Inhibition of Human Immunodeficiency Virus-1 (HIV-1) Replication After Transduction of Granulocyte Colony-Stimulating Factor–Mobilized CD34⁺ Cells From HIV-1–Infected Donors Using Retroviral Vectors Containing Anti–HIV-1 Genes

GENE THERAPY was initially conceived as a method for the correction of congenital diseases. However, infectious diseases, including human immunodeficiency virus-1 (HIV-1), may also be candidates for a gene-therapy approach.1,2 Using retroviral vectors to insert genes that block viral replication may lead to protection from virus-induced cytotoxicity, limit HIV-1 spread, and thereby allow protected cells to assume their normal function within the immune system. To mediate long-term HIV-1 resistance, anti-HIV-1 genes may need to be inserted into human hematopoietic stem cells (HSC). HSC produce all the cells of the immune system which are involved in acquired immunodeficiency syndrome (AIDS) pathogenesis, such as CD4⁺ T lymphocytes, monocytes/macrophages, dendritic cells, and microglia cells. Hypothetically, newly arising T lymphocytes and cells of the monocyte/macrophage lineage would be protected from the potentially destructive events associated with HIV-1 infection if they carry anti–HIV-1 genes inherited from transduced HSC.

Several strategies have been developed to construct retroviral vectors transferring anti–HIV-1 activity, leading to suppression of HIV-1 in transduced cells. In this report, three different Moloney murine leukemia virus–based retroviral vectors were tested for their anti–HIV-1 properties in an in vitro system: a vector acting as an RRE-decoy, over-expressing the Rev responsive element,3 a vector encoding a double hammerhead ribozyme, cleaving the tat and rev transcript of HIV-1,4,5 and a vector carrying the message for a trans-dominant mutant Rev M10 protein competing with the wild-type HIV-1 Rev protein.6,7 

Hematopoietic stem cells are normally present in postnatal bone marrow (BM) and neonatal umbilical cord blood; the HSC may be enriched by immunoselection for cells expressing the CD34 antigen. Another method to readily obtain human HSC is mobilization of CD34⁺ cells into the peripheral blood (PB) using granulocyte-colony stimulating factor (G-CSF ). Apheresis of white blood cells can then be applied to obtain a sufficient number of CD34⁺ PB stem cells (PBSC).

It has been shown in our laboratory that retroviral vectors may be used to successfully transfer anti–HIV-1 genes into BM or cord blood CD34⁺ cells from uninfected individuals.3 Long-term BM cultures (LTBMC) established from these transduced CD34⁺ cells and challenged with the macrophage-tropic HIV-1 isolate JR-FL8 exhibit strong suppression of HIV-1 replication.3 Based on these results, we tested the feasibility of applying gene therapy to individuals already infected with HIV-1. HIV-1–infected volunteers were subjected to mobilization of PBSC using G-CSF, followed by apheresis and CD34⁺ cell separation. CD34⁺ cells were transduced with the retroviral vectors carrying genes conferring anti–HIV-1 activity. The clonogenic ability of transduced CD34⁺ cells was evaluated, and LTBMC were established. Subsequent challenges with HIV-1_JR-FL and a primary HIV-1 isolate were performed to evaluate suppression of HIV-1 replication in transduced cells.

Cells and virus.PBSC were obtained from six HIV-1–infected volunteers at the City of Hope Medical Center after approval by the Institutional Review Board at the City of Hope, and evaluation of the medical status according to strict exclusion criteria. Subjects were HIV-1 seropositive adults (>18 years of age) with CD4 counts > 200/μL and no AIDS-defining conditions. To mobilize PBSC, volunteers were treated with G-CSF (10 μg/kg) for 5 days; apheresis was performed and the CD34⁺ cell fraction was separated using the CEPRATE SC Stem Cell Concentrator (CellPro, Seattle, WA). After overnight storage at room temperature, separated CD34⁺ cells were transported to the Gene Therapy Laboratory at Childrens Hospital Los Angeles. PBSC were grown in I20 (Iscove's modified Dulbecco medium [Irvine Scientific, Santa Ana, CA], 20% fetal bovine serum [FBS; Bio-Whittaker, Walkersville, MD], 1% deionized bovine serum albumin [Sigma, St Louis, MO], 2 mmol/L L-glutamine [Irvine Scientific], 50 U/mL penicillin [Irvine Scientific], 50 μg/mL streptomycin [Irvine Scientific], 10⁻⁶ mol/L hydrocortisone [Sigma], and 10⁻⁴ mol/L 2-mercaptoethanol [Sigma]).

The amphotropic packaging cell line PA3179 was provided by A. D. Miller (Fred Hutchinson Cancer Center, Seattle, WA) and maintained in D10 (Dulbecco's Modified Eagle Medium, high glucose [Irvine Scientific], 10% FBS, 50 U/mL penicillin, 50 μg/mL streptomycin, 2 mmol/L L-glutamine). The murine stromal line S17 was provided by Kenneth Dorshkind (University of California, Riverside),10 and maintained in D10. Primary human BM stroma was grown in DOM (Dexter's original medium) consisting of Iscove's modified Dulbecco medium, 15% FBS, 15% horse serum, 2 mmol/L L-glutamine, 50 U/mL penicillin, 50 μg/mL streptomycin, 10⁻⁶ mol/L hydrocortisone, and 10⁻⁴ mol/L 2-mercaptoethanol. All of the above cells were maintained at 37°C and 5% CO₂ in humidified incubators.

HIV-1_JR-FL viral stock8 was initially purchased from Applied Biotechnologies, Inc (Columbia, MD) and had a titer of 10⁵ TCID₅₀/mL (according to manufacturer, titer obtained on primary human monocytes). To expand the viral stock, HIV-1_JR-FL was inoculated into human PB mononuclear cell (PBMC) cultures stimulated for 2 days with phytohemagglutinin (PHA); supernatant was monitored for sufficient viral production by repeated HIV-1 core protein (p24) assays. At the peak of p24 production, culture supernatant was obtained, aliquoted, and frozen in liquid nitrogen. A frozen aliquot was subsequently thawed and titered on human PB monocytes/macrophages. The titer of the HIV-1_JR-FL used for the studies was 10⁵ TCID₅₀/mL. A low-passage, primary HIV-1 isolate was obtained by coculturing PBMC from subject 3 with 2-day PHA-stimulated PBMC in a 1:10 ratio. Essentially, the same method as described before was applied to harvest, store, and titer HIV-1–containing culture supernatant. Titers of the primary isolate in PB monocytes were 1 × 10⁶ TCID₅₀/mL. Estimated multiplicities of infection (MOI) in LTBMC infection assays were based on these TCID₅₀ values, divided by the total number of cells in culture, regardless of the frequency of susceptible cells.

Anti–HIV-1 gene retroviral vectors.The control retroviral vector used for this study was the vector LN,11 containing the neomycin resistance gene driven by the long terminal repeat (LTR) of the Moloney murine leukemia virus (MoMuLV). Anti–HIV-1 gene vectors include the vector L-RRE-neo, containing a small fragment of the HIV-1 Rev Responsive Element (RRE) (bp 7813 to 7853 from the SF2-isolate) acting as a decoy for REV binding3 subcloned into the Bcl I site 5′ of the neo gene in the LN vector; the vector L-TR/TAT-neo, containing the message for two hammerhead ribozymes,4,5 cleaving in the tat and rev region of the HIV-1 RNA, cloned into the Bcl I site immediately 5′ of the neo gene in the LN vector; and the vector L-M10-SN containing the gene for the trans-dominant rev mutant M10,6 cloned in the LXSN vector,11 where the neomycin resistance gene is under transcriptional control of the SV40 promoter (see Fig 2).

Establishment of stromal cell layers to support gene transfer and LTBMC.BM stroma cells from HIV-1–infected and uninfected individuals were grown by diluting 1 mL of fresh BM aspirate in DOM and letting the cells adhere overnight to the plastic surface of a 75-cm² tissue culture flask (Costar, Cambridge, MA). After removal of the nonadherent cell fraction, adherent cells were allowed to expand for approximately 14 days. Cells were passaged at least four times before being used as support for gene transfer. One day before retroviral transductions, stroma cells were mobilized by trypsinization (Irvine Scientific), irradiated (20 Gy), and 1 × 10⁵ cells were plated into 25-cm² tissue culture flasks.

G418-resistant S17 mouse stromal cells were produced by retroviral transduction of S17 cells in log-growth phase by exposure to high titer (2 × 10⁶ colony-forming units [CFU]/mL) LN vector-containing supernatant. After overnight exposure, cells were G418 selected (0.75 mg/mL active concentration) for 14 days. Selected cells were aliquoted and frozen for later use in G418 selection of transduced CD34⁺ cells.

Establishment of vector-producing cell lines and collection of high-titer supernatant.Clones of high-titer supernatant-producing cell lines were produced as described previously.3,12 High-titer vector-containing supernatant was collected after conditioning of D10 for 48 hours on a confluent layer of PA317 producer cells maintained at 32°C and 5% CO₂ . A sensitive reverse transcriptase (RT) assay was performed on fresh supernatants to estimate the number of vector particles; only collections with sufficient RT activity were aliquoted and frozen at −70°C. After thawing, supernatants were filtered (0.45 μm) and titers were determined on 3T3 cells by assessment of the amount of neomycin resistance transfer. Titers in all vector supernatants ranged from 1 × 10⁶ to 5 × 10⁶ CFU/mL. Vector supernatants were free of replication-competent retrovirus, assayed as described previously.13 

Production of vector-transduced monocytic cells derived from PBSC.Isolated CD34⁺ cells (1 × 10⁵/mL) were transduced for 72 hours in 10 mL of a 1:1 mixture of vector-containing supernatant and cytokine-containing 2× transduction medium14 on irradiated (20 Gy) allogeneic BM stromal layers, from HIV-1–infected or uninfected individuals. 2× transduction medium consisted of I20 plus stem cell factor (SCF, 100 ng/mL), human interleukin-6 (IL-6) (50 ng/mL), and human IL-3 (10 ng/mL). MOI for vector particles ranged from 5 to 25. CD34⁺ cells were collected, pelleted, and resuspended in fresh transduction medium plus vector supernatant five additional times (twice daily) over a period of 3 days. After transduction, CD34⁺ cells were washed and cultured on irradiated, G418-resistant S17 cells (LN-S17) in I20 medium containing 50 ng/mL SCF, 25 ng/mL IL-6, and 5 ng/mL IL-3. Selection for CD34⁺ cells containing the neomycin resistance gene was performed over a time period of 12 days at an active G418 concentration of 0.75 mg/mL. Culture medium was replaced every 4 days. After the selection period and before HIV-1 challenge, cells were allowed to recover for 4 days in medium without G418.

Evaluation of colony-forming ability and gene transfer into CD34⁺ cells.To determine the number of colonies formed by progenitor cells and the percentage of progenitor cells transduced by the retroviral vectors, CD34⁺ cells were grown in colony assays in the presence and absence of G418.15 The active G418 concentration was 0.9 mg/mL. Aliquots of cells (1 × 10³ and 2 × 10³) were plated in quadruplicates immediately after the 72-hour transduction period into semisolid methylcellulose medium supporting the growth of progenitor cells. After 14 days, colonies containing more than 50 cells were counted. The numbers of colonies formed in the presence of G418 divided by the numbers of colonies formed in the absence of G418 multiplied by 100 reflects the percentage of colonies displaying neomycin resistance. The colony-forming ability and gene transduction of progenitor cells obtained either from cultures following G418 selection or from LTBMC propagated over a period of 5 weeks in the absence of G418 were evaluated by plating 0.5 × 10⁵ and 1 × 10⁵ cells into colony assays in the presence and absence of G418. Colony counts and percentages of G418-resistant colonies were determined as described above.

HIV-1 infection of monocytic cells derived from transduced CD34⁺ cells.Monocytic cells derived from transduced, G418-selected LTBMC were infected with HIV-1 by exposing 1 × 10⁶ cells to HIV-1_JR-FL stock at an MOI of 0.01 TCID₅₀ or primary isolate at an MOI of 0.1 TCID₅₀ in a total volume of 200 μL; after 1 hour of incubation at 37°C and 5% CO₂ , 300 μL of I20 culture medium was added. The cells were then incubated overnight under the same conditions. On the following day, the infected cells were washed three times with IMDM, resuspended in 5 mL of I20 containing 50 ng/mL SCF, 25 ng/mL IL-6, and 5 ng/mL IL-3, and seeded into 25-cm² tissue-culture flasks containing an irradiated layer of S17 murine stromal cells. Supernatant samples for HIV-1 core protein (p24) evaluation were obtained weekly and stored at −70°C.

HIV-1-p24 antigen immunoassay.Coulter p24 assay test kits were purchased from Immunotech (Westbrook, ME), and the assay was performed according to manufacturer's instructions. Assay plates were evaluated on a Dynatech MR5000 ELISA reader (Dynatech, Chantilly, VA); samples with optical density (OD) values exceeding the range of the standard curve obtained with the provided p24 standard were retested after serial dilutions in I20 to obtain readable values.

Reverse transcriptase (RT) assay.A sensitive RT assay protocol was applied to detect RT activity in MoMuLV-based vector supernatants, and culture supernatants of HIV-1–infected cells. Briefly, a reaction mix of 50 μL contained 50 mmol/L Tris HCl (pH 7.8), 62.5 mmol/L KCl, 4.2 mmol/L MgCl₂ , 3.3 mmol/L dithiothreitol (DTT), 0.83 mmol/L EGTA (pH 7.8), 0.083% NP40 (Pierce, Rockford IL), 4.2 μg/mL Poly AdT15_mer (Boehringer-Mannheim, Chicago IL), and 15 mmol/L (α³²P) dTTP (stock 3,333 Ci/mmol) (Amersham, Arlington Heights, IL). Ten microliters of supernatant to be tested was added to the reaction mix and incubated for 2 hours at 37°C. Ten microliters of incubated reaction mix and sample were spotted onto DE81 ion exchange paper strips (Whatman, Maidstone UK) and dried at room temperature. The paper strips were washed four times in 2X saline-sodium citrate (SSC), fixed in 96% ethanol, and dried again at room temperature. The amount of radioactivity remaining on the strips was quantified using a liquid scintillation counter, as a measurement of the amount of RT activity present in the sample.

Semi-quantitative polymerase chain reaction (PCR).PCR samples were assayed from whole-cell lysates for the presence of HIV-1 provirus.16 Cell samples were pelleted and lysed in 10 mmol/L Tris-HCl (pH 8.0), 50 mmol/L KCl, 1.25 mmol/L MgCl₂ , 0.45% NP40, 0.45% Tween, and 0.75 μg/μL Proteinase K. Samples were digested overnight at 37°C, followed by heat inactivation for 10 minutes at 95°C. After digestion, DNA concentrations were quantitated using a TK0100 DNA Mini Fluorometer (Hoefer Scientific Instruments, San Francisco, CA). PCR was performed in a solution with a final concentration of 20 μmol/L each dNTP, 1 mmol/L MgCl₂ , 1× PCR buffer (Perkin Elmer, Emeryville CA), 1 pmol/μL of each primer from the gag region of HIV-1_JR-FL sequence (A: 5′-GTGCGAGAGCGTCAGTATTA-3′ and B: 5′-TGGCTTGCTCTTCCTCTATC-3′ ), and 2.5 U Taq polymerase. Each reaction consisted of a final volume of 50 μL with 300 ng of DNA. Samples were processed for an initial cycle at 94°C for 5 minutes, 63°C for 2 minutes, and 72°C for 2 minutes, followed by 29 cycles at 94°C for 1 minutes, 63°C for 1 minute, and 72°C for 1 minute. PCR products were stored at 4°C until separated on a 2% agarose gel. DNA was transferred to a nylon membrane via capillary transfer and hybridized with an internal oligonucleotide (5′-A AAC GTT CTA GCT CCC TGC TTG CCC-3′ ) end-labeled with ³²P-γ-dATP using deoxynucleotide kinase.

PCR products were quantitated for HIV-1 frequency by comparison with a logarithmic standard curve, derived from ACH-2 cells. ACH-2 cells contain one proviral copy of the integrated HIV-1 virus and were diluted serially with CEM cells to obtain PCR products with a frequency of 100%, 10%, 1%, 0.1%, 0.01%, and 0.001% HIV-1⁺ cells. The standard curve was made by serial dilution using a total of 1 × 10⁸ cells per dilution. Cell mixtures were subsequently lysed and digested as described above. For every PCR run, a complete set of standards was processed in parallel to the patient samples.

Development of anti–HIV-1 retroviral vectors.The retroviral vectors L-RRE-neo, L-TR/TAT-neo, and L-M10-SN were established as reported previously.3-6 Each vector was packaged using the amphotropic packaging cell line PA317. Before applying these retroviral vectors for the transduction of PBSC, comparisons in transduced CEM T lymphocytes were performed demonstrating their capability of suppressing HIV-1 replication.3-6 Also, CD34⁺ cells from HIV-1⁻ individuals were transduced with these vectors and tested for their ability to suppress HIV-1 replication.3 

In vitro model for the evaluation of inhibition of HIV-1 replication in human myelomonocytic cells derived from CD34⁺ cells of HIV-1–infected individuals.Based on the findings in the HIV-1 challenge experiments performed on CD34⁺ cells from HIV-1⁻ individuals,3 the possibility of suppression of HIV-1 replication in transduced CD34⁺ cells derived from HIV-1–infected individuals was further evaluated. CD34⁺ cells were obtained from HIV-1–infected volunteers at the City of Hope Medical Center after mobilization with G-CSF for 5 days and subsequent apheresis (Fig 1). CD34⁺ cells were isolated using the CEPRATE SC Stem Cell Concentrator. After overnight storage, CD34⁺ cells were transduced with three different retroviral vectors mediating neomycin resistance in conjunction with anti–HIV-1-activity, or a control vector mediating neomycin resistance only. Mock transductions with medium only were also performed. After the transduction period, cell aliquots were plated in colony assays in the presence and absence of G418 for evaluation of transduction efficiency and potential toxic effects from transduction. Aliquots of the transduced cells were used to initiate LTBMC without G418. Samples were analyzed after 5 weeks of LTBMC in colony assays, with or without G418, to assess the persistence of transduced progenitors. The remaining cell aliquots were subjected to G418 selection on a layer of mouse stromal cells, S17, known to support growth of CD34⁺ cells and not to be susceptible to HIV-1 infection, since infection of human stroma with HIV-1 may impair hematopoiesis.17 To resist the toxic effects of G418, the stromal layer consisted of S17 cells previously transduced with the LN vector and G418 selected. Cytokine-containing culture medium was completely replaced every 4 days, and new G418 was added. After 12 days of selection and 4 days of recovery in G418-free medium, highly enriched populations of G418-resistant cells, also carrying anti–HIV-1 activity, were obtained. Cells were then evaluated for the percentages of G418-resistant progenitors and for the efficacy of the anti–HIV-1 genes to inhibit HIV-1 replication. To measure the relative increase in the amount of neomycin-resistant cells, samples were assayed in colony assays with or without G418. To evaluate the efficacy of the transferred anti–HIV-1 activity, the selected cell populations were exposed to HIV-1_JR-FL or a low-passage, primary HIV-1 isolate. After overnight exposure to the HIV-1, cells were washed and grown in LTBMC on a layer of S17 stromal cells, and monitored for p24 output weekly for up to 60 days.

Analysis of clonogenic activity of CD34⁺ cells.To assess potential toxicities from retroviral vectors, CD34⁺ cells were plated in colony assays 72 hours after transduction to estimate the number of hematopoietic progenitor cells giving rise to colonies of mature blood cells. In seven experiments, similar numbers of CFU-GM were observed in the absence of G418 for an individual donor from CD34⁺ cells transduced with either LN, L-RRE-neo, L-TR/TAT-neo, and L-M10-SN, compared with mock transduced CD34⁺ cells (Table 1).

To evaluate potential toxic long-term effects from the retroviral vectors, the ability of progenitor cells to differentiate along the myelomonocytic lineage after 5 weeks of LTBMC were evaluated by plating in myeloid colony assays. Among four experiments, very similar numbers of colonies were produced from the vector-transduced cultures and the mock transduced controls (Table 2).

Assessment of transduction efficiency.Because all of the vectors contained the neomycin resistance gene (Fig 2), transduction efficiency may be estimated by determining the percentage of G418-resistant CFUs. Among seven experiments, the extent of gene transfer into clonogenic CD34⁺ cells ranged from 9% to 75% G418-resistant CFU-GM, with each subject showing a characteristic susceptibility to transduction that did not vary with each of the vectors (Table 1). There were only slight variabilities in transduction efficiency among the vector groups, with the average transduction for all samples between 25% and 28% (Fig 3). Transduction was also performed for two samples using stromal cells derived from the marrow of HIV-1⁺ donors; the efficiencies of transduction for these samples were within the range seen using stroma from uninfected donors.

To evaluate the percentages of gene-containing progenitor cells which remain clonogenic after LTBMC, cells were obtained after 5 weeks of LTBMC, plated in colony assays, and their abilities to form colonies in the presence and absence of G418 were evaluated. Among four experiments, the percentage of G418-resistant progenitors ranged from 17% to 33% (Table 2). Again, the average percentages of gene-containing progenitors present after 5 weeks of LTBMC were similar with each of the vectors (23% to 28%) and showed no decline compared with the percentage of the progenitors containing the vectors when evaluated immediately after transduction (Fig 3). These results show that the presence of the vectors in the progenitor cells is not detrimental to their survival in long-term culture, nor to their ability to subsequently form CFU-GM colonies.

HIV-1 challenge of transduced, G418-selected cultures.From all six HIV-1⁺ volunteers participating in this study, CD34⁺ cells could be obtained, transduced, G418 selected, and challenged with HIV-1. Colony assays performed on the G418-selected CD34⁺ cells showed 74% to 90% G418-resistant colonies, demonstrating the efficacy of drug selection (Table 3).

Five challenge experiments with HIV-1_JR-FL and three with the HIV-1 primary isolate were successfully performed and were followed for up to 60 days. In the control cultures, mature myelomonocytic cells produced in LTBMC from LN transduced, G418-selected CD34⁺ cells were capable of supporting vigorous growth of HIV-1_JR-FL (Fig 4A). HIV-1 core protein (p24) output reached into the nanogram/milliliter levels. In the LTBMC established from CD34⁺ cells transduced with the anti–HIV-1 vectors and G418 selected, delayed growth of HIV-1_JR-FL and significantly reduced peak output levels of p24 were observed (Fig 4A). In five HIV-1_JR-FL challenge experiments, 44.2% to 99.9% reductions of p24 output were achieved using the RRE-decoy vector, 75.4% to 99.9% reductions were seen with the ribozyme vector, and 98% to 99.9% reductions of p24 output were observed in the group with the M10 vector, compared to the cultures with the LN vector (Table 4).

The challenges with the HIV-1 primary isolate were done simultaneously with the HIV-1_JR-FL challenges, using portions of the same G418-selected LTBMC. The control cultures with the LN vector were capable of producing p24 output into the nanogram/milliliter levels from the primary isolate (Fig 4B). LTBMC established from CD34⁺ cells transduced with the anti–HIV-1 vectors showed significantly delayed growth of the HIV-1 primary isolate, and strongly decreased peak p24 output (Fig 4B). In three challenges with the HIV-1 primary isolate, 73.7% to 99.9% reductions of p24 output could be seen in the RRE-decoy group, 99.6% to 99.9% reductions in the ribozyme group, and 98.6% to 99.9% reductions in the M10 group (Table 4).

Assessment of cell viability in HIV-1–infected LTBMC.To rule out the possibility that poor cell survival of G418-selected LTBMC was the cause of the reduced p24 output, nonadherent cells obtained from cultures 30 to 40 days after the start of HIV-1 challenges were counted for viability, and plated in colony assays. Good viability (>90% trypan blue exclusion) and clonogenic ability of transduced, G418-selected cells was observed (70 to 110 CFU-GM per 1 × 10⁵ cells). Seventy-five percent to 85% of the colonies derived from these cells were resistant to G418. No significant differences could be observed by microscopy, comparing the cell densities and appearances of the transduced and G418-selected adherent cells (data not shown).

Assessment of HIV-1 replication in cells during transduction and after viral challenge.To determine whether HIV-1 replication occurred during the period of transduction, culture supernatants taken at the completion of the third transduction cycle were evaluated for the presence of HIV-1 core protein (p24) using a p24 antigen enzyme-linked immunosorbent assay (ELISA). In all six experiments, p24 could not be detected (sensitivity ≥5 pg/mL). To further evaluate the amount of HIV-1 present in the purified CD34⁺ fractions, PCR was performed on three samples of the CD34⁺ cells before and after transduction using primers to the HIV-1 gag gene. In cells from two of the three subjects tested, HIV-1 proviral DNA could not be detected (sensitivity <1 HIV-1–containing cell/10⁵ cells) either before or after transduction. In the CD34⁺ cells from the third subject tested, HIV-1–containing cells were detected at a frequency of 1 HIV-1–containing cell/10⁴ cells before transduction, but were not detectable after transduction.

To determine whether some cells expressing the anti–HIV-1 genes became infected by HIV-1 and persist harboring the provirus, cell samples at the completion of one HIV-1 challenge experiment were assayed for the presence of HIV-1 provirus by DNA PCR. Cells transduced by the control vector, LN, and challenged with HIV-1_JR-FL showed the presence of HIV-1 provirus at a level of approximately 1 cell per 100. In contrast, cells expressing the anti–HIV-1 genes did not have detectable HIV-1 provirus (<1/10,000), which may indicate that the numbers of cells infected by the low MOI inoculum remained below the limit of detection when viral spread was inhibited or may reflect loss of cells initially infected by HIV-1.

A promising method to permanently suppress HIV-1 replication is the insertion of anti–HIV-1 genes into hematopoietic stem cells. We have previously modeled this procedure by using CD34⁺ cells from the BM or umbilical cord blood of uninfected individuals, demonstrating strong suppression of HIV-1 replication after transduction with retroviral vectors.3 As a prerequisite for the initiation of clinical trials, it is necessary to show that CD34⁺ cells from HIV-1–infected individuals can be successfully transduced with retroviral vectors carrying anti–HIV-1 activity, and that HIV-1 growth in mature cells derived from these transduced CD34⁺ cells can be effectively suppressed. Also, the insertion of anti–HIV-1 genes must not produce toxicity in the transduced cell populations. The data presented here show that three retroviral vectors encoding an RRE-decoy, a double hammerhead ribozyme, or a trans-dominant rev protein meet the criteria of efficacy and nontoxicity in vitro in primary human hematopoietic cells from HIV-1–infected individuals.

Effective inhibition of HIV-1 replication was consistently achieved with all three vectors carrying the anti–HIV-1 genes. The MOI used for infection with HIV-1_JR-FL was 0.01 TCID₅₀ , a value calculated using the number of infectious viral particles divided by the total number of cells in culture. The actual MOI based on the number of cells susceptible to HIV-1 infection may have been at least 20-fold higher, since only 1% to 5% of the cells in culture may express the CD4 receptor.3 It is difficult to know how these viral concentrations correspond to those which would be encountered by cells in lymphoid tissue or circulation in vivo.

An important issue in the evaluation of anti–HIV-1 genes is to show that they confer resistance to primary isolates of HIV-1, which represent viral strains more closely related to the types found in HIV-1–infected individuals than laboratory-adapted strains, such as HIV-1_JR-FL . In our case, a low-passage HIV-1 isolate was obtained from one of the subjects enrolled in the study. After determination of the number of infectious viral particles by titration on PB monocytes/macrophages, the MOI most suitable for challenge of LTBMC was found to be 0.1 TCID₅₀ , an MOI 10-fold higher than the one used with HIV-1_JR-FL . Peak viral output in the nonprotected cultures (transduced with the LN vector) ranged into the nanogram/milliliter levels of p24, while suppression of viral replication in three experiments typically exceeded 98% with the RRE decoy vector, the double hammerhead ribozyme vector, and the rev trans-dominant vector. These data provide strong evidence that these anti–HIV-1 genes can inhibit replication not only of HIV-1_JR-FL , but also the replication of a primary HIV-1 isolate in monocytic cells derived from CD34⁺ cells of HIV-1–infected individuals.

A concern for the use of retroviral vectors transferring anti–HIV-1 genes in clinical protocols is the potential for interference with normal cellular functions. For hematopoietic progenitor cells, the primary functions that can be observed in vitro are their ability to produce multi-lineage colonies. In our experiments, colony assays performed immediately after completing transduction and again after 5 weeks of LTBMC showed no difference in growth and differentiation of mature cell colonies derived from CD34⁺ cells transduced by any of the anti–HIV-1 vectors, compared with mock transduced CD34⁺ cells. Additionally, the percentages of progenitors which contained and expressed the vectors remained stable during the 5 weeks of LTBMC, demonstrating that the presence of the gene products does not lead to loss of clonogenic potential. These observations imply that the expression of the RRE decoy, the ribozymes, or the rev trans-dominant sequences does not interfere with normal cellular processes in hematopoietic progenitors needed for proliferation and maturation.

The trans-dominant rev mutant M10 gave the most consistent inhibition of replication of both HIV-1_JR-FL and the primary isolate, always producing at least greater than 98% inhibition. Because M10 is the only anti–HIV-1 gene product from among the three examined here which is active as a protein rather than as an RNA, there may be, in effect, an amplification of the vector transcripts by translation, yielding more inhibitory molecules per vector transcript. However, it is not possible to definitively identify the most effective inhibitory strategy from these in vitro studies. Factors such as emergence of viral resistance or immune responses to the anti–HIV-1 gene product may occur in vivo, but would not be demonstrable in vitro.

This study shows that mobilized PBSC from HIV-1–infected individuals may be successfully transduced with retroviral vectors encoding anti–HIV-1-genes without loss of cellular function. Mature blood cells carrying these anti–HIV-1 genes produced in LTBMC showed strong inhibition of HIV-1 replication. To facilitate long-term resistance to HIV-1 in vivo, anti–HIV-1 genes need to be delivered into pluripotent stem cells capable of sustained engraftment after BM transplantation, a small fraction of the CD34⁺ cell population. Clinical trials with retroviral vectors have demonstrated low transduction into reconstituting stem cells, in the range of 0.1% to 1%.18-21 It is presently unknown if anti–HIV-1 genes introduced into such a small fraction of stem cells would confer any clinical benefit to HIV-1–infected individuals receiving gene therapy. In two clinical trials using retroviral vectors to transfer the adenosine deaminase (ADA) cDNA into CD34⁺ cells of patients with severe combined immunodeficiency, selective accumulation of T lymphocytes carrying the newly introduced ADA gene could be observed.18,21 It is conceivable that monocytes/macrophages or T cells protected from cytopathicity associated with HIV-1 infection may also have a selective advantage over unprotected cells.22 Mature, protected cells may therefore accumulate in the circulation and maintain their normal function within the immune system. Clinical trials with HIV-1–infected individuals will be required to further investigate this possibility.