVIII constructs (5.109-5.371 kb) were generated that utilized combinations of the following regulatory elements: alphaI microglobulin/bikunin enhancer and insulin-like growth factor binding protein (lGFBP) promoter OR a version of the hepatic control region of the human apoliprotein gene locus and human alpha-I anti-trypsin promoter (hAAT) In addition, we compared constructs with or without an intron and with different SV40 polyA signals (134bp and 263bp) cFVIII expression from these plasmid constructs was assessed in vitro by transient transfection ofHepG2 cells and in vivo by hydrodynamic infusion into HA mice These studies demonstrated that one hAAT construct (5.211 kb) followed by two IGFBPconstructs, IGFBP-I (5.209 kb) and IGFBP-2 (5.242 kb), consistently displayed the highest levels of cFVIII activity (as assayed by Coatest and aPTI') relative to the other constructs These constructs had cFV111 expression greater than or equal to the original 5.6 kb construct AAV-8 vector was produced from these three 5.2 kb constructs and the original 5.6 kb construct HA mice (n=4 mice/group) wcre injected in the tail vein with 5x IOlllgc/mouse At weeks post vector administration, the levels of cFVIII activity determined by Coatest assay were IlOo/o±23.91 (original), 126o/o±9 (hAAT), 106o/o±9 (lGFBP- J) and IIIo/o±33 (lGFBP-2) These results suggest that the smaller single chain vectors (-5.2 kb) express high levels of cFVIII that is similar to the levels of expression observed with the early 5.6 kb version Further evaluation of the packaging efficiency of these single chain FVIII AAV vectors will determine their utility in large animal models 952 Long-Term FVIII Expression Via Sleeping Beauty in Liver Sinusoidal Endothelial Cells of Transgenic Mice Betsy 1' Kren, I Gretchen M Unger.' Alycia A Trossen ,' Mark T Reding; Robert P Hebbel ,' Clifford J Steer.' I Department ofMedicine University ofMinnesota Minneapolis, MN; 'Genesegues, lnc., Chaska, MN The use of nonviral vectors for gene therapy has been hindered by the lack of adequate in vivo delivery systems Although hydrodynamic has been used extensively for hepatic plasmid delivery, this method delivers DNA nonselectively to liver cells One approach for specific cell type delivery is to target receptor(s) that are either unique or highly expressed by that cell Liver sinusoidal endothelial cells (LSEC) express hyaluronan receptors (HAr) in high abundance providing an ideal target for ligand-mediated receptor uptake The aim of this study was to determine if LSEC specific delivery of DNA could be achieved in vivo using HA targeting to the HAr Using a novel dispersion atomization method that forms sub 50 nm capsules coated with the receptor ligand, a red fluorescent protein (DsRed2) controlled by a constitutive promoter or ~-galactosidase (lacZ) expressed by the hepatocyte-specific albumin (Alb) promoter were encapsulated using HA for LSEC uptake Eight-week (wk) old mice received 100 ug of HA encapsulated plasmid via tail vein injection and were sacrificed I wk post-injection Liver, spleen , kidneys, lung, heart & brain were excised and processed for histology, protein extracts and DNA Immunohistochemical identification of LSECs in cryosections was done using anti-CD 14 antibody (Ab), a marker specific for the discontinuous endothelial cells in liver and a Cy5 labeled secondary Ab The merged confocal micrographs demonstrated colocalization of DsRcd2 and the LSEC specific CDI4 marker, Presence of the DsRed2 protein was confirmed by western blot analysis of total liver protein extracts When the Alb promoter was used, no liver lacZ expression was Molecular 'Therapy Volume 15 Supplement ~ br 2007 Copyright © The American Soci ety o r Gene "1l1f:r:lpy observed although the plasmid was readily detectable by PCR No plasmid was detected by PCR in the DNA isolated from the other tissues Transgenic hemophilia A (hemA) mice were injected with HA nanocapsules containing a Sleeping Beauty (SB) transposon (Tn) construct expressing B-domain deleted coagulation factor (F) VIII in cis with SB transposase for genomic insertion The mice were bled 2,5, II, and 44 wks post injection and plasma activated partial thromboplastin time (aPTT) determined The treated mice had aP1Ts of 25.5 ± 3.1 (2 wks) that remained low 28.8 ± 3.7 sec (44 wks) and were not significantly different from the age matched wild type (wt) aPTI's of23.5 ± 1.3 sec (2 wks) and 27.9 ± 1.6 sec (44 wks) In contrast, untreated hemA mice had aPTTs of 65.7 ± 9.6 sec (P < 0.00 I from treated & wt mice) We also targeted a cis SB-Tn expressing a l-antitrypsin (a IAT) to the LSEC using the HA nanocapsules The wt mice were bled I, and wks after tail vein injection, the blood clotted and spun , and the serum levels determined using thee l-antitrypsin enzyme immunoassay fromALPCO diagnostic The levels observed were 375 ± 45 nglml and increased slightly to 410 ± 75 nglml by wks, In conclusion, HA targeted nanocapsules can deliver plasm ids in vivo with high specificity to LSEC SB-Tns in HA nanocapsules targeted to LSEC provided longterm expression of clinically relevant gene products and the FV111 improved the phenotype of hemA transgenic mice 953 High-Level, Erythroid-Specific Human Factor IX Expression in Hematopoietic Chimeras Using Non-Myeloablative Conditioning? Alex H Chang, Michel Sadelain ' I Laboratory ofGene Transfer and Gene Expression, Sloan-Kettering Institute , Nell' }vrk, Nf Hematopoietic stem cells (HSCs) arc essential target cells for gene therapy because oftheir self-renewing properties and their ability to generate all blood cell types The broad usc of genetically modified HSCs is currently limited by the risk of insertional mutagenesis associated with LTR-driven garnma-retroviral vectors and the toxicity of the conditioning regimens needed to achieve high-level engraftment of transduced cells We have previously developed an HSC-based tissue-specific gene expression platform for long-term and systemic therapeutic protein delivery, using an crythroid-spccific lentiviral vector (Chang, et al Nature Biotechnology, 2006) We demonstrated that therapeutic levels of human factor IX (hFIX) can be expressed at low average vector copy (VC) number (250350 ng/mL serum hFIX with 0.5 VC per cell) in radiation chimeras To make this promising therapeutic paradigm more acceptable for patient therapy, we examined the levels of erythroid-specific protein delivery obtained after engraftment under nonmyeloablative conditioning using Busulfan, followed by in vivo selection for mcthylguanine mcthyltransferase (MGMT)-mediatcd drug resistance In non-myeloablated recipients conditioned with Busulfan (20 mg/kg, given twice 48 hand 24 h before transplantation), the initial levels ofhFIX expression were from negligible to low (37.1 I ± 28.11 nglml) Three cycles of selection were administered using BGIBCNU (30 mg/kgand mg/kg, respectively) at 5-week intervals hFIX expression levels steadily increased after each selection The average hFIX levels reached 610.07 ± 194.32 nglml weeks after the second selection, a level corresponding to and above 10% of the human physiological level and expected to be curative in hemophilia B patients The hFIX levels in these chimeras continue to rise after the third selection, eventually reaching 1879.66 ± 165.22 nglml (37.5% of the normal human level) The high levels ofhFIX expression were maintained in secondary chimeras for 16 weeks (712.26 ± 155.50 nglml) These results indicate that BG/BCNU selection enhances hematopoietic reconstitution by lentivirustransduced long-term progenitor cells stably engrafted following non-myeloablative conditioning Altogether, our findings suggest S363 .. .that plateaued at 50% activity (Sarkar et al., 2004) However, the inability to produce high titer AAV vector limits application ofsuch a large transgene construct in large animals We hypothesized... high levels of cFVIII that is similar to the levels of expression observed with the early 5.6 kb version Further evaluation of the packaging efficiency of these single chain FVIII AAV vectors will... generated that utilized combinations of the following regulatory elements: alphaI microglobulin/bikunin enhancer and insulin-like growth factor binding protein (lGFBP) promoter OR a version of