Pseudotyping and Culture Conditions Affect Efficiency and Cytotoxicity of Retroviral Gene Transfer to Human Corneal Endothelial Cells

Abstract

Purpose.: To evaluate retroviral vectors as a tool to transduce normal human corneal endothelial cells (HCECs) and to optimize transduction to increase gene transfer efficiency.

Methods.: Enhanced green fluorescent protein (EGFP) encoding retroviral vectors based on HIV-1 or murine leukemia virus (MLV), pseudotyped with either vesicular stomatitis virus glycoprotein (VSV-G) or a modified foamy virus envelope protein (FV Env), and prototype foamy virus (PFV) were produced. Transduction was performed in four HCEC culture media that were previously described for specific cultivation of HCECs or organ culture of donor corneas, namely enriched HCEC growth medium F99_HCEC, its unsupplemented basal medium F99, MEM + 2% fetal calf serum (FCS) (MEM), and Human Endothelial-SFM (SFM). Transduction efficiency was evaluated by marker gene transfer assay, and cytotoxic effects of virus infection were evaluated by means of resazurin conversion assay.

Results.: PFV- and HIV-1–based vectors showed superior transduction efficiency compared with MLV-based vectors. Pseudotyping with a modified FV Env increased transduction efficiency compared with pseudotyping with VSV-G. In medium SFM, transduction efficiency of PFV, HIV-1–/FV Env, and MLV-based vectors was markedly reduced compared with the other culture media. When cells were cultured in F99-based media, cell viability was reduced by retroviral transduction compared with uninfected or mock infected controls, but remained unaffected when cells were cultured in SFM and was even increased when cells were cultured in MEM.

Conclusions.: HIV-1–based vectors pseudotyped with FV Env can efficiently be used to transduce primary HCECs in vitro. However, transduction efficiency is dependent on culture conditions and impairs metabolic activity and viability of HCECs in vitro.

The human corneal endothelium has an extremely reduced proliferative capacity in vivo, hence the endothelial layer cannot regenerate. The potential risk of endothelial decompensation as a response to an immunologic attack after keratoplasty or in case of corneal endothelial dystrophies made several research groups investigate the genetic manipulation of corneal endothelial cells. Considering a possible future therapeutic application of genetically manipulated corneal endothelial cells, the aims of these studies were either to induce proliferation, suppress cell loss during corneal organ storage, or to express genes that modulate immune responses after keratoplasty, as recently reviewed. ^{1 –4} While proof of principle was shown in vitro and in animal transplantation studies, the systems used so far are not applicable to human patients. In recent years, viral transduction of corneal cells came into focus, because viruses, especially retroviruses, developed strategies to effectively insert their genetic information into host cells. Furthermore, their tropism can be manipulated, e.g., by the designed exchange of glycoproteins in viral vectors, termed pseudotyping, using heterologous viral glycoproteins to fit the needs of gene transfer into specific target cell types. Adenoviral vectors are attractive for transducing corneal endothelial cells, because they can be produced in high titers, are able to infect quiescent cells, and can transfer large transgenes. ⁵ It was shown that corneal endothelial cells in situ can selectively be transduced using adeno- and adeno-associated viruses. ^{6 –8} However, production of adenoviruses is laborious and time-consuming, and they are only suitable if transient transgene expression is desired because they do not integrate into the host genome. ^9,10 Furthermore it was reported that high titers of adenoviruses were cytotoxic to corneal endothelial cells and led to a dramatic decrease of cell density. ⁵  

Retroviral vectors integrate their genome into the host and allow a stable transgene expression, which is useful in long-term applications like treatment of genetic disorders, e.g., endothelial dystrophies. Retroviruses that allow transduction of quiescent cells, like lentiviral vectors, are particularly interesting candidates for gene transfer into growth arrested human corneal endothelial cells (HCECs) in situ. ^11,12 Furthermore, they have been shown to efficiently transduce corneal endothelial cells after injection into the anterior chamber and to fail to induce an immune response after intraocular delivery. ^13,14 They can be produced more easily and in shorter time than adenoviruses. Retroviral vectors described for genetic manipulation of corneal endothelial cells were mainly based on different lentiviruses like HIV, equine infectious anemia virus (EIAV) or feline immunodeficiency virus (FIV) ^{4,11,12,15 –18} or gammaretroviruses like murine leukemia virus (MLV) ¹¹ pseudotyped with vesicular stomatitis virus glycoprotein (VSV-G). Using these vectors a stable transduction of corneal endothelial cells could be achieved in vitro in cell culture and in organ culture of whole corneas. First experimental trials in animal models were already performed, using either genetically manipulated corneas for transplantation or direct intravitreal/intracameral application of the viral vectors. ^{11 –13,15,18}  

While the cytotoxic effects of adenoviral transduction on corneal endothelial cells were already described, such information about retroviral transduction and the implication of pseudotyping on efficiency and cytotoxic effects of retroviral gene transfer into human corneal endothelial cells (HCECs) is still missing. We therefore aimed to develop an in vitro culture system for optimal and efficient retroviral gene transfer for human corneal endothelial cells and examined the infectivity and cytotoxicity of five different retroviral vectors, HIV-1 and MLV, each pseudotyped with either VSV-G or a modified foamy virus envelope protein (FV Env), ¹⁹ and prototype foamy virus (PFV). Furthermore, the influence of nutritional condition on gene transfer efficiency and cytotoxicity was examined by infecting and culturing the cells in four different media, because we could recently show that the type of culture medium has a major effect on cell survival under stress conditions. ²⁰ Although a therapeutic application of the retroviral transduction system was not a primary goal of this study, it was designed with respect to a prospective therapeutic application of genetically manipulated HCEC in vitro or in situ. 

Methods

Cell Culture

Cells were maintained at 37°C in a humidified atmosphere containing 5% CO₂. Primary HCECs were isolated from four organ-cultured corneas (mean donor age 19.75 ± 26.02 years, range 0 to 64 years) that were unsuitable for transplantation using the method of Engelmann et al. ²¹ Briefly, corneas were rinsed in PBS, and a solution of 0.5% collagenase IV (Sigma Aldrich, Munich, Germany) in medium F99 (Ham's F12/Medium 199; Biochrom AG, Berlin, Germany) was applied to the endothelial side at 37°C for 90 minutes, followed by incubation in 0.04% collagenase IV for 17 hours. Subsequently, HCECs were collected by several rinses of F99 containing 5% fetal calf serum (FCS) administered to the endothelial side using a sterile syringe and a narrow gauge needle. Cells were centrifuged at 100g for 5 minutes, resuspended in optimized HCEC growth medium F99_HCEC (F99 supplemented with 5% FCS (Gibco Invitrogen, Karlsruhe, Germany), 20 μg/mL ascorbic acid (Sigma Aldrich), 10 ng/mL human recombinant FGF-2 (Gibco Invitrogen), 20 μg/mL human recombinant insulin (Sigma Aldrich), 2.5 μg/mL amphotericin-B and 50 μg/mL gentamicin (Biochrom AG), ^22,23 and plated onto 24- or 12-well plates precoated with laminin (10 μg/mL; Gibco Invitrogen) and chondroitin sulfate (10 mg/mL; Sigma Aldrich) solubilized in F99. Medium was changed every other day. 

Cells were passaged at subconfluence using trypsin/EDTA (0.05%/0.02%; Gibco Invitrogen) with trypsin activity being quenched with medium F99_HCEC. Cells were pelleted by centrifugation at 100g for 5 minutes and plated at a density of 2000 cells/cm² in precoated T25 tissue culture flasks. Primary HCECs of passages P2-P8 were used for experiments. Purity of HCEC cultures was examined by immunohistochemical staining using the HCEC-specific mab 9.3.E ²⁴ (1:50 in PBS + 1% BSA) for 1 hour at room temperature, followed by a phycoerythrin (PE)-conjugated polyclonal donkey anti-mouse IgG as secondary antibody (Jackson Immunoresearch Laboratories, Suffolk, UK) diluted 1:100 in PBS + 1% BSA for 1 hour at room temperature in the dark. In controls, the primary antibody was omitted. Cells were washed twice in PBS, pelleted, and resuspended in 120 μL FACS (fluorescence-aided cell scanning) buffer (PBS + 1% FCS) for flow cytometric analysis (FACSCalibur flow cytometer; Becton Dickinson, Heidelberg, Germany). Data recording and processing was performed (BD CellQuest 3.3, Becton Dickinson; Microsoft Excel 2003, Microsoft Deutschland GmbH, Unterschleissheim, Germany; GraphPad Prism S.03, GraphPad Software, La Jolla, CA; and Adobe Photoshop CS2, Adobe Systems GmbH, Munich, Germany software). 

The human kidney cell line 293T (virus producer cells ²⁵ ) and the human fibrosarcoma cell line HT1080 (transduction control cells ²⁶ ) were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco Invitrogen) supplemented with 10% fetal calf serum and penicillin/streptomycin (Gibco Invitrogen) in T75 or T175 tissue culture flasks. Cells were subcultured using trypsin/EDTA with split ratios of 1:10 or 1:20 every other day. 

Virus Production

Viral vectors based on HIV-1 or MLV were pseudotyped with either VSV-G or a modified FV envelope protein. ^19,27 Vectors were produced using the 3-component-system in 293T producer cells. Replication defective HIV-1 vectors encoding for enhanced green fluorescent protein (EGFP) were produced using p6NST50 ¹⁹ and pCD/NL-BH ²⁸ encoding for Gag/Pol-plasmids. Replication defective MLV vectors encoding for EGFP were produced using pczCFG2fEGN ²⁹ and pHIT60 ³⁰ encoding for Gag/Pol. Both retroviral vectors were pseudotyped with VSV-G using pczVSV-G or a modified foamy virus envelope protein using pczPFVenv PE01. ¹⁹ Prototype FV (FV/FV) vectors were likewise produced in 293T producer cells by means of the 4-component-system using the EGFP expressing transfer vector puc2MD9 SFFVU3WPRE and the expression optimized packaging constructs, pcoPP for PFV Pol, pcoPG4 for PFV Gag, and pcoPE for PFV Env. ^31,32  

Briefly, 293T producer cells were plated in 10-cm dishes at a number of 5 × 10⁶ cells/dish. Virus-containing supernatants were produced by cotransfection of 293T cells with a total of 15 μg DNA using polyethyleneimine (PEI) transfection reagent. For production of HIV-1– or MLV-based vectors equal amounts (5 μg) of Env-encoding plasmid, Gag/Pol-encoding plasmid, and viral vector plasmid were used. For production of PFV vectors an optimized mix ratio of 10.4 μg of puc2MD9SFFVU3WPRE, 1.2 μg of pcoPP, 2.6 μg of pcoPG4, and 0.8 μg of pcoPEwt were used. After 24 hours sodium butyrate was added to the cultures at a final concentration of 10 mol/L for 6 hours, followed by a medium change. After an additional 18 hours virus-containing culture supernatants were harvested and centrifuged at 2200 rpm for 10 minutes to remove cell debris. Supernatants were then quick-frozen in aliquots and stored at −80°C until further use. 

Transduction of HT1080 and HCEC

Cells were plated at a density of 2 × 10⁴ cells per well in 12-well plates, either 24 hours (HT1080, transduction control cells) or 72 hours (HCEC) before infection. HT1080 were infected using 1 mL per well of virus supernatant diluted 1:10, 1:100, 1:1000, and 1:10,000 in Dulbecco's modified Eagle's medium + 10% FCS. HCECs were infected using 1 mL per well of virus supernatant diluted 1:10 and 1:100 in one of four different culture media: (1) HCEC growth medium F99_HCEC, ²² which is enriched with several supplements and specially designed to promote primary HCEC proliferation in vitro, (2) unsupplemented basal medium F99, ^22,33 which was investigated as a basal medium for in vitro cultivation of HCEC and organ cultivation of human donor corneas, (3) standard corneal organ cultivation medium minimal essential medium (Biochrom AG) supplemented with 2% FCS (MEM), ³⁴ and (4) serum-free medium (SFM) (Human Endothelial-SFM; Gibco Invitrogen) supplemented with 10 ng/mL FGF-2, ^20,33,35,36 which is currently being tested in a multicenter clinical trial for corneal organ cultivation (ClinicalTrials.gov, identifier NCT00623584). Uninfected HCECs served as control (uninf). Supernatant from 293T cells that were transfected with the prokaryotic cloning vector pUC19 was used on primary HCECs to check if traces of the transfection reagents may influence their metabolic activity (mock). Medium was changed 4 hours after transduction and cells were subjected to flow cytometric analysis after 72 hours to determine the percentage of marker gene expressing cells as described below. 

Marker Gene Transfer Assay

The percentage of EGFP-positive cells was determined by FACS analysis 72 hours after infection. Transduced cells were trypsinized, pelleted by centrifugation at 1200 rpm for 5 minutes and resuspended in 120 μL FACS buffer on ice for subsequent flow cytometric analysis (FACSCalibur flow cytometer; Becton Dickinson). For each cell type a specific gate was defined on the base of size and granularity of untreated cells (forward/sideward scatter). Fluorescence readings after transduction were taken from 10,000 cells within the defined gate. Data recording and processing was performed (BD CellQuest 3.3, Microsoft Excel 2003, GraphPad Prism S.03, and Adobe Photoshop CS2 software). Titers were calculated from samples with 0.1 to 80% EGFP positive cells. 

Cytotoxicity Assay

Primary HCECs of passages P2-P8 were seeded into 96-well plates (TPP AG, Trasadingen, Switzerland; with an ascending cell density of 10–10,000 cells per well or 30–30,000 cells/cm², respectively) and allowed to attach and spread for 24 hours. Cells were then exposed to test media (F99_HCEC, F99, MEM, and SFM), each containing either virus-free supernatant from 293T cells that were transfected with pUC19 (mock) or virus-containing supernatant (HIV-1/FV or HIV-1/VSV-G) at a dilution of 1:10. Untreated cells served as controls (uninf). After 24 hours test media were replaced by the respective virus-free media and cells were cultured for up to 9 days with medium changes 3 times per week. Cytotoxicity was determined 3 days and 9 days after infection (CellTiter-Blue cell viability assay; Promega, Mannheim, Germany) according to the manufacturer's instructions. Conversion of blue resazurin to pink fluorescent resorufin was measured fluorometrically (SPECTRAFluor Plus plate reader; Tecan, Crailsheim, Germany) using a 545/595nm excitation/emission filter pair, at 3 hours after resazurin addition. Data recording and processing was performed (Magellan 4, Tecan; and Microsoft Excel 2003 software). 

Statistical Analysis

Values are given as mean ± SD. Statistical analysis was performed by univariate one-way ANOVA with correction for post hoc multiple comparisons according to Bonferroni (SPSS 17.1; SPSS, Chicago, IL). Data were grouped by culture medium or by the type of viral vector used, and comparisons were made between the groups as well as between the data within each group. Alpha error was adjusted appropriately and significance was accepted at P < 0.05. 

Results

Transduction Efficiency of the Retroviral Vector Systems

Five different retroviral vectors were produced in 293T producer cells, either based on MLV or HIV-1 and pseudotyped with either VSV-G or FV, and PFV (FV/FV). HT1080 cells and primary HCECs were infected with each of the five retroviral vectors using plain virus-containing 293T supernatant and dilutions thereof. In some experiments, an aliquot of the transduced HCECs was additionally stained with the HCEC-specific mab 9.3.E. Double fluorescence of 9.3.E positive and transduced (EGFP-labeled) cells was proven by flow cytometry, confirming that the EGFP-labeled cells were HCECs and not contaminating trabecular meshwork or corneal stromal cells (Fig. 1). 

Figure 1.

 

Flow cytometric analysis of 9.3.E-stained HCEC cultures. Primary HCECs were transduced with an EGFP-encoding retroviral vector (MLV/VSV-G), stained with the HCEC-specific mab 9.3.E followed by a phycoerythrin (PE)-conjugated secondary antibody, and analyzed by FACS. Left: all treated cells show positive staining for 9.3.E antigen and can be divided into an infected (EGFP-positive) and a noninfected (EGFP-negative) subpopulation. Right: control, the primary antibody was omitted. The secondary antibody gave no unspecific staining reaction and all cells appeared PE-negative. Again, the culture can be divided into an infected and a noninfected subpopulation after treatment with the retroviral vector. FACS analysis displayed as dot plot; data from one representative experiment.

Relative infection rates of the vectors were evaluated using the marker gene transfer assay and titers were calculated. In HT1080 control cells (Fig. 2) titers of HIV-1 vectors were significantly higher than those of MLV vectors, and FV pseudotyped HIV-1 or MLV vectors yielded significantly higher titers than their VSV-G pseudotyped counterparts. Titers ranged from 3.0 × 10⁵ (MLV/VSV-G) to 1.95 × 10⁷ (HIV-1/FV). These observations were even more pronounced in HCECs, with titers being slightly lower than in HT1080. Overall, HIV-1/FV titers were significantly higher than those of the other vectors, and titers of MLV/VSV-G were significantly lower than those of HIV-1/FV and FV/FV (Fig. 3A). Transduction efficiency in primary HCECs was also depending on the medium in which the cells were cultured for infection (Fig. 3B). Overall titers obtained in medium SFM (ranging from 1.5 × 10³ for MLV/VSV-G to 9.3 × 10⁵ for HIV-1/VSV-G) were significantly lower than in growth medium F99_HCEC (ranging from 6.3 × 10⁴ for MLV/VSV-G to 4.6 × 10⁶ for HIV-1/FV) or in unsupplemented basal medium F99 (ranging from 1.6 × 10⁴ for MLV/VSV-G to 5.4 × 10⁶ for HIV-1/FV). 

Figure 2.

 

Infection rates of retroviral vectors in HT1080. HT1080 were infected with virus-containing 293T supernatants, analyzed by marker gene transfer assay and titers were calculated. Viral vectors used were HIV-1 and MLV pseudotyped with either VSV-G or FV Env, and plain PFV vectors (FV/FV). Mean ± SD from n = 8; curly brackets: grouped data, squared brackets: data pairs with significant difference (P < 0.05).

Figure 3.

 

Infection rates of retroviral vectors in primary HCEC. Primary HCECs were infected with virus-containing 293T supernatants of the same virus production lot that was used to infect HT1080 in preceding experiments. Transduction efficiency was determined using the marker gene transfer assay and titers were calculated. Viral vectors used were HIV-1 and MLV pseudotyped with either VSV-G or FV Env, or plain PFV vectors (FV/FV). HCECs were cultured in four different media and transduction efficiency was determined in dependence on vector and culture medium. (A) Titers in primary HCECs, data grouped by type of retroviral vector. (B) Titers in primary HCECs, data grouped by culture media. Mean ± SD from n = 8 (four donors); curly brackets: grouped data, squared brackets: data pairs with significant difference (P < 0.05).

In addition, HCECs that were transduced in medium F99_HCEC using a 1:10 dilution of recombinant vector supernatant were examined by fluorescence microscopy after 72 hours (Fig. 4 micrographs). The percentage of fluorescent cells was higher in cultures that were transduced with the modified FV Env pseudotyped vectors than in those cultures that were transduced with VSV-G pseudotyped vectors (Fig. 4 bar chart). Furthermore, HCECs transduced with HIV-1/VSV-G or HIV-1/FV pseudotypes were subcultured and transgene expression was observed to be stably retained for up to 9 passages (data not shown). 

Figure 4.

 

EGFP fluorescence of transduced HCECs after subcultivation in medium F99_HCEC. Primary HCECs were transduced with HIV-1 or MLV pseudotyped with either FV or VSV-G, or were transduced with plain PFV vector (FV/FV) using 1:10 dilutions of virus-containing 293T supernatant. Successful transduction of cells was visualized by EGFP fluorescence in the cells 72 hours after infection by fluorescence microscopy and flow cytometry. The micrographs depict cells at 72 hours after transduction viewed by differential interference contrast (DIC) (left) and fluorescence microscopy (right). The bar chart shows percentages of EGFP-positive cells. Transduction with MLV/VSV-G yielded significantly fewer EGFP-positive cells than transduction with HIV-1–based vectors and pseudotyping with FV Env. Scale bar 100 μm. Mean ± SD from n = 8 (four donors); curly brackets: grouped data, squared brackets: data pairs with significant difference (P < 0.05).

Cytotoxicity of Retroviral Transduction

Plain PFV vectors showed higher macroscopic cytotoxicity than the other vectors, as seen by considerable formation of syncytia (data not shown), and were therefore not evaluated further. HIV-1 based vectors yielded higher infection rates than MLV-based vectors. For this reason they were considered to be the most promising candidates for future analyses. Thus the effects of transduction with HIV-1 based vectors on HCEC survival were further characterized using the resazurin conversion assay for cytotoxicity. Preliminary experiments revealed that repeated exposure of primary HCECs to resazurin impaired cell viability, so that the assay was used for end point measurements only. In a first experimental set-up, cells were seeded at densities of 10–10,000 cells per well in 1:2 serial dilutions and resazurin conversion was measured in uninfected cells after a 9-day cultivation period in the four different test media. Resazurin conversion correlated in a linear fashion with cell seeding density in a culture medium-dependent fashion, ranging from 312 to 2500 (F99_HCEC) to 2500–10,000 (MEM) cells per well with a coefficient of determination R ² > 0.98 (Fig. 5). A cell seeding density of 2500 cells per well was chosen for infection experiments, because correlation of resazurin conversion was linear to this cell seeding density in all tested culture media. 

Figure 5.

 

Resazurin conversion by primary HCECs in different culture media. Primary HCECs were seeded at densities of 10–10,000 cells per well (1:2 serial dilutions) in medium F99_HCEC in a 96-well plate. Medium was switched to the four test media after 24 hours. Cells were cultured for an additional 9 days and then subjected to resazurin conversion assay. Black lines represent the range where resazurin conversion is linearly correlated to cell seeding density (linear regression, coefficient of determination R ² > 0.98 for each medium). Mean ± SD from n = 8 (four donors).

For assaying retroviral cytotoxicity, primary HCECs were seeded into 96-well plates at a density of 2500 cells per well in medium F99_HCEC. After 24 hours cells were exposed to the four different test media and infected with HIV-1/FV or HIV-1/VSV-G, using a 1:10 dilution of recombinant vector supernatant. HCEC treated likewise with virus-free 293T supernantant (mock) or remaining uninfected (uninf) served as controls. Metabolic activity of all cells was examined after 3 and 9 days (Fig. 6). HCECs cultured in nutrient-rich media (F99_HCEC and SFM) had a persistent significantly higher metabolic activity than HCECs cultured in nutrient-poor media (F99, MEM). Uninfected HCECs survived well in F99_HCEC and SFM, but died in MEM during the observation period. After mock infection, HCECs in F99_HCEC, F99, or MEM showed a considerable increase in metabolic activity compared with their uninfected counterparts, but only a small increase in SFM. Metabolic activity of HCECs cultured in F99_HCEC decreased below the level of untreated cells after transduction with HIV-1/FV, but not after transduction with HIV-1/VSV-G, pointing to a cytotoxic effect of HIV-1/FV in this medium. In the other three media, VSV-G pseudotyped vectors showed a greater cytotoxicity than the FV Env pseudotyped vectors. More strikingly, metabolic activity of HCECs cultured in MEM was significantly increased after transduction with HIV-1/FV and markedly increased after transduction with HIV-1/VSV-G compared with uninfected controls, indicating that the cells seemed to benefit from retroviral transduction under starvation conditions. HCECs cultured in SFM showed only moderate changes in metabolic activity after retroviral transduction compared with uninfected or mock transduced controls, indicating that the cells seem to be unaffected by retroviral treatment in this medium. 

Figure 6.

 

Metabolic activity of HCECs after transduction in different culture media. Primary HCECs were seeded at densities of 2500 cells per well in 96-well plates. After 24 hours, cells were incubated with the respective fresh culture medium (uninf), medium supplemented with 10% virus-free 293T supernatant (mock), or medium supplemented with 10% virus-containing 293T supernatant (HIV-1/VSV-G or HIV-1/FV). Metabolic activity was determined by resazurin conversion assay after (A) 3 and after (B) 9 days of culturing. Mean ± SD from n = 8 (four donors); curly brackets: grouped data, squared brackets: data pairs with significant difference (P < 0.05).

Discussion

It was previously shown that the corneal endothelium can be transduced in situ by retroviral gene transfer, e.g., to counteract the risk of endothelial decompensation after keratoplasty and thus to reduce the need for second transplants. Retroviral transduction may also be applied in vitro on cultured HCECs and can be combined with tissue engineering techniques such as corneal endothelial sheet design or cell transplantation onto denuded donor corneas. ^37,38 Because HCECs are growth-arrested in situ but show considerable proliferative capacity in vitro, we chose MLV and PFV, which transduce only proliferating cells, and HIV-1, which transduces proliferating as well as quiescent cells, as vectors for transduction of HCECs. Contrasting to MLV and HIV, PFV has been shown to reside in quiescent cells for longer times without losing its capacity to complete a productive infection, hence yielding efficient transduction rates also in cells with reduced proliferative activity such as hematopoietic stem cells. ³⁹ In our experiments HIV-1 based vectors and PFV vectors showed markedly higher transduction efficiencies than the MLV based vectors. Furthermore we could successfully improve transduction efficiency by pseudotyping the retroviral vectors with a modified foamy virus envelope protein. ¹⁹  

We also examined the effect of producer cell supernatant on HCECs, because the supernatant is depleted of nutrients and may contain residual substances of the transfection mix, which might exert an additional negative effect on cell survival after retroviral transduction. However, we observed a beneficial effect of the producer cell supernatant on HCECs especially under serum-reduced conditions. These results indicate that exposure to a 10% 293T-supernatant promoted cell viability and proliferation of the cells similar to a conditioned medium, and that a possible adverse effect of residual reagents of the transfection mix on HCECs could be excluded. 

Formerly, our own studies focusing on the influence of different culture media on HCEC in vitro demonstrated that culture media can have a major effect on cell survival and their ability to counteract stress inducers. ^20,35 We could show that HCECs cultured in MEM + 2% FCS, a standard corneal organ culture medium, are prone to apoptotic cell death, while HCECs cultured in Human Endothelial-SFM, a serum-free but nutrient-rich medium, were protected from cell death even when the apoptosis inducer staurosporine was administered. ²⁰ Hence it was hypothesized and thereafter proved by the data presented here, that culture conditions may also have a marked effect on transduction efficiency and cell survival after retroviral transduction. We observed that the FV Env pseudotyped lentiviral vectors were less cytotoxic than the VSV-G pseudotyped vectors under reduced nutritional or serum-free conditions. This makes them especially applicable in situ in serum-reduced or serum-free media, e.g., media that are used for corneal organ cultivation. It can therefore be assumed that pseudotyping lentiviral vectors with FV Env greatly improves retroviral transduction of HCECs, because FV Env increases transduction efficiency while reducing cytotoxicity of the vector. On the long way toward a therapeutic application of genetically manipulated cells, all aspects of a transduction system, like type of vector and glycoprotein, culture medium, and cell type need to be addressed. Then such a system can be a useful tool to transduce cells for a number of research purposes, e.g., the characterization of functional features or cell biological aspects, or the development of strategies to improve transplant survival. 

Another interesting application of retroviral gene transfer could be the insertion of proliferation-promoting genes into corneal endothelial cells to overcome their postmitotic status. However, insertion of proliferation-promoting genes requires transient systems to avoid tissue hyperplasia. Recently, transient retroviral systems were described which are based on either MLV lacking the retroviral primer binding site or integrase-defective HIV-1 viruses. ^40,41 These vectors cannot integrate into the host genome, but allow transgene expression as a consequence of direct translation of the viral RNA or episomal circular DNA. The methods described here may be a suitable basis to establish retroviral-based methods for such future therapeutic goals. 

Footnotes

 Supported by Grants from the BMBF (01ZZ0102) and DFG (Li621/4-2, Li621/6-1) (DL) and DFG (CRTD seed grant) (DL and KE).

Footnotes

 Disclosure: M. Valtink, None; N. Stanke, None; L. Knels, None; K. Engelmann, None; R.H.W. Funk, None; D. Lindemann, None