Gt$$$$957p

Gene Therapy (2000) 7, 910–913
 2000 Macmillan Publishers Ltd All rights reserved 0969-7128/00 $15.00
www.nature.com/gt
VIRAL TRANSFER TECHNOLOGY
BRIEF COMMUNICATION
Production and concentration of pseudotyped HIV-1-based gene transfer vectors J ReiserDevelopmental and Metabolic Neurology Branch, National Institute of Neurological Disorders and Stroke, National Institutes ofHealth, Bethesda, MD; and Louisiana State University Gene Therapy Program, Louisiana State University School of Medicine, NewOrleans, LA, USA Strategies to generate highly concentrated HIV-1 vector ing titers of up to 109 transducing units per milliliter. There pseudotypes involving different envelope (Env) proteins was no loss in titer with any of the pseudotypes tested. Thus, including the vesicular stomatitis virus (VSV) G glycoprotein, like lentiviral vectors pseudotyped with VSV-G, HIV-1-based the Moloney murine leukemia virus (MLV) 4070A ampho- vectors pseudotyped with the MLV 4070A amphotropic Env tropic Env and the rabies G glycoprotein were established.
and the rabies G glycoprotein resist inactivation during con- Virus stocks were prepared by transient transfection using centration. This opens up the possibility to generate highly standard cell culture media or serum-free media. Such concentrated HIV-1 vector stocks carrying alternative Env stocks were concentrated 50- to 300-fold by ultracentrifug- proteins on a large scale. Gene Therapy (2000) 7, 910–913.
ation or by ultrafiltration using Centricon Plus-80 units yield- Keywords: lentiviral vectors; pseudotypes; gene therapy
Lentiviral vectors including HIV-1-based vectors are This paper documents the fact that HIV-1 vectors being increasingly used for gene delivery in vitro. They pseudotyped with alternative Env proteins including the are also promising for long-term gene expression in vivo MLV amphotropic Env and the rabies G glycoprotein in cells of the central nervous system, hematopoietic sys- retain full infectivity after ultracentrifugation. The paper tem, retina, muscle and liver, lung and pancreatic islets.1,2 also extends our original protocol with emphasis on Lentivirus-based vectors are traditionally produced by scale-up of vector production, mode of concentration transient cotransfection of human embryonic kidney 293T (ultracentrifugation versus ultrafiltration) and compo- cells using three to four different plasmids encoding sition of the medium (serum-containing media versus helper (packaging), envelope (Env) and gene transfer protein-free media). A novel 293-based cell line express- functions, respectively. Typical virus titers range from 106 ing the amphotropic MLV Env protein allowing the gen- to 107 transducing units per milliliter. Increased titers can eration of high-titer amphotropic HIV-1 pseudotypes is be achieved by physical concentration. Pseudotyped len- tiviral particles containing the VSV-G glycoprotein have The influences of time and composition of the cell cul- been shown to be quite stable and to withstand concen- ture medium on vector production were evaluated first.
tration by ultracentrifugation without significant loss in 293T cells were cotransfected with the previously titer,3,4 thus opening up the way to generate highly con- described HIV-neo⌬E vector7 encoding G418 resistance centrated vector stocks for in vivo applications. Ultracen- together with the C-Help packaging construct7 and the trifugation protocols to concentrate HIV-1 pseudotypes LTR-G envelope construct encoding VSV-G6 (Figure 1).
involving alternative Env proteins have not been Twelve hours later the medium was removed and described so far, in part because it was assumed that such replaced with DMEM containing 10% fetal bovine serum pseudotypes would be unstable. This was previously or with media lacking serum (UltraCULTURE). Media observed with MLV-based vectors involving Env pro- were replaced again 24, 40, 60 and 80 h after initiating transfection and the culture supernatants were collected Alternative concentration protocols were established and frozen immediately. For titration the samples were for HIV-1-based vectors. Reiser et al6 have developed a thawed simultaneously and end point titers (G418- facile ultrafiltration procedure to concentrate HIV-1 resistant colony-forming units (c.f.u.) per milliliter) were pseudotypes. Good recoveries were obtained with HIV- determined on human osteosarcoma (HOS) cells follow- 1 particles containing VSV-G or the MLV amphotropic ing G418 selection. HOS cells were primarily used because they have been shown in the past to be readilyinfectable presented in Figure 2 show that there was an increase in Correspondence: J Reiser, LSU Gene Therapy Program, LSU School of titer up to 40 h in serum-containing and serum-free Medicine, MEB 3205, 1901 Perdido Street, New Orleans, LA 70112, USA media with titers reaching 2 × 107 c.f.u./ml. Subsequent Received 2 September 1999; accepted 15 February 2000 to that virus production dropped. This was particularly Production of HIV-1 vectors
J Reiser
Figure 2 Effects of cell culture medium on vector production. For the
preparation of HIV-1 pseudotypes 5 ␮g each of C-Help packaging plasmid
DNA,7 LTR-G plasmid DNA6 and HIV-neo⌬E vector plasmid DNA7
were co-transfected into sub-confluent 293T cells by using the calcium
phosphate precipitation method.11 Approximately 4 × 105 cells were seeded
into six-well plates 24 h before transfection. Chloroquine (25 ␮m final
concentration) was added to the cells immediately before transfection and
Figure 1 Components of the HIV-1-based gene transfer system. (a) Trans-
the medium was replaced 12 h later with 2 ml per well of fresh DMEM ducing vector constructs. The HIV-neo⌬E vector construct was described supplemented with 10% FBS or serum-free UltraCULTURE medium before.7 Boxes interrupted by jagged lines contain partial deletions. SV, (BioWhittaker, Walkersville, MD, USA) containing glutamine (2 mm) SV40 early promoter; CMV, human cytomegalovirus immediate–early and penicillin/streptomycin. Media were replaced four more times 24, 40, promoter; Ori, 168-bp fragment encompassing SV40 origin of replication;6 60, and 83 h after initiation of transfection and aliquots of the culture PPT, 133-bp fragment harboring HIV-1 polypurine tract sequence.12 (b) supernatants were immediately frozen on dry ice and stored at −80°C. Helper (packaging) construct. The C-Help packaging plasmid was All aliquots were thawed simultaneously and end-point titers (mean ± described previously.7 The open triangle symbolizes a deletion affecting s.d.) determined on HOS cells using G418 selection (Life Technologies; the packaging signal between the 5Ј splice donor site and the beginning 0.35–0.5 mg of active drug per milliliter) as described in the legend to of the gag sequence. The polyA site was derived from the bovine growth hormone gene. (c) Env expression constructs. The LTR-G, LTR-RVG andALF expression constructs encode the VSV-G glycoprotein, the rabies Gglycoprotein and the MLV amphotropic (4070A) Env protein, respect- determined on HOS cells following G418 selection. Titers ively.6–8 VSV-G and rabies G gene expression are driven by the HIV-1 remained relatively constant (1–2 × 107 c.f.u./ml) at 37°C LTR. The polyA site was derived from the SV40 late region. The 4070A from 40 h up to 86 h in DMEM/10% FBS and in Ultra- Env coding region is driven by the Friend FB29 MLV LTR.8 CULTURE medium. At 33°C titers in both media reacheda plateau at 60 h and remained above 107 c.f.u./ml upto 86 h (data not shown). Taken together with the data obvious with cells kept in DMEM/10% while cells kept presented in Figure 2, these results indicate that HIV-1 in serum-free UltraCULTURE medium showed a more vectors pseudotyped with VSV-G are stable for at least prolonged phase of virus production. Thus, serum-free 86 h in serum-containing and in serum-free media at media such as UltraCULTURE allow multiple collections for extended periods of time. This is an attractive feature HIV-1 vector particles pseudotyped with the MLV for virus production at a preparative scale. Also, the amphotropic Env have previously been produced using absence of serum in the final virus preparation may be three-component transient transfection protocols with tit- advantageous in certain circumstances where serum ers around 105 transducing units per milliliter.3,7 With a interferes with the physiology of the target cell to be view toward increasing the titers of HIV-1 vectors pseudotyped with the amphotropic Env, stable cell lines To investigate the stability of VSV-G-pseudotyped expressing the 4070A Env protein were generated.
HIV-1 vector particles during production, the accumu- 293 Cells were transfected using the the ALF plasmid lated vector titers in DMEM/10% FBS and in serum-free encoding the 4070A Env8 and clones were obtained fol- medium were investigated at 37°C and at 33°C. Trans- lowing selection with phleomycin (50 ␮g/ml final fected cells were incubated at 37°C and 33°C and aliquots concentration). Some of the best clones yielded vector tit- of the culture supernatants were collected 40, 60, and 86 h ers up to 3 × 106 c.f.u./ml following transient transfection after the start of the transfection and immediately frozen.
of such clones with the C-Help and HIV-neo⌬E plasmids.
Media were not changed during collection. All aliquots These titers are are about five- to 10-fold higher than the were thawed simultaneously and end point titers were Production of HIV-1 vectors
Increases in viral titers were achieved by concentration of the culture supernatants by ultracentrifugation. Ultra-centrifugation protocols to generate high-titer HIV-1 vec-tor stocks pseudotyped with VSV-G have been describedbefore.3,4 However, the stability of HIV-1 vectors pseudo-typed with alternative Env proteins has not been investi-gated. The results shown in Table 1 demonstrate thatHIV-neo⌬E pseudotypes carrying the MLV 4070A Env orthe rabies G glycoprotein could be concentrated some 70-to 100-fold by ultracentrifugation with good recovery asjudged from the virus titers. These findings suggest thatHIV-1-based vectors pseudotyped with the MLV 4070Aamphotropic Env or the rabies G glycoprotein resist inac-tivation during ultracentrifugation similar to other retro-viral vectors pseudotyped with VSV-G. Other HIV-1-based vector constructs also lend themselves to concen-tration by ultracentrifugation. The NL-EGFP/P vectorencoding EGFP was pseudotyped with VSV-G and con-centrated about 300-fold using ultracentrifugation. TheNL-EGFP/P and HIV-neo⌬E vectors differ in terms ofthe reporter genes, internal promoters and presence orabsence of the tat and rev coding regions (Figure 1a). Itis apparent from the results presented in Figure 3 thatunconcentrated vector stocks reached titers between 5 ×106 and 1 × 107 transducing units per milliliter in T-150flasks and that subsequent concentration by ultracentri- Figure 3 Production and concentration of vector pseudotypes encoding
fugation yielded titers above 109 transducing units per EGFP. Transfections into 293T cells were carried out as described in the milliliter. The recovered virus titer after centrifugation legend to Table 1, but using the NL-EGFP/P vector construct (Figure 1a). Thirty milliters of virus supernatant were collected 60 h after transfectionand filtered through a 0.45 ␮m membrane. Ultracentrifugation (25 000 Ultrafiltration was used as an alternative to ultracentri- r.p.m., 15°C, 115 min) was carried out using a Beckman SW28 rotor. fugation to generate concentrated virus stocks. Centricon The viral pellet was resuspended in 0.1 ml of DMEM/10% FBS at 4°C Plus-80 units with a 100 kDa molecular weight cut-off for 1 h. HOS cells in six-well plates (1–2 × 105 cells per well) were trans- filter were used. HIV-neo⌬E vectors pseudotyped with duced with different amounts of the virus in 0.5 ml of DMEM/10% FBS VSV-G, the MLV 4070A Env or the rabies G glycoprotein containing polybrene (8 ␮g/ml) for 16 h at 37°C. The concentrated virus were prepared in T-75 or T-150 cell culture flasks using was diluted 300-fold and different amounts of the diluted virus were usedfor transduction. Four days later the cells were detached from the plateusing trypsin/EDTA (Life Technologies) and collected into DMEM/10% Concentration of virus stocks by ultracentrifugation FBS and subsequently washed with Hank’s balanced salt solution (LifeTechnologies) containing 2% FBS (Hank’s/FBS). The cells were washedwith Hank’s/FBS and fixed in 2% paraformaldehyde for 5 min and resus- pended in Hank’s/FBS, and then subjected to FACS analysis. The graph shows the percentage of EGFP-positive cells relative to the volume of virus added (unconcentrated virus or concentrated virus which was sub-sequently diluted). To calculate titers the number of target cells was multi- plied by the percentage of EGFP-positive cells divided by the volume of Transfections into 293T cells10 were done in T-150 flasks using 30ml of DMEM/10% FBS in the presence of chloroquine (25 ␮m final 15 ml or 30 ml, respectively of DMEM/10% FBS per flask.
concentration). Thirty micrograms each of helper, Env and vector The cell-free supernatants were then concentrated 60- to DNA were combined with 250 mm CaCl2 in a total volume of 3 ml.
70-fold by ultrafiltration (Table 2). Surprisingly, the titers An equal volume of 2 × HEPES-buffered saline (HBS)11 was sub-sequently added by vortexing. Media were changed 12 h after increased up to 350-fold. Titers around 5 × 107 c.f.u./ml transfection. Thirty milliliters of virus supernatant were collected were observed for rabies G pseudotypes and up to 109 60 h after transfection and filtered through a 0.45 ␮m membrane.
c.f.u./ml for VSV-G and 4070A pseudotypes. This sug- Ultracentrifugation (25 000 r.p.m., 15°C, 105 min) was carried out gests that the ultrafiltration procedure may have using a Beckman SW28 rotor. Virus titers were determined by limit- removed an inhibitor of viral infection. Higher than ing dilution in the following way. Target cells were split into six- expected virus titers following ultrafiltration were well plates the day before infection to give approximately 50% con-fluence at the time of infection. Infections were performed with ser- observed before with spleen necrosis virus vector prep- ial dilutions of virus stocks in a total of 0.5 ml of medium containing arations.9 Ultrafiltration is expected to bypass resuspen- 8 ␮g of polybrene per milliliter. After 3 to 6 h at 37°C, the virus sion problems that are encountered with virus stocks was removed and 2 ml of medium was added and the plates incu- concentrated by ultracentrifugation. However, for in vivo bated at 37°C for an additional 3 days. The medium was then aspir- experiments the use of serum-free medium during virus ated, and 2 ml of medium supplemented with G418 (0.35 to 0.5 mg production is recommended in order to avoid serum pro- of active drug per milliliter; Life Technologies Inc, Gaithersburg,MD, USA) was placed into each well. The medium was changed teins that tend to accumulate during ultrafiltration.
every 3 to 4 days, and the colonies were counted on days 10 or 14 A variety of different Env proteins are currently being after staining with crystal violet (0.2% in 20% ethanol).
tested with regard to their capacity to pseudotype lentivi- Production of HIV-1 vectors
J Reiser
Table 2 Concentration of virus stocks by ultrafiltration using
1 Naldini L. Lentiviruses as gene transfer agents for delivery to non-dividing cells. Curr Opin Biotechnol 1998; 9: 457–463.
2 Naldini L, Verma IM. Lentiviral vectors. In: Friedmann T (ed).
The Development of Human Gene Therapy. Cold Spring HarborLaboratory Press: Cold Spring Harbor, 1999, pp 47–60.
3 Naldini L et al. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 1996; 272: 263–
4 Bartz SR, Rogel ME, Emerman M. Human immunodeficiency virus type 1 cell cycle control: Vpr is cytostatic and mediates G2accumulation by a mechanism which differs from DNA damage Transfections into 293T cells were done in T-75 flasks (results checkpoint control. J Virol 1996; 70: 2324–2331.
presented on first three lines) or T-150 flasks (results presented on 5 Burns JC et al. Vesicular stomatitis virus G glycoprotein pseudo- bottom line) using 15 ml and 30 ml, respectively, of DMEM/10%FBS in the presence of chloroquine (25 ␮m final concentration). For typed retroviral vectors: concentration to very high titer and T-75 flasks 15 ␮g each of helper, Env and vector DNA were com- efficient gene transfer into mammalian and nonmammalian cells. Proc Natl Acad Sci USA 1993; 90: 8033–8037.
volume of 2 × HBS was subsequently added by vortexing. For T- 6 Reiser J et al. Transduction of nondividing cells using pseudo- 150 flasks all volumes were doubled. Sixty milliliters of combined typed defective high-titer HIV type 1 particles. Proc Natl Acad supernatants were collected 60 h after transfection filtered through Sci USA 1996; 93: 15266–15271.
a 0.45 ␮m membrane and concentrated using Centricon Plus-80 7 Mochizuki H et al. High-titer immunodeficiency type 1-based units (Millipore, Bedford, MA, USA) by centrifugation at 1500 vector systems for gene delivery into nondividing cells. J Virol r.p.m. for 2–2.5 h at 15°C in a Beckman GPR centrifuge (Beckman 1998; 72: 8873–8883.
Instruments, Palo Alto, CA, USA). Titers were determined as 8 Cosset FL et al. High-titer packaging cells producing recombi- nant retroviruses resistant to human serum. J Virol 1995; 69:
7430–7436.
9 Chu TH, Dornburg R. Toward highly efficient cell-type-specific gene transfer with retroviral vectors displaying single-chain ral vectors. The protocols described in this study will be antibodies. J Virol 1997; 71: 720–725.
useful to generate high-titer stocks of such pseudotypes.
10 DuBridge RB et al. Analysis of mutation in human cells by using an Epstein–Barr virus shuttle system. Mol Cell Biol 1987; 7:
379–387.
11 Pear WS, Nolan GP, Scott ML, Baltimore D. Production of high- titer helper-free retroviruses by transient transfection. Proc Natl I am grateful to Yasuhiro Takeuchi for providing plasmid Acad Sci USA 1993; 90: 8392–8396.
ALF. I thank Simon Tang and Hideki Mochizuki for their 12 Charneau P et al. HIV-1 reverse transcription. A termination step help during the early phase of this work.
at the center of the genome. J Mol Biol 1994; 241: 651–662.

Source: http://fuel1.biochem.dote.hu/LRB_Protocols/VirusConc2000.pdf