Scholarly article on topic 'Biology of AAV Serotype Vectors in Liver-Directed Gene Transfer to Nonhuman Primates'

Biology of AAV Serotype Vectors in Liver-Directed Gene Transfer to Nonhuman Primates Academic research paper on "Biological sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Mol Ther
OECD Field of science
Keywords
{""}

Academic research paper on topic "Biology of AAV Serotype Vectors in Liver-Directed Gene Transfer to Nonhuman Primates"

Biology of AAV Serotype Vectors in Liver-Directed Gene Transfer to Nonhuman Primates

Guangping Gao,* You Lu,* Roberto Calcedo, Rebecca L. Grant, Peter Bell, Lili Wang, Joanita Figueredo, Martin Lock, and James M. Wilson1"

Gene Therapy Program, Division of Medical Genetics, Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, PA 19104, USA

*These authors contributed equally to this work.

1To whom correspondence and reprint requests should be addressed at TRL, Suite 2000, 125 S. 31st Street, Philadelphia, PA 19104, USA. Fax: +1 215 898 6588. E-mail: wilsonjm@mail.med.upenn.edu.

Available online 10 October 2005

Vectors based on adeno-associated viruses (AAVs) show promise for the treatment of genetic diseases. This study evaluates the biology of AAV-mediated gene transfer to liver in nonhuman primates (NHPs) using vectors based on AAV serotypes 2, 7, and 8. Transgenes encoding self-proteins were selected to minimize the confounding development of transgene-specific immune responses. These included the (3 subunit of choriogonadotropic hormone (bCG) and erythropoietin (Epo), both derived from cDNAs from rhesus macaques. Experiments were performed with bCG in rhesus macaques and Epo in cynomolgus macaques. We demonstrated the previously untested hypothesis that preexisting immunity to a natural infection does substantially diminish the efficacy of gene transfer with a vector derived from an endogenous virus. Route of vector administration clearly has an impact on the development of immune responses to self-antigens. In general, efficiency of gene transfer to liver with AAV7 and 8 vectors was higher than what was achieved with AAV2, although a variety of host factors may influence this important parameter, such as preexisting immunity, gender, and transgene immunity.

Key Words: novel AAVs, vector biology, NHP model for liver gene transfer, gene expression, molecular status of vector genome, preexisting immunity, safety profiles

Introduction

Vectors based on adeno-associated viruses (AAVs) have shown promise in preclinical and clinical models of gene therapy. To date, more than 110 different primate AAVs have been reported, with 9 of them demonstrating unique serological properties [1-7]. While AAV1, AAV2, AAV3, AAV4, and AAV6 were isolated as contaminants in adenovirus preparations and AAV5 was isolated from a human condylomatous wart [1-4], AAV7, AAV8, and AAV9 were recovered from primate tissues by a polymer-ase chain reaction (PCR)-based molecular rescue method [5,7]. As a naturally emerged hybrid of AAV1 and AAV2, AAV6 did not appear to be a distinct serotype [4,7]. Natural infection of AAV serotypes in primates was documented based on the existence of latent genomes but no clinical sequelae have been associated with AAV infections [8,9].

AAV2 is the first serotype developed as a vector for gene transfer studies. A majority of the preclinical and clinical studies of AAV-mediated gene therapy were based on

AAV2 [10-13]. However, further development of AAV2 as a vector has been limited because of its restricted tissue tropism and low efficiency of gene transfer, particularly in liver [13].

Our laboratory has long been interested in developing vector systems for liver-directed gene therapy. Vectors based on serotypes other than AAV2, generated using the transcapsidation method, have demonstrated serotype-specific vector biology and improved gene transfer efficiency in preclinical studies, mainly in rodents [1418]. Comparative studies of AAV serotypes 1- to 8-based vectors for liver-directed gene transfer in different murine models have established serotype-specific liver tropism [5,13,19,20]. In particular, vectors based on AAV7 and AAV8 demonstrated 10- to 100-fold higher gene transfer efficiency than AAV2 in murine liver and achieved long-term phenotypic correction in several murine disease models [5,20-22]. In a more recent study, Davidoff et al. compared AAV2, 5, and 8 vectors in both rodent and rhesus macaque models [23].

Studies of novel AAV serotype vectors in rodent models provide unique opportunities to screen and select lead vector candidates for further evaluation. We believe it is important to test lead gene transfer vectors for their biological potency and safety in nonhuman primate models. In the present study, we focused on evaluation of the biology of AAV serotypes 2, 5, 7, and 8 vectors in liver-directed gene transfer to two species of NHPs.

Results and Discussion

The goal of our study was to evaluate vector biology of novel AAV serotypes in liver-directed gene transfer to NHPs. We considered two aspects of host responses that could influence directly the outcome of the study: adaptive B cell immunity to the transgene product, which would blunt or diminish detection of the transgene product, and serotype-specific preexisting immunity to the capsid proteins of AAV vectors that would decrease gene transfer efficiency. To overcome the first obstacle, we selected the h chain of rhesus chorionic gonadotropin (rhCG), a self-antigen, as a reporter gene and rhesus macaques (Macaca mulatta) as the recipients for a side-by-side comparison of AAV2, AAV5, AAV7, and AAV8 vectors. In a second set of studies, we used rhesus erythropoietin (rhEpo) as a reporter gene injected into the portal circulation of cynomolgus macaques, a species of NHPs with less prevalence of preexisting immunity to the novel AAVs. We used rhEpo as the reporter gene based on concerns that there might be adaptive B cell response against rhCG in cynomolgus macaques. Unlike the Epo gene, which has total sequence identity between rhesus and cynomolgus macaques, rhesus CG shares only 94% sequence identity with the cynomolgus gene [7,26].

Study 1: Assessment of AAV2-, 5-, 7-, and 8-Derived Vectors Expressing RhCG in Rhesus Macaques

To eliminate possible B cell immunity directed to the transgene product, we administered AAV serotype vectors expressing rhCG driven by the human thyroid-hormone-binding globulin (TBG) promoter for liver-specific transduction intraportally at a dose of 1 x 1013 genome copies (GC)/kg to the rhesus macaques that were prescreened for neutralizing antibodies (NABs) against the vectors. A summary of these experiments is provided in Table 1. We have monitored the animals for up to 172 weeks after gene transfer;most of them are still alive.

The kinetics of transgene expression varied within the different serotypes (Fig. 1). Both animals who received AAV2 realized an acute increase in rhCG followed by a gradual decline to an average of 8 x 102 rU/ml. We observed basically similar patterns for AAV2/5 and AAV2/ 7, although for each vector there were variations between the two animals in each group and the current levels of transgene exceeded those in the AAV2 group by 4- and 11fold for AAV2/5 and 14- and 68-fold for AAV2/7. We saw substantial differences between the two AAV8 animals 1 week after gene transfer, which in each case immediately declined; one of the two animals died of unrelated causes. The most recent rhCG data from each animal were 5- and 11-fold higher than the data for AAV2.

We considered several hypotheses for what could be contributing to variation in transgene expression between cohorts and between animals within a cohort. One potential explanation is the presence of NABs against the vector. To determine if serologic responses to natural AAV infections would indeed have an impact on gene transfer to liver, we performed additional studies with two animals who did demonstrate AAV8 NABs of 1:320 and 1:1280 based on our in vitro assay (Table 1). The animal with the higher level of preexisting NABs failed to show rhCG

TABLE 1: Summary of AAV serotype vector-mediated rhCG gene transfer to the liver of rhesus macaques

Weight Age ROAb NABc rhCG level (> <103 rU/ml)d

Animal ID (kg) (years) Gender Serotype3 Pre d30 Max/tp Stable/tp Animal statuse

97E104 4.25 3.7 M 2 ip <1:20 1/1,280 10.7/week 8 1/week 172 Alive/week 172

98E067 4.00 2.7 F 2 ip <1:20 1/1,280 7.2/week 1 0.8/week 172 Alive/week 172

97E082 4.55 3.8 M 5 ip <1:20 1/20,480 19/week 1 8.9/week 71 Alive/week 71

98E103 3.5 2.8 F 5 ip <1:20 1/163,840 19/week 1 3.6/week 71 Alive/week 71

RQ4324 3.4 2.5 F 7 ip <1:20 1/640 140/week 1 11/week 87 Alive/week 89

RQ4350 4.15 4.5 M 7 ip <1:20 1/320 210/week 4 54/week 87 Alive/week 89

99E149 2.85 2.7 F 8 ip <1:320 1/1,0240 0.27/week 1 0.1/week 20 Died/week 20

99E042 2.5 2.4 F 8 ip <1:1280 1/1,0240 0/week 34 0/week 1 Alive/week 168

RQ4338 4.7 4.3 F 8 ip <1:20 1/1,280 7.7/week 1 4/week 87 Necro/week 100

RQ4400 4.1 3.5 F 8 ip <1:20 1/1,280 22/week 2 8.6/week 30 Died/week 31

a All animals received the same dose of the corresponding vector (1 x 1013 genome copies/kg of animal weight). b ROA, route of administration; ip, intraportal.

c NAB, antibody. All NAB titers listed are titers against the serotype vector used in a particular animal.

d Max/tp, the peak of expression in terms of rhCG/time point at which it was observed; stable/tp, expression at the most recent time point/the most recent time point. e Animals 99E149 and RQ4400 died of unrelated causes at 20 and 31 weeks of the study, respectively. Animal RQ4338 was necropsied on day 702. All other animals except for 99E042 were recycled for ongoing vaccine studies while serum rhCG levels in those animals are continuously monitored.

fig. 1. Kinetics of rhCG expression in different AAV serotype vector-mediated gene transfers to the NHP liver. The study animals were bled at different time points post-vector infusion for measurement of serum rhCG levels. Animal ID numbers and vectors received are indicated in each graph.

expression, while that with the intermediate NAB level had rhCG expression 50- to 500-fold lower than animals without NABs. This result confirms the previously untested hypothesis that natural infections with wildtype virus can impact gene transfer against the corresponding vector. One concern is whether the in vitro assay is sufficiently sensitive to detect NABs that could have an impact on transgene expression in vivo. This is particularly problematic in human and nonhuman primate populations in which infection with a diverse array of AAVs seems to occur, leading to a complex mixture of cross-reacting antibodies of varying affinities. Additionally, the relatively low levels of transduction achieved with the newer AAV vectors in vitro limits the sensitivity of the NAB assay.

A second hypothesis is variability in the development of antibodies against the transgene product. Analyses of sera for rhCG-specific antibodies by ELISA and Western blots were negative (data not shown). The presence of rhCG antibodies at levels below the sensitivity of these assays cannot be ruled out, however.

The other confounding variable was the gender of the animals. A number of investigators have now demonstrated substantially diminished AAV-mediated gene transfer in female mice as opposed to male mice [27]. It is interesting to note that variations within the AAV2/5 and AAV2/7 groups was associated with higher expression in the male than in the female and that the overall disappointing results with AAV2/8 were in a cohort of only females.

fig. 2. Comparative analysis of the transgene expression and liver function test (LFT) data in the first 2 months of the AAV-mediated rhCG gene transfer to the NHP liver. AST (the left y axes) and rhCG (the right y axes) levels at different time points after vector infusion from each of eight study subjects are presented.

We monitored serum alanine aminotransferase and aspartate aminotransferase (AST) activities of the NHP animals enrolled in this study. RhCG gene transfer to rhesus liver by AAV vectors did not appear to cause any remarkable transaminase elevation other than some low-level, transient AST spikes in a few animals immediately after vector infusion (Fig. 2). The data from hematological and other clinical chemistry analyses showed no consistent differences from the baselines (data not shown).

Study 2: Evaluation of AAV2, 5, 7, and 8 Vectors for Rhepo Gene Transfer to Cynomolgus Liver

Our seroepidemiology data revealed that preexisting immunity to primate AAVs is more prevalent and potent in the rhesus than in cynomolgus populations (data not shown). We repeated the serotype comparison study in another species of NHPs, cynomolgus macaques (Macaca fascicularis) using a self-antigen, rhEpo, as the reporter. This would also allow us to investigate if previously reported autoimmune anemia in AAVrhEpo vector-treated NHPs was associated with the route of administration or target tissues [28,29].

We dosed serologically prescreened cynomolgus macaques with AAV2, AAV5, AAV7, and AAV8 TBGrhEpo vectors through intraportal (ip) or intravenous (iv) injections (Table 2). We followed the animals for gene expression up to 62 weeks after gene transfer at which time we euthanized and necropsied them. Following vector injection, we monitored serum rhEpo levels and clinical pathology data at 6, 12, and 24 h postprocedure, weekly for the first month, biweekly for the next 3 months, and monthly for the remaining time of the study. Interpretation of any vector-related effects in the hematological system is confounded by the effects of Epo. We examined hematocrits of the experimental animals weekly for the duration of the study. For any animal with a hematocrit over 65%, we performed a therapeutic phlebotomy procedure.

All recipient animals started showing elevated hema-tocrit (HCT) levels within 2 weeks (Fig. 3). Unlike previous studies of Epo gene transfer to lung and muscle of NHPs, no animals administered vector to the liver developed autoimmune anemia. While all AAV TBGrhEpo vector-treated animals achieved superphysiological levels of rhEpo expression, kinetics of the rhEpo gene expression in the different AAV serotype cohorts were quite different. For animals that received AAV2 vectors, we observed variable onset of rhEpo gene expression and levels of stable expression (Fig. 3 and Table 2). Animals that received the other AAV vectors showed more complex kinetics of gene expression characterized by a peak 10- to 100-fold higher than the highest expression observed with AAV2, followed by a rapid decline to steady-state levels that were remarkably consistent between animals within a group and between cohorts. Similar results were obtained when AAV2/8 was injected iv or ip.

Biological mechanisms behind the rapid onset and biphasic gene expression profiles (i.e., a high expression phase followed by a stabilization phase at lower levels) of new serotype vectors in NHP liver remain to be elucidated. A recent study by Thomas et al. suggested that rapid uncoating and self-annealing of single-stranded vector genomes is the key for efficient transduction of mouse liver by AAV8 [20]. This could explain the high levels of rapid gene expression achieved by new AAV serotype vectors in the early stages of NHP liver gene transfer. However, to maintain long-term transgene expression, double-stranded AAV genomes appear to undergo a complex series of modifications such as circularization and concatemerization. These processes occur through intra- and intermolecular recombination, which requires host cell factors involved in DNA replication and double-strand break repair [30]. One hypothesis could be that the capability of novel AAV capsids to target liver cells efficiently and deliver large loads of vector genomes in a short time frame could lead to a

TABLE 2: Summary of AAV-serotype vector-mediated rhEpo gene transfer to the liver of cynomolgus macaque

ROAb NABc Epo level (> < 103 mU/ml)d

Animal ID Weight (kg) Age (years) Gender Serotype" Pre Post (day 30) Max/tp Stable/tp

17086 2.35 3 F 2 ip <1:20 1:1,280 3.7/week 7 1.3/week 62

17080 2.10 3 F 2 ip <1:20 1:320 14/week 1 6.8/week 62

17130 2.55 4.4 M 5 ip <1:20 1:40,960 59/week 2 4.2/week 62

17132 1.95 3 M 5 ip <1:20 1:20,480 100/week 1 8.5/week 62

17093 1.95 4.5 F 7 ip <1:20 1:10,240 120/week 1 8.7/week 62

17095 1.95 4.3 F 7 ip <1:20 1:5,120 100/week 1 7.7/week 62

17099 2.32 3 F 8 ip <1:20 1:640 180/week 1 13/week 62

17045 2.20 3 M 8 ip <1:20 1:320 64/week 1 5.7/week 62

17119 2.10 4.4 F 8 iv <1:20 1:640 100/week 2 9.8/week 15

17155 2.15 3 M 8 iv <1:20 1:320 72/week 2 4.6/week 62

a All animals received the same dose of the corresponding vector (1 x 1013 genome copies/kg of animal weight). b ROA, route of administration; ip, intraportal; iv, intravenous.

c NAB, neutralizing antibody. All NAB titers listed are titers against the serotype vector used in a particular animal. d Duration of the study for all animals was 436 days except for animal 17119, who died of unrelated causes at day 120.

fig. 3. RhEpo expression profiles and hematocrit levels in the cynomolgus macaques that received different AAV serotype vectors. The study animals were bled at different time points post-vector infusion for measurement of serum rhEpo and hematocrit levels. Animal ID numbers and vectors received are indicated in each graph.

transient depletion of these critical cellular factors and loss of some unprocessed linear genomes.

Lack of adverse immunological reaction to long-term Epo expression in NHP liver by AAV-mediated gene transfer is in sharp contrast to the results of two previous studies by us and others. AAV vector-mediated rhEpo gene transfer to NHP muscle and lung led to the production of potent Epo inhibitors and autoimmune anemia in 50 to 100% of animals, respectively, with most of the serotypes tested [28,29]. Interestingly, none of the 10 animals with five different vectors in this liver study demonstrated any sign of such adverse responses (Fig. 3). It is worth noting that in our previous muscle experiment, a CMV promoter was used. However, in both lung and liver studies, the transgene Epo expression was directed from tissue-specific promoters, CC10 for lung and TBG for liver, respectively. Our studies suggest that the route of administration and the target cell/tissue type may have profound impacts on outcomes of viral vector-mediated

gene therapy [31]. Based on data available to date, one could argue that liver may be the most effective and safe target tissue of AAV gene transfer for natural and ectopic production of secreted factors based on transgene-specific B cell responses.

We investigated the molecular status of AAV TBGrhEpo vector genomes in the target liver cell by DNA hybridization analysis. When we analyzed total liver DNAs from the animals necropsied at day 436 without restriction endonuclease digestion, the abundance of vector genomes was ranked as follows: AAV8 > AAV7 > AAV5 > AAV2 (Fig. 4A). Liver DNAs from all animals of both the AAV7 and the AAV8 cohorts were the most abundant in vector genomes, but we detected very weak hybridization signals in the livers of the AAV2 cohort. This was confirmed by analyses of Taq-Man PCR and DNA hybridization with liver DNAs treated with a double cutter in the vector genome, Hpal (Fig. 4C).

fig. 4. Molecular status of AAV serotype vector genomes in cynomolgus macaque livers. Liver tissues were harvested at the end of the study by necropsy for total DNA extraction. Ten micrograms each of DNA from different animal livers, which were either (A) untreated or (B) digested with Xhol, a single cutter, or (C) Hpal, a double cutter in the vector genome, was subjected to Southern blot DNA hybridization analysis (RDC, relaxed double-strand circles; MSC, monomeric supercoiled double-strand double circles; LDC, linearized monomeric double-strand circles; HMW, high molecular weight; H-T, head-to-tail; T-T, tail-to-tail). The vector genome plasmid, pAAVTBGrhEpo, was digested with Hpal and used as a copy number control. A 1.2-kb HpaI fragment from vector genome plasmid was used as the probe for hybridization. The data of vector genome copy numbers per cell as quantified by TaqMan real-time PCR and serum rhEpo levels at the peak expression and the end of the study are also presented.

The dominant molecular form of uncut DNA from NHP liver seemed to be the relaxed double-strand monomeric circles but other forms, including monomeric supercoil double-strand circles and high-molecular-weight concatemers, were also identified (Fig. 4A). We performed further analysis of vector genome structure with DNA digested with Xhol, which has only one site within the genome (Fig. 4B). Analysis of these digests revealed two major bands: a dominant 2.6-kb band at the size of uni-length genomes could represent both linearized monomeric circles and head-to-tail concatemeric forms;a weaker 3.5-kb band could be released from tail-to-tail concatemeric forms. In a separate analysis, we treated liver DNAs with the Plasmid-Safe ATP-dependent exonuclease followed by TaqMan PCR quantification, rolling-circle amplification, and single-cutter digestion. This analysis specifically amplifies double-stranded circular genomes. The data suggest that double-strand circular molecules constitute a major portion of the persistent genomes as suggested by others (data not shown [30,32,33]).

We followed the animals for toxicity through clinical observations and sequential evaluation of blood hematol-ogies and chemistries. No other noticeable abnormality stood out in clinical pathology tests of all 10 recipients, except for a two- to threefold elevation in AST levels for the first month or two (data not shown and Fig. 5). We present serum AST and rhEpo levels of each animal in the first 2 months of the study for analysis in Fig. 5. Mild AST elevations in the cynomolgus macaques that received AAV TBGrhEpo vectors were quite common and appeared as either monophasic or biphasic (Fig. 5);the first spike observed within 24 h was probably associated with the laparotomy and vector infusion. The second peak of the AST activity coincided with the start of rhEpo transgene expression. Most noticeably, AST levels in three animals (19080, 17093, and 17132) from the AAV2, AAV7, and AAV5 cohorts, respectively, increased three- to fourfold above baseline after the rhEpo expression reached peak at day 7. The increases in AST activity were followed by rapid declines of serum rhEpo levels in these animals, although a cause and effect between these two observations has not been established (Fig. 5).

Innate immunity to viral capsids in gene therapy recipients is another safety concern in the systemic delivery of viral vectors. We measured serum levels of five major inflammatory cytokines (TNF-a, IFN-g, IL-6, IL-10, and IL-12) in all animals in this study at 6, 12, and 24 h immediately after vector dosing. Serum from a rhesus macaque at the same time points before and after intra-vascular infusion of an Ad5 vector was used as a positive control [25]. Our data revealed that systemic administration of different AAV serotype vectors at a dose of 1 x 1013 GC/kg led to no elevations of the inflammatory cytokines assayed in any of the study animals (data not shown) [25].

Histopathological analysis of liver sections of all animals necropsied at day 436 showed no apparent

abnormality (data not shown). We noticed blood congestion in some livers, but it seemed to be cardiac in nature, probably associated with the necropsy procedure rather than extramedullary hemacenteroiesis caused by the high level of erythropoietin in the animals (data not shown). The same analysis performed in 10 other organs, including brain, bone marrow, colon, heart, kidney, lung, lymph node, small bowel, spleen, and ovary/testes, provided no evidence of histopathology.

To study vector dissemination and biodistribution, we extracted total DNAs from peripheral blood samples collected at early time points and 11 different tissues harvested at the end of the study for real-time PCR and DNA hybridization analyses. In terms of pharmacoki-netics of AAV vectors in blood, our data demonstrated that high copy numbers of AAV vectors remained in the peripheral blood circulation for at least 24 h but decreased over time. At the same dose, 10- to 100-fold more AAV5, 7, and 8 vectors were detected in blood compared to AAV2 (data not shown). Our data seem to suggest that AAV2 vector is cleared from blood more rapidly than other serotypes. However, most of those liver-targeted genomes were gradually lost in the process of genome conversion to form transcriptionally active molecules [23].

Biodistribution analysis of vector genomes of the vectors delivered ip or iv in 11 tissues revealed that the primary target tissue was liver, as expected, followed by spleen (Fig. 6). This TaqMan-based assay has a detection limit of 10 vector genomes per 0.1 Ag of tissue DNA, which corresponds to 1.5 x 10~3 copies per diploid genome. There was some animal-to-animal variation in the same vector cohorts but no other tissue(s) with a strong vector affinity stood out. When vector abundance was low in liver in the case of the AAV2 cohort (0.1-1 GC/ cell), the vector also targeted other tissues poorly. Highlevel targeting of liver observed with AAV2/7 and AAV2/8 (~10 GC/cell) was associated with proportionately higher targeting outside the liver. The situation with AAV2/5 was quite different, with substantially higher levels of extrahepatic distribution and persistence relative to the levels found in liver.

Our study demonstrated the feasibility and safety of targeting novel AAV serotype vectors to NHP liver to express rhCG and rhEpo efficiently. A few important principles did emerge from these studies. The previously untested hypothesis that preexisting immunity to a natural AAV infection will diminish AAV-mediated gene transfer to liver was indeed confirmed. Epo gene transfer demonstrated that toxic humoral immune responses to the transgene product are much less likely to occur when vector is transferred to liver than to muscle or to lung. In general, the observation in mice that vectors based on the novel AAVs 7 and 8 effect more efficient gene transfer to liver than AAV2 was relevant to some of the NHP studies. The issue of gene transfer efficiency, however, is a

0 10 20 30 40 so 60 700 10 20 30 40 50 60 70

Days Days

fig. 5. Comparative analysis of the transgene expression and LFT data in the first 2 months of the AAV-mediated rhEpo gene transfer to the NHP liver. AST (the left y axes) and rhEpo (the right y axes) levels at different time points after vector infusion from each of 10 study subjects are presented.

fig. 6. Biodistribution of vector genomes following AAV TBGrhEpo serotype vector infusion into cynomolgus livers. Ten distant organs and targeted liver tissue were harvested from the nine study animals at the end of the study by necropsy. The quantity of vector genomes within each tissue was analyzed by TaqMan PCR using a probe/primer set to amplify the vector genome. The abundance of vector genomes is illustrated as copies per diploid cell (F, female; M, male). BM, bone marrow; B, brain; C, colon; H, heart; K, kidney; LI, liver; Lu, lung; LN, lymph node; GD, gonadal tissue; SB, small bowel; S, spleen.

complicated issue potentially impacted by a variety of host-related factors such as preexisting immunity, gender, and transgene-specific immune responses. The challenges of studying the biology of a vector derived from a virus endogenous to the recipient are highlighted in these experiments. Our previous studies on liver-directed gene transfer in mice focused on AAV8 as a preferred vector. The current studies suggest AAV7 should also be considered.

Materials and Methods

AAV vector production. AAV2TBGrhCG and AAV2TBGrhEpo vector genomes were transcapsidated with capsid proteins of different serotypes and purified by CsCl gradient sedimentation as previously described [5]. All vector preparations used for the NHP studies were subjected to extensive quality control tests including four repeated vector genome titrations by TaqMan PCR, SDS-PAGE analysis for vector purity, LAL, and transgene expression analysis in mice. Liver-specific expression of rhCG and rhEpo was directed by the human TBG gene promoter [24].

NHP experiments. Rhesus macaques (Indian origin and captive bred; 2.45-4.7 kg; AAV2, 97E104 and 98E067; AAV2/5, 97E082 and 97E103; AAV7, RQ4324 and RQ4350; AAV8, RQ4338 and RQ4400) used in the rhCG gene transfer experiment were treated and cared for at the NHP facility of University of Pennsylvania. Cynomolgus macaques (Philippine origin and captive bred; 1.95-2.55 kg; AAV2, 17086 and 17080; AAV5, 17130 and 17132; AAV7, 17093 and 17095; AAV8 ip, 17099 and 17045; AAV8 iv, 17119 and 17155) in the rhEpo gene transfer experiment were treated and cared for at a contract NHP facility (LABS of Virginia, Yemassee, SC, USA) during the study. Both experiments were performed following study protocols approved by the Environment Health and Radiation Safety Office, the Institutional Biosafety Committee, and the Institutional Animal Care and Use Committee (IACUC) of the University

of Pennsylvania and the IACUC of LABS of Virginia. The vector dose for both rhCG (via a mesenteric tributary of portal vein) and rhEpo (via either a mesenteric tributary of portal vein or a saphenous vein) experiments was 1 x 1013 GC/kg of animal weight. Blood samples were taken via venipuncture of the saphenous vein. In the rhEpo gene transfer experiment, therapeutic phlebotomy was performed for any animal with a hematocrit over 65% and all animals were necropsied at day 436 of the study. Eleven tissues including the target tissue liver and 10 other distant organs (brain, bone marrow, colon, heart, gonadal tissue, kidney, lung, lymph nodes, small bowel, and spleen) were collected for histopathology and molecular analysis.

Sample analysis. All clinical pathology tests of blood samples were conducted by Laboratory Corp. of America, Inc., including complete blood counts and differentials and complete clinical chemistries. Serum concentrations of cytokines including IL-6, IL-10, IL-12, IFN-a, and IFN-g were analyzed using ELISA kits as instructed by the manufacturer (BioSource International, Camarillo, CA, USA). Titers of neutralizing antibodies against different AAV capsids before and after vector administration were determined as described previously [6]. ELISA-based assays were performed to quantify serum levels of rhCG and rhEpo proteins, as previously described [5,7].

Tissue sample fixation, processing, and staining for histopathology evaluation were performed as previously described [25]. For molecular analysis, total blood DNAs were prepared using a Blood & Cell Culture DNA kit as instructed by the manufacturer (Qiagen, Inc., Valencia, CA, USA). Tissue DNAs were extracted as previously described [5]. Copy numbers of the vector genomes in blood and tissue samples were quantified by TaqMan PCR using a primer/probe set that targets the poly(A) portion of the vector (forward primer, 5'-TCTAGTTGCCAGCCATCTGTTGT-3'; reverse primer, 5'-TGGGAGTGGCACCTTCCA-3'; and probe, 6FAM-TCCCCCGTGCCT-TCCTTGACC-TAMRA) under the conditions previously described [25]. The molecular status of vector genomes in blood and tissue DNAs was analyzed by Southern blot as previously described [6] using a 1.21-kb HpaI fragment from the vector genome as the probe.

Acknowledgments

We greatly appreciate the technical support provided by Julio Sanmiguel, Joe McLaughlin, and Brian Murphy. The contributions of the Vector, Bioassay, Immunology, and Cell Morphology Cores and NHP Program of the Gene Therapy Program are appreciated. This work was supported by the NIH (NIDDK P30 DK47757 and NHLBI P01 HL59407), the Cystic Fibrosis Foundation, the Juvenile Diabetes Research Foundation, and GlaxoSmith-Kline Pharmaceuticals. J.M.W. is an inventor on patents licensed to various commercial entities.

RECEIVED FOR PUBLICATION APRIL 4, 2005; REVISED AUGUST 24, 2005; ACCEPTED AUGUST 24, 2005.

References

1. Atchison, R. W., Casto, B. C., and Hammon, W. M. (1965). Adenovirus-associated defective virus particles. Science 149: 754-756.

2. Hoggan, M. D., Blacklow, N. R., and Rowe, W. P. (1966). Studies of small DNA viruses found in various adenovirus preparations: physical, biological, and immunological characteristics. Proc. Natl. Acad. Sci. USA 55: 1467-1474.

3. Mayor, H. D., and Melnick, J. L. (1966). Small deoxyribonucleic acid-containing viruses (picodnavirus group). Nature 210: 331 -332.

4. Rutledge, E. A., Halbert, C. L., and Russell, D. W. (1998). Infectious clones and vectors derived from adeno-associated virus (AAV) serotypes other than AAV type 2. J. Virol. 72: 309-319.

5. Gao, G. P., Alvira, M. R., Wang, L., Calcedo, R., Johnston, J., and Wilson, J. M. (2002). Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc. Natl. Acad. Sci. USA 99: 11854-11859.

6. Gao, G., et al. (2003). Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc. Natl. Acad. Sci. USA 100: 6081 -6086.

7. Gao, G., et al. (2004). Clades of adeno-associated viruses are widely disseminated in human tissues. J. Virol. 78: 6381 -6388.

8. Blacklow, N. R., Hoggan, M. D., and Rowe, W. P. (1968). Serologic evidence for human infection with adenovirus-associated viruses. J. Natl. Cancer Inst. 40: 319-327.

9. Blacklow, N. R., Hoggan, M. D., Kapikian, A. Z., Austin, J. B., and Rowe, W. P. (1968). Epidemiology of adenovirus-associated virus infection in a nursery population. Am. J. Epidemiol. 88: 368 -378.

10. Samulski, R. J., Srivastava, A., Berns, K. I., and Muzyczka, N. (1983). Rescue of adeno-associated virus from recombinant plasmids: gene correction within the terminal repeats of AAV. Cell 33: 135 -143.

11. Kotin, R. M. (1994). Prospects for the use of adeno-associated virus as a vector for human gene therapy. Hum. Gene Ther. 5: 793-801.

12. Rabinowitz, J. E., and Samulski, J. (1998). Adeno-associated virus expression systems for gene transfer. Curr. Opin. Biotechnol. 9: 470-475.

13. Grimm, D., et al. (2003). Preclinical in vivo evaluation of pseudotyped adeno-associated virus vectors for liver gene therapy. Blood 102: 2412-2419.

14. Xiao, W., Chirmule, N., Berta, S. C., McCullough, B., Gao, G., and Wilson, J. M. (1999). Gene therapy vectors based on adeno-associated virus type 1. J. Virol. 73: 3994-4003.

15. Auricchio, A., etal. (2001). Exchange of surface proteins impacts on viral vector cellular

specificity and transduction characteristics: the retina as a model. Hum. Mol. Genet. 10: 3075-30781.

16. Chao, H., Monahan, P. E., Liu, Y., Samulski, R. J., and Walsh, C. E. (2001). Sustained and complete phenotype correction of hemophilia B mice following intramuscular injection of AAV1 serotype vectors. Mol. Ther. 4: 217-222.

17. Davidson, B. L., et al. (2000). Recombinant adeno-associated virus type 2, 4, and 5 vectors: transduction of variant cell types and regions in the mammalian central nervous system. Proc. Natl. Acad. Sci. USA 97: 3428-3432.

18. Zabner, J., et al. (2000). Adeno-associated virus type 5 (AAV5) but not AAV2 binds to the apical surfaces of airway epithelia and facilitates gene transfer. J. Virol. 74: 3852-3858.

19. Mingozzi, F., et al. (2002). Improved hepatic gene transfer by using an adeno-associated virus serotype 5 vector. J. Virol. 76: 10497-10502.

20. Thomas, C. E., Storm, T. A., Huang, Z., and Kay, M. A. (2004). Rapid uncoating of vector genomes is the key to efficient liver transduction with pseudotyped adeno-associated virus vectors. J. Virol. 78: 3110-3122.

21. Lebherz, C., Gao, G., Louboutin, J. P., Millar, J., Rader, D., and Wilson, J. M. (2004). Gene therapy with novel adeno-associated virus vectors substantially diminishes atherosclerosis in a murine model of familial hypercholesterolemia. J. Gene Med. 6: 663-672.

22. Sarkar, R., et al. (2004). Total correction of hemophilia A mice with canine FVIII using an AAV 8 serotype. Blood 103: 1253-1260.

23. Davidoff, A. M., et al. (2005). 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. Mol. Ther. 11: 875-888.

24. Wang, L., Zoppe, M., Hackeng, T. M., Griffin, J. H., Lee, K. F., and Verma, I. M. (1997). A factor IX-deficient mouse model for hemophilia B gene therapy. Proc. Natl. Acad. Sci. USA 94: 11563-11566.

25. Schnell, M. A., et al. (2001). Activation of innate immunity in nonhuman primates following intraportal administration of adenoviral vectors. Mol. Ther. 3(5 Pt 1): 708-722.

26. Wilken, J. A., Matsumoto, K., Laughlin, L. S., Lasley, B. L., and Bedows, E. (2002). Comparison of chorionic gonadotropin expression in human and macaque (Macaca fascicularis) trophoblasts. Am. J. Primatol. 56: 89-97.

27. Lonning, S. M., et al. (2002). Influence of gender-specific hormones on adeno-associated virus vector gene transfer and expression. Mol. Ther. 5: S2.

28. Chenuaud, P., et al. (2004). Autoimmune anemia in macaques following erythropoie-tin gene therapy. Blood 103: 3303-3304.

29. Gao, G., et al. (2004). Erythropoietin gene therapy leads to autoimmune anemia in macaques. Blood 103: 3300-3302.

30. Choi, V. W., Samulski, R. J., and McCarty, D. M. (2005). Effects of adeno-associated virus DNA hairpin structure on recombination. J. Virol. 79: 6801 -6807.

31. Flotte, T. R. (2004). Immune responses to recombinant adeno-associated virus vectors: putting preclinical findings into perspective. Hum. Gene Ther. 15: 716-717.

32. Schnepp, B. C., Clark, K. R., Klemanski, D. L., Pacak, C. A., and Johnson, P. R. (2003). Genetic fate of recombinant adeno-associated virus vector genomes in muscle. J. Virol. 77: 3495-3504.

33. Yan, Z., Zak, R., Zhang, Y., and Engelhardt, J. F. (2005). Inverted terminal repeat sequences are important for intermolecular recombination and circularization of adeno-associated virus genomes. J. Virol. 79: 364-379.