INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The potential inhibitory effects of antioxidative agents, including glutathione (GSH) and its precursors, against human immunodeficiency virus type I (HIV-1) have been investigated over the last several years. In early studies, reducing compounds such as D-penicillamine, 2,3-dimercapto-1-propanolol, and N-acetylcysteine (NAC) were found to inhibit HIV-1 long terminal repeat (LTR)-directed viral gene transcription (5, 16, 18, 28). In parallel with these initial basic studies, reduction of GSH levels in plasma, peripheral blood cells, and lung epithelial-lining fluid has been reported in HIV-1-infected-individuals (3, 9, 30). GSH is known not only as a major intracellular antioxidant but also as a modulator of the immune system (11). Hence, altering the GSH deficiency of HIV-1-infected individuals by glutathione precursors has been hypothesized to be one of the rational therapeutic strategies to prevent HIV-1 propagation in vivo (2, 4, 27). In this manner, the inhibitory effects of GSH prodrugs, such as NAC, against HIV-1 have been further characterized. These compounds have been shown to be capable of inhibiting HIV-1 gene transcription, which is induced by tumor necrosis factor alpha (TNF-) or phorbol 12-myristate 13-acetate (PMA), from latent proviruses. This is a model for the cellular latent stage of HIV-1 infection (16, 25). Notably, in a recent report, enhancement of HIV-1 growth by NAC was described in peripheral blood mononuclear cells (PBMCs) in direct contact with U1 cells, indicating some complexity in the anti-HIV-1 property of these agents (6).

Recent studies have demonstrated that the replication of HIV-1 is continuously active in lymphoreticular tissues (10, 24). Therefore, it may be difficult to significantly alter HIV-1 infection only by keeping latent HIV-1-proviruses in a nonreplicative state, without shutting off the massive virus production from so-called late-phase cells and subsequent further rounds of infection. It has been implied that, to retard the progression of AIDS, removal of the late-phase cells may be critical (14). Here, we report on a unique GSH prodrug, -glutamylcysteine ethyl ester (-GCE; TEI-2306) (31), which is shown to have novel effects against HIV-1. -GCE possesses a long plasma half-life (more than fivefold longer than that of GSH) and high membrane permeability (more than five times higher than that of GSH) and has been reported to be effective against heart and liver reperfusion injury (13, 17, 20-22), asthma (15), and cataracts (23). In the present study, this compound is shown to possess a unique anti-HIV-1 activity in both chronically and acutely infected cells, and even for free viruses, as opposed to inhibiting oxidative stress-induced increases of HIV-1 transcription.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Cells and viruses. The uninfected human T-lymphoid cell line H-9 and H-9 cells infected with an HIV-1 strain, HIV-1_IIIB (H-9/IIIB) (26), were maintained in RPMI 1640 tissue culture medium supplemented with 10% fetal bovine serum (FBS). PBMCs were isolated via Ficoll-Hypaque centrifugation, as described previously (35). Isolated PBMCs were prestimulated with phytohemagglutinin (PHA) (10 µg/ml) and human interleukin 2 (IL-2; 50 µg/ml) for 2 days and subjected to infection experiments. For acute-infection experiments, a complete infectious strain of HIV-1, NL_4-3 (1), was used as described in "HIV-1 inhibition assay of acute infections."

Materials. -GCE (TEI-2306) was provided by the Teijin Institute for Biomedical Research. GSH and NAC were purchased from Sigma Chemical Co. (St. Louis, Mo.). The structure of -GCE and its property as a GSH prodrug are described in Fig. 1.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Structural formula of -GCE and its property as a GSH prodrug. A double line represents the cell membrane. The figure illustrates that -GCE permeates the membrane into the cell (closed arrow) and is metabolized to GSH intracellularly (open arrow) through -glutamyl cysteine (-GC).

Cytotoxicity assays. H-9 or H-9/IIIB cells were seeded in 24- or 96-well tissue culture plates at a density of 2 × 10⁵ to 2.5 × 10⁵/ml, with or without a variety of concentrations of -GCE, GSH, or NAC. To monitor cell growth and viability, viable cells were counted every 2 or 3 days by the trypan blue exclusion method. Cell viability was represented as the percentage of viable cells in the total cell population. At the same time points, 50 to 80% of the cell suspensions were removed, and the same amounts of fresh medium and the same concentrations of the compound to be analyzed were added for further incubation.

HIV-1 inhibition assay in chronically infected cells. For studies of high doses of the compounds and short-term experiments, H-9/IIIB cells were washed thoroughly with phosphate-buffered saline (PBS) and resuspended at a density of 10⁵/ml in RPMI 1640 medium-10% FBS with or without 1.25 or 2.5 mM -GCE, GSH, or NAC. After 2 and 4 days, culture supernatant was removed, centrifuged to remove cells and debris, and evaluated by HIV-1 p24 antigen quantification. Measurement of the HIV-1 p24 antigen was performed by a sensitive enzyme-linked immunosorbent assay (ELISA; DuPont). For evaluation of low concentrations of the compounds, H-9/IIIB cells were washed with PBS and seeded at a density of 2 × 10⁵/ml in RPMI 1640 medium-10% FBS with or without 0.2, 0.4, or 0.8 mM -GCE. Three and 6 days after initiation of experiments, culture supernatants were sampled and assayed as described above. On day 3, 72% of each cell suspension was removed and fresh medium with a corresponding GSH prodrug was added. Total HIV-1 p24 antigen production values on day 6 were deduced by multiplying raw values on day 6 by the dilution factor on day 3.

HIV-1 inhibition assay of acute infections. Virus stocks of the HIV-1 strain NL_4-3 were prepared and titers were determined as described previously (8). H-9 cells (3 × 10⁵ cells) were suspended in 1 ml of growth medium with or without -GCE at a variety of concentrations and seeded in a chamber of a 24-well tissue culture plate. For the initial infection, virus suspension containing 30 pg of HIV-1 p24 antigen was added to each chamber. Infection was carried out overnight at 37°C, and then the cells were washed once with PBS and once more with growth medium. The supernatant of the final wash was saved as a day 0 time point sample. Washed cells were resuspended in the same media for further incubation at 37°C. Once in 3 days, cell suspensions and supernatants were harvested for cell counting and HIV-1 p24 antigen quantification. Measurements of cell growth, viability, and virus production were performed as described above. On the days of harvest, 200 µl of cell suspension was left in each well and 800 µl of fresh medium with the same concentrations of -GCE was added to support cell growth, except on day 18, when 400 µl of cell suspension was left and 600 µl of medium was added to each well. Cell numbers at each time point were determined accordingly, based on the accumulation of dilution factors and the frequency of passage.

HIV-1 direct inactivation assay. An HIV-1 virus stock (NL_4-3) containing 6 ng of HIV-1 p24 antigen was preincubated in the presence or absence of -GCE in 20 µl of RPMI 1640 medium-10% FBS for 4 h at 37°C. Afterwards, acute infection of H9 cells, using 10 µl of each pretreated virus suspension in 1 ml of cell suspension without -GCE, was carried out as described in "HIV inhibition assay of acute infections." Sampling and maintenance of cell culture were also performed, following the same procedure as that described above, without adding any -GCE through passage.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Differential cytotoxic effects of -GCE. (i) Lack of cytotoxicity of -GCE for uninfected H-9 cells at 2.5 mM. The effects of -GCE on the viability and cell growth of uninfected H-9 cells, in comparison with those of GSH and NAC, were initially evaluated. At a concentration of 2.5 mM, -GCE demonstrated no negative effects on cell viability and growth (Fig. 2A). These results are quite similar to the findings with GSH and NAC. Through 6 days of incubation, cell viability never decreased below 90% in any case, with the lowest value being 91.75% ± 0.45% (NAC-treated cells on day 4), also suggesting the low toxicity of -GCE for uninfected T-lymphocytic cells. Similar results were obtained with another human T cell line, CEM (data not shown).

View larger version (10K): [in this window] [in a new window]   FIG. 2.   (A) Effects of 2.5 mM -GCE and other GSH prodrugs on the growth of uninfected H-9 cells. Symbols: closed circles, -GCE; triangles, GSH; squares, NAC; open circles, control. Error bars may not be visible in cases in which variations among experiments were very small. These results are the mean values of two independent experiments. (B) Effects of 800 µM -GCE on growth of H-9/IIIB cells. Growth of H-9/IIIB cells in 800 µM -GCE (closed circles) and control experiments without -GCE (open circles) are shown. Error bars are indicated. The data are the mean values of two independent experiments. Total cell numbers (in millions) were deduced by adjusting raw values with accumulated dilution factors along with the passage of cell culture.

(ii) Potent cytotoxicity of -GCE for H-9/IIIB cells at 800 µM. Experiments similar to those described above were carried out with a chronically infected line, H-9/IIIB, which highly expresses HIV-1. In contrast to uninfected H-9 cells, surprisingly, H-9/IIIB cells displayed a remarkably high sensitivity to the cytotoxic effects of -GCE. As shown in Fig. 2B, cell growth was dramatically impaired at a concentration of -GCE as low as 800 µM. On day 6, cell growth was reduced to a level of less than 40% of that of the control. Eventually, after 12 days of treatment, no significant cell growth was observed. The cell viability also decreased below 40% after day 9 (data not shown), which suggests that -GCE has relatively high cytotoxicity against HIV-1-producing cells.

(iii) Differential thresholds of the toxic doses of -GCE for H-9 and H-9/IIIB cells. To compare the cytotoxic effects of -GCE on H-9 and H-9/IIIB cells, which give similar growth profiles in the absence of -GCE, a variety of concentrations of -GCE were tested to determine two doses for each cell line, which gave equivalent profiles. It was observed that the effect of 5 mM -GCE on H-9 cells was approximately equivalent to that demonstrated by 400 µM -GCE on H-9/IIIB cells (Fig. 3). At those cytostatic doses, both cell lines showed some decreases in relative cell growth, and minimal decreases in viability, after 6 days of -GCE treatment. In conclusion, H-9/IIIB cells are shown to be approximately 12.5 times more sensitive to -GCE than uninfected cells; this difference represents a selective toxicity of -GCE for HIV-1-producing cells.

View larger version (25K): [in this window] [in a new window]   FIG. 3.   (A) Effects of 5 mM -GCE on growth and viability of uninfected H-9 cells; (B) effects of 400 µM -GCE on growth and viability of H-9/IIIB cells. Columns: 1, without -GCE; 2, with -GCE. After 6 days of treatment, the relative cell number and viability were monitored for each case. Relative cell numbers were calculated, with the mean number of the control cells at the same time point defined as 100%. These data are the mean values of two independent experiments.

Effect of -GCE on constant HIV-1 production at doses which selectively impair the growth and viability of HIV-1-producing H-9 cells. At a concentration between 800 µM and 2.5 mM, -GCE was shown to be capable of altering the viability of H-9/IIIB cells while not affecting uninfected H-9 cells. As a next step, the efficacy of such a selective toxicity for limiting HIV-1 production was evaluated. Since H-9/IIIB cells are constantly producing a large quantity of virus particles, stimulation with factors such as TNF- or PMA were not required for carrying out these experiments. Two doses were examined for -GCE, GSH, and NAC. Neither GSH nor NAC could overwhelm the vigorous HIV-1 production by H-9/IIIB cells. Rather, at 2.5 mM, they increased the production of virus. However, 1.25 or 2.5 mM -GCE almost completely shut off the overall logarithmic production of HIV-1 from H-9/IIIB cells, apparently through a cytotoxic effect (Fig. 4). These data demonstrate a novel property of -GCE as an anti-HIV-1 agent.

View larger version (14K): [in this window] [in a new window]   FIG. 4.   Inhibition of HIV-1 production from H-9/IIIB cells by high concentrations of -GCE. -GCE, GSH, and NAC were used at concentrations of 1.25 (A) and 2.5 (B) mM. The mean values of two independent experiments are illustrated. Symbols: closed circles, -GCE; triangles, GSH; squares, NAC; open circles, control.

Low concentrations of -GCE are capable of repressing virus production without affecting the viability of H-9/IIIB cells. Since -GCE is a potential antioxidant, inhibition of HIV-1 LTR-directed transcription within a nontoxic range of -GCE was also predicted. However, it was also presumed that it may be difficult for this type of compound to inhibit such constant and high retroviral production. Treatment of H-9/IIIB cells with low concentrations of -GCE is illustrated in Fig. 5A. As expected, -GCE significantly inhibited the vigorous HIV-1 production by H-9/IIIB cells at low concentrations (200 and 400 µM), whereas higher concentrations of GSH or NAC did not alter virus expression (Fig. 4 and 5). With such -GCE treatment, cell viability was not significantly affected at either dose (>86%). Slower cell growth was observed only at 400 µM (as illustrated in Fig. 3B), not at 200 µM (data not shown). The antiviral effect was dose dependent up to 800 µM, at which concentration the selective cytotoxic effects appeared (Fig. 2B and 5B). These findings demonstrate the efficacy of -GCE as a GSH prodrug against active and continuous HIV-1 production, even at low concentrations.

View larger version (24K): [in this window] [in a new window]   FIG. 5.   (A) Inhibitory activity of low concentrations of -GCE against HIV-1 production from H-9/IIIB cells. The effects of 400 µM (closed circles) and 200 µM (squares) on viral production are shown with control experimental data (open circles). The data are the mean values of two independent experiments. Error bars may not be visible in cases in which variations were very small. (B) Dose-dependent inhibition of HIV-1 production from H-9/IIIB cells by -GCE. The data from panel A on day 6 are comparatively summarized with the results of parallel experiments with 800 µM -GCE, at which concentration the cytotoxic effects become evident (see Fig. 2B). Relative values were calculated, with the mean value without -GCE defined as 100%.

Inhibiting acute HIV-1 infection with -GCE. If -GCE is able to remove HIV-1-producing cells effectively, it is expected to prevent HIV-1 from spreading over the entire population of cells during acute infection. To verify this hypothesis, acute-infection studies were carried out.

View larger version (13K): [in this window] [in a new window]   FIG. 6.   Inhibition of acute HIV-1 infection by -GCE. Cell growth (A), cell viability (B), and release of HIV-1 p24 antigen (C) were monitored along the course of acute HIV-1 infection of H-9 cells in the presence or absence of -GCE. The data shown are from a representative of two independent series of experiments. Symbols: closed circles, 1.6 mM -GCE; closed squares, 800 µM -GCE; closed triangles, 400 µM -GCE; open circles, control without -GCE. Cell numbers (A) were deduced by multiplying raw values by dilution factors, which had been accumulating during maintenance of growing cells, as described in Materials and Methods.

View larger version (14K): [in this window] [in a new window]   FIG. 7.   Inhibitory effects of -GCE on acute HIV-1 infection in human primary PBMCs at higher viral input. HIV-1 p24 antigen was monitored, along the time course of HIV-1 infection, with 170-fold more virions per cell than that used in the H9 experiments. Symbols: closed circles, 2.5 mM -GCE; open circles, control without -GCE.

Direct inactivation of HIV-1 by -GCE. The results of the acute HIV-1 infection study raised another possibility, that -GCE may inactivate the infectivity of HIV-1 before or along with the infection process. For the examination of such an infection-inactivating effect of -GCE, we designed and carried out another series of experiments. After the HIV-1 (NL_4-3) was pretreated with a variety of concentrations of -GCE for 4 h, acute infection was carried out to monitor the propagation of the pretreated virus in the absence of -GCE (Fig. 8). Since -GCE was diluted extensively upon infection (25 µM), the effect of -GCE on host cells could be discounted. Surprisingly, -GCE treatment at 1.25 or 2.5 mM for 4 h caused the total loss of infectivity of HIV-1, with no viral production even 15 days postinfection. Even at the lowest concentration (625 µM), -GCE caused some attenuation of the propagation of HIV-1 in H-9 cells, indicating partial inactivation of virus particles. However, preincubation with the same concentrations of -GCE for 1 h did not show significant interference with viral infectivity (data not shown). These data indicate a direct inactivation effect of -GCE on HIV-1 particles, which may be involved in the inhibitory effect observed in the HIV-1 acute infection experiments.

View larger version (13K): [in this window] [in a new window]   FIG. 8.   Effect of preincubation of HIV-1 (strain-NL4-3) with -GCE on viral infectivity. Acute-infection experiments using H-9 cells were performed, and viral production was monitored, after preincubation of the virus in the presence of the indicated concentrations of -GCE for 4 h. Symbols: closed circles (all on x axis), 2.5 mM; triangles (all on x axis), 1.25 mM; squares, 625 µM; open circles, control without -GCE. These results are representative of two independent sets of experiments.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The anti-HIV-1 effects of -GCE have been demonstrated in two distinct manners. At relatively high concentrations, -GCE displayed a selective cytotoxicity against HIV-1-producing cells and potently interfered with acute HIV-1 infection. Direct inactivation of HIV-1 by -GCE was also observed at this range of concentrations, which may contribute to the observed blockade of the acute infection. At concentrations below cytotoxic levels, repressive effects on HIV-1 production from such cells were observed. Although a vast number of anti-HIV-1 chemotherapeutic candidates have been described, these have included no compounds which display a selective killing of HIV-1-infected cells. Similarly, although there have been various reports which describe the anti-HIV-1 effects of a variety of antioxidants (5, 16, 18, 28), no such compound was reported to block acute HIV-1 infection in specific cells or to disable HIV-1 infectivity. Therefore, compounds such as -GCE may lead to a new strategy to combat AIDS.

-GCE was capable of inhibiting the viral production of H-9/IIIB cells at a low concentration (200 µM). In previous reports, other GSH prodrugs, such as NAC, have been shown to inhibit viral production from latently infected cells which had been stimulated by PMA or TNF- (16, 28). However in the present studies, even GSH and NAC concentrations of 1.25 and 2.5 mM did not show any inhibitory effects against HIV-1 production from H-9/IIIB cells. These findings imply that the effect of such compounds may not be sufficient to interfere with a vigorous HIV-1 production from late-phase cells, although they may be able to prevent the stimulation of latently infected cells.

Recently, it has been suggested that the turnover of HIV-1-infected cells is strikingly rapid in vivo. This suggests that many cells are newly infected, and many infected cells proceed to the late infection stage to produce numerous virus particles before their death (12, 33). If so, to shut off HIV-1 propagation, three critical effects must take place. The acquisition of new infection must be blocked, infected cells must be prevented from proceeding to the late phase of the infection, and the late-phase virus producers must be removed as quickly as possible. As such, -GCE seems to possess excellent properties as an anti-HIV-1 agent. Namely, it inactivates HIV-1 itself, blocks acute infection, represses virus production from infected cells, and kills HIV-1-producing cells selectively. If an appropriate drug delivery system is established, -GCE itself might be directed towards chemotherapy against AIDS. Of note, 100 mg of -GCE per kg was infused into dogs and gave a positive effect as a cardio-protective agent (16a); this dosage should yield a millimolar concentration in plasma at its peak. Although additional investigation of a drug delivery system to maintain such concentrations in vivo seems to be required at present, it may be possible to maintain such effective doses in vivo, as used in the present study. Anti-HIV-1 effects of -GCE derivatives are also critical to evaluate.

In addition to the experiments described, further analysis must be performed with PBMCs under a variety of conditions. Starting with the reevaluation of the applicable doses in PBMCs, such extensive studies are currently ongoing. Especially, the effect of -GCE on HIV-1 production from initially quiescent PBMCs should be evaluated, since NAC was found to be stimulatory under certain conditions. A detailed evaluation of the effect of -GCE on HIV-1 infection, using both stimulated and unstimulated PBMCs, will follow this initial study. It is also important to reevaluate such anti-HIV-1 effects with primary isolates of HIV-1.

It is clear that the direct inactivating effect of -GCE may play a role in the blockade of the acquisition of acute HIV-1 infection. However, it may not be the only anti-HIV-1 action of -GCE to shut off an acute infection. In fact, 1 h of pretreatment did not show efficient blockade of the infection, even at a concentration of 2.5 mM, indicating that 1 h is not long enough for -GCE to inactivate HIV-1 particles. In the acute-infection experiments, it is suggested that initial infection may be initiated within 1 or a few hours before the virus is inactivated by -GCE. Therefore, it is suspected that inhibition of acute HIV-1 infection may be accomplished via the direct inhibitory potential and other anti-HIV-1 effects, possibly by the removal of infected cells or by the blockade of infectious virus particle production from initially infected cells.

The less pronounced anti-HIV-1 effects observed in the PBMC experiments implies that inactivation of higher inputs of virus may require increased time for inactivation by -GCE or that the cellular sensitivity to -GCE among different types of infected cells may differ. Detailed analyses with PBMCs should also provide further data in determining which stage of HIV-1 infection is the major target for -GCE during the acute-infection process.

The mechanism(s) of the selective cytotoxic effect of -GCE is at present unclear. However, it can be hypothesized that some specific changes of cell membranes that are caused by HIV-1 infection could play a role in altering cells to become more sensitive to -GCE. The major viral factors to interact directly with host cell membranes are the envelope (env) gene products. Of note, the gp41 transmembrane glycoprotein is known to alter the cell membrane structure and the permeability of infected cells (19, 29). In gp41, there are two small domains with typical amphipathic helical structures in the cytoplasmic tail which have been demonstrated to form pores on cell membranes (7, 32). Since -GCE is composed of hydrophilic groups of the -glutamylcysteine residue and a hydrophobic ethyl ester group that does not exist in GSH or NAC, interactions with the amphipathic helixes of gp41 may be possible. Such interactions would specifically increase the influx of -GCE into the cells or may accelerate the pore formation on the membrane and eventually lead to cell death by altering the intracellular redox regulatory system or by injuring the cell membrane. Interestingly, a recent study has revealed the same gp41-induced pore formation and increased permeability on the HIV-1 virion surface as well (34). Since such an unusual feature on the surface is shared by HIV-1-producing cells and the HIV-1 virus itself, it is expected that -GCE may affect the structure and function of HIV-1 particles. In accordance with these data, the HIV-1-inactivating potential of -GCE was confirmed, further suggesting the role of gp41 in the antiviral effect of -GCE. Thus, examination of the cooperative cytotoxic effects of gp41 and -GCE is now ongoing in our laboratories. However, possible interactions of -GCE and other viral proteins cannot be ruled out at the present time.

The mechanism(s) of the repressive effect on HIV-1 replication of -GCE at low concentrations is predicted to be similar to that of NAC and other antioxidants. Presumably, -GCE may be inhibiting HIV-1 gene transcription by restraining the oxyradical-mediated activation of the cellular nuclear factor-kappa B (NF-B) system (27). By utilizing a bacterial -galactosidase gene construct as a reporter gene, which is driven by the HIV-1 LTR, experiments to verify the repressive effect of -GCE on NF-B-mediated stimulation of the HIV-1 gene transcription are also in preparation.

Selective removal of infected cells and subsequent inactivation of viruses released from such cells constitute the major defensive strategy of the in vivo immune system to combat viral infection. However, no chemotherapeutic agent which uses this antiretroviral strategy has been available. In this study, -GCE was shown to behave as a unique anti-HIV-1 agent in vitro, mimicking the antiviral immune reaction that is impaired in AIDS patients. Therefore, -GCE may represent a possible new approach in the chemotherapy of AIDS.