Download Scalable Recombinant Adeno-Associated Virus Production

HUMAN GENE THERAPY 20:861–870 (August 2009)
ª Mary Ann Liebert, Inc.
DOI: 10.1089=hum.2009.004
Scientific Article
Scalable Recombinant Adeno-Associated Virus Production
Using Recombinant Herpes Simplex Virus Type 1
Coinfection of Suspension-Adapted Mammalian Cells
Darby L. Thomas,1 Lijun Wang,1 Justine Niamke,1,2 Jilin Liu,1 Wen Kang,1 Marina M. Scotti,1
Guo-jie Ye,1 Gabor Veres,1 and David R. Knop1
Abstract
Recombinant adeno-associated virus (rAAV) production systems capable of meeting clinical or anticipated
commercial-scale manufacturing needs have received relatively little scrutiny compared with the intense research
activity afforded the in vivo and in vitro evaluation of rAAV for gene transfer. Previously we have reported a highly
efficient recombinant herpes simplex virus type 1 (rHSV) complementation system for rAAV production in
multiple adherent cell lines; however, production in a scalable format was not demonstrated. Here we report
rAAV production by rHSV coinfection of baby hamster kidney (BHK) cells grown in suspension (sBHK cells),
using two ICP27-deficient rHSV vectors, one harboring a transgene flanked by the AAV2 inverted terminal repeats
and a second bearing the AAV rep2 and capX genes (where X is any rAAV serotype). The rHSV coinfection of sBHK
cells produced similar rAAV1=AAT-specific yields (85,400 DNase-resistant particles [DRP]=cell) compared with
coinfection of adherent HEK-293 cells (74,600 DRP=cell); however, sBHK cells permitted a 3-fold reduction in
the rHSV-rep2=capX vector multiplicity of infection, grew faster than HEK-293 cells, retained specific yields
(DRP=cell) at higher cell densities, and had a decreased virus production cycle. Furthermore, sBHK cells were able
to produce AAV serotypes 1, 2, 5, and 8 at similar specific yields, using multiple therapeutic genes. rAAV1=AAT
production in sBHK cells was scaled to 10-liter disposable bioreactors, using optimized spinner flask infection
conditions, and resulted in average volumetric productivities as high as 2.41014 DRP=liter.
Introduction
R
ecombinant adeno-associated viral (rAAV) vectors
continue to hold immense promise as gene transfer vehicles for a variety of gene therapy applications. Numerous
preclinical and human clinical studies have been undertaken
with rAAV, employing several of the identified serotypes
to leverage their differing tissue tropism to correct a broad
spectrum of diseases (Aucoin et al., 2008). AAV is a dependovirus whose full life cycle is complemented by a helper virus,
most often adenovirus or herpes simplex virus (Muzyczka and
Berns, 2001). In addition, AAV wild-type infections lack pathological effects, conferring an added measure of safety to
rAAV use relative to other potential gene therapy viral vectors (Hermonat and Muzyczka, 1984; Tratschin et al., 1984;
Muzyczka and Berns, 2001).
The production of rAAV can be accomplished by a variety
of methods, whether in suspension or in adherent cell lines.
The predominant method involves transient transfection of
mammalian cells with multiple plasmids. Historically, three
plasmids are employed to deliver all the necessary elements
for rAAV production: one plasmid contains the gene of interest flanked by the AAV inverted terminal repeats (ITRs),
the minimal cis-acting sequences necessary for AAV replication and packaging; the second plasmid expresses the rep
(Rep78, 68, 52, and 40 proteins) and cap (structural proteins
VP1, VP2, and VP3 from any AAV serotype) genes; and the
third plasmid expresses the adenovirus helper functions
(E1a, E1b, E2a, E4orf6, and virus-associated RNA I genes).
Multiple variations of the transient transfection method have
been pursued, including plasmid transfection to deliver the
AAV genes and the gene of interest followed by adenovirus
infection, and infection=transfection of stable cell lines expressing either rep and cap or harboring the ITR-flanked gene
of interest, and have been reviewed (Buning et al., 2008).
Despite the advantageous characteristics of rAAV and the
extensive research into preclinical applications, production
scale-up continues to limit rAAV use in large clinical trials
1
Applied Genetic Technologies Corporation (AGTC), Alachua, FL 32615.
Present address: 65 West Watkins Mill Road, Gaithersburg, MD 20878.
2
861
862
that require even moderate doses of vector. Suspension cell
culture can facilitate scale-up of viral vector production;
however, reports of rAAV production on mammalian cells
adapted to grow in suspension have been limited to transient
transfection of HEK-293 cells (Park et al., 2006; Durocher et al.,
2007; Feng et al., 2007; Hildinger et al., 2007) and amplification of an rAAV carrying the gene of interest on A549 packaging cells that expresses rep and cap genes by coinfection of
the rAAV with adenovirus (Farson et al., 2004). In contrast,
extensive work has been described using baculovirus infection of insect cells for rAAV production. Baculovirus-based
suspension systems employ simultaneous infection of insect
cells with three or four baculoviral expression vectors (BEVs)
to deliver the ITR-flanked gene of interest and the AAV rep
and cap genes (Urabe et al., 2002). Stability of BEV rAAV
production systems during serial expansion of helper constructs continues to hamper its clinical applicability, despite
the finding that the multiplicity of infection (MOI) input can
be significantly reduced (Kohlbrenner et al., 2005; Negrete
et al., 2007; Cecchini et al., 2008).
The herpes simplex virus (HSV) coinfection method of
rAAV production is an alternative to the baculovirus system,
transient transfection, and adenovirus infection. HSV has been
widely used to complement rAAV replication and packaging,
with the HSV helicase–primase complex (UL5, UL8, and UL52),
DNA polymerase, and DNA-binding protein (UL29) genes
providing the essential AAV helper functions (Buller et al.,
1981; Weindler and Heilbronn, 1991; Alazard-Dany et al., 2009).
Although wild-type HSV and several recombinant deletion
mutants of HSV have been examined for rAAV production,
ICP27-deleted HSV vectors have demonstrated the most
promise regarding specific yields in mammalian cells (Conway
et al., 1997, 1999; Zhang et al., 1999; Feudner et al., 2001; Mistry
et al., 2002; Wustner et al., 2002; Booth et al., 2004; Toublanc et al.,
2004; Alazard-Dany et al., 2009). We have previously reported a
highly efficient rHSV-based rAAV complementation system
that uses two rHSV vectors: one harboring the ITR-flanked
gene of interest and the second bearing the rep and cap genes
(Kang et al., 2009). The helper vectors were demonstrated to be
both stable during serial helper expansion (Kang et al., 2009)
and scalable to 3.5-liter packed-bed bioreactors (Knop and
Harrell, 2007). The rHSV-based rAAV production platform
used adherent HEK-293 cells infected at about 1106 viable
cells (VC)=ml in 10-stack Nunclon D cell factories (Thermo
Fisher Scientific, Waltham, MA) at MOIs of 12 and 2 for rHSVrep2=capX and rHSV-GOI (with GOI representing any transgene of interest), respectively, and vector was harvested at
52 hr postinfection (Kang et al., 2009). Here we present significant improvements to rAAV production by rHSV coinfection
by employing baby hamster kidney cells adapted to grow in
suspension (sBHK). This platform provides ease of scale-up
due to suspension cell culture, abbreviated cell expansion and
vector production cycles due to faster cellular growth rates and
metabolism, an increase in volumetric productivity, and a decrease in total rHSV input by >50% during coinfection.
Materials and Methods
Cell culture, herpes simplex virus production,
and infectious titer determination
V27, HEK-293, and C12 cells were obtained from the
Powell Gene Therapy Center of the University of Florida
THOMAS ET AL.
(Gainesville, FL). Suspension baby hamster kidney (sBHK)
cells were purchased from the European Collection of Cell
Cultures (ECACC, Porton Down, UK) and HeLaRC32 cells
were purchased from the American Type Culture Collection
(ATCC, Manassas, VA). All cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; HyClone, South
Logan, UT) containing 10% fetal bovine serum (FBS; HyClone) and either 1% (v=v) penicillin–streptomycin (Cellgro=
Mediatech, Manassas, VA) for sBHK, HEK-293, and HeLaRC32 culture or Geneticin (50 mg=ml; Invitrogen) for V27 and
C12 cells.
The rHSV-rep2=capX (with X representing any AAV capsid
serotype) and rHSV-GOI vectors were constructed as described previously (Kang et al., 2009). The rHSV vectors were
propagated on the ICP27-complementing V27 cell line (Rice
and Knipe, 1990), in packed-bed bioreactors or cell factories as
previously described (Knop and Harrell, 2007), and infectious
vector was harvested 72 hr postinfection by recovering culture supernatant and cell-associated vector. The resulting
process fluids were combined and clarified by depth filtration
(1.2 mm) and absolute filtration (0.8=0.45 mm), and concentrated by hollow fiber filter (HFF) tangential flow filtration
(TFF). The infectious titers (plaque-forming units [PFU]=ml)
of rHSV vector stocks were determined by a modified plaque
assay. Briefly, V27 cells in 24-well plates were infected with
serial dilutions of rHSV. The inoculum was exchanged with
culture medium containing 0.2% human g-globulins (SigmaAldrich, St. Louis, MO) at 6 hr postinfection. At 3 days postinfection, the culture medium was aspirated and the cells
were fixed with ice-cold methanol. The cells were then incubated with a horseradish peroxidase (HRP)-labeled anti-HSV
antibody (Dako, Glostrup, Denmark) and rHSV plaques were
visualized by Vector VIP staining (Vector Laboratories, Burlingame, CA) as per the manufacturer’s instructions.
rHSV coinfection method
Mammalian cells were simultaneously coinfected with
both rHSV vectors. At 2–4 hr postinfection infectious medium
was exchanged with serum-free DMEM equivalent to the
preinfection culture volume. At the time of harvest, MgCl2
was added to a final concentration of 2 mM, 10% (v=v) Triton
X-100 (octyl phenol ethoxylate; J.T. Baker, Phillipsburg, NJ)
was added to a final concentration of 1% (v=v), and Benzonase
(EMD Chemicals, Gibbstown, NJ) was added to a final concentration of 25–50 U=ml. This mixture was stirred (25-ml
spinner flasks) or rocked (disposable bioreactors) at 378C for
2–4 hr. Before processing, 5 M NaCl was added to a final
concentration of 1.0 M and the mixture was frozen at 808C
(25-ml spinner flasks) or clarified (disposable bioreactors).
Assays
DNase-resistant particles (DRP) were quantified by realtime quantitative polymerase chain reaction (qPCR) in an
iCycler iQ 96-well block format thermocycler (Bio-Rad, Hercules, CA). Crude samples were incubated in the presence of
DNase I (100 U=ml; Promega, Madison, WI) at 378C for
60 min, followed by proteinase K (Invitrogen, Carlsbad, CA)
digestion (10 U=ml) at 508C for 60 min, and then denatured
at 958C for 30 min. Linearized plasmid pDC-67=þSV402 was
used to generate standard curves. The primer–probe set
SUSPENSION rAAV PRODUCTION USING rHSV COINFECTION
was specific for the simian virus 40 (SV40) poly(A) sequence
(rAAV-F, 5 0 -AGCAATAGCATCACAAATTTCACAA-3 0 ;
rAAV-R, 50 -CCAGACATGATAAGATACATTGATGAGTT-30 ;
rAAV-Pr, 50 -6-FAM-AGCATTTTTTTCACTGCATTCTAGTT
GTGGTTTGTC-TAMRA-30 ). Amplification of the PCR product was achieved with the following cycling parameters: 1
cycle at 508C for 2 min, 1 cycle at 958C for 10 min; 40 cycles of
958C for 15 sec, and 608C for 60 sec.
Infectious particle (ip) titering was performed on rAAVGFP stocks, using a green cell assay. C12 cells (Clark et al.,
1995, 1996) were infected with serial dilutions of rAAV-GFP
plus saturating concentrations of human adenovirus type 5
(reference material, VR-1516; ATCC) to provide AAV replication helper functions. At 48–72 hr postinfection, the number of fluorescing green cells (each cell representing one
infectious event) was counted and used to calculate the titer
(ip=ml) of the virus sample.
Infectivity of rAAV particles harboring a gene of interest
(rAAV-GOI) was determined in a TCID50 (tissue culture infectious dose at 50%) assay. Eight replicates of rAAV were
serially diluted in the presence of human adenovirus type 5
and used to infect HeLaRC32 cells (Salvetti et al., 1998) in a 96well plate. Three days postinfection, lysis buffer (final concentrations of 1 mM Tris-HCl [pH 8.0], 1 mM EDTA, 0.25%
[w=v] deoxycholate, 0.45% [v=v] Tween 20, 0.1% [w=v] sodium dodecyl sulfate, and proteinase K at 0.3 mg=ml [Promega, Madison, WI]) was added to each well then incubated
at 378C for 1 hr, 558C for 2 hr, and 958C for 30 min. The lysate
from each well (2.5-ml aliquot) was assayed in the DRP qPCR
assay described previously. Wells with Ct values lower than
the value of the lowest quantity of plasmid of the standard
curve were scored as positive. TCID50 infectivity per milliliter (TCID50=ml) was calculated with the Kärber equation,
using the ratios of positive wells at 10-fold serial dilutions.
rHSV-specific assays for quantifying DNA and protein were
executed as previously described and human a1-antitrypsin
expression in C57BL=6 mice sera was quantified by enzymelinked immunosorbent assay (ELISA), as previously described
(Kang et al., 2009).
Enumeration of viable cells (VC) and nonviable cells was
performed with a Guava EasyCyte mini system (Guava
Technologies=Millipore, Hayward, CA) using the Guava
ViaCount assay and reagents in accordance with the manufacturer’s instructions.
Bioreactor production of rAAV
Disposable bioreactors were controlled with a Sartorius
BIOSTAT CultiBag RM 20=50 basic control unit and platform
(Sartorius Stedim Systems, Springfield, MO). Cultures were
seeded at 2.5–3.5105 VC=ml into disposable bioreactors in
DMEM–10% (v=v) FBS, and coinfected with rHSV 24–48 hr
postseeding at densities of 1.4–2.2106 VC=ml. Scaled-up
runs (5 or 10 liters) were initially seeded at 2 or 4 liters, diluted
to 5 or 10 liters at 24 hr postseeding, respectively, and infected
24 hr later. Culture pH was maintained at 7.2 by headspace
CO2 addition, dissolved oxygen (DO) was monitored until
< 50% saturation, and then headspace oxygen enrichment
was used to maintain the DO at 50%. Agitation was maintained at 20 rocks=min at an angle of 6.88 (5 or 10 liters) or 7.28
(1 liter). Temperature was maintained at 378C. A total headspace gas flow rate of 0.1 volume gas=volume culture=min
863
(vvm) was used. Data acquisition was accomplished with the
Sartorius MFCS=DA software package.
Medium exchange at 2–4 hr postinfection was accomplished by centrifugation (640g, 4 min), with an integrated
perfusion membrane (100 ml=min for volume reduction=
perfusion), or via a Sartorius Sartocon Slice 1000 flat
sheet TFF holder (Sartorius Stedim Systems) and 0.2- and
0.45-mm nominal pore size Hydrosart membranes (cat. nos.
3051860701W-SG and 3051860601O-SG, respectively). Flat
sheet TFF employed recirculation rates of 350–500 ml=min
and the transmembrane pressure (TMP) was maintained at
approximately 0.5 psig. Volume was reduced 5- to 10-fold and
culture was diluted back to the preinfection volume with
DMEM lacking FBS.
rAAV vector purification and in vivo evaluation
Crude lysate from disposable bioreactors was clarified by
depth filtration (1.2 mm) and absolute filtration (0.8=0.45 mm)
and stored for 18–96 hr. Clarified lysate was concentrated
and buffer was exchanged with either a Sartorius Sartocon
Slice 1000 flat sheet holder using 100-kDa MWCO Hydrosart
membranes (cat. no. 3051446801E-SG) or GE Healthcare Life
Sciences (Piscataway, NJ) 300-kDa MWCO hollow fiber filter
(HFF) devices (cat. nos. UFP-300-C-5A and UFP-300-C-9A).
A recirculation rate of 0.5–1.5 liters=min was used and the
TMP was maintained at 8.5 1.5 psig.
Concentrated, buffer-exchanged rAAV lysate was purified
by two-column chromatography. Capture chromatography
employed an anion-exchange column. The anion-exchange
column peak was loaded directly onto an affinity column
and eluted with a pH gradient. The affinity column eluent
was neutralized to pH 7.0–8.0 and buffer was exchanged to
lactated Ringer’s solution, using a 300-kDa MWCO hollow
fiber cartridge (cat. no. UFP-300-E-MM01A; GE Healthcare
Life Sciences) for TFF and diafiltration (DF).
Animal experiments with C57BL=6 mice were conducted
as previously described (Kang et al., 2009). Purified rAAV1=
AAT was administered to C57BL=6 mice (11011 DRP=mouse)
and serum samples were collected from mice 1 month postinjection as previously described (Kang et al., 2009).
Results
BHK cell line and suspension growth
Several cell lines have been widely used for rAAV manufacture, most notably HEK-293 and HeLa cell lines. Despite
receiving comparably little use, baby hamster kidney cells
(clone 21) have been employed to produce rAAV via rHSV
vector coinfection (Booth et al., 2004; Kang et al., 2009). The
lineage and safety of mammalian-derived cell lines are important for regulatory filings in support of biotherapeutic
manufacturing methods for clinical trials. BHK cells have
been shown to generate (Albu and Holmes, 1973) and release
(Bergmann and Wolff, 1981) endogenously produced viruslike particles (VLPs); however, the BHK cell line has been
used both in a Huntington’s disease phase 1 clinical trial to
deliver ciliary neurotrophic factor (CNTF) (Bloch et al., 2004)
and for the manufacture of recombinant factor VIII to treat
hemophilia A. The latter is a U.S. Food and Drug Administration (FDA)-licensed, commercial process (Boedeker, 2001),
validating the use of BHK cell-based manufacturing as a
864
A
THOMAS ET AL.
UL
US
RL RS
RL
Stu I
BamH I
UL22
UL23 (TK)
UL24
RS
UL54
(ICP27)
UL53
UL55
p40 p19 p5
B
UL22
TK
rep2
capX
TK
∆UL54
(ICP27)
XbaI
XbaI
UL55
UL53
rHSV-rep2/capX where X = 1, 2, 5, 8
or
C
UL22
TK
hGFP
neo
R
KpnI
TK
rHSV-GFP
KpnI
PCMV
SV-40
polyA
intron
hAAT
FIG. 1. The relevant genetic
sequences of rHSV-rep2=capX
and rHSV-GOI used for rAAV
production. (A) Structure of
the ICP27-negative HSV parental vector, highlighting the
ICP27 deletion (between BamHI and StuI) and the thymidine kinase (tk) gene. (B) The
rep2=capX cassette insertion
into the tk gene, where capX
can be the capsid gene from
AAV1, 2, 5, or 8. (C) The
AAV2 ITR-GOI cassette insertion into the tk gene. The
structures of the GFP and GOI
(either AAT or sFLT01) expression cassettes are shown.
or
UL22
TK
KpnI
ECMV
PCBA
SV-40
polyA
TK
rHSV-GOI where
GOI = AAT or sFLT01
KpnI
viable method for the production of human therapeutics.
Therefore, BHK cells adapted to grow in suspension (sBHK)
were examined as an rAAV production platform because of
the high specific yield of their adherent counterpart and their
favorable regulatory profile. sBHK cells were cultured in
spinner flasks (80–120 rpm) in DMEM–FBS in water-jacketed
incubators (5% CO2, 378C) and initially maintained between
cell densities of 2105 and 1106 VC=ml. The sBHK cells had
an average doubling time of 12 2 hr (n ¼ 33) under these
conditions, in contrast to adherent HEK-293 cells, which had
a doubling time of about 24 hr in T-flasks.
rAAV production optimization: Harvest timing,
cell density, and MOI
rAAV production by rHSV coinfection was initially optimized in 25-ml spinner flask cultures by examining rAAV
harvest timing after coinfection, infection cell density, and
infection MOI. The rHSV constructs used for coinfection
production of rAAV on sBHK cells are shown in Fig. 1. sBHK
cells in spinner flasks were coinfected with rHSV-rep2=cap2
(MOI of 12) and rHSV-GFP (MOI of 2). Samples were taken at
the indicated intervals and the infectious titer was determined. Across the three time points examined, rAAV2=GFP
specific yields of 6500–7500 ip=cell were obtained (Fig. 2A),
which were similar to previous results with adherent HEK293 and BHK cells. On the basis of these results, the harvest
of rHSV-coinfected sBHK cells could be executed as early as
24 hr postinfection (Fig. 2A), or about two preinfection doubling times for the sBHK cells.
We next examined the effect of cell density at the time of
infection on rAAV vector production during rHSV coinfection
of sBHK cells. sBHK cells were grown in 25-ml spinner cultures to the densities indicated in Fig. 2B and coinfected
with rHSV-rep2=cap2 (MOI of 12) and rHSV-GFP (MOI of 2).
The cultures were harvested 24 hr postinfection and processed by in situ lysis as described in Material and Methods.
The rAAV2=GFP specific yields were insensitive to infection
densities between 1.6106 and 3.8106 VC=ml (Fig. 2B), permitting 2- to 3-fold increases in volumetric productivity by
increasing the cell density at the time of infection.
Finally, the input MOI for rHSV vector coinfection production of rAAV was examined. The rHSV-rep2=capX MOI of
12 used for HEK-293 cells required potentially prohibitive
quantities of rHSV for optimal rAAV production, making
rHSV MOI reduction a principal consideration in scale-up
(Kang et al., 2009). Here, we did simultaneous coinfection of
adherent HEK-293 cells in 75-cm2 flasks and sBHK cells in 25ml spinner flask cultures at various MOI ratios, using rHSVrep2=cap1 and rHSV-AAT vectors to produce rAAV1=AAT.
The rHSV-rep2=cap1 MOI was 12, 8, or 4, whereas the rHSVAAT MOI was held constant at 2. As expected, adherent HEK293 cell rAAV1=AAT production was sensitive to the
rHSV-rep2=cap1 MOI, decreasing from 74,600 to 42,100 DRP=
cell with an MOI decrease from 12 to 4 (Fig. 3). By contrast,
rAAV1=AAT production in sBHK cells was insensitive to
SUSPENSION rAAV PRODUCTION USING rHSV COINFECTION
9,000
rAAV1/AAT production (DRP/cell)
rAAV2/GFP production (ip/cell)
A
8,000
7,000
6,000
5,000
4,000
3,000
865
100,000
80,000
60,000
40,000
20,000
0
12
2,000
24
48
69
time of harvest (hpi)
rAAV2/GFP production (ip/cell)
B
8
4
2
1
MOI of rHSV-rep2/cap1
FIG. 3. rAAV1=AAT production as a function of rHSVrep2=cap1 vector MOI. The rHSV-rep2=cap1 MOI was varied
(12, 8, or 4) while holding the rHSV-AAT vector MOI constant
at 2 in 75-cm2 flask cultures of HEK-293 cells (1107 cells,
open columns) and 25-ml spinner flask cultures of sBHK cells
(1106 VC=ml, shaded columns). Error bars indicate standard
deviation; n ¼ 3 for spinner flask sBHK rAAV production and
n ¼ 4 for 75-cm2 flask HEK-293 rAAV production.
10,000
9,000
8,000
7,000
6,000
5,000
4,000
3,000
2,000
1.6
2.2
3.2
3.8
rHSV-sFLT01 at an MOI of 2 (Ye et al., 2006). The results are
shown in Fig. 4, and demonstrate that high specific yields of
rAAV (DRP=cell) can be consistently achieved in sBHK cells
independent of the AAV serotype and packaged transgene,
without genetic modification of either the AAV genes or
the producer cell line. Furthermore, the rAAV DRP=ip ratios
were low when infectivity of representative samples was
determined in an end-point dilution TCID50 assay. rHSV
coinfection-generated rAAV1=AAT, rAAV2=sFLT01, and
infection density (×
× 10 vc/mL)
6
rHSV-rep2=cap1 vector MOI inputs of 12, 8, and 4, resulting in rAAV1=AAT specific yields of 77,200, 93,400, and
85,100 DRP=cell, respectively (Fig. 3). As expected, continued
reduction of the rHSV-rep2=cap1 MOI to 2 and 1 at a constant
rHSV-AAT MOI of 2 showed marked reduction in the
rAAV1=AAT specific yield, resulting in only 31,400 and
17,800 DRP=cell, respectively. Furthermore, a 4:2 MOI ratio
(rHSV-rep2=cap1:rHSV-AAT) resulted in about 80% of the
rAAV1=AAT DRP=cell production observed for a 5:5 MOI
ratio (data not shown).
Production of multiple rAAV
serotype–transgene combinations
Production of rAAV serotypes 1, 2, 5, and 8 in sBHK cells
was investigated by coinfection of the corresponding rHSVrep2=capX vector at an MOI of 4 with either rHSV-AAT or
120,000
rAAV production (DRP/cell)
FIG. 2. Optimization of harvest timing and infection density for rAAV production in sBHK cells. (A) rAAV2=GFP
infectious particles per cell as a function of time postinfection, (B) rAAV2=GFP infectious particles per cell as a function of cell density at the time of infection. Error bars indicate
standard deviation and n ¼ 3 for each.
100,000
80,000
60,000
40,000
20,000
0
rAAV1/
AAT
rAAV2/
sFLT01
rAAV5/
sFLT01
rAAV8/
AAT
FIG. 4. rAAV production yields for various serotype–
transgene combinations on sBHK cells. sBHK cells (1–2106
VC=ml) in 25-ml spinner flask cultures were coinfected with
rHSV-rep2=capX (where X is AAV cap serotype 1, 2, 5, or 8)
and rHSV-GOI (where GOI is either AAT or sFLT01) at MOIs
of 4 and 2, respectively. Error bars indicate the standard
deviation and n ¼ 5, 8, 6, and 3 for rAAV1=AAT, rAAV2=
sFLT01, rAAV5=sFLT01, and rAAV8=AAT, respectively.
THOMAS ET AL.
80,000
60,000
40,000
20,000
0
1L
10 L
set A
10 L
set B
FIG. 5. Disposable bioreactor production of rAAV1=AAT
by rHSV coinfection. Error bars indicate the standard deviation for 1-liter (n ¼ 3), 10-liter (n ¼ 4 with combination A of
rHSV stocks), and 10-liter (n ¼ 3 with combination B of rHSV
stocks) working volume scales.
rAAV5=sFLT01 viral lysates had average DRP=ip ratios of
110 30, 6 4, and 82 25 to 1 (n ¼ 3 for each), respectively,
with the difference between serotypes most likely reflecting
their in vitro infectivity variation (see Discussion). The DRP=ip
ratio generated here for the prototypical rAAV2 (6 4) is
highly favorable when compared with other production
methods, such as transient transfection (Aucoin et al., 2008).
Bioreactor production of rAAV
The production of rAAV in sBHK cells by rHSV coinfection
was executed in disposable bioreactors to exploit the increased scalability of suspension cells relative to adherent
HEK-293 cell rAAV production. Initial sBHK growth and
rAAV production experiments were performed in 1-liter
working volume Sartorius BIOSTAT CultiBag RM Wave
disposable bioreactors and resulted in high rAAV specific
yields, using the optimized spinner flask production parameters. Cells were seeded at 2.5105 VC=ml and fed with a
bolus of 300 ml of 5DMEM (containing 1salts) to prevent lglutamine depletion as needed. The doubling time was
13 1.7 hr (n ¼ 7). The cultures were coinfected with rHSVrep2=cap1 (MOI of 4) and rHSV-AAT (MOI of 2) at cell densities ranging from 1.4106 to 2.1106 VC=ml. The resulting
rAAV1=AAT vector production was ascertained at 24–26 hr
postinfection and at harvest (49 hr postinfection) and found to
be identical, confirming the vector production time postinfection to be the same in bioreactors as in spinner flasks (data not
shown). The rAAV1=AAT vector generated was harvested by
in situ lysis as described in Materials and Methods. The resulting average rAAV1=AAT specific yield was 75,600 DRP=
cell (Fig. 5), similar to that observed in 25-ml spinner cultures
(85,100 DRP=cell) at this MOI ratio, and afforded a volumetric
productivity of 1.41014 5.61013 DRP=liter (n ¼ 3).
rAAV production was scaled to 5- and 10-liter disposable
bioreactors and resulted in similar cell growth and rAAV
A
B
C
3.0
100%
7.5
2.4
95%
6.0
1.8
90%
4.5
1.2
85%
0.6
80%
1.5
75%
0.0
0.0
-48
-24
0
3.0
ammonium (mM)
100,000
viability
120,000
specific yields as at the 1-liter scale. The initial seed volume
was 2 or 4 liters and the cultures were diluted to 5 or 10 liters,
respectively, at 24 hr postseeding. Cells were grown (doubling
time of 12 1 hr, n ¼ 7) to a density of 1.6–2.0106 VC=ml,
coinfected with rHSV, and medium was exchanged 2 hr
postinfection to DMEM lacking FBS. The average specific
yields of 69,000 7500 DRP=cell (n ¼ 4) and 113,000 16,400
DRP=cell (n ¼ 3) for 10-liter runs (Fig. 5) were similar to 25-ml
spinner flask and 1-liter disposable bioreactor specific yields,
resulting in volumetric productivities of 1.31014 2.01013
and 2.41014 5.51013 DRP=liter. The approximately 2-fold
variation in 10-liter disposable bioreactor specific yield was
rHSV stock specific, as indicated in Fig. 5. Infectivity was
determined for several 10-liter bioreactor productions of
rAAV1=AAT and resulted in a DRP=ip ratio of 210 60
(n ¼ 5), within 2-fold of the 25-ml spinner flask results.
Medium exchange was accomplished by one of three
methods: centrifugation, flat sheet tangential flow filtration
(TFF), or integrated membrane perfusion of the reactor, the
latter two being relevant to scale-up. Cells grown in 5-liter
disposable bioreactors were collected by centrifugation
(640g, 4 min) and resuspended in DMEM lacking FBS.
The recovery of viable cells was 84 11% (n ¼ 10). A scalable
process for medium exchange in 5- and 10-liter disposable
bioreactors was pursued initially by flat sheet TFF in which
the postinfection volume was reduced 5- to 10-fold and then
diluted to the original volume with DMEM lacking FBS. The
TFF medium exchange operational parameters are listed in
Materials and Methods. Recovery of viable cells by TFF was
99 1% (n ¼ 5). Medium exchange was also examined using 10-liter disposable bioreactors with integrated perfusion
membranes (1.2 mm), permitting medium exchange without
additional equipment. Two to 4 hr postinfection, the bioreactor volume was reduced 3-fold by removing medium
through the perfusion membrane (100 ml=min). The concentrated cell culture was then perfused (100 ml=min) at
total viable cells (×1010)
rAAV1/AAT production (DRP/cell)
866
24
t (hpi)
FIG. 6. Typical growth, viability, and ammonium concentrations during 10-liter bioreactor sBHK cell cultivation
and rAAV production by rHSV coinfection. Arrows indicate
time of (A) seed, (B) dilution, and (C) medium exchange.
Circles=solid line, total viable cells; triangles, percent viability; diamonds=dashed line, ammonium concentration (mM).
SUSPENSION rAAV PRODUCTION USING rHSV COINFECTION
constant volume with 3 volumes of DMEM lacking FBS and
then diluted back to the initial volume with the same. The
recovery of viable cells was 93 4% (n ¼ 7).
Typical cell growth, viability, and ammonium concentrations for 10-liter disposable bioreactor sBHK cell growth and
rAAV production by rHSV coinfection are shown in Fig. 6.
Viability at harvest was typically 90–93%. Contrary to a previous report, cell growth did not appear to be substantially inhibited by ammonium accumulation of nearly 4 mM
(Christie and Butler, 1999). Christie and Butler have reported
that ammonium accumulation as low as 3 mM can significantly inhibit sBHK growth in ammoniagenic medium; however, no apparent cell growth inhibition was observed here
(e.g., constant doubling time throughout the cell growth
period).
Purification and in vivo evaluation of rAAV1=AAT
10-liter production runs
rAAV produced in 10-liter Wave bioreactors was prepared
for column chromatography by clarification followed by
concentration via TFF and buffer exchange via diafiltration
(DF). Clarification was accomplished via depth filtration
(1.2 mm) followed by dead end filtration (0.8=0.45 mm) and
resulted in an average recovery of 98 4% (n ¼ 10). TFF and
DF were executed with both hollow fiber filter (HFF) cartridges (0.2 or 1.15 m2) and flat sheet membranes (40.1 m2
membranes). Protein was routinely observed to precipitate
during HFF DF, regardless of scale. As a result, recovery of
rAAV by HFF TFF and DF was correspondingly low (Table 1).
Flat sheet TFF was also examined for concentration and buffer
exchange of rAAV and found to result in substantially higher
recovery of rAAV, without protein precipitation, when the
same operational parameters (recirculation rate, TMP, etc.) as
used in HFF TFF and DF were employed (Table 1). The large
protein reduction (29-fold) and standard deviation (s ¼ 16)
observed during HFF TFF=DF was most likely a result of the
protein precipitation observed in the diafiltered product. Flat
sheet TFF=DF resulted in high recovery of vector (96%) with a
substantial reduction in total protein (7.9-fold).
In vitro and in vivo evaluation of purified rAAV1=AAT
vector generated nearly identical results to those previously
seen for rHSV-produced rAAV1=AAT. Clarified, concentrated, and buffer-exchanged vector was purified by twocolumn chromatography, followed by buffer exchange into
lactated Ringer’s solution by TFF=DF. The resulting purified
vector was characterized by AAV- and HSV-specific assays,
and administered to C57BL=6 mice (11011 DRP=mouse) to
ascertain biological activity. Table 2 shows the results for the
in vitro and in vivo evaluation of purified rAAV1=AAT produced by the rHSV coinfection of a 10-liter Wave disposable
bioreactor sBHK culture. The residual HSV DNA (11 ng=ml)
and protein (3.2 ng=ml) were nearly identical to the values
obtained for rAAV1=AAT produced by rHSV coinfection of
adherent HEK-293 cells in cell factories and purified by threecolumn chromatography (Kang et al., 2009). In addition, the
serum hAAT level observed by 1 month in C57BL=6 mice
(2.8104 ng=ml serum) was within the experimental error of
that previously observed for rAAV1=AAT made by rHSV
coinfection of HEK-293 cells (3.1104 ng=ml serum) at the
same dose (11011 DRP=mouse) at 1 month postinjection
(Kang et al., 2009).
867
Table 1. TFF=DF Recoveries for Hollow Fiber
and Flat Sheet Concentration and Buffer
Exchange Prior to Column Chromatography
TFF=DF parameter
TFF fold concentration
Diafiltration volumes
Fold protein reduction
Overall rAAV recovery
Hollow fibera
Flat sheeta
4.4 1.0
7.7 2.7
29 16
59 10%
7.8 0.6
5.4 0.4
7.9 3.6
96 7%
Abbreviations: DF, diafiltration; TFF, tangential flow filtration.
a
n ¼ 6.
Discussion
Manufacturing technologies for rAAV continue to be a
limiting factor in its application to gene therapy protocols. The
surface area required for the culturing of adherent mammalian cell lines and subsequent rAAV production is a daunting
obstacle to gene delivery protocols aimed at treatment of
systemic diseases with high dosing requirements such as
Duchenne muscular dystrophy or a1-antitrypsin deficiency.
Historic methodologies for rAAV production that use adherent cell lines will not be able to accomplish vector production on the required scale. Therefore, manufacture of
rAAV in a suspension format is paramount to the future
success of gene therapy disease indications that require largescale vector production.
Here we have extended our rHSV coinfection platform
production of rAAV by application to mammalian cells
adapted to grow in suspension. The sBHK cells described in
this paper produced similar specific yields (DRP=cell) of
rAAV as both adherent HEK-293 and adherent BHK cells
when coinfected with HSV=AAV hybrid vectors (Kang et al.,
2009), but sBHK cells were easily scaled in disposable bioreactors to volumes necessary for clinical manufacturing
requirements. sBHK cells are attractive from a process perspective because of their short doubling time of about 12–
13 hr. For example, a seed train initiated with 1107 total
viable cells can be expanded to a 10-liter culture with a density
of 2109 viable cells=liter in 5 to 6 days. The sBHK system
Table 2. Characterization of Two-Column
Purified rAAV1=AAT Produced by rHSV
Coinfection of sBHK Cells
Characteristic
rAAV titer (DRP=ml)
Infectivity (TCID50=ml)
DRP:TCID50
Residual HSV protein
(ng=ml)
Residual HSV DNA
(ng=ml)
hAAT expression
(ng=ml serum)a
Assay
Result
qPCR
TCID50
qPCR=TCID50
ELISA
1.21012
1.31010
95
3.2
qPCR
11
ELISA
2.8104
Abbreviations: DRP, DNase-resistant particles; hAAT, human
a1-antitrypsin; HSV, herpes simplex virus; qPCR, quantitative polymerase chain reaction; rAAV, recombinant adeno-associated virus.
a
In C57BL=6 mice 1 month postinjection of 11011 DRP, n ¼ 6,
s ¼ 6.7103.
868
allowed 2- to 4-fold increases in cell density relative to adherent cell culture of HEK-293 cells, resulting in a similar increase in volumetric productivity, while allowing a decrease
in rHSV vector input. The 400-fold scale-up from 25-ml
spinner flasks to 10-liter bioreactors resulted in the same
doubling time (12 hr) and similar specific yields, depending
on the rHSV stocks used. The resulting average volumetric
productivity for 10-liter bioreactors runs was as high as 2.4
1014 DRP=liter. These data are promising for further scale-up
of the rHSV coinfection of sBHK cells for production of rAAV
to meet increasing clinical requirements for human gene
transfer. Finally, rAAV1=AAT produced by rHSV coinfection
of sBHK cells in 10-liter disposable bioreactors and purified
by two-column chromatography was similar to adherent
cell factory HEK-293-produced and three column-purified
rAAV1=AAT with respect to both residual HSV components
and biological activity of the rAAV1=AAT produced (Kang
et al., 2009).
Previous reports of rAAV production in suspensionadapted mammalian cell lines have been limited to HEK-293
cells (Park et al., 2006; Durocher et al., 2007; Feng et al., 2007;
Hildinger et al., 2007) and the A549 packaging cell line, which
expresses the AAV rep and cap genes (Farson et al., 2004). The
HEK-293 cell systems employed the transient transfection
method, which remains technically difficult to execute, results
in low specific yields (DRP=cell), and contaminates rAAV
with microgram quantities of exogenous DNA. A disadvantage of the A549 packaging cell line platform is that a new
packaging cell line must be developed for each serotype–
gene-of-interest combination, and often requires large numbers of potential clones to be screened (e.g., 1 rep=cap-positive
cell per 1300 cells approximately) (Farson et al., 2004).
The sBHK rAAV production platform resulted in equivalent rAAV specific yields as adherent HEK-293 cells, but used
only one third of the rHSV-rep2=capX vector input and was
amenable to scale-up. The Poisson distribution is often used to
characterize viral infection probability; however, it does not
account for transport resistances or cell–virus contact limitations observed in T-flask culture of mammalian cells, making
its application to adherent HEK-293 cells difficult. In addition,
the kinetics of infection with multiple viruses can increase the
difficulty in predicting viral infection by Poisson analysis even
in suspension culture (Mena et al., 2007). Therefore, we employed an empirical approach to rAAV specific yield optimization in the sBHK system via MOI (total and ratio)
modulation. An MOI ratio of 12:2 for the rHSV-rep2=cap2 and
rHSV-GFP vectors, respectively, resulted in the maximal
specific yield (ip=cell) of rAAV2=GFP in adherent HEK-293
cell culture (Kang et al., 2009). Reduction in the rHSVrep2=capX MOI from 12 to 4 on sBHK cells did not reduce
the specific yield of rAAV1=AAT (Fig. 3), and permitted the
total MOI to be reduced from 14 to 6, or a 60% decrease. This
most likely reflects the rHSV vector transport kinetics in the
respective cultures, since the Poisson distribution predicts
greater than 98% of cells should still be infected at an MOI
of 4. The MOI ratio of 4:2 for rHSV coinfection production of
rAAV1=AAT produced about 80% as much rAAV1=AAT as
at a 5:5 MOI ratio (data not shown), and agreed closely with
theoretical expectation with respect to infected cells (about
85%) because these MOIs would be predicted to infect 98 and
86% of the cell population, respectively. The MOI ratio of 4:2
was attractive from an rHSV raw material usage and purifi-
THOMAS ET AL.
cation burden standpoint, permitting the total rHSV MOI to
be reduced 40%, from 10 (5:5) to 6 (4:2), with only a 20% loss of
productivity (DRP=cell).
rHSV coinfection of sBHK cells for rAAV production
compares favorably with the baculovirus infection production of rAAV in suspension insect cells. Cecchini and colleagues demonstrated production of 21014 DRP=liter at a
10-liter scale in a stirred-tank bioreactor at 72 hr postinfection,
using three recombinant baculoviruses (Cecchini et al., 2008).
By comparison, the average volumetric productivities we
achieved with the rHSV coinfection of suspension BHK cells
in 10-liter disposable bioreactors was as high as 2.41014
DRP=liter; however, fewer recombinant viruses were required and the sBHK cells grow faster than insect cells. Another advantage of the rHSV coinfection platform is that all
AAV capsid serotypes inserted into the rHSV-rep2=capX
vector tested to date remain infectious; unlike the baculovirus
system, which requires alteration of the capsid sequences
for serotypes other than serotype 2 to remain infectious on
mammalian cells. For example, rAAV1 produced by baculovirus infection of insect cells requires replacement of the
N-terminal portion of VP1 with the analogous portion from
serotype 2 to generate a functional phospholipase domain
required for infectivity (Kohlbrenner et al., 2005). rAAV
vectors of various serotypes were produced here by rHSV
coinfection of sBHK cells at a low DRP=ip ratio, with the
differences between serotypes most likely reflecting the differences in infectivity on the HeLaRC32 cell line used for infectious titering. This difference is exemplified by comparing
the infectivity between serotypes 2 and 5, where expression
of the platelet-derived growth factor receptor (a or b) was
previously required for increased infectivity of serotype 5 on
HeLa cells (Di Pasquale et al., 2003).
The advent of suspension rAAV manufacture provides a
method for generating therapeutic quantities of vector for the
treatment of a variety of diseases, inclusive of systemic vector
applications requiring high DRP doses. Packaging=producer
cell lines that use adenovirus infection and insect cell lines
that use baculoviruses have previously been employed for
suspension rAAV production. We have now extended this
list to include the rHSV coinfection of sBHK cells for suspension production of rAAV, demonstrating high specific and
volumetric productivities, short production cycle times, and
substantially reduced helper vector requirements. These improvements should establish rHSV coinfection production of
rAAV as another important production technology for preclinical and clinical rAAV gene therapy protocols. Future research will focus on removal of animal-derived products from
the upstream process and additional characterization of rHSV
stocks to ascertain attributes (particle-to-infectivity ratio, concentration, formulation, etc.) that enhance or diminish rAAV
production.
Acknowledgment
The authors thank Ms. Robin Griffis for technical discussions and contributions to this work.
Author Disclosure Statement
The authors all hold share options in AGTC and have a
conflict of interest to the extent that this work potentially
increases their personal financial interests.
SUSPENSION rAAV PRODUCTION USING rHSV COINFECTION
References
Alazard-Dany, N., Nicolas, A., Ploquin, A., Strasser, R., Greco,
A., Epstein, A.L., Fraefel, C., and Salvetti, A. (2009). Definition
of herpes simplex virus type 1 helper activities for adenoassociated virus early replication events. PLoS Pathog. 5,
e1000340.
Albu, E., and Holmes, K.V. (1973). Isolation and preliminary
characterization of the RNA-containing R-type, virus-like
particle of BHK-21 cells. J. Virol. 12, 1164–1172.
Aucoin, M.G., Perrier, M., and Kamen, A.A. (2008). Critical assessment of current adeno-associated viral vector production
and quantification methods. Biotechnol. Adv. 26, 73–88.
Bergmann, D.G., and Wolff, D.A. (1981). Production of R-type
virus-like particles in hamster cells. Intervirology 16, 61–70.
Bloch, J., Bachoud-Levi, A.C., Deglon, N., Lefaucheur, J.P., Winkel, L., Palfi, S., Nguyen, J.P., Bourdet, C., Gaura, V., Remy, P.,
Brugieres, P., Boisse, M.F., Baudic, S., Cesaro, P., Hantraye, P.,
Aebischer, P., and Peschanski, M. (2004). Neuroprotective gene
therapy for Huntington’s disease, using polymer-encapsulated
cells engineered to secrete human ciliary neurotrophic factor:
Results of a phase I study. Hum. Gene Ther. 15, 968–975.
Boedeker, B.G. (2001). Production processes of licensed recombinant factor VIII preparations. Semin. Thromb. Hemost. 27,
385–394.
Booth, M.J., Mistry, A., Li, X., Thrasher, A., and Coffin, R.S. (2004).
Transfection-free and scalable recombinant AAV vector production using HSV=AAV hybrids. Gene Ther. 11, 829–837.
Buller, R.M., Janik, J.E., Sebring, E.D., and Rose, J.A. (1981).
Herpes simplex virus types 1 and 2 completely help adenovirus-associated virus replication. J. Virol. 40, 241–247.
Buning, H., Perabo, L., Coutelle, O., Quadt-Humme, S., and
Hallek, M. (2008). Recent developments in adeno-associated
virus vector technology. J. Gene Med. 10, 717–733.
Cecchini, S., Negrete, A., and Kotin, R.M. (2008). Toward exascale production of recombinant adeno-associated virus for
gene transfer applications. Gene Ther. 15, 823–830.
Christie, A., and Butler, M. (1999). The adaptation of BHK cells
to a non-ammoniagenic glutamate-based culture medium.
Biotechnol. Bioeng. 64, 298–309.
Clark, K.R., Voulgaropoulou, F., Fraley, D.M., and Johnson, P.R.
(1995). Cell lines for the production of recombinant adenoassociated virus. Hum. Gene Ther. 6, 1329–1341.
Clark, K.R., Voulgaropoulou, F., and Johnson, P.R. (1996). A
stable cell line carrying adenovirus-inducible rep and cap genes
allows for infectivity titration of adeno-associated virus vectors. Gene Ther. 3, 1124–1132.
Conway, J.E., Zolotukhin, S., Muzyczka, N., Hayward, G.S., and
Byrne, B.J. (1997). Recombinant adeno-associated virus type 2
replication and packaging is entirely supported by a herpes
simplex virus type 1 amplicon expressing Rep and Cap. J. Virol.
71, 8780–8789.
Conway, J.E., Rhys, C.M., Zolotukhin, I., Zolotukhin, S., Muzyczka, N., Hayward, G.S., and Byrne, B.J. (1999). High-titer
recombinant adeno-associated virus production utilizing a
recombinant herpes simplex virus type I vector expressing
AAV-2 Rep and Cap. Gene Ther. 6, 986–993.
Di Pasquale, G., Davidson, B.L., Stein, C.S., Martins, I., Scudiero,
D., Monks, A., and Chiorini, J.A. (2003). Identification of
PDGFR as a receptor for AAV-5 transduction. Nat. Med. 9,
1306–1312.
Durocher, Y., Pham, P.L., St-Laurent, G., Jacob, D., Cass, B., Chahal, P., Lau, C.J., Nalbantoglu, J., and Kamen, A. (2007). Scalable
serum-free production of recombinant adeno-associated virus
869
type 2 by transfection of 293 suspension cells. J. Virol. Methods
144, 32–40.
Farson, D., Harding, T.C., Tao, L., Liu, J., Powell, S., Vimal, V.,
Yendluri, S., Koprivnikar, K., Ho, K., Twitty, C., Husak, P.,
Lin, A., Snyder, R.O., and Donahue, B.A. (2004). Development
and characterization of a cell line for large-scale, serum-free
production of recombinant adeno-associated viral vectors.
J. Gene Med. 6, 1369–1381.
Feng, L., Guo, M., Zhang, S., Chu, J., and Zhuang, Y. (2007).
Optimization of transfection mediated by calcium phosphate
for plasmid rAAV-LacZ (recombinant adeno-associated virusb-galactosidase reporter gene) production in suspensioncultured HEK-293 (human embryonic kidney 293) cells.
Biotechnol. Appl. Biochem. 46, 127–135.
Feudner, E., De Alwis, M., Thrasher, A.J., Ali, R.R., and Fauser,
S. (2001). Optimization of recombinant adeno-associated virus
production using an herpes simplex virus amplicon system.
J. Virol. Methods 96, 97–105.
Hermonat, P.L., and Muzyczka, N. (1984). Use of adeno-associated
virus as a mammalian DNA cloning vector: Transduction of
neomycin resistance into mammalian tissue culture cells. Proc.
Natl. Acad. Sci. U.S.A. 81, 6466–6470.
Hildinger, M., Baldi, L., Stettler, M., and Wurm, F.M. (2007).
High-titer, serum-free production of adeno-associated virus
vectors by polyethyleneimine-mediated plasmid transfection
in mammalian suspension cells. Biotechnol. Lett. 29, 1713–1721.
Kang, W., Wang, L., Harrell, H., Liu, J., Thomas, D.L., Mayfield,
T.L., Scotti, M.M., Ye, G.J., Veres, G., and Knop, D.R. (2009). An
efficient rHSV-based complementation system for the production of multiple rAAV vector serotypes. Gene Ther. 16, 229–239.
Knop, D.R., and Harrell, H. (2007). Bioreactor production of
recombinant herpes simplex virus vectors. Biotechnol. Prog.
23, 715–721.
Kohlbrenner, E., Aslanidi, G., Nash, K., Shklyaev, S., CampbellThompson, M., Byrne, B.J., Snyder, R.O., Muzyczka, N.,
Warrington, K.H., Jr., and Zolotukhin, S. (2005). Successful
production of pseudotyped rAAV vectors using a modified
baculovirus expression system. Mol. Ther. 12, 1217–1225.
Mena, J.A., Ramirez, O.T., and Palomares, L.A. (2007). Population kinetics during simultaneous infection of insect cells with
two different recombinant baculoviruses for the production of
rotavirus-like particles. BMC Biotechnol. 7, 39.
Mistry, A.R., De Alwis, M., Feudner, E., Ali, R.R., and Thrasher,
A.J. (2002). High-titer stocks of adeno-associated virus from
replicating amplicons and herpes vectors. Methods Mol. Med.
69, 445–460.
Muzyczka, N., and Berns, K.I. (2001). Parvoviridae: The viruses
and their replication. In Fields Virology. D.M. Knipe and P.M.
Howley, eds. (Lippincott, Williams & Wilkins, New York)
pp. 2327–2360.
Negrete, A., Yang, L.C., Mendez, A.F., Levy, J.R., and Kotin,
R.M. (2007). Economized large-scale production of high yield
of rAAV for gene therapy applications exploiting baculovirus
expression system. J. Gene Med. 9, 938–948.
Park, J.Y., Lim, B.P., Lee, K., Kim, Y.G., and Jo, E.C. (2006).
Scalable production of adeno-associated virus type 2 vectors
via suspension transfection. Biotechnol. Bioeng. 94, 416–430.
Rice, S.A., and Knipe, D.M. (1990). Genetic evidence for two
distinct transactivation functions of the herpes simplex virus a
protein ICP27. J. Virol. 64, 1704–1715.
Salvetti, A., Oreve, S., Chadeuf, G., Favre, D., Cherel, Y.,
Champion-Arnaud, P., David-Ameline, J., and Moullier, P.
(1998). Factors influencing recombinant adeno-associated
virus production. Hum. Gene Ther. 9, 695–706.
870
Toublanc, E., Benraiss, A., Bonnin, D., Blouin, V., Brument, N.,
Cartier, N., Epstein, A.L., Moullier, P., and Salvetti, A. (2004).
Identification of a replication-defective herpes simplex virus
for recombinant adeno-associated virus type 2 (rAAV2) particle assembly using stable producer cell lines. J. Gene Med. 6,
555–564.
Tratschin, J.D., West, M.H., Sandbank, T., and Carter, B.J. (1984).
A human parvovirus, adeno-associated virus, as a eucaryotic
vector: Transient expression and encapsidation of the procaryotic gene for chloramphenicol acetyltransferase. Mol. Cell.
Biol. 4, 2072–2081.
Urabe, M., Ding, C., and Kotin, R.M. (2002). Insect cells as a factory to produce adeno-associated virus type 2 vectors. Hum.
Gene Ther. 13, 1935–1943.
Weindler, F.W., and Heilbronn, R. (1991). A subset of herpes
simplex virus replication genes provides helper functions for
productive adeno-associated virus replication. J. Virol. 65,
2476–2483.
Wustner, J.T., Arnold, S., Lock, M., Richardson, J.C., Himes,
V.B., Kurtzman, G., and Peluso, R.W. (2002). Production of
recombinant adeno-associated type 5 (rAAV5) vectors using
recombinant herpes simplex viruses containing rep and cap.
Mol. Ther. 6, 510–518.
THOMAS ET AL.
Ye, G.-J., Drogemuller, C., Thomas, D.L., Scotti, M.M., Kang, W.,
Knop, D.R., Boye, S., Pechan, P., Peterson, J., Hauswirth, W.,
Scaria, A., and Wadsworth, S. (2006). A novel herpes simplex
virus helper based production of adeno-associated virus vectors for treatment of retinal angiogenesis. Mol. Ther. 13, S194.
Zhang, X., De Alwis, M., Hart, S.L., Fitzke, F.W., Inglis, S.C.,
Boursnell, M.E., Levinsky, R.J., Kinnon, C., Ali, R.R., and
Thrasher, A.J. (1999). High-titer recombinant adeno-associated
virus production from replicating amplicons and herpes vectors deleted for glycoprotein H. Hum. Gene Ther. 10, 2527–
2537.
Address correspondence to:
Dr. David R. Knop
Applied Genetic Technologies Corporation
11801 Research Dr., Suite D
Alachua, FL 32615
E-mail: [email protected]
Received for publication January 12, 2009;
accepted after revision May 6, 2009.
Published online: June 8, 2009.