gene transfer properties and structural modeling of human stem cell-derived aav

original article© The American Society of Gene & Cell Therapy

Adeno-associated virus (AAV) vectors are proving to be remarkably successful for in vivo gene delivery. Based upon reports of abundant AAV in the human marrow, we tested CD34+ hematopoietic stem cells for the presence of natu-ral AAV. Here, we report for the first time, the presence of novel AAV variants in healthy CD34+ human peripheral blood stem cells. The majority of healthy peripheral blood stem cell donors were found to harbor AAV in their CD34+ cells. Every AAV isolated from CD34+ cells mapped to AAV Clade F. Gene transfer vectors derived from these novel AAVs efficiently underwent entry and postentry process-ing in human cord blood stem cells and supported stable gene transfer into long-term, in vivo engrafting human HSCs significantly better than other serotypes. AAVHSC-transduced human CD34+ cells engrafted in vivo and gave rise to differentiated transgene- expressing progeny. Importantly, gene-marked CD34+ stem cells persisted long term in xenograft recipients, indicating transduction of primitive progenitors. Notably, correlation of structure with function permitted identification of potential cap-sid components important for HSC transduction. Thus, AAVHSCs represent a new class of genetic vectors for the manipulation of HSC genomes.

Received 24 March 2014; accepted 29 May 2014; advance online publication 15 July 2014. doi:10.1038/mt.2014.107

INTRODUCTIONThe ability of adeno-associated virus (AAV) to establish latent infection in the absence of helper virus coinfection has led to their widespread use as in vivo gene delivery vectors.1–4 Their capac-ity to transduce nondividing primary tissues and direct sustained transgene expression in the absence of toxicity in vivo has led to remarkable observations of therapeutic efficacy, including gene delivery for the treatment of retinal diseases, Parkinson’s Disease, hemophilia B, and heart failure.5–8 Recently, Glybera, an AAV-based vector for the treatment of lipoprotein lipase deficiency, became the first approved gene therapy drug in the western world.9

Identification of novel AAV genomes from primate tissues has led to the creation of a repertoire of vectors with wide tis-sue tropisms, likely based on interactions with specific cell sur-face receptors10–16 and allowed homology-based grouping of AAV

into clades (A to F), potentially revealing evolutionary relation-ships.17–19 These novel vectors have greatly expanded the ability to deliver genes to targeted tissues, including those previously thought to be refractory to gene transfer and have provided a means of circumventing highly prevalent preexisting immunity to AAV2.20,21

We and others have previously reported AAV transduction of human CD34+ cells, a population enriched for hematopoietic stem cells (HSCs).22–30 Although transduction of in vivo engrafting CD34+ stem cells was noted, the efficiencies were modest. The use of AAV capsids with site-directed mutagenesis of surface exposed tyrosine residues led to a substantial increase in the level of trans-duction of in vivo engrafting stem cells.24,26,27 Similarly, the use of pseudotyped recombinant AAV using capsids of other serotypes was also found to result in higher transduction efficiencies,26 sug-gesting that significant improvement in gene transfer efficiency could be addressed through capsid-based strategies.

Based upon the identification of the human bone marrow as an abundant source of natural AAV,17 we hypothesized that AAV isolates may be present in human CD34+ HSCs and that their capsids would have tropism for CD34+ cells. Here, we report for the first time, the presence of natural AAV variants in CD34+ human peripheral blood stem cells (PBSC) from healthy adults. The isolation and sequencing of a panel of CD34+ cell-derived AAV genomes (AAVHSC) revealed that the novel capsids were unique. Every AAV sequence isolated from CD34+ cells mapped to AAV Clade F.17 Novel AAVHSC vectors transduced human HSCs in vitro and also efficiently transduced long-term in vivo engrafting multipotential human HSCs in a xenotransplantation model. Transduction of CD34+ cells with AAVHSCs was found to be significantly more efficient than with serotypes mapping to other clades, including AAV2, AAV7, and AAV8 and for some also more than AAV9, the representative member of Clade F. Finally, serial in vivo bioluminescent imaging of mice transplanted with AAVHSC-transduced human CD34+ cells revealed that AAVHSC supported long-term in vivo transgene expression without con-comitant toxicity, suggesting their utility for stem cell gene trans-fer. Additionally, correlation of stem cell transduction capacities with limited structural differences between the AAVHSCs allow for mapping of AAV virion components important for genetic modification of CD34+ cells.

15July2014

1625

1634

Novel Stem Cell-derived AAV Vectors

Molecular Therapy

10.1038/mt.2014.107

original article

00sep2014

22

9

24March2014

29May2014

© The American Society of Gene & Cell Therapy

Correspondence: Saswati Chatterjee, Department of Virology, Beckman Research Institute of City of Hope, 1500 E Duarte Rd, Duarte, California 91010, USA. E-mail: [email protected]

Gene Transfer Properties and Structural Modeling of Human Stem Cell-derived AAVLaura J Smith1, Taihra Ul-Hasan1, Sarah K Carvaines1, Kim Van Vliet2, Ethel Yang1, Kamehameha K Wong, Jr3, Mavis Agbandje-McKenna2 and Saswati Chatterjee1

1AAV Laboratory, Department of Virology, Beckman Research Institute of City of Hope, Duarte, California, USA; 2Department of Biochemistry and Molecular Biology, Center for Structural Biology, The McKnight Brain Institute University of Florida, Gainesville, Florida, USA; 3Divisions of Hematology/Stem Cell Transplantation, Beckman Research Institute of City of Hope, Duarte, California, USA

Vector engineering and delivery

Molecular Therapy vol. 22 no. 9, 1625–1634 sep. 2014 1625

http://www.nature.com/doifinder/10.1038/mt.2014.107

mailto:[email protected]

© The American Society of Gene & Cell TherapyNovel Stem Cell-derived AAV Vectors

RESULTSNovel AAV from human CD34+ cells map to Clade F and package recombinant AAV genomesWe first tested the hypothesis that CD34+ human HSCs harbored endogenous AAV. Using a sensitive quantitative PCR (qPCR) assay specific for the highly conserved AAV rep genes, we screened high molecular weight genomic DNA from purified cytokine-primed human CD34+ PBSCs from 71 healthy stem cell donors

for the presence of endogenous AAV. Approximately 70% of the donors screened were positive for endogenous AAV sequences (Table 1). The frequency of AAV in the CD34+ cell population was normalized to the single copy Apo B gene. Forty-two percent of the samples had a frequency of 0.1 to 0.9 copies, 25% had 1 to 10 copies, and 1 sample (1.41%) had >10 copies per 1000 cells. Thirty-one percent of donors were negative for AAV. While it is possible that some of the AAV isolates originated from contami-nating cells (probability: 0.02), the probability that any given iso-late originated from a CD34+ cell is 0.98. Thus while we cannot rule out that some individual AAV isolates may have been derived from contaminating cells, the probability that every AAVHSC isolate originated from a contaminating cell is extremely low at 0.02 (ref. 17). Conversely, the probability that some, if not most of these AAV isolates originated from CD34+ cells is very high. In addition, every isolate from one of the two donors analyzed has a common signature amino acid change, indicating a common ori-gin. Thus, we conclude that endogenous AAV occurs frequently in CD34+ PBSCs of healthy adult donors and that natural AAV infection of CD34+ cells is not rare.

Table 1 The frequency of naturally occurring AAV in human cytokine-primed peripheral blood stem cell samples

Frequency of endogenous AAVNumber

of donorsPercent

of donors

>10 copies/1,000 cells 1 1.41%

1–10 copies/1,000 cells 18 25.35%

0.1–0.9 copies/1,000 cells 30 42.25%

AAV negative 22 30.99%

Total 71 100%

Figure 1 Nucleotide sequence alignment phylogram of AAVHSC. Tree analysis was performed using the neighbor joining algorithm. Neighbor joining tree analysis of nucleotide sequence homology between the novel AAVHSC capsids and previously described AAV isolates showed that the new viruses were more closely related to each other and to the other members of Clade F. AAV isolates derived from the same donor show preferential clustering.

AAVHSC62112

5

3

83

0

1

2

0

43

10425

4

7

7

1

AAVHSC8

AAVHSC4

AAVHSC5

AAVHSC2

AAVHSC10

AAVHSC8

AAVHSC1

AAVHSC7

AAVHSC9

AAVHSC11

AAVHSC13

AAVHSC14AAVHSC15

AAVHSC6AAVHSC7

0.001

16

1243

31

AAVHSC17AAVHSC16

AAVHSC12

AAVHSC4

AAVHSC3

AAVHSC10

AAVHSC2

AAVHSC5

AAVHSC1

AAVHSC11

AAVHSC9AAVHSC8

AAV10

AAV8

AAV11

AAV2

Clade E

Clade CClade D

Clade A

Clade BAAV1

AAV6

AAV7AAV3

AAV4AAV5100

65

37

87

948580

89

100

95

AAVHSC3

AAVhu32AAVhu31

100

0.600

Clade F

9

8

0

0

9

82

7

1

100

AAV9

AAVHSC17

AAVHSC13AAVHSC15

AAVHSC12

AAVHSC16

AAVHSC14

1626 www.moleculartherapy.org vol. 22 no. 9 sep. 2014


Using primers complementary to conserved regions of the AAV genome and a high fidelity polymerase, we amplified full-length capsid genes from PBSC samples from two donors. Sequence analysis of AAV isolates revealed the presence of 18 novel AAV capsids as well as wild type AAV9. All novel sequences shared the highest homology with AAV9 (Figure 1), a member of AAV antigenic Clade F, differing from this serotype by only one to four amino acids. All AAV isolates from PB0130 (AAVHSC1 to 11) had amino acid changes in one or more of the three capsid viral proteins (VP)s, VP1, 2, and 3, while all six novel isolates from the second donor, PB0201, only had changes in VP3 (AAVHSC12-17) (Table 2). Every AAV capsid isolated from PB0201 had a signature amino acid change at residue 505. Neighbor joining tree analysis of nucleotide sequence homology between the novel AAVHSC capsids and previously described AAV isolates showed that the new viruses were more closely related to each other and to the other members of Clade F (Figure 1). AAV isolates derived from the same donor show preferential clustering, possibly suggesting a common evolutionary origin.

Five of the novel amino acids in AAVHSCs from PB0130, mapped to the VP1 unique (VP1u) region of the capsid VP which has thus far eluded structural characterization. Most of the VP1u changes were conservative (Table 2). The only exception was a nonpolar to polar change at position 2 of AAVHSC1. While most of the VP1u changes did not fall within specifically defined motifs, one amino acid change, V65I in AAVHSC11, mapped to the last of the four recently described PDZ-binding motifs, while another, K77R in AAVHSC5, is located within the phospholipase A2 (PLA2) region.31 AAVHSC5 and AAVHSC7, both with changes in the VP1u, were capable of transduction suggesting that the

conservative amino acid changes did not affect function including virus escape from the endoslysosomal system after cellular entry.

To test the gene transfer capacities of the AAVHSCs, we pack-aged recombinant self-complementary (sc) and single-stranded (ss) AAV2 genomes in AAVHSC capsids (Table 2). Rep and cap functions were provided in trans from plasmids encoding a hybrid AAV2 rep (nt151 to 1890) along with the novel capsid gene. The hybrid rep gene consisted of nt151 to 1890 of AAV2. The remain-der of the rep gene was derived from the novel AAVHSCs. The use of the hybrid rep resulted in recombinant viral titers indis-tinguishable from those obtained with AAV2 rep. AAVHSC2, AAVHSC11, and AAVHSC14 packaged recombinant AAV only to low titers, ranging from 106 to 5 × 109 vector genomes per ml (vg/ml), while the rest of the capsids packaged rAAV at robust titers ranging from 1011 to >1013 vg/ml (Table 2).

AAVHSC vectors transduce CD34+ cells in vitroSince the AAVHSCs were derived from CD34+ stem cells, we reasoned that they should be capable of transducing HSCs. Thus, we evaluated the ability of AAVHSC vectors to mediate gene transfer into CD34+ cells in vitro. Purified pooled cord blood CD34+ cells were transduced with a self-complementary AAV vector encoding EGFP under the control of a chicken beta actin (CBA) promoter at an multiplicity of infection of 20,000. Efficient transgene expression following in vitro transduction of CD34+ cells confirmed the ability of the AAVHSCs to negotiate receptor-mediated entry, and postentry processing, including transgene expression. Certain isolates including AAVHSC1, AAVHSC5, and AAVHSC12, transduced CD34+ cells 2.3- to 4.3-fold more efficiently than AAV2, possibly due to efficient entry and/or

Table 2 Amino acid changes and packaging capacities of AAVHSC relative to AAV9

Capsid

Novel amino acids Silent amino acids Titersa

VP1 VP2 VP3 VP1 VP2 VP3 scAAVEGFP ssAAVLuc

AAVHSC1 A2T — R312Q — — — 2 × 1012 6 × 1011

AAVHSC2 — — D626, E718G — — 505 1.34 × 109 N/A

AAVHSC3 — G161D — — — 348 — —

AAVHSC4 F119L — P468S — — 603 8.4 × 1011 3.14 × 1011

AAVHSC5 K77R — E690K — — 468, 716 5.5 × 1010 1.4 × 1012

AAVHSC6 — — Q590R — — — — —

AAVHSC7 A68V — — — — 308 8.6 × 1010 1.2x1013

AAVHSC8 — Q151R — — — 686 — —

AAVHSC9 — — C206G 119 — — 5.2 × 1011 1 × 1012

AAVHSC11 V65I — D626Y — — — 7 × 106 N/A

AAVHSC12 — — R296H, S464N, G505R, V681M — 193 502 9.54 × 1010 2.15 × 1011

AAVHSC13 — — G505R — — — 4 × 1010 1 × 1012

AAVHSC14 — — G505R, L687R — — — 2.4 × 108 5.3 × 109

AAVHSC15 — — T346A, G505R — — — 6.4 × 1010 2 × 1012

AAVHSC16 — — F501I, G505R, Y706C — — 571 8 × 1010 3.7 × 1012

AAVHSC17 — — G505R 60 — — 5.9 × 1011 2.4 × 1012

AAV9 — — — — — — 6.2 × 1.010 1 × 1012

aThe titers shown are representative for each recombinant AAV vector.

Molecular Therapy vol. 22 no. 9 sep. 2014 1627


intracellular processing (Figure 2a,b). AAVHSC4, AAVHSC15, AAVHSC16, and AAVHSC17 transduced at intermediate levels comparable to AAV2 and AAV9, while AAVHSC7, AAVHSC9, and AAVHSC13 transduced at low levels. Thus, we concluded that AAVHSCs were capable of receptor-mediated entry, nuclear translocation, uncoating, second strand synthesis, and transgene expression in human cord blood-derived CD34+ cells.

AAVHSC vectors transduce long-term in vivo engrafting human CD34+ cells more efficiently than other AAV serotypesNext, we directly queried the capacity of AAVHSCs to transduce long-term engrafting, multipotential human HSC using a xeno-transplantation model in immunodeficient nonobese diabetic/severe combined immunodeficiency (NOD/SCID) mice. Highly

Figure 2 Transduction of cord blood-derived human CD34+ cells with AAVHSC vectors in vitro. (a) Scatter plots of flow cytometric analysis of enhanced green fluorescent protein (EGFP) expression in cord blood (CB) CD34+ cells 24 hours after transduction. Purified pooled CB CD34+ cells were transduced with a self-complementary AAV vector encoding EGFP under the control of a CBA promoter at an MOI of 20,000. Representative samples are shown. (b) EGFP expression in AAV-transduced human CD34+ cells 24 hours posttransduction. Data were compiled from five experi-ments, each performed with pools of three or more CB samples. Bars represent standard deviations of the means. CBA, chicken beta actin.

100

0

AAVHSC1

AAVHSC5

AAVHSC12

AAVHSC15

AAVHSC16

AAVHSC4

AAVHSC17

AAVHSC9

AAVHSC13

AAVHSC7AAV2

AAV9AAV8

AAV7AAV6

10

AAV strain

20

30

40

% S

peci

fic tr

ansg

ene

expr

essi

on

50

60

70

10,000 20,000 30,000 10,000 20,000 30,000 10,000 20,000 30,000 10,000 20,000 30,000 10,000 20,000 30,000

101

102

103

104

100

101

102

103

104100

101

Control AAVHSC1 – 59.11% AAVHSC4 – 20.19% AAVHSC5 – 64.19% AAVHSC7 – 7.24%

AAVHSC9 – 20.87% AAVHSC12 – 44.45% AAVHSC13 – 18.89% AAVHSC15 – 44.65% AAVHSC16 – 36.25%

AAVHSC17 – 24.21% AAV2 – 6.34% AAV7 – 11.87% AAV8 – 7.10% AAV9 – 3.27%

102

103

104

a

b



purified pooled CD34+ cells were transduced with AAVHSCs at multiplicity of infection: 20,000 prior to transplantation into sublethally irradiated immune-deficient NOD/SCID mice. The AAV vectors used encoded the single-stranded firefly luciferase gene (ssLuc) under the control of the ubiquitous CBA promoter to permit serial in vivo bioluminescent monitoring of trans-gene expression. Serial quantification of luciferase expression by

bioluminescent imaging was initiated at 4 weeks posttransplan-tation and continued biweekly until the time of harvest, 20–26 weeks posttransplantation. Luciferase expression at any given time, likely represents the expression in the transduced circulat-ing hematopoietic progeny of a limited number of transduced CD34+ progenitor cells as well as the human marrow-resident progenitor and differentiated hematopoietic cells. While all AAV

Figure 3 In vivo transgene expression in NOD/SCID mice transplanted with AAVHSC-transduced CD34+ cells. (a) Representative serial biolumines-cent images of xenotransplant recipients of AAV transduced human cord blood (CB) CD34+ cells. Transgene expression is shown at different time intervals posttransplantation. Purified pooled CB CD34+ cells were transduced with AAV at MOI:20,000 prior to transplantation into sub-lethally irradiated immune-deficient NOD/SCID mice. Luciferase expression was serially quantified by bioluminescent imaging initiated at 4 weeks posttransplantation and continued biweekly until the time of harvest, 20–26 weeks posttransplantation. (b) Serial in vivo bioluminescent imaging of luciferase expression in NOD/SCID mice xenografted with AAVHSC-transduced CB CD34+ cells. Imaging was performed biweekly. Results are shown for 3–5 mice per group. Each mouse was transplanted with transduced CD34+ cells pooled from 1 to 5 CB samples. *Luciferase expression for AAVHSC7, AAVHSC15 relative to AAV9 is significant to P ≤ 0.01 and ** for AAVHSC17 relative to AAV9 is significant to P ≤ 0.005, respectively. NOD/SCID, nonobese diabetic/severe combined immunodeficiency.

AAVHSC1

AAVHSC4

AAVHSC5

AAVHSC7

AAVHSC12

AAVHSC13

7.00 × 105

4 6 8 10 12 14

Weeks post-transplantation

AAVHSC17**

AAVHSC16

AAVHSC15*

AAVHSC13

AAVHSC4

AAVHSC7*

AAVHSC5AAVHSC1

AAVHSC12AAV9

AAV2AAV7AAV8

16 18 20 22

9.00 × 105

1.10 × 106

1.30 × 106

Flu

x (p

hoto

ns/s

ec/c

m2 )

1.50 × 106

AAVHSC15

AAVHSC16

AAVHSC17

AAV2

AAV8

5,000

4,000

3,000

2,000

1,000Color barMin = 1,000Max = 5,000

4 weeks 8–12 weeks 16–20 weeks 4 weeks 8–12 weeks 16–20 weeks

AAV9

a

b



serotypes tested supported long-term in vivo transduction, Clade F viruses were significantly more efficient than AAV2, AAV7, and AAV8 (Figure 3a,b), confirming the hypothesis that AAVHSCs have enhanced tropism for CD34+ cells. Notably, both initial in vivo transduction levels as well as sustained long-term transgene expression were higher with every AAVHSC tested as compared to AAV2 (P < 0.001), AAV7, and AAV8. The use of pooled CD34+ cells from multiple donors obviated differences due to donor vari-ability.25,30 AAV9, also a member of Clade F, supported transgene expression at levels comparable to several AAVHSCs (Figure 3). AAVHSC17, AAVHSC15, and AAVHSC7 were found to sup-port the highest level of long-term luciferase expression, in that order, throughout the 6-month period of analysis and were sig-nificantly higher than AAV9 (P ≤ 0.005 for AAVHSC17, P ≤ 0.01 for AAVHSC15 and AAVHSC7).

AAVHSC vector genomes are detected long-term in all human hematopoietic lineages recovered from the marrow of xenograft recipients including CD34

+ cells

The effects of AAVHSC transduction on the engraftment and dif-ferentiation potentials of human CD34+ HSCs were evaluated in human hematopoietic lineages in the marrow of xenotransplant recipients at 20–26 weeks posttransplantation. Human cell engraft-ment was found to be stable long term (Table 3). Differentiated human lymphomyeloid, erythroid, and CD34+ stem/progenitor cells were detected in the marrow of transplant recipients up to 26 weeks posttransplantation, the end of the experiment (Table 3), sug-gesting that AAVHSC-transduced cells were capable of supporting long-term engraftment, with no associated toxicity. On the basis of the purity of the input CD34+ cells, we concluded that the differenti-ated human hematopoietic cells were derived from the transplanted CD34+ cells, indicating their capacity for multilineage engraftment.

The frequency of AAV genomes in human hematopoietic cells was quantified in purified human lineages isolated from the marrow of xenotransplant recipients at the time of harvest. AAV genomes were detected by qPCR and normalized to the single copy human gene ApoB. Every recipient analyzed showed the presence of AAVHSC-marked cells in multiple hematopoietic lineages

(Table 4). AAVHSC-transduced human CD19+ B cells, CD33+ myeloid cells, and glycophorin A+ erythroid cells were readily detected. Importantly, AAVHSC-marked CD34+ cells were identi-fied in every xenotransplant, indicating the capacity of AAVHSC-marked progenitors to persist and/or self-renew in vivo. AAVHSC genomes were found to persist in a stable fashion within the divid-ing and differentiating progeny of transduced human CD34+ cells long-term. Thus, we conclude that AAVHSC vectors efficiently transduced CD34+ multipotential HSC and that transduced cells differentiated normally into all expected human lineages.

DISCUSSIONRecombinant vectors derived from natural AAVs isolated from human and nonhuman primates have transformed gene deliv-ery and have led to promising clinical successes.1–9 Our survey of adult human CD34+ PBSCs revealed that the majority of healthy donors had detectable AAV sequences. The frequency of AAV in CD34+ cells indicated that infection of adult PBSCs was not rare. The mapping of every stem cell-derived AAV to Clade F, indi-cated that this clade may have primary tropism for CD34 cells. This was further supported by the observation that every Clade F virus tested, including the AAVHSCs and AAV9, transduced long-term, multipotential in vivo engrafting human CD34+ HSCs significantly better than other AAV serotypes.

While most AAVHSCs packaged recombinant AAV2 genomes and produced viable transducing virus, two variants, AAVHSC2 and AAVHSC11, packaged DNA to substantially lower titers than others. Both capsids had amino acid changes at the same site, resi-due D626G and D626Y, respectively. This residue is positioned inside the capsid under the icosahedral threefold axis of the AAV capsid (Figure 4a,b) at a VP:VP interface and is involved in polar interactions with amino acids from symmetry related VP mono-mers. This region also contains the single structurally ordered nucleotide reported in a number of AAV capsid structures.32 Thus, it is possible that the change at this site reduces virion formation or stability or may affect genome stabilizing interactions following packaging which are affected when mutated. AAVHSC2 has an

Table 3 Average engraftment and human hematopoiesis in the bone marrow of xenotransplant recipients

CD19+ CD33+ Glyco A+ CD34+ CD45+

AAVHSC1 49.01 3.16 1.35 7.07 28.82

AAVHSC4 52.69 8.15 2.40 11.22 32.14

AAVHSC5 51.05 4.39 1.31 12.73 27.42

AAVHSC7 52.10 4.24 1.21 5.63 57.00

AAVHSC12 35.64 3.44 8.32 5.71 6.43

AAVHSC13 40.17 7.51 0.73 6.33 24.72

AAVHSC15 41.20 9.69 4.25 4.84 41.77

AAVHSC16 37.96 3.15 1.00 3.86 21.28

AAVHSC17 38.56 14.05 5.05 2.51 23.58

AAV9 55.33 6.29 3.26 7.43 39.03

Numbers represent the percent of total CD45+ human cells represented in the human CD19+, CD33+, glycophorin A+ and CD34+ lineages. The CD45+ column represents the frequency of human cells in the xenotransplanted mouse marrow at the end of the experiment.

Table 4 Average frequency of AAV genomes in cells of human hematopoietic lineages in the bone marrow of xenotransplant recipients

CD19+ CD33+ GlycoA+ CD34+

AAVHSC1 0.8a 0.0 4.6 0.8

AAVHSC4 0.8 0.075 6.3 1.5

AAVHSC5 6.7 1.4 0.0 7.0

AAVHSC7 0.2 6.0 10.4 0.37

AAVHSC12 2.8 7.4 191.4 7.5

AAVHSC13 0.2 2.2 1.0 8.7

AAVHSC15 0.2 2.8 35.9 1.1

AAVHSC16 0.5 0.4 7.7 0.6

AAVHSC17 0.3 0.2 0.1 0.03

AAV9 0.8 0.2 6.1 16.3aAAV genomes were amplified using vector-specific primers and probes and normalized to the single copy human gene ApoB. Numbers represent vector genome copies per 1000 cells of each lineage at 22–26 weeks posttransplantation.



additional amino acid change, E718G, also at a VP:VP interface, on the raised capsid region between the depressions that surround the two- and fivefold axes (Figure 4a,c). This residue is involved in intramolecular VP interactions, and thus may play a role in VP stabilization; a role in genome packaging is less likely based on its capsid surface location. AAVHSC11 also had the V65I change within VP1u mentioned above. Interestingly, none of the packag-ing-defective isolates had changes in the VP2 coding region that overlaps with the assembly activation protein, AAP, responsible for initiation of virion formation.33 The remainder of the novel AAV capsids successfully interacted with AAV2 ITRs, packaged recombinant genomes, and assembled infectious virus particles at titers comparable to wild-type AAV9.

PB0130 variants AAVHSC1, AAVHSC4, and AAVHSC5 with normal packaging and assembly phenotypes, had single non-conservative amino acid changes in VP3 compared to AAV9 at positions R312Q, P468S, E690K, respectively (Figure 4). Residue R312 is located in a β-strand inside the capsid with its side-chain pointing towards the interior surface of the capsid. Residue P468 is located on the outside wall of the protrusions surrounding the threefold axis (not facing the threefold axis) and the change could cause a conformational change in the loop containing this resi-due. Residue E690K is located on the inside surface of the capsid close to the twofold interface and interacts with R296 and R300 (Figure 4b,d). A change to lysine would dramatically alter these interactions, possibly altering the stability of the capsid.

Figure 4 Capsid surface location of AAVHSC variant amino acid changes. Exterior (a) and interior (b) capsid surface images of the AAV9 struc-ture showing the positions of amino acid residues for AAVHSC variants arising from donor PB0130 (green), donor PB0201 (yellow), silent mutations (orange), AAV9 galactose binding (brown), and AAV2 heparan sulfate binding (blue). The triangle depicts the viral asymmetric unit bounded by a fivefold (filled pentagon) axis and two threefold (filled triangles) axes divided by a line through a twofold (filled oval) axis. (c,d) depict a viral asym-metric unit as a “Roadmap”39 with the surface area occupied by capsid surface amino acids delineated and labeled. The variant residues are colored as in panel A. The images are viewed down an icosahedral twofold axis.

E718

V681

T346

S348

E690

L687

E686

R312

D626

N716 P468

Y706 Gal binding site

Q590

A589

R485

R488

K533

G505

a b

55.0

65.0

φ

φ

θ θ75.0

85.0

95.0

55.0

65.0

75.0

85.0

95.0

62.0

72.0

82.0 92.0

102.

0

112.

0

118.

0

62.0

72.0

82.0

92.0

102.0

112.0

118.0

55.0

65.0

φ

φ

θ θ75.0

85.0

95.0

55.0

65.0

75.0

85.0

95.062.0

72.0

82.0 92.0

102.

0

112.

0

118.

0

62.0

72.0

82.0

92.0

102.0

112.0

118.0

c d



Interestingly, every isolate from PB0201 had changes only within the VP3 common sequence and each had a nonconserved signature G505R change, indicating that this mutation pre-ceded the others and suggested a common evolutionary origin. Although the sequence of the acquisition of the additional diver-sity is unknown, given its replication defective status and the use of a higher fidelity cellular DNA polymerase by AAV,34 it is less likely for this diversity to have arisen postinfection. The initial infection may have occurred with a swarm of closely related but diverse virions, or helper virus coinfection may have led to repli-cation and subsequent generation of diversity.

The common G505R change is located on the outside wall of the protrusions that surround the threefold axis of the AAV9 capsid (Figure 4a,c). It is located on a loop, which together with two other loops from a threefold symmetry-related VP monomer, forms the protrusions. The other VP3 changes were distributed throughout the capsid, with residue R296 located inside the cap-sid at an icosahedral twofold interface; residue T346 located in a β-strand on the inside surface of the capsid, residues S464 and F501 located on the same wall of the protrusion as residue G505 and P468; residues V681 and L687 in a β-strand on the inside sur-face of the capsid; and Y706 on the capsid surface close to the twofold axis (Figure 4). All of these variants were capable of cap-sid assembly and genome packaging indicating that these changes did not affect these properties. With the exception of AAVHSC13, all the variants had improved in vitro transduction efficiencies relative to the wild-type serotypes and all had improved in vivo transduction. This suggested that the changes imparted improved fitness, possibly in the form of improved capsid stability, recep-tor attachment, cellular trafficking, or other cellular interaction, in CD34+ cells. Consistent with these possibilities, these mutations surround the known AAV2 heparan sulfate and AAV9 galactose binding residues (Figure 4). Furthermore, Y706 is structurally equivalent to Y705 of AAV2 which when mutated to a phenyla-nine improves transduction by reducing capsid phosphorylation and subsequent ubiquitination for degradation by the proteasome.

In addition to the coding mutations observed for the AAVHSC variants arising from donors PB0130 and PB0201, several nucleo-tide changes occurred which were silent with respect to amino acid changes (Table 2). Three of these residues, 60, 119, and 193, were within the VP1u and VP1/2N regions for which there is cur-rently no structure information as stated above. Interestingly, the mutations occurring within the VP3 common region, 308, 348, 502, 505, 571, 603, and 716 (Table 2), are at the same or close to the residues where coding changes occur (Figure 4). This suggests that the capsid region containing these residues may constitute a mutational “hot spot”. For example, while a change at the 505 codon is silent for AAVHSC2 from PB0130, the residue is altered in all PB0201 variants. Significantly, the nucleotide changes at the codons for residues 464, 468, and 501 resulting in amino acid changes (as discussed above) together with the silent 502 and non-silent 505 mutations surround the AAV9 galactose binding site (Figure 4). The need for altered receptor attachment and/or tis-sue tropism may have provided selective pressure leading to the observed silent and coding changes.

Efficient in vitro transduction of CD34+ cells confirmed the ability of the AAVHSCs to negotiate receptor-mediated entry, and

postentry processing leading to transgene expression. Mechanistic differences in CD34+ cell transduction efficiencies of AAVHSCs are important to explore in the context of the limited sequence variability in the capsid genes and should help delineate capsid elements important for transduction of CD34+ cells.

All Clade F vectors tested, including the AAVHSCs and AAV9, supported long-term in vivo transduction better than other AAV serotypes, indicating the CD34-tropism of this clade. AAVHSC17, AAVHSC15, and AAVHSC7 were the most efficient at transduc-ing long-term in vivo engrafting stem cells, with AAVHSC17 supporting in vivo transduction at the highest long-term level. Based upon the dynamics of reconstitution and hematopoiesis, it is likely that systemic luciferase expression observed in xeno-transplant recipients represents the transduced circulating prog-eny of CD34+ progenitor cells as well as marrow-resident human hematopoietic cells. Interestingly, the G505R change is common between AAVHSC17, AAVHSC15, but is missing in AAVHSC7, which contained a single A68V residue change in the VP1u. This suggests that different mechanisms are involved in the improved transduction properties of these variants.

The presence of AAV vector genomes in each human hema-topoietic lineage harvested from the marrow of xenotransplanted mice, including CD34+ cells, demonstrated that AAVHSC-transduced stem/progenitor cells survive long term in vivo and give rise to transduced differentiated progeny and that transduction did not impede normal hematopoiesis. Importantly, the presence of transduced CD34+ cells in the marrow of xenograft recipients long-term, suggested that Clade F viruses are capable of transduc-ing primitive HSCs without affecting their engraftment potential or their ability to persist and/or self-renew in vivo. The presence of AAV vector genomes long term, in in vivo engrafting CD34+ cells and their dividing progeny likely represents chromosomal integra-tion events. We and others have previously demonstrated integra-tion of AAV vector sequences in CD34+ cells by fluorescent in situ hybridization, Southern blots, and sequence analysis.25,35–37 Deep sequencing studies are in progress to further evaluate the status of the long-term surviving AAV vector genomes.

In summary, endogenous AAVHSCs cloned from CD34+ cells of adult stem cell donors efficiently transduce long-term in vivo engrafting, multipotential human HSCs and support long-term transgene expression with no evidence of toxicity or pathogenic-ity. Thus the AAVHSCs constitute a family of safe, novel gene transfer vectors for potential use in human HSCs.

MATERIALS AND METHODSCell and DNA isolation. Cytokine-primed PBSC were obtained from healthy donors at the City of Hope Medical Center, and umbilical CB was obtained from Stemcyte (Covina, CA) by informed consent under IRB approved protocols. CD34+ cells were isolated from mononuclear cells by two successive rounds of immunomagnetic selection using CD34+ Indirect isolation kits (Miltenyi Biotech, Auburn, CA) to a final purity of 98–99%.

Detection of natural AAV frequency in genomic DNA. The frequency of endogenous AAV was determined by real-time qPCR of high molecular weight DNA from purified CD34+ cells using RepR1b primer and probe set.19 The single-copy human ApoB gene and endogenous AAV were each quantified by TaqMan qPCR as previously described22,24 to determine human cell equivalents and ensure integrity of the template.



Amplification, cloning, and analysis of full-length AAV capsid genes. Full-length capsid genes were amplified from PBSC genomic DNA samples determined to be positive for endogenous AAV by nested amplification. The first round was amplified using primers previously described.19 The second round was amplified using primers complementary to conserved regions of the rep and cap genes internal to the first round primers, (McapF3SpeI, 5′-ATCGATACTAGTCCATCGACGTCAGACGCGGAAG-3′ and Mcap-R1NotI, 5′ATCGATGCGGCCGCAGTTCAACTGAAACGAATCAACC GGT-3′). A high fidelity polymerase (Hot Star High Fidelity Polymerase; Qiagen, Valencia, CA), with an error rate of 1 in 300,000 bases was used for the amplifications. Correctly sized capsid genomes were excised and purified using QIAquick Gel Extraction Kit (Qiagen) prior to sequenc-ing. Amplified capsids genes were cloned into pBluescript SK. Packaging plasmids were cloned by amplifying nucleotides 151 to 1890 of AAV2 Rep from pTZAAV38 using primers AAV2RepF, 5′-GATCATATCGATGG TGGAGTCGTGACGTGAATTACG-3′ and AAV2RepR 5′-GATCATAA GCTTCCGCGTCTGACGTCGATGG-3′. Rep amplicons were cloned into each plasmid containing a full-length novel HSC capsid clone. The result-ing packaging plasmids were then sequenced using degenerative primers to confirm identity and sequence integrity. Phylogenetic tree analysis of the sequences was done with the CLC Main Workbench 6.8.2 (CLC Bio, Cambridge, MA) using the neighbor joining algorithm.

rAAV production, purification, and titration. Self-complementary enhanced green fluorescent protein (scEGFP) or single stranded firefly luciferase (ssLuc) genes were packaged into the novel AAVHSC capsids in herpes simplex virus-infected 293 cells as previously described.38 The resulting recombinant AAV vectors were purified through two rounds of CsCl2 density centrifugation gradients and titers were determined using qPCR with transgene-specific primers and probes.22

AAV transductions. CD34+ were transduced with the novel rAAV vari-ants as previously described for other AAVs.24,25,27 Briefly, purified cord blood CD34+ cells from up to five samples were pooled and transduced with rAAV at a multiplicity of infection of 20,000 particles per cell in Iscove’s Modified Dulbecco’s Medium (IMDM) (Irvine Scientific, Santa Ana, CA) containing 20% FCS, 100 ug/ml streptomycin, 100 U/ml peni-cillin, 2mmol/l l-glutamine, IL-3 (10 ng/ml; R&D Systems, Minneapolis, MN), IL-6 (10 ng/ml; R&D Systems), and stem cell factor (1 ng/ml; R&D Systems) and incubated overnight in humidified CO2 at 37 °C. To deter-mine EGFP expression, cells were harvested at 24 hours posttransduction, washed, and analyzed by flow cytometry. For in vivo assays, cells trans-duced with AAV-ssLuc were harvested at 24 hours posttransduction, and washed three times prior to transplantation into NOD/SCID mice as previ-ously described.22,24

Xenotransplantation of CD34+ cells. All animal care and experiments were performed under protocols approved by the Institutional Animal Care and Use Committee, City of Hope. Transplantations were performed as previ-ously described.22,24 Briefly, 6–8-week old male NOD.CB17-Prkdcscid/NCrCrl (NOD/SCID) mice were maintained in a specific pathogen free facility at the Animal Resources Center, Beckman Research Institute of the City of Hope. Mice were placed on sulfamethoxazole and trimethoprim oral pedi-atric antibiotic (Hi-Tech Pharmacal (Amityville, NY), 10ml/500ml H2O) for at least 48 hours before transplant. Mice were irradiated with a sublethal dose of 350cGy from a 137Cs source and allowed to recover for a minimum of 4 hours prior to transplantation. For the transplants, 7 × 105–1.1 × 106 transduced CD34+ cells were infused via the tail vein in a total volume of 150–200 µl. About 3–5 mice were transplanted per vector group. Bone mar-row from femurs and tibiae were harvested for analysis from each mouse at 14–24 weeks posttransplantation.

Serial bioluminescent analysis of luciferase expression. Luciferase expres-sion in xenografted mice was monitored by serial biweekly bioluminescent imaging using a Xenogen In vivo Imaging System (Caliper Life Sciences,

Hopkinton, MA). Mice were anesthetized with oxygen containing 4% isoflu-rane (Phoenix Pharmaceuticals, Burlingame, CA) for induction, and 2.5% for maintenance. Luciferin (Caliper Life Sciences) was injected intraperito-neally at a dose of 0.15 mg/g of mouse weight. Photons were accumulated over a 5-minute exposure from the ventral aspect, 10 minutes postinjection. Living Image 3.0 software (Caliper Life Sciences) was used to calculate light emission.

Flow cytometric analysis. In vitro expression was analyzed 24 hours after rAAV-EGFP transduction on a Cyan ADP Flow Cytometer (Dako, Carpinteria, CA). Specific EGFP was quantified following the subtrac-tion of autofluorescence. In vivo engraftment of human cells in both the bone marrow and spleen of xenografted mice was analyzed as described previously.22 Lineage distribution was assessed in bone marrow and spleen cell suspensions following staining with human specific antibodies: FITC-conjugated anti-CD45, fluorescein isothiocyanate (FITC)- or allo-phycocyanin (APC)-conjugated anti-CD34, APC-conjugated anti-CD33 and anti-CD14, anti-Glycophorin A, phycoerythrin (PE)-conjugated anti-CD19, and FITC-, PE-, and APC-conjugated IgG controls (Becton Dickinson, Mountain View, CA). Bone marrow lineages were sorted by fluorescence-activated cell sorting (FACS) (Becton Dickinson) using FITC-CD34, APC-CD33, PE-CD19, and Glycophorin A-APC, as well as the appropriate controls. FITC and PE fluorescence was excited by a 488 nm laser, and APC fluorescence was excited by a 670 nm laser. Flow cytometry data was then analyzed for specific populations with FlowJo software (Treestar, Ashland, OR).

AAV frequency detection. High molecular weight DNA was extracted from flow sorted human lineages isolated from the xenograft marrow using standard methods. Vector-specific sequences were amplified by qPCR analysis using vector-specific primers and probe on a 7900HT Sequence Detection System (Applied Biosystems, Foster City, CA) as previously described.22,24 Samples were also evaluated for the single-copy human gene ApoB, which served to quantify human cell equivalents and as a template integrity control.22,24

ACKNOWLEDGMENTSThis research was supported in part by NIH grants R01HL087285 (to S. Chatterjee) and R01GM082946 and P01HL51811 (to M. Agbandje-McKenna). Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under grant number P30CA033572. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. City of Hope core facilities utilized for this study include Small Animal Imaging, Analytical Cytometry and DNA Sequencing Cores. We thank John A. Zaia and Steven J. Forman for sup-port; Manasa Chandra, for thoughtful discussions; Supriya Bautista, for assistance with manuscript preparation. The authors declare no conflicts of interest.

REFERENCES 1. Daya, S and Berns, KI (2008). Gene therapy using adeno-associated virus vectors. Clin

Microbiol Rev 21: 583–593. 2. Asokan, A, Schaffer, DV and Samulski, RJ (2012). The AAV vector toolkit: poised at the

clinical crossroads. Mol Ther 20: 699–708. 3. Pacak, CA and Byrne, BJ (2011). AAV vectors for cardiac gene transfer: experimental

tools and clinical opportunities. Mol Ther 19: 1582–1590. 4. High, KA (2012). The gene therapy journey for hemophilia: are we there yet? Blood

120: 4482–4487. 5. Hajjar, RJ (2013). Potential of gene therapy as a treatment for heart failure. J Clin Invest

123: 53–61. 6. MacLaren, RE, Groppe, M, Barnard, AR, Cottriall, CL, Tolmachova, T, Seymour, L et

al. (2014). Retinal gene therapy in patients with choroideremia: initial findings from a phase ½ clinical trial. Lancet 383: 1129–1137.

7. Muramatsu, S, Fujimoto, K, Kato, S, Mizukami, H, Asari, S, Ikeguchi, K et al. (2010). A phase I study of aromatic L-amino acid decarboxylase gene therapy for Parkinson’s disease. Mol Ther 18: 1731–1735.

8. Nathwani, AC, Tuddenham, EG, Rangarajan, S, Rosales, C, McIntosh, J, Linch, DC et al. (2011). Adenovirus-associated virus vector-mediated gene transfer in hemophilia B. N Engl J Med 365: 2357–2365.

9. Haddley, K (2013). Alipogene tiparvovec for the treatment of lipoprotein lipase deficiency. Drugs Today (Barc) 49: 161–170.



10. Shen, S, Bryant, KD, Brown, SM, Randell, SH and Asokan, A (2011). Terminal N-linked galactose is the primary receptor for adeno-associated virus 9. J Biol Chem 286: 13532–13540.

11. Bell, CL, Vandenberghe, LH, Bell, P, Limberis, MP, Gao, GP, Van Vliet, K et al. (2011). The AAV9 receptor and its modification to improve in vivo lung gene transfer in mice. J Clin Invest 121: 2427–2435.

12. Asokan, A, Hamra, JB, Govindasamy, L, Agbandje-McKenna, M and Samulski, RJ (2006). Adeno-associated virus type 2 contains an integrin alpha5beta1 binding domain essential for viral cell entry. J Virol 80: 8961–8969.

13. Summerford, C and Samulski, RJ (1998). Membrane-associated heparan sulfate proteoglycan is a receptor for adeno-associated virus type 2 virions. J Virol 72: 1438–1445.

14. Qing, K, Mah, C, Hansen, J, Zhou, S, Dwarki, V and Srivastava, A (1999). Human fibroblast growth factor receptor 1 is a co-receptor for infection by adeno-associated virus 2. Nat Med 5: 71–77.

15. Akache, B, Grimm, D, Pandey, K, Yant, SR, Xu, H and Kay, MA (2006). The 37/67-kilodalton laminin receptor is a receptor for adeno-associated virus serotypes 8, 2, 3, and 9. J Virol 80: 9831–9836.

16. Weller, ML, Amornphimoltham, P, Schmidt, M, Wilson, PA, Gutkind, JS and Chiorini, JA (2010). Epidermal growth factor receptor is a co-receptor for adeno-associated virus serotype 6. Nat Med 16: 662–664.

17. Gao, G, Vandenberghe, LH, Alvira, MR, Lu, Y, Calcedo, R, Zhou, X et al. (2004). Clades of Adeno-associated viruses are widely disseminated in human tissues. J Virol 78: 6381–6388.

18. Gao, GP, Alvira, MR, Wang, L, Calcedo, R, Johnston, J and Wilson, JM (2002). Novel adeno-associated viruses from rhesus monkeys as vectors for human gene therapy. Proc Natl Acad Sci U S A 99: 11854–11859.

19. Gao, G, Alvira, MR, Somanathan, S, Lu, Y, Vandenberghe, LH, Rux, JJ et al. (2003). Adeno-associated viruses undergo substantial evolution in primates during natural infections. Proc Natl Acad Sci U S A 100: 6081–6086.

20. Wang, L, Calcedo, R, Bell, P, Lin, J, Grant, RL, Siegel, DL et al. (2011). Impact of pre-existing immunity on gene transfer to nonhuman primate liver with adeno-associated virus 8 vectors. Hum Gene Ther 22: 1389–1401.

21. Calcedo, R, Vandenberghe, LH, Gao, G, Lin, J and Wilson, JM (2009). Worldwide epidemiology of neutralizing antibodies to adeno-associated viruses. J Infect Dis 199: 381–390.

22. Santat, L, Paz, H, Wong, C, Li, L, Macer, J, Forman, S et al. (2005). Recombinant AAV2 transduction of primitive human hematopoietic stem cells capable of serial engraftment in immune-deficient mice. Proc Natl Acad Sci USA 102: 11053–11058.

23. Paz, H, Wong, CA, Li, W, Santat, L, Wong, KK and Chatterjee, S (2007). Quiescent subpopulations of human CD34-positive hematopoietic stem cells are preferred targets for stable recombinant adeno-associated virus type 2 transduction. Hum Gene Ther 18: 614–626.

24. Kauss, MA, Smith, LJ, Zhong, L, Srivastava, A, Wong, KK Jr and Chatterjee, S (2010). Enhanced long-term transduction and multilineage engraftment of human hematopoietic stem cells transduced with tyrosine-modified recombinant adeno-associated virus serotype 2. Hum Gene Ther 21: 1129–1136.

25. Chatterjee, S, Li, W, Wong, CA, Fisher-Adams, G, Lu, D, Guha, M et al. (1999). Transduction of primitive human marrow and cord blood-derived hematopoietic progenitor cells with adeno-associated virus vectors. Blood 93: 1882–1894.

26. Song, L, Li, X, Jayandharan, GR, Wang, Y, Aslanidi, GV, Ling, C et al. (2013). High-efficiency transduction of primary human hematopoietic stem cells and erythroid lineage-restricted expression by optimized AAV6 serotype vectors in vitro and in a murine xenograft model in vivo. PLoS One 8: e58757.

27. Song, L, Kauss, MA, Kopin, E, Chandra, M, Ul-Hasan, T, Miller, E et al. (2013). Optimizing the transduction efficiency of capsid-modified AAV6 serotype vectors in primary human hematopoietic stem cells in vitro and in a xenograft mouse model in vivo. Cytotherapy 15: 986–998.

28. Schuhmann, NK, Pozzoli, O, Sallach, J, Huber, A, Avitabile, D, Perabo, L et al. (2010). Gene transfer into human cord blood-derived CD34(+) cells by adeno-associated viral vectors. Exp Hematol 38: 707–717.

29. Veldwijk, MR, Sellner, L, Stiefelhagen, M, Kleinschmidt, JA, Laufs, S, Topaly, J et al. (2010). Pseudotyped recombinant adeno-associated viral vectors mediate efficient gene transfer into primary human CD34(+) peripheral blood progenitor cells. Cytotherapy 12: 107–112.

30. Ponnazhagan, S, Mukherjee, P, Wang, XS, Qing, K, Kube, DM, Mah, C et al. (1997). Adeno-associated virus type 2-mediated transduction in primary human bone marrow-derived CD34+ hematopoietic progenitor cells: donor variation and correlation of transgene expression with cellular differentiation. J Virol 71: 8262–8267.

31. Popa-Wagner, R, Porwal, M, Kann, M, Reuss, M, Weimer, M, Florin, L et al. (2012). Impact of VP1-specific protein sequence motifs on adeno-associated virus type 2 intracellular trafficking and nuclear entry. J Virol 86: 9163–9174.

32. Ng, R, Govindasamy, L, Gurda, BL, McKenna, R, Kozyreva, OG, Samulski, RJ et al. (2010). Structural characterization of the dual glycan binding adeno-associated virus serotype 6. J Virol 84: 12945–12957.

33. Naumer, M, Sonntag, F, Schmidt, K, Nieto, K, Panke, C, Davey, NE et al. (2012). Properties of the adeno-associated virus assembly-activating protein. J Virol 86: 13038–13048.

34. Ni, TH, McDonald, WF, Zolotukhin, I, Melendy, T, Waga, S, Stillman, B et al. (1998). Cellular proteins required for adeno-associated virus DNA replication in the absence of adenovirus coinfection. J Virol 72: 2777–2787.

35. Fisher-Adams, G, Wong, KK Jr, Podsakoff, G, Forman, SJ and Chatterjee, S (1996). Integration of adeno-associated virus vectors in CD34+ human hematopoietic progenitor cells after transduction. Blood 88: 492–504.

36. Han, Z, Zhong, L, Maina, N, Hu, Z, Li, X, Chouthai, NS et al. (2008). Stable integration of recombinant adeno-associated virus vector genomes after transduction of murine hematopoietic stem cells. Hum Gene Ther 19: 267–278.

37. Maina, N, Han, Z, Li, X, Hu, Z, Zhong, L, Bischof, D et al. (2008). Recombinant self-complementary adeno-associated virus serotype vector-mediated hematopoietic stem cell transduction and lineage-restricted, long-term transgene expression in a murine serial bone marrow transplantation model. Hum Gene Ther 19: 376–383.

38. Chatterjee, S, Johnson, PR and Wong, KK Jr (1992). Dual-target inhibition of HIV-1 in vitro by means of an adeno-associated virus antisense vector. Science 258: 1485–1488.

39. Xiao, C and Rossmann, MG (2007). Interpretation of electron density with stereographic roadmap projections. J Struct Biol 158: 182–187.


gene transfer properties and structural modeling of human stem cell-derived aav

Documents