Packaging of an AAV vector encoding human acid α-glucosidase for gene therapy in glycogen storage disease type II with a modified hybrid adenovirus-AAV vector

Packaging of an AAV vector encoding human acid α-glucosidase for gene therapy in glycogen storage disease type II with a modified hybrid adenovirus-AAV vector

ARTICLE doi:10.1016/S1525-0016(03)00022-4 Packaging of an AAV Vector Encoding Human Acid ␣-Glucosidase for Gene Therapy in Glycogen Storage Disease ...

434KB Sizes 0 Downloads 18 Views

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

Packaging of an AAV Vector Encoding Human Acid ␣-Glucosidase for Gene Therapy in Glycogen Storage Disease Type II with a Modified Hybrid Adenovirus-AAV Vector Baodong Sun, Y.-T. Chen, Andrew Bird, Fang Xu, Yang-Xun Hou,* Andrea Amalfitano, and Dwight D. Koeberl† Division of Medical Genetics, Duke University Medical Center, Durham, North Carolina 27710, USA *Present address: Toxicology Research Division, HPFB, Health Canada, Ottawa, Ontario KIA OL2 Canada. †

To whom correspondence and reprint requests should be addressed. Fax: (919) 684-2362. E-mail: [email protected]

We have developed an improved method for packaging adeno-associated virus (AAV) vectors with a replication-defective adenovirus-AAV (Ad-AAV) hybrid virus. The AAV vector encoding human acid ␣-glucosidase (hGAA) was cloned into an E1, polymerase/preterminal protein-deleted adenovirus, such that it is packaged as an Ad vector. Importantly, the Ad-AAV hybrid cannot replicate during AAV vector packaging in 293 cells, because of deletion of polymerase/preterminal protein. The residual Ad-AAV in the AAV vector stock was reduced to <1 infectious particle per 1010 AAV vector particles. These modifications resulted in ⬃30-fold increased packaging of the AAV vector for the hybrid Ad-AAV vector method as compared with standard transfection-only methods. Similarly improved packaging was demonstrated for pseudotyping the AAV vector as AAV6, and for AAV vector packaging with a second Ad-AAV vector encoding canine glucose-6phosphatase. Liver-targeted delivery of either the Ad-AAV hybrid or AAV vector particles in acid ␣-glucosidase-knockout (GAA-KO) mice revealed secretion of hGAA with the Ad-AAV vector, and sustained secretion of hGAA with an AAV vector in hGAA-tolerant GAA-KO mice. Further development of hybrid Ad-AAV vectors could offer distinct advantages for gene therapy in glycogen storage diseases. Key Words: Gene therapy, adeno-associated virus vectors, glycogen storage disease, Pompe disease, acid maltase deficiency, acid ␣-glucosidase, hybrid adenoviral-adeno-associated virus vector

INTRODUCTION Limitations of adeno-associated virus (AAV) vectors include inefficient production methods, packaging size constraints (⬃5 kb), and a high level of immunity to AAV among adults (although AAV infection is not associated with any disease). Initially, AAV vectors were produced by transfection of 293 cells with two plasmids (an AAV vector plasmid and an AAV helper plasmid), and infection with adenovirus [1]. This method provided the essential elements needed for AAV vector production, including AAV terminal repeat (TR) sequences flanking a gene of interest, AAV helper functions consisting of the rep and cap genes, and Ad genes. Improvements to the basic method have included delivery of Ad genes by transfec-

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy 1525-0016/03 $30.00

tion to eliminate contaminating Ad [2,3], construction of first-generation packaging cell lines containing the AAV rep and cap genes [4 –7], and delivery of AAV vector sequences within an Ad-AAV hybrid vector to increase vector production [8,9]. Glycogen storage disease type II (GSD II) is a classic lysosomal storage disorder, characterized by accumulation of glycogen in lysosomes and tissue damage, primarily in muscle and heart [10]. Intriguingly, administration of an Ad vector encoding mouse GAA or hGAA that was targeted to mouse liver reversed the glycogen accumulation in a GAA-KO mouse model [11] for Pompe disease within days, although the effect diminished with time [12–14]. By contrast, liver-targeted or intravenous administration of AAV vectors has reversed the abnormalities in

467

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

FIG. 1. Hybrid Ad-AAV vector containing the CB promoter driving the hGAA cDNA. (A) The hybrid vector, Ad-AAV-CBGAApA, was constructed with the bacterial recombination system with modifications [18,40]. (B) Southern blot of DNase I-resistant hybrid Ad-AAV vector particles (Ad-AAV hybrid), the plasmid containing the AAV vector sequences before bacterial recombination to produce Ad particles (pShuttle-AAV), and the AAV vector plasmid (pAAV-CBGAApA), hybridized with an hGAA cDNA probe. (C) Cesium chloride gradient of hybrid Ad-AAV vector particles. (D) Southern blot of a representative Ad-AAV vector preparation, hybridized with an hGAA probe. Each sample was extracted from 10 ␮l of vector stock. Lanes 11–16 contain pAAV-CBGAApA DNA equivalent to the indicated number of double-stranded (ds) Ad particles. (E) Hybrid Ad-AAV vector packaging method for AAV vector purification.

mouse models for hemophilia B [15], Sly disease [16], and Fabry disease [17], with long-term benefits from expression of the introduced gene primarily in liver. Similar studies with AAV vectors have not been reported to date in the GAA-KO mouse model for GSD II. As a first step toward this goal, we have observed markedly increased AAV vector packaging with a hybrid AdAAV vector, and have used a modified Ad such that no Ad-AAV particles were replicated during AAV packaging. Furthermore, we demonstrated hGAA expression for both the Ad-AAV and AAV vectors encoding hGAA in the GAA-KO mouse model for GSD II.

RESULTS Markedly Enhanced AAV Vector Packaging with a Hybrid Ad-AAV Vector We have cloned an AAV vector sequence into a multiply deleted, replication-deficient adenovirus (Ad), such that it is packaged as an Ad vector (Fig. 1A). The hybrid vector plasmid, Ad-AAVCBGAApA, was constructed with the bacterial recombination system [18]. The plasmid containing the adenovirus genome was modified by deletion of the E1, polymerase, and preterminal protein genes, such that the Ad-AAV vector carried the deletion and was

468

completely replication-deficient in 293 cells [19]. The presence of the AAV vector sequences within the recombinant Ad-AAV genome was demonstrated by Southern blot analysis (Fig. 1B). After two rounds of equilibrium density cesium chloride centrifugation, two Ad-AAV bands were visible; however, no vector DNA could be isolated from the upper, lower density Ad-AAV band in the cesium gradient (Fig. 1C), as detected by Southern blot analysis with an hGAA probe (Fig. 1D) or an Ad5 probe (not shown). Furthermore, no DNA was present for the upper Ad-AAV band by ethidium staining of the agarose gel before Southern blotting. Consistent with the lack of vector DNA from the upper Ad-AAV band, HeLa cells transduced with this fraction had no detectable hGAA activity above the baseline level (not shown). AAV TR sequences were intact in the Ad-AAV vector DNA from the lower band, because restriction sites within the TR sequences were intact as determined by restriction enzyme digestion with AhdI and BssHII (Figs. 1B and 1D). DNase I-resistant vector particles (DRP) of hybrid Ad-AAV were quantitated from the lower band by Southern blot analysis versus a standard curve of plasmid DNA (Fig. 1D). The titer for the initial hybrid Ad-AAV vector stock was 5 ⫻ 109 plaque-forming units (pfu)/ml by pfu assay, and thus the vector stock contained ⬃1012 DRP/ml,

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

doi:10.1016/S1525-0016(03)00022-4

ARTICLE

FIG. 2. AAV vector packaging with an Ad-AAV hybrid vector. (A) Southern blot of a representative AAV vector preparation, hybridized with an hGAA probe. Each sample represented 6 ⫻ 105 293 cells. Lanes 7–11 contained vector plasmid, digested with BglII to release the ds AAV vector sequences, representing the indicated number of single-stranded (ss) AAV vector particles. (B) Packaging AAV-CBGAApA with different Ad helpers and AAV vector packaging plasmids. The data for each condition represent four independent experiments. (C) 293 cells were transfected with an AAV vector plasmid encoding AP, pAAV-CBAPpA, and stained for AP [38]. The transfection control was not transduced with Ad-AAV, whereas the other conditions included transduction with the indicated number of particles of the Ad-AAV vector. The proportion of 293 cells stained for AP (AP⫹) are shown as percentages, and the standard deviation for each condition is shown. A total of three 150-mm tissue culture dishes were analyzed for each condition, and five areas of 1 mm2 in a 1-cm2 grid were quantified for AP⫹ and unstained cells in each dish. (D) Southern blot analysis of AAV vector DNA extracted from AAV vector stocks to quantitate contaminating Ad-AAV genomes. Lanes as follows: (1) untreated 293 cells, (2) transfection of pAAV-CBGAApA ⫹ pMTRep ⫹ pCMVCap ⫹ pLNX-CORF6, (3) hybrid Ad-AAV transduction plus transfection of pMTRep ⫹ pCMVCap, (4) wild-type Ad5 infection plus transfection of pACG2 ⫹ pAAV ⫺ CBGAApA, (5) hybrid Ad-AAV transduction plus transfection of pACG2, (6) no sample, and (7–11) Ad5-containing plasmid (pAdEasy [18]) representing the indicated number of ds Ad particles. Each sample represented 6 ⫻ 105 293 cells.

consistent with a particle/infectivity ratio of 1:200. Thus, we confirmed that the full-length hybrid Ad-AAV vector DNA contained both the AAV vector sequences and the Ad helper genes required to package an AAV vector (Fig. 1E). Analysis of Different Packaging Conditions for an AAV Vector in 293 Cells The AAV vector packaging efficiency increased ⬃30-fold for the hybrid Ad-AAV method (as shown in Fig. 1E), because the yield for this particular vector increased from 1300 DRP/cell with a transfection-only method [20] to 38,000 DRP/cell (Fig. 2A). Southern blot analysis of AAV vector DNA demonstrated an identical signal for the vector DNA fragment packaged with either the hybrid AdAAV or transfection-only method (Fig. 2A).

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

The relative efficiency of AAV vector packaging with four different helper-gene combinations in 293 cells was evaluated to elucidate variables that may influence the efficiency of the hybrid Ad-AAV vector-based production method (Fig. 2B). High levels of AAV vector packaging was observed for transduction with the hybrid Ad-AAV method (Fig. 2B, Hybrid), which again showed a 33-fold increase in vector particles per cell compared with the transfection-only method [20]. Transfection with pACG2 [21] (an AAV packaging plasmid) and the pAAV-CBGAApA vector plasmid, accompanied by infection with Ad5 (Fig. 2B, Traditional) or by transduction with AdAAVCBGAApA (Fig 2B, Hybrid Ad-AAV ⫹ pACG2), demonstrated an intermediate efficiency of AAV vector packaging. Northern blot analysis of E4orf6 and E1A

469

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

TABLE 1: Small-scale comparison of packaging of two AAV vectors by the hybrid Ad-AAV method or a transfection-only method (DRP/cell)a AAV-CBGAApA b

AAV-cG6PpA c

Transfection-only

Ad-AAV Hybrid

AAV2

580 ⫾ 200

27,000 ⫾ 14,000

AAV6

4700 ⫾ 2000

b

Ad-AAV Hybridc

Transfection-only

12,000 ⫾ 5300

1400 ⫾ 600

23,000 ⫾ 10,000

12,000 ⫾ 7200

29,000 ⫾ 10,000

a

Transfections done in triplicate; average and standard deviation shown. DRP, DNase I-resistant vector particles. b The Ad helper was pHelper [23], and the AAV packaging plasmid was pAAV-RC [23]. For pseudotyping as AAV6, the Ad helper was pHelper [23], and the AAV packaging plasmids were pMTRep [42] and pCMVCap6 [27]. c Ad helper was the hybrid Ad-AAV, and AAV packaging plasmids were pMTRep (for AAV2 and AAV6), pCMVCap [42] (for AAV2), and pCMVCap6 [27] (for AAV6).

transcripts showed equivalent levels of these RNAs under the four conditions analyzed (not shown), consistent with a critical function for another Ad gene in the packaging of AAV-CBGAApA. The Ad gene in question was provided by the modified hybrid Ad-AAV or wild-type Ad5, but not by transfection with pLNCorf6 [22] during the all-transfection method. The effects of different Ad helpers (plasmid or hybrid Ad-AAV) and AAV packaging plasmids were directly compared in a series of small-scale preparations of AAV-CBGAApA and AAV-cG6PpA (Table 1). The plasmid pHelper contains Ad genes including the E2A, E4, and VA genes [23]. Along with pAAV-RC, pHelper comprises an all-transfection method for AAV vector packaging in 293 cells that features a full complement of Ad genes that are important to AAV packaging [23]. In parallel transfections, the combination of pHelper and pAAV-RC was ⬃46-fold less efficient than the hybrid Ad-AAV method for packaging AAV-CBGAApA (Table 1). Similarly, 16-fold lower production was noticed for packaging of a second AAV vector, AAV-cG6PpA, by the all-transfection method with pHelper and pAAV-RC as opposed to the hybrid Ad-AAV method, although AAVcG6PpA was packaged more efficiently by the all-transfection method than was AAV-CBGAApA (Table 1). Intriguingly, pseudotyping of AAV-CBGAApA or AAV-cG6Pase as AAV6 with the AAV packaging plasmids pMTRep and pCMVCap6 was almost equivalent to packaging the vector as AAV2, employing either pHelper or a hybrid AdAAV as the Ad helper (Table 1). To evaluate further the effect of Ad-AAV transduction on transfection efficiency, and potentially upon AAV vector packaging efficiency, a series of transfection control experiments were done. An AAV vector plasmid encoding AP, pAAV-CBAPpA, was transfected in 293 cells with and without transduction by the hybrid Ad-AAV vector. The transfection efficiency ranged from ⬃60% to 80%, within experimental variation between specific conditions, and it did not reflect a positive effect of Ad-AAV transduction upon the transfection efficiency of the AAV vector plasmid containing the AP gene (Fig. 2C). Therefore, although the majority of 293 cells were transfected under each condition, a difference in transfection efficiency was not

470

the underlying cause for increased AAV vector packaging after Ad-AAV transduction. Contaminating Ad-AAV particles in the AAV vectorcontaining cell lysates were quantitated by Southern blot analysis. Southern blot quantitation revealed high-molecular-weight Ad-AAV DNA at 0.5 DRP/cell for AAV-CBGAApA packaged with the Ad-AAV hybrid (Fig. 2D, lanes 3 and 5), after probing with Ad5 genome sequences (pAdEasy from [18]). For comparison, AAV-CBGAApA packaged with a wild-type Ad5 helper contained ⬃6000 contaminating Ad DNase-resistant particles/cell during AAV vector packaging (Fig. 2D, lane 4). Thus, the level of contaminating Ad was reduced markedly by the use of a second-generation hybrid Ad-AAV vector that did not replicate in 293 cells to provide Ad helper functions. We compared the yield for large-scale preparations of AAV-CBGAApA with the hybrid Ad-AAV packaging method to previous results for a transfection-only method [20] (Table 2). The number of particles per cell for AAVCBGAApA was elevated ⬃14-fold for the hybrid Ad-AAV packaging method (Ad-AAV-CBGAApA ⫹ pMTRep ⫹ pCMVCap) compared with the transfection-only method (pLNX-CORF6 ⫹ pMTRep ⫹ pCMVCap ⫹ pAAV-CB-

TABLE 2: Large-scale comparison of packaging of AAVCBGAApA with different Ad helpers and AAV packaging plasmids (DRP/cell)a Ad helper AAV packaging plasmid(s)

PLNX-CORF6b (with pAAVCBGAApA)

pMTRepc ⫹ pCMVCapc

1800 ⫾ 800

25,000 ⫾ 3700

Not done

11,000 ⫾ 1800

Not done

40,000 ⫾ 20,000

pRepCap6d pMTRep ⫹ pCMVCap6 c

e

Ad-AAVCBGAApA

a

Average and standard deviation for three preparations of 40 ⫻ 150 mm tissue culture dishes. b Ref.[22]. c Ref.[42]. d Ref.[43]. e Ref.[27].

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

FIG. 3. Analysis of large-scale AAV vector packaging with an Ad-AAV hybrid vector. Southern blot hybridized with an hGAA probe of samples from steps during AAV-CBGAApA purification, quantified versus a standard curve of vector plasmid DNA. The samples represent vector DNA extracted from 25 ␮l of sample. Lanes represent the following samples: (1) crude cell lysate; (2) 40% iodixanol fraction; (3) heparin-agarose (HA) column flowthrough; (4) HA column wash; (5) HA column eluate fraction (ef) 1; (6) HA column ef 2; (7) HA column ef 3; (8) HA column ef 4; (9) HA column ef 2 after dialysis; (10) HA column ef 2 plus 2.5 ⫻ 1010 particles AAV vector plasmid; (11) HA column ef 2 plus 2.5 ⫻ 1010 particles AAV vector plasmid, no DNase I added; (12) blank; (13)–(18) vector plasmid, digested with BglII to release the ds AAV vector sequences, representing the indicated number of ss AAV vector particles.

GAApA). The yield of AAV-CBGAApA for the transfectiononly method was 3-fold higher than expected on the basis of the small-scale vector preparations described earlier (Fig. 2B). These variations could reflect higher efficiency transfections with the transfection-only method during large-scale vector preparations. The packaging of AAVCBGAApA as AAV6 was improved 4-fold by driving rep and cap6 genes with heterologous promoters in pMTRep and pCMVCap6, as opposed to the endogenous promoters in the pRepCap6 plasmid. Somewhat remarkably, pseudotyping the AAV vector as AAV6 was as efficient as packaging it as AAV2. AAV-CBGAApA packaged with the hybrid Ad-AAV method (Fig. 1E) was purified by the heparin-agarose column method [24] and quantified by Southern blot analysis (Fig. 3). AAV-CBGAApA particles were recovered efficiently from the 40% iodixanol fraction (Fig. 3, lane 2) by heparin-agarose column purification (Fig. 3, lane 6). The DNase I resistance of AAV vector particles was confirmed by the elimination of the signal for 2.5 ⫻ 1010 added

vector plasmid particles when DNase I was present (Fig. 3, lane 10 versus lane 11). Residual Ad-AAV in the AAV-CBGAApA vector preparation was quantified by Southern blot analysis and by pfu assay. For Southern blot analysis, the signal for contaminating Ad-AAV in the vector preparation was reduced markedly to less than the limit of detection by column purification of the AAV vector (⬍0.5 Ad-AAV particle/ cell), and overexposure of the autoradiograph hybridized with an Ad5 probe did not reveal a signal for Ad-AAV genomes in the purified AAV vector stock (not shown). For pfu assay on C7 cells, residual Ad-AAV in the AAV vector stock after column purification was absent, which revealed ⬍1 pfu Ad-AAV/1010 DRP of AAV-CBGAApA. For in vitro analyses of hGAA expression with the transgene in the hybrid Ad-AAV and AAV vectors, cultured cells were transduced with AAV-CBGAApA and AAV-AdCBGAApA. Enzyme analysis of hGAA expression in transduced HeLa cells showed slightly higher levels for the Ad-AAV vector than for the AAV vector (Table 3), when cells were transduced with 1000 DRP/HeLa cell of each vector in parallel. Higher hGAA production was observed for the AAV vector after transduction with 50,000 DRP/ HeLa cell. AAV-CBGAApA-transduced GSD II patient fibroblasts had clearly detectable hGAA compared with untransduced patient cells, demonstrating hGAA replacement by western blot analysis (Fig. 4A). Hence, the transgene delivered hGAA detected by both enzyme analysis and western blot analysis of transduced human cells. hGAA Expression with Ad-AAV and AAV Vectors in GAA-KO Mice The hybrid Ad-AAV vector encoding hGAA was administered in vivo to detect hGAA expression with this vector. The hybrid Ad-AAV vector encoding hGAA was administered intravenously to GAA-KO mice, equivalent to 1 ⫻ 108 pfu or 2 ⫻ 1010 DRP/GAA-KO mouse. Secretion of hGAA was demonstrated in plasma by western blot analysis on day 3 after vector administration (Fig. 4B); however, no hGAA was detected in plasma by western blotting on day 7 after vector administration (Fig. 4B). The hGAA levels were sufficiently elevated in liver to generate detectable plasma levels of the ⬃110-kDa precursor hGAA, as

TABLE 3: Enzyme analysis of transduced HeLa cells (nmol/hour/mg protein)a 102 Vector particles/cell

34.5 ⫾ 17.7

N.A.b

N.A.

N.A.

N.A.

Ad-AAV-CBGAApA

N.A.

43.4 ⫾ 2.5

189.3 ⫾ 27.

N.D.c

N.D.

AAV-CBGAApA

N.A.

N.D.

106.3 ⫾ 25.5

394.2 ⫾ 79.4

No vector

103 Vector particles/cell

74.4 ⫾ 10.2

104 Vector particles/cell

5 ⫻ 104 Vector particles/cell

No vector

Vector

a

Two 150-mm tissue culture dishes of HeLa cells (1 ⫻ 106 cells plated) were analyzed in two independent experiments for each condition. The results shown reflect the average and standard deviation. b Not applicable. c Not done.

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

471

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

FIG. 4. Human GAA delivery with an AAV or an Ad-AAV hybrid vector in GAA-KO mice. (A) Western blotting analysis of HeLa and GSD II patient fibroblasts transduced with AAV-CBGAApA (1000 DRP/cell). Recombinant hGAA is shown for comparison, and the ⬃110 kDa, ⬃76 kDa, and ⬃67 kDa forms were increased in transduced HeLa cells compared with untransduced HeLa cells. (B) Western blot analysis of plasma from GAA-KO mice at 3 days and 7 days after intravenous administration of the hybrid Ad-AAV vector encoding hGAA (2 ⫻ 1010 vector particles/mouse), and from untreated, GAA-KO mice (controls). Each lane represents an individual mouse for the two groups, and each group was analyzed on day 3 and day 7 after Ad-AAV vector administration. (C) Western blot analysis of liver after intravenous injection of an Ad-AAV or AAV vector. Untreated, affected GAA-KO mouse liver is shown for comparison (No vector, n ⫽ 2). (D) PCR detection of vector DNA in mouse liver DNA. Liver DNA was extracted from the liver samples analyzed by western blot in panel (C). The control sample was liver DNA from a GAA-KO mouse that did not receive either vector.

have been reported for an Ad vector encoding hGAA [12]. The number of hybrid Ad-AAV vector particles administered was within the range given in previous studies (from 2.5 ⫻ 108 to ⬃1 ⫻ 1011 vector particles by optical density) [25]. AAV vector stocks were administered intravenously to GAA-KO mice for in vivo analysis of the AAV vector encoding hGAA. However, no hGAA was detectable in plasma samples by western blotting at 1, 2, or 6 weeks after administration of the AAV vector (not shown). Western blot detection of hGAA production in liver with the Ad-AAV (2 ⫻ 1010 DRP) and the AAV vector (4 ⫻ 1010 DRP) revealed similar, low levels at 6 weeks after vector administration (Fig. 4C). Untreated, affected GAA-KO mouse liver was analyzed in parallel, and no signal corresponding to hGAA was detected by western blot analysis (Fig. 4C, No vector). Enzyme analysis of the liver after administration of Ad-AAV or AAV vectors (2 ⫻ 1010 DRP or 4 ⫻ 1010 DRP) showed levels near background GAA activity at this time point (Table 4). In an attempt to

472

increase GAA production with AAV-CBGAApA, liver-targeted administration of a higher number of AAV vector particles (1 ⫻ 1012 DRP) produced markedly higher levels of human GAA in liver at 6 weeks after vector administration than for a lower number of Ad-AAV or AAV vector particles at that time point (Fig. 4C). Enzyme analysis of liver hGAA activity 6 weeks after administration of the higher number of AAV vector particles (1 ⫻ 1012 DRP) showed the GAA concentration approached that detected in the livers of wild-type mice (⬃100 nmol /hour/mg protein [13]). Although hGAA expression in liver was not analyzed during the first several days following AAV-Ad vector administration, hGAA levels in liver were probably much higher then than at 2 weeks after vector administration as seen after administration of modified Ad vectors encoding hGAA [13]. PCR detection of the hGAA cDNA delivered by the Ad-AAV or AAV vector in liver DNA revealed vector DNA at 2 and 6 weeks after vector administration, whereas no signal was present for control GAA-KO mouse liver DNA

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

ARTICLE

doi:10.1016/S1525-0016(03)00022-4

TABLE 4: Enzyme analysis of GAA-KO mouse liver (nmol/hour/mg protein) Time after intravenous vector administrationa

Ad-AAVb, 2 ⫻ 1010 (n ⫽ 3)

AAVc, 4 ⫻ 1010 (n ⫽ 3)

AAVc, 1 ⫻ 1012 (n ⫽ 2)

2 weeks

5.02 ⫾ 2.30

2.09 ⫾ 0.35

Not done

6 weeks

1.55 ⫾ 0.48

3.07 ⫾ 0.77

71.28 ⫾ 2.60

a

hGAA in age-matched, 3-month-old GAA-KO mouse liver ⫽ 1.35 ⫾ 0.15 [units?] (n ⫽ 2). Ad-AAV-CBGAApA, hybrid AAV-Ad vector (DRP). AAV-CBGAApA, AAV vector (DRP).

b c

(Fig. 4D). The copy number for the AAV vector at these time points was ⬃1 copy/liver cell, as shown by a standard curve of vector plasmid DNA in GAA-KO mouse liver DNA (Fig. 4D, lanes 8 –15 and lanes 17–20). AAV-Ad vector DNA was present at both 2 and 6 weeks after vector administration, and the copy number ranged from 0.01 to 0.1 copies/liver cell. Hence, we demonstrated delivery and persistence for vector DNA in the liver for both Ad-AAV and AAV vectors. hGAA Secretion After Portal Vein Injection with AAV Vectors in GAA-KO/SCID Mice To maximize the transduction of liver with the AAV vector and increase the likelihood of secretion of hGAA, we delivered the AAV vector by portal vein injection in GAAKO/SCID mice. Anti-hGAA antibodies had earlier abbreviated the secretion of hGAA with Ad vectors in the liver of GAA-KO mice, and anti-hGAA antibodies were not expected after vector administration in GAA-KO/SCID mice [13]. Western blot analysis of plasma demonstrated hGAA at 2 and 4 weeks after portal vein administration of AAV-CBGAApA packaged either as AAV2 or as AAV6 (Fig. 5A), and was still detected 3 months after injection of the AAV vector (not shown). Liver production of hGAA was elevated ⬃10-fold compared with wild-type concentrations (1110 ⫾ 120 nmol /mg/hour) for a GAA-KO/SCID mouse at 3 months after AAV2 vector administration. For the vector pseudotyped as AAV6, hGAA was elevated to approximately twice the concentration seen in normal mice. Untreated GAA-KO/SCID control mice had low GAA activity in liver (1.4 ⫾ 0.3 nmol/mg/hour). Distal uptake of hGAA at the 3-month time point was demonstrated in mice by GAA analysis of spleen, heart, diaphragm, and the gastrocnemius, where GAA was clearly above the background activity for untreated, GAA-KO/ SCID mice (Fig. 5B). The potential benefit from hGAA delivery to the heart, a primary site of pathology in infantile GSD II (also known as Pompe disease) was shown by reduced glycogen staining for the mouse that received the AAV2 vector, as compared with an untreated control (Fig. 5C). Furthermore, glycogen content in heart was reduced to 0.21 mmol glucose/g protein (range 0.14 – 0.28), compared with 1.2 ⫾ 0.25 mmol glucose/g protein for untreated controls.

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

DISCUSSION The hybrid Ad-AAV vector developed here has shown promise for AAV vector packaging with a nonreplicating Ad-AAV vector, and column purification eliminated residual Ad-AAV from the AAV vector stocks below the limit of detection (⬍1 pfu Ad-AAV in 1010 AAV particles). The improved packaging of an AAV vector encoding hGAA with the hybrid Ad-AAV allowed liver-targeted delivery of a high number of AAV vector particles. After portal vein injection of the AAV vector, hGAA secretion into plasma with accompanying receptor-mediated uptake by other tissues was observed. Earlier work had shown that the secretion of GAA with liver-targeted adenoviral vectors corrected glycogen storage in skeletal muscle and the heart, prompting hope for curative therapy in GSD II [12–14,25]. Recently, titration of an (E1-, pol-)Ad vector containing the cytomegalovirus (CMV) immediate early promoter/enhancer to drive hGAA expression demonstrated low-level hGAA secretion into plasma in GAA-KO mice at 3 days after administration of only 6.3 ⫻ 108 vector particles; however, liver hGAA activities required to allow determination of the lower threshold for hGAA expression that will drive secretion into plasma were not presented [25]. However, Raben et al. have demonstrated with an inducible-GAA transgene in GAA-KO mice that liver secretion of GAA required overexpression of GAA in mouse liver at 25-fold higher than wild-type levels [26]. We found that sustained hGAA expression at levels slightly above wild type drove secretion with an AAV vector, and that glycogen storage in the heart was reduced in one mouse in association with hGAA secretion. Therefore, it remains to be evaluated whether low, persistent hGAA secretion with an AAV vector might generate sufficient GAA transfer from liver to skeletal muscle and heart to achieve a therapeutic effect reproducibly in GSD II. The improved packaging of an AAV vector encoding hGAA with the hybrid Ad-AAV method compared to a transfection-only method [20] could reflect the role of adenoviral helper functions other than E1 or E4orf6. E2a and VA RNA have earlier been implicated to have important Ad functions in AAV vector packaging [1], and E2a may have a critical activity in optimizing the efficiency of AAV vector packaging [23]. The lack of replication of the

473

ARTICLE

FIG. 5. Human GAA secretion and uptake after portal vein injection of an AAV vector in GAA-KO mice. (A) Western blot analysis of plasma from GAA-KO/ SCID mice at the indicated times after portal vein injection of the indicated AAV vector encoding hGAA, and from untreated, GAA-KO/SCID mice (controls). Each lane represents an individual mouse. (B) GAA analysis for tissues after portal vein injection of an AAV vector. GAA-KO/SCID mice received the vector packaged as AAV2 (n ⫽ 1) or AAV6 (n ⫽ 1). Controls were age-matched, untreated GAA-KO/SCID mice (n ⫽ 2). The GAA level was analyzed twice, independently, and the average and range are shown. (C) Periodic acid Schiff (PAS) of the heart for a GAA-KO/SCID mouse that received an AAV vector and for an untreated GAA-KO/SCID mouse. Magnification ⫻100.

adenovirus vector in 293 cells might also contribute to more efficient AAV vector packaging for this hybrid (E1-, polymerase-, preterminal protein-) Ad-AAV vector, similar to all-transfection methods where Ad replication is absent. When compared to an all-transfection method that employed an Ad helper plasmid containing the E2a, E4, and VA genes, and an AAV vector packaging plasmid that contained the intact AAV2 rep and cap genes [23], the hybrid Ad-AAV method was more efficient. We evaluated

474

doi:10.1016/S1525-0016(03)00022-4

the role of transfection efficiency in determining the efficiency of AAV vector packaging for the hybrid Ad-AAV system, and transfection was neither increased by Ad-AAV transduction nor too low to allow packaging by the alltransfection methods to which the hybrid system was compared. The purified AAV vector stock was analyzed for residual contaminating Ad-AAV particles by pfu assay, and ⬍1 pfu Ad-AAV is present in 1010 AAV vector particles. Given the particle/infectivity ratio of 200:1, ⬍1 Ad-AAV particle was present in 108 AAV vector particles. Before column purification, ⬃1 Ad-AAV particle/50,000 AAV vector particles was present. Thus, the level of residual contaminating Ad-AAV was very low after AAV vector purification, and this represented a ⬎200-fold reduction in residual AdAAV after column purification. As expected, we showed that an Ad-AAV vector could be used with different AAV packaging plasmids to package an AAV2 vector as an alternative AAV serotype. Somewhat surprisingly, pseudotyping two different AAV vectors as AAV6 with hybrid Ad-AAV vectors produced similar numbers of vector particles per cell as packaging the same vectors as AAV2, whereas such pseudotyping had earlier been found to be ⬃3 logs less efficient than packaging an AAV vector as AAV2 [27]. Hybrid Ad-AAV (alternatively, Ad/AAV) vectors have earlier been applied in packaging cell systems for improved AAV vector packaging [9,28]. Although the number of AAV vector particles packaged by those systems was higher than we found for the method described here (䡠 106 vector particles/cell), these packaging cell lines required co-infection with a replicating Ad, and consequently, high levels of contaminating Ad were packaged with the AAV vector. We constructed a replication-deficient AdAAV vector that produced only small residual amounts of Ad during vector packaging (⬍0.5 Ad particle/cell), and this contaminating Ad was readily eliminated by column purification to ⬍1 infectious Ad-AAV pfu in 1010 AAV vector particles. A replication-deficient Ad was earlier described to reduce Ad contamination during AAV vector packaging, although the level of residual contaminating (preterminal protein-, E3-) Ad was not directly quantitated in AAV vector stocks packaged by that method [29]. Other applications of Ad-AAV vectors include delivery of an integrating AAV vectors by introduction of Rep78 [30], or by the isolation of Ad-AAV vectors deleted for all Ad genes during construction of a first-generation AAV-Ad vector [31,32]. We did not observe the deleted Ad-AAV vector band after cesium chloride gradient centrifugation as earlier reported [31]. We observed an upper, lower density band upon purification of the replicationdeficient hybrid Ad-AAV vector. We ascertained that this upper band fraction did not contain AAV or Ad vector DNA by Southern blot analysis, nor did this fraction transduce cultured cells. We surmise that the production of deleted Ad-AAV genomes as described by Lieber et al.

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

doi:10.1016/S1525-0016(03)00022-4

[31] could require recombination of the Ad-AAV vector after co-transfection of 293 cells, whereas we employed the bacterial recombination system to generate hybrid Ad-AAV genomes [18] and did not observe this phenomenon. The current results suggest that hybrid (E1-, polymerase-, preterminal protein-) Ad-AAV vectors could be used for high-efficiency AAV vector packaging, especially if a completely Ad-based method that delivered AAV rep and cap genes with AAV particles were employed [33]. Importantly, an all-adenovirus method would allow scaleable AAV packaging. The AAV and Ad-AAV vectors could prove valuable for gene therapy in animal models for GSD, either together or separately, such as in the canine model for GSD Ia in which both early and prolonged transgene expression may be required for efficacious gene therapy [34].

MATERIALS

AND

METHODS

Cell culture. 293 cells, C-7 cells [35], and GSD II patient fibroblasts were maintained in Dulbecco’s modified Eagle medium supplemented with 10% FBS, 100 U penicillin/ml, and 100 ␮g streptomycin/ml at 37°C in a 5% CO2-air atmosphere. C-7 cells were grown in the presence of hygromycin, 50 ␮g/ml. HeLa cells were maintained in minimum essential medium Eagle supplemented with 10% FBS, 100 U penicillin/ml, and 100 ␮g streptomycin/ml at 37°C in a 5% CO2-air atmosphere. Transduction of cultured cells. HeLa cells and GSD II fibroblasts were plated at 1 ⫻ 106 cells per 150-mm tissue culture dish. Cells were transduced the next day by adding a volume of the respective vector stock containing 1000, 10,000, or 50,000 DNase I-resistant vector particles (as defined by Southern blot analysis) per cell. Cells were harvested 5 days later for hGAA measurement and western blotting analysis. Construction of an AAV vector plasmid encoding hGAA, human placental alkaline phosphatase (AP), and canine glucose-6-phosphatase. The hGAA cDNA was subcloned with the CMV promoter from pcDNA3-hGAA [36] into an AAV vector plasmid, as an NruI-EcoRV fragment, upstream of the human growth hormone intron 4 and polyadenylation sequence [37]. The resulting transcriptional unit was flanked by the AAV2 TR sequences in pAAV-ChGAAGH. A 530-bp deletion spanning the human growth hormone intron 4 was generated by EcoRV and partial PvuII digestion and then blunting of ends with T4 DNA polymerase and ligation with T4 DNA ligase to generate pAAV-ChGAAG(⫺). The hybrid CMV enhancer/chicken ␤-actin (CB) promoter was amplified by PCR from pTriEx1 (Novagen, Madison, WI) with primers that introduced unique upstream XbaI and downstream KpnI restriction sites, and the CB promoter was subcloned as a KpnI-XbaI fragment to replace the CMV promoter in pAAV-ChGAAG(⫺), to generate pAAV-CBhGAAG(⫺). Next, to reduce the packaging size further, the plasmid pAAV-CBhGAAG(⫺) was linearized at a unique AflII site in the 3⬘-untranslated sequence of the hGAA cDNA and partially digested with NspI to introduce a 411-bp deletion in the 3⬘-untranslated sequence of the hGAA cDNA, and then blunting of ends with T4 DNA polymerase and ligation with T4 DNA ligase to generate pAAV-CBGAApA. Finally, the vector sequences from pAAV-CBGAApA were isolated as a 4.4-kbp fragment from a partial BglII digest, and ligated with the calf intestinal alkaline phosphatase-dephosphorylated BglII site of pShuttle [18]. The AP cDNA was cloned by PCR of the AP cDNA from pAAV-LAPSN [38] with primers containing an EcoRV site in the upstream primer and an EcoRI site in the downstream primer. Purification of the AP cDNA-containing fragment, followed by digestion with EcoRV and EcoRI and ligation with an EcoRI-EcoRV fragment containing the CB promoter and human

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

ARTICLE

growth hormone polyadenylation sequences described earlier and flanked by the AAV2 TR sequences, generated pAAV-CBAPpA. The promoter sequence from the canine glucose-6-phosphatase cG6Pase gene was cloned by PCR of the 5⬘ sequences of the cG6Pase gene (GenBank accession no. U91844) from a canine genomic library [23], and then isolation of an XbaI-KpnI fragment from position ⫺1372 to 11, and ligation with the KpnI-XbaI fragment from pAAV-CcG6PGH [39]. The resulting AAV vector plasmid, pAAV-cG6PpA, contained the cG6Pase promoter to drive a cG6Pase cDNA, followed by a human growth hormone polyadenylation signal, and flanked by the AAV2 TR sequences. Construction of a hybrid [E1-, polymerase-, preterminal protein-] AAV-Ad vector encoding hGAA. Kanamycin-resistant shuttle plasmids were constructed to contain the CB promoter ⫹ hGAA cDNA ⫹ poly(A) transgene cassette flanked by the AAV2 TR sequences within the Ad E1 region. The shuttle plasmid [18] was digested with PmeI, and electroporated into the BJ5183 recombinogenic strain of Escherichia coli with the pAd(E1-, polymerase-, preterminal protein-) plasmid [40]. Recombinant kanamycin-resistant clones were screened by restriction enzyme digestion (BstXI) to confirm successful generation of the full-length recombinant Ad vector genomes. These clones were digested with PacI and transfected as described earlier into the E1, and E2b expressing cell line, C-7 [19]. The vectors were amplified and confirmed to have the correct construction by restriction enzyme mapping of vector genomes, and subsequent functional assays in vitro and in vivo. Once isolated, the respective Ad vectors are serially propagated in increasing numbers of C-7 cells [19]. Infected cell pellets were harvested 48 hours after infection by low-speed centrifugation, resuspended in 10 mM Tris-HCl, pH 8.0, vector released from the cells by repeated freeze-thawing (three times) of the lysate and by ultrasonification, and the vector-containing supernatant subjected to two rounds of equilibrium density CsCl centrifugation [19]. Two virus bands were visible. The virus bands were then removed, dialyzed extensively against 10 mM Tris-HCl, pH 8.0 (or PBS), sucrose added to 1%, and aliquots stored at ⫺80 °C. The number of vector particles was quantified based on the OD260nm of vector contained in dialysis buffer with SDS disruption [13], and by DNase I digestion, DNA extraction, and Southern blot analysis. Vector titers were determined by pfu assay of serial dilutions of the vector preparations on C-7 cells [19]. Hybrid Ad-AAV vector DNA analysis consisted of vector DNA isolation and restriction enzyme digestion and subsequent Southern blotting to verify the presence of intact AAV vector sequences, including restriction enzymes that demonstrate the presence of AAV terminal repeat sequences flanking the transgene (AhdI and BssHII). A second Ad-AAV vector, Ad-AAV-cG6PpA, was constructed and purified by isolating a BglII fragment containing the AAV vector sequences from pAAV-cG6PpA, and proceeding as described earlier for the construction of Ad-AAVCBGAApA. Preparation of AAV vectors. AAV vector stocks were prepared as described with modifications [20]. Briefly, 293 cells were transduced with the hybrid Ad-AAV vector (2000 DNase I-resistant vector particles/cell as quantitated by Southern blot analysis) containing the AAV vector sequences 15–30 minutes before transfection with a AAV packaging plasmids containing the AAV2 rep (pMTRep) and cap (pCMVCap) genes driven by heterologous promoters, which typically generate no detectable replication-competent AAV (rcAAV) [20]. For the hybrid Ad-AAV method of packaging AAVCBGAApA as AAV6, pCMVCap6 [27] was substituted for pCMVCap. For the initial transfection-only method (modified from [20]), pLNXCORF6 [22] provided E4orf6 gene, which is an essential Ad helper function for AAV packaging, and no Ad or Ad-AAV vector was required. For the second transfection-only method, the plasmid pHelper (Stratagene, La Jolla, CA) contains Ad genes including the E2A, E4, and VA genes [23], and 293 cells were co-transfected with pHelper, the AAV packaging plasmid, pAAV-RC (Stratagene, La Jolla, CA), and pAAV-CBGAApA. Cell lysate was harvested 48 hours after infection and freeze-thawed three times, isolated by iodixanol step gradient centrifugation before heparin affinity column purification [24], and aliquots were stored at ⫺80 °C. For purification of vectors pseudotyped as AAV6, heparin affinity column purification was modified as described [27]. The number of vector DNA-

475

ARTICLE

containing-particles was determined by DNase I digestion, DNA extraction, and Southern blot analysis. Contaminating replication-competent AAV particles were detected by a sensitive PCR assay using primers spanning the junction between the rep and cap genes. The amount of replication-competent AAV was ⬍1 particle in 105 AAV vector particles. Contaminating Ad-AAV vector was quantified by pfu assay of serial dilutions of the AAV vector preparations on C-7 cells. All viral vector stocks were handled according to Biohazard Safety Level 2 guidelines published by the National zInstitutes of Health (NIH). In vivo administration of hybrid Ad-AAV and AAV vector stocks. The vector was administered intravenously (through the retroorbital sinus) into 6-week-old GAA-KO mice [11]. Alternatively, 2 ⫻ 1010 DNase I-resistant Ad-AAV, 4 ⫻ 1010 AAV, or 1 ⫻ 1012 AAV vector particles were injected per animal. At the respective time points after injection, plasma or tissue samples were obtained and processed as described later. All animal procedures were done in accordance with Duke University Institutional Animal Care and Use Committee-approved guidelines. Portal vein injections were administered in 3-month-old GAA-KO/ SCID mice (described by F. Xu et al., manuscript in preparation). Determination of hGAA activity. Liver hGAA activity was measured after removal of the liver from control or treated mice, flash-freezing on dry ice, homogenization and sonication in distilled water, and pelleting of insoluble membranes and proteins by centrifugation. The protein concentrations of the clarified suspensions were quantified by the Bradford assay. hGAA activity in the liver was determined as described [12]. Cellular hGAA was measured in transduced and control HeLa cells after scraping, washing with PBS, suspension in and sonication in distilled water, and pelleting of insoluble membranes and proteins by centrifugation. The protein concentrations of the clarified suspensions were quantified by the Bradford assay, and hGAA activity was determined as described [12]. Western blotting analysis of hGAA. For direct detection of hGAA in liver, liver homogenate samples (100 ␮g of protein) were electrophoresed overnight in a 6% polyacrylamide gel to separate proteins, and transferred to a nylon membrane. The blots were blocked with 5% nonfat milk solution, incubated with primary and secondary antibodies, and visualized with the enhanced chemiluminescence (ECL) detection system (Amersham Pharmacia, Piscataway, NJ) [13]. Semiquantitation of hGAA vector DNA by PCR. Genomic DNA was extracted from GAA-KO mouse livers, and PCR was carried out in a 50-␮l reaction containing 500 ng of liver DNA from each mouse, 2.5 units of Taq DNA polymerase with 1 ⫻ PCR buffer (Qiagen, Valencia, CA), and 150 ng each of the sense and antisense primers. Gene-specific primers for hGAA (sense 5⬘-AGTGCCCACACAGTGCGACGT-3⬘, nucleotide 672– 692, and antisense 5⬘-CCTCGTAGCGCCTGTTAGCTG-3⬘, nucleotide 998 –1018, GenBank accession no. NM 000152), and for mouse ␤-actin (sense 5⬘AGAGGGAAATCGTGCGTGAC-3⬘ and antisense 5⬘-CAATAGTGATGACCTGGCCGT-3⬘ [41] were used for each reaction. Samples were denatured at 94 °C for 3 minutes, then subjected to 30 cycles (25 cycles for ␤-actin, internal control) of 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 45 s. Plasmid DNA corresponding to 0.01 copy to 10 copies of human GAA gene (in 500 ng genomic DNA) was mixed with 500 ng of genomic DNA from control (mock) GAA-KO mouse as the standards for semiquantitative assay. The reaction was terminated with a 10-minute extension at 72 °C. Aliquots of 20 ␮l of each PCR reaction were electrophoretically separated on 1.2% agarose gel with ethidium bromide and photographed.

ACKNOWLEDGMENTS We acknowledge the support of research grants from Synpac and Genzyme corporations. D. D. K. was supported by the Muscular Dystrophy Association and by departmental funds provided by the Howard Hughes Medical Institute. We thank Joe Zeidner, Eric Bonno, Ayn Schneider, and Ibrahim Bori for technical support. A. A. was supported by R01-DK 52925 and the Muscular Dystrophy Association.

476

doi:10.1016/S1525-0016(03)00022-4

REFERENCES 1. Muzyczka, N. (1992). Use of adeno-associated virus as a general transduction vector for mammalian cells. Curr. Top. Microbiol. Immunol. 158: 97–129. 2. Xiao, X., et al. (1998). Production of high-titer recombinant adeno- associated virus vectors in the absence of helper adenovirus. J. Virol. 72: 2224 –2232. 3. Grimm, D., et al. (1998). Novel tools for production and purification of recombinant adeno-associated virus vectors. Hum. Gene Ther. 9: 2745–2760. 4. Yang, Q., et al. (1994). Characterization of cell lines that inducibly express the adenoassociated virus Rep proteins. J. Virol. 68: 4847– 4856. 5. Clark, K.R., et al. (1995). Cell lines for the production of recombinant adeno-associated virus. Hum. Gene Ther. 6: 1329 –1341. 6. Tamayose, K., et al. (1996). A new strategy for large-scale preparation of high-titer recombinant adeno-associated virus vectors by using packaging cell Lines and sulfonated cellulose column chromatography. Hum. Gene Ther. 7: 507–513. 7. Inoue, N., and Russell, D.W. (1998). Packaging cells based on inducible gene amplification for the production of adeno-associated virus vectors. J. Virol. 72: 7024 –7031. 8. Gao, G.-P., et al. (1998). High-titer adeno-associated viral vectors from a Rep/Cap cell line and hybrid shuttle virus. Hum. Gene Ther. 9: 2353–2362. 9. Liu, X.L., et al. (1999). Production of recombinant adeno-associated virus vectors using a packaging cell line and a hybrid recombinant adenovirus. Gene Ther. 6: 293–299. 10. Hirschhorn, R., and Reuser, A.J. (2001). Glycogen storage disease type II: acid ␣-glucosidase (acid maltase) deficiency. In The Metabolic and Molecular Basis for Inherited Disease (C.R. Scriver, A.L. Beaudet, W.S. Sly, and D. Valle, Eds.), pp. 3389 –3419. McGraw-Hill, New York. 11. Raben, N., et al. (1998). Targeted disruption of the acid ␣-glucosidase gene in mice causes an illness with critical features of both infantile and adult human glycogen storage disease type II. J. Biol. Chem. 273: 19086 –19092. 12. Amalfitano, A., et al. (1999). Systemic correction of the muscle disorder glycogen storage disease type II after hepatic targeting of a modified adenovirus vector encoding human acid-␣-glucosidase. Proc. Natl. Acad. Sci. USA 96: 8861– 8866. 13. Ding, E.Y., et al. (2001). Long-term efficacy after [E1-, polymerase-] adenovirus-mediated transfer of human acid-␣-glucosidase gene into glycogen storage disease type II knockout mice. Hum. Gene Ther. 12: 955–965. 14. Pauly, D.F., et al. (2001). Intercellular transfer of the virally derived precursor form of acid ␣-glucosidase corrects the enzyme deficiency in inherited cardioskeletal myopathy Pompe disease. Hum. Gene Ther. 12: 527–538. 15. Snyder, R.O., et al. (1999). Correction of hemophilia B in canine and murine models using recombinant adeno-associated viral vectors. Nat. Med. 5: 64 –70. 16. Daly, T.M., et al. (1999). Neonatal gene transfer leads to widespread correction of pathology in a murine model of lysosomal storage disease. Proc. Natl. Acad. Sci. USA 96: 2296 –2300. 17. Jung, S.-C., et al. (2001). Adeno-associated viral vector-mediated gene transfer results in long-term enzymatic and functional correction in multiple organs of Fabry mice. Proc. Natl. Acad. Sci. USA 98: 2676 –2681. 18. He, T.-C., et al. (1998). A simplified system for generating recombinant adenoviruses. Proc. Natl. Acad. Sci. USA 95: 2509 –2514. 19. Amalfitano, A., et al. (1998). Production and characterization of improved adenovirus vectors with the E1, E2b, and E3 genes deleted. J. Virol. 72: 926 –933. 20. Allen, J.M., et al. (2000). Improved adeno-associated virus vector production with transfection of a single helper adenovirus gene, E4orf6. Mol. Ther. 1: 88 –95 doi: 10.1006/mthe.1999.0010. 21. Li, J., et al. (1997). Role for highly regulated rep gene expression in adeno-associated virus vector production. J. Virol. 71: 5236 –5243. 22. Scaria, A., et al. (1995). Complementation of a human adenovirus early region 4 deletion mutant in 293 cells using adenovirus-polylysine-DNA complexes. Gene Ther. 2: 295–298. 23. Matsushita, T., et al. (1998). Adeno-associated virus vectors can be efficiently produced without helper virus. Gene Ther. 5: 938 –945. 24. Zolotukhin, S., et al. (1999). Recombinant adeno-associated virus purification using novel methods improves infectious titer and yield. Gene Ther. 6: 973–985. 25. Ding, E., et al. (2002). Efficacy of gene therapy for a prototypical lysosomal storage disease (GSD-II) is critically dependent on vector dose, transgene promoter, and the tissues targeted for vector transduction. Mol. Ther. 5: 436 – 446 doi:10.1006/ mthe.2002.0557. 26. Raben, N., et al. (2001). Conditional tissue-specific expression of the acid ␣-glucosidase (GAA) gene in the GAA knockout mice: implications for therapy. Hum. Mol. Genet. 10: 2039 –2047. 27. Halbert, C. L., et al. (2001). Adeno-associated virus type 6 (AAV) vectors mediate efficient transduction of airway epithelial cells in mouse lungs compared to that of AAV2 vectors. J. Virol. 75: 6615– 6624. 28. Liu, X.L., et al. (1996). Production of recombinant adeno-associated virus vectors using a packaging cell line and a hybrid recombinant adenovirus. Gene Ther. 6: 293–299. 29. Maxwell, I.H., et al. (1998). An adenovirus type 5 mutant with the preterminal protein gene deleted efficiently provides helper functions for the production of recombinant adeno-associated virus. J. Virol. 72: 8371– 8373.

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

doi:10.1016/S1525-0016(03)00022-4

30. Recchia, A., et al. (1999). Site-specific integration mediated by a hybrid adenovirus/ adeno-associated virus vector. Proc. Natl. Acad. Sci. USA 96: 2615–2620. 31. Lieber, A., et al. (1999). Integrating adenovirus-adeno-associated virus hybrid vectors devoid of all viral genes. J. Virol. 73: 9314 –9324. 32. Shayakhmetov, D.M., et al. (2002). A high-capacity, capsid-modified hybrid adenovirus/adeno-associated virus vector for stable transduction of human hematopoietic cells. J. Virol. 76: 1135–1143. 33. Zhang, X., and Li, C.-Y. (2001). Generation of recombinant adeno- associated virus vectors by a complete adenovirus-mediated approach. Mol. Ther. 5: 787–792, doi: 10.1016/mthe.2001.0306. 34. Kishnani, P.S., et al. (2001). Canine model and genomic structural organization of glycogen storage disease type Ia (GSD Ia). Vet. Pathol. 38: 83–91. 35. Amalfitano, A., and Chamberlain, J.S. (1997). Isolation and characterization of packaging cell lines that coexpress the adenovirus E1, DNA polymerase, and preterminal proteins: implications for gene therapy. Gene Ther. 4: 258 –263. 36. Van Hove, J.L., et al. (1996). High-level production of recombinant human lysosomal acid ␣-glucosidase in Chinese hamster ovary cells which targets to heart muscle and corrects glycogen accumulation in fibroblasts from patients with Pompe disease. Proc. Natl. Acad. Sci. USA 93: 65–70.

MOLECULAR THERAPY Vol. 7, No. 4, April 2003 Copyright © The American Society of Gene Therapy

ARTICLE

37. Brinster, R.L., et al. (1988). Introns increase transcriptional efficiency in transgenic mice. Proc. Natl. Acad. Sci. USA 85: 836 – 840. 38. Koeberl, D.D., et al. (1997). Persistent expression of human clotting Factor IX from mouse liver after intravenous injection of adeno-associated virus vectors. Proc. Natl. Acad. Sci. USA 94: 1426 –1431. 39. Beaty, R.M., et al. (2002). Delivery of glucose-6-phosphatase in a canine model for glycogen storage disease, type Ia, with adeno-associated virus (AAV) vectors. Gene Ther. 9: 1015–1022. 40. Hodges, B.L., et al. (2000). Multiply deleted [E1, polymerase-, and pTP-] adenovirus vector persists despite deletion of the preterminal protein. J. Gene Med. 2: 250 –259. 41. Overbergh, L., et al. (1999). Quantification of murine cytokine mRNAs using real-time quantitative reverse transcriptase PCR. Cytokine 11: 305–312. 42. Allen, J.M., et al. (1997). Identification and elimination of replication-competent adenoassociated virus (AAV) that can arise by nonhomologous recombination during AAV vector production. J. Virol. 71: 6816 – 6822. 43. Rutledge, E.A., et al. (1998). Infectious clones and vectors derived from adeno-associated virus (AAV) serotypes other than AAV type 2. J. Virol. 72: 309 –319.

477