Novel Function of Prothymosin Alpha as a Potent Inhibitor of Human Immunodeficiency Virus Type 1 Gene Expression in Primary Macrophages

Abstract

CD8⁺ T lymphocytes control human immunodeficiency virus type 1 (HIV-1) infection by a cytotoxic major histocompatibility complex-restricted pathway as well as by secretion of noncytotoxic soluble inhibitory factors. Several components of CD8⁺ cell supernatants have been identified that contribute to the latter activity. In this study we report that prothymosin alpha (ProTα), a protein found in the cell culture medium of the herpesvirus saimiri-transformed CD8⁺ T-cell line, K#1 50K, has potent HIV-1-inhibitory activity. Depletion of native ProTα from an HIV-1-inhibitory fraction of CD8⁺ cell supernatants removes the inhibitory activity, supporting its role in inhibition via soluble mediators. ProTα is an abundant, acidic peptide that has been reported to be localized in the nucleus and associated with cell proliferation and activation of transcription. In this report we demonstrate that ProTα suppresses HIV-1 replication, its activity is target cell specific, and inhibition occurs following viral integration. Native and recombinant ProTα protein potently inhibit HIV-1 long terminal repeat (LTR)-driven gene expression in macrophages. Furthermore studies using different promoters in lentiviral vectors (cytomegalovirus and phosphoglycerate kinase) revealed that suppression of viral replication by ProTα is not HIV LTR specific.

Control of viral replication in human immunodeficiency virus type 1 (HIV-1)-infected individuals requires cooperative functioning of the innate and adaptive arms of the immune system (9, 27). CD8⁺ T cells derived from HIV-1-positive and -negative donors are thought to contribute to both innate and adaptive immune responses to HIV. In addition to cytotoxic T-lymphocyte activity, CD8⁺ T cells have been shown to secrete soluble molecules that inhibit HIV-1 replication in vitro (3, 16, 31, 49).

The β-chemokines RANTES, MIP-1α, and MIP-1β, which competitively inhibit HIV-1 entry of R5 viruses into susceptible cells, are present in the supernatant of cultured CD8⁺ cells (5). Macrophage-derived cytokine (MDC), another chemokine isolated from CD8⁺ cell-conditioned medium, has activity against both R5 and X4 strains of HIV-1, although chemically synthesized MDC has more restricted inhibitory activity (8, 36, 38). Interleukin-16 (IL-16), as well as gamma interferon, can be found in variable amounts in CD8⁺ T-cell supernatants and has associated anti-HIV-1 activity (1, 18, 32). However, a number of groups have shown that the anti-HIV-1 activity found in CD8⁺ T-cell supernatants is not fully accounted for by known β-chemokines, IL-16, MDC, or gamma interferon (13, 30, 31).

The amino-terminal fragment of the urokinase-type plasminogen activator and antithrombin III, both present in CD8⁺ T-cell-conditioned medium, have been shown to have HIV-1-inhibitory activity in tissue culture (12, 48). In this study we report that prothymosin alpha (ProTα), a protein found in the cell culture medium of the herpesvirus saimiri (HVS)-transformed CD8⁺ T-cell-line K#1 50K, previously demonstrated to have anti-HIV activity (3), has potent HIV-1-inhibitory activity that is target cell specific and inhibits steps following viral integration. Native and recombinant ProTα proteins inhibit HIV-1 long terminal repeat (LTR)-driven gene expression in infected macrophages.

MATERIALS AND METHODS

Isolation of HIV-1-suppressive factors from HVS-transformed CD8⁺ cell lines.

K#1 50K CD8⁺ cells (3) were cultured at a concentration of 1 × 10⁶ to 1.5 × 10⁶ cells/ml in phenol red-free RPMI 1640 (Life Technologies, Rockville, MD) supplemented with IL-2 (50 U/ml), 100 U/ml of penicillin, 10 μg/ml of streptomycin, and 2 mM l-glutamine without serum for 48 h. The cells were removed by centrifugation, and the supernatant was clarified by filtration though 0.45-μm cellulose acetate filters. A final volume of 1.3 liters of clarified cell-conditioned medium was diluted 12-fold with MilliQ water buffered at pH 8.0 with 20 mM Tris. The diluted cell-conditioned medium at a stable pH of 8.0 was applied to a sedimented bed of Streamline QXL resin (Amersham Biosciences, Piscataway, NJ) in a Streamline 25 column (Amersham Biosciences, Piscataway, NJ) at a rate of 10 ml/min using an Δkta Explorer 100 chromatography system (Amersham Biosciences Piscataway, N.J.). The flow rate used was sufficient to expand the sedimented bed to 1.5 times its sedimented volume. When the diluted cell-conditioned medium had all been applied to the QXL resin, the flow was paused, the resin was allowed to sediment, and the adaptor of the Streamline 25 column was lowered to the top of the sedimented bed. The flow was then reversed and the resin washed with 10 resin volumes of Tris, pH 8.0, followed by a linear gradient of 0 to 1 M NaCl with peak capture. Peaks that showed HIV-1-suppressive activity were brought up to 1.8 M (NH₄)₂SO₄. These were applied to a Resource phenyl Sepharose hydrophobic interaction column (Amersham Biosciences, Piscataway, NJ) at 1 ml/min. Bound material was eluted with a linear gradient of 1.8 M to 0 M (NH₄)₂SO₄ with peak capture.

HIV-suppressive peaks from the hydrophobic interaction chromatography step were further fractionated using a Superdex Peptide H 10/30 gel filtration column (Amersham Biosciences, Piscataway, NJ). The peaks were applied without modification in 250-μl aliquots using phosphate-buffered saline solution at pH 7.4 as the eluent, and peaks were captured. Aliquots of the fractions were analyzed by polyacrylamide gel electrophoresis using the PhastSystem with Phasgel precast 8 to 25% gels (Amersham Biosciences, Piscataway, NJ) run according to the manufacturer's instructions. The proteins were visualized in a gel by silver staining using Silver Stain Plus (Bio-Rad, Hercules, CA).

Protein sequencing.

Equivalent fractions from successive gel filtration chromatography steps were pooled. Aliquots of each pool were digested with trypsin and analyzed by using an LCQ ion trap mass spectrometer (Thermo-Finnigan, San Jose, CA) fitted with a custom-built nanoflow-high-pressure liquid chromatography microelectrospray ionization source (43a). Tandem mass spectrometry spectra were searched against the human nonredundant database (NCBI, Bethesda, MD), de novo sequenced, and manually validated.

Reagents.

Synthetic thymosin alpha 1 (Tα1) was purchased from Accurate Chemical and Scientific Corporation (Westbury, NY), and biological activity (in vivo mouse protection assay against Candida albicans) of the peptide was tested and provided in the data sheet by the company. Recombinant human ProTα and antibody to ProTα were purchased from Alexis Biochemicals (San Diego, CA). Antibodies against STAT1 and P-STAT1 were purchased from Cell Signaling (Beverly, MA). The integrase inhibitor L-731988 was a gift from Merck (19).

Cells.

CD8⁺ K#1 50K cells were derived from an adult healthy donor and transformed by HVS as described previously (31). HeLa CD4 cells were obtained from the National Institutes of Health AIDS Research and Reference Reagent Program. Primary macrophages were isolated by adhesion from Ficoll-Hypaque (Sigma)-purified peripheral blood mononuclear cells after 10 to 14 days in culture with Dulbecco modified Eagle medium containing 20% fetal bovine serum. Primary CD4⁺ cells were purified by positive or negative selection using magnetic beads from Miltenyi Biotech (Auburn, CA) and cultured in RPMI containing 10% fetal bovine serum and 50 U IL-2. H9 and CEM cells were purchased from ATCC (Manassas, VA).

Viruses.

HIV-1_BaL and HIV-1_IIIB were purchased from ABI (Columbia, Maryland). HIV-1 primary isolates 92BR028 and 92BR030-R5 and 92RW009-R5/X4 were obtained from the National Institutes of Health AIDS Research and Reference Reagent Program. The viral construct mC2,C3 BaL (kindly provided by A. J. Henderson, Penn State University) was used for generation of an infectious virus stock by Lipofectamine 2000 (Invitrogen) transfection of 293T cells as described previously (20, 33).

Inhibition of HIV-1 replication by ProTα.

To screen for inhibitors that worked after viral entry, the purified macrophages were first incubated with the R5 isolate of HIV-1_BaL for 2 h at a multiplicity of infection (MOI) of either 0.1 or 0.01 and then washed and subsequently cultured in the presence of CD8⁺ cell-conditioned medium (10% by volume) or different fractions derived from serum-free conditioned medium. The medium was changed every 3 to 4 days, and fresh medium with 10% conditioned medium or fractions were added. HIV-1 replication was monitored by measuring HIV-1 p24 antigen in cell supernatants using an enzyme-linked immunosorbent assay (ELISA) (SAIC, Frederick, MD). Both 10- to 15-day-old primary macrophages and primary CD4⁺ cells separated by magnetic beads using a Miltenyi Biotech CD4⁺ T-cell isolation kit were plated in 96-well plates. Macrophages were plated at concentrations of 0.2 × 10⁵ to 0.3 × 10⁵ cells per well, and phytohemagglutinin (PHA)-activated CD4⁺ lymphocytes were plated at 1.5 × 10⁵ to 2 × 10⁵ cells per well. The R5 virus HIV-1_BaL, originally isolated from lung tissue and passaged in primary macrophages, as well as the laboratory-adapted strain HIV-1_IIIB, was used to infect cells at an MOI of 0.01 or 0.1. Virus was washed off at 2 hours followed by the addition of ProTα at the concentrations indicated. One-half of the medium was replaced every 3 to 4 days with new ProTα. HIV-1 p24 antigen production was measured by ELISA using the SAIC Frederick kit (National Cancer Institute, Maryland).

Statistics.

Calculations were performed using Student's t test. P values of <0.05 were considered significant.

RESULTS

ProTα is an HIV-1-inhibitory molecule present in a chromatographic fraction derived from HVS-transformed CD8⁺ T cells.

Cell-free, serum-free, phenol red-free RPMI 1640 conditioned medium from 1.3 liters of a 48-h culture of the HVS/CD8⁺ T-cell line K#1 50K (31) was collected. A multistep chromatographic separation strategy was utilized to capture and separate fractions with anti-HIV-1 activity. An expanded bed format (Streamline Q XL; Amersham, New Jersey) was used to concentrate all HIV-1-suppressing activity by ion-exchange capture. The flowthrough, unretained fraction had no HIV-1-suppressing activity (data not shown). Active peaks were further fractionated on phenyl Sepharose (Amersham, New Jersey), and samples of the fractions were assayed for anti-HIV activity against HIV-1_BaL in primary macrophages as previously reported (31). To further fractionate under biologically mild conditions, we used gel filtration to separate the components of active phenyl Sepharose fractions (Superdex peptide; Amersham, New Jersey) and the HIV-1-suppressive activity of each resultant fraction was assayed.

The macrophages isolated from peripheral blood mononuclear cells by adhesion were infected with HIV-1_BaL for 2 h, the residual virus was removed by washing, and the cells were incubated with growth medium, which was supplemented with 3 ng/ml of fractionated protein for the entire assay period. Viral replication was determined by quantitation of HIV-1 p24 Gag protein in cell supernatants at day 7 by ELISA.

The de novo sequence (acSDAAVDTSSEITTKDLK, where ac represents acetylation of the N terminus) of a tryptic peptide derived from an aliquot of an active fraction was an exact match to N-terminal residues 1 to 17 of ProTα. An aliquot of the ProTα-containing fraction diluted to 2 ng/ml of protein produced a 95% reduction of viral p24 protein compared with untreated control (Fig. 1A). A similar aliquot from the same fraction had no protective activity for PHA-stimulated CD4⁺ T cells infected with HIV-1_BaL and HIV-1_IIIB (Fig. 1B).

FIG. 1.

Effects of ProTα-containing chromatographic fraction 29 and recombinant ProTα on HIV-1 replication. In each panel the indicated virus at the indicated MOI was incubated with cells. Unbound virus was washed out, and the cells were incubated with 2 ng/ml of protein from fraction 29 containing ProTα or recombinant ProTα protein (at the indicated concentration). The results (means ± standard deviations for triplicate wells) for HIV-1 Gag p24 protein are from a single experiment at day 14 postinfection. (A) Primary macrophages treated with fraction 29 as described above after incubation with HIV-1_BaL (MOI of 0.01) for 2 h. (B) PHA-activated primary CD4⁺ T cells treated with fraction 29 after 1 h of incubation with HIV-1_BaL (MOI of 0.01) or HIV-1_IIIB (MOI of 0.01). (C) Primary macrophages infected with HIV-1_BaL at an MOI of 0.01 and treated with different concentrations of recombinant ProTα. (D) Primary macrophages were incubated with HIV-1_JRFL for 2 h and treated with indicated concentrations of recombinant ProTα; the luciferase assay was performed 48 h posttreatment. (E) Primary macrophages were incubated with R5 and R5/X4 HIV-1 primary isolates at an MOI of 0.1 as described above and treated with 200 ng/ml of recombinant ProTα. The p24 ELISA was done on day 14 postinfection. (F) HeLa CD4⁺/CCR5⁺ cells were incubated with HIV-1_JRFL for 2 h and treated with 200 ng/ml of ProTα or 20% CD8⁺ T-cell supernatant from the K#1 50K cell line after residual virus was washed out; the luciferase assay was performed 48 h posttreatment. Similar results were obtained in three independent experiments (*, P < 0.01 compared to medium result). med, medium; RLU, relative light units.

Recombinant ProTα potently suppresses HIV-1 replication in primary macrophages but not in CD4⁺ T cells.

To determine if ProTα alone has HIV-1-suppressive activity, the same assays were performed with recombinant ProTα. Recombinant ProTα suppressed HIV-1_BaL in primary macrophages in a dose-dependent manner, reducing viral p24 by >80% at a concentration of 2 ng/ml (equal to the protein concentration of the fraction in Fig. 1A) (Fig. 1C). To limit the assay to a single round of viral infection, replication-defective JRFL envelope-pseudotyped HIV-1 (HIV-1_JRFL) containing a luciferase gene under the control of the HIV-1 LTR was used to infect primary macrophages. ProTα inhibited the expression of the luciferase reporter gene in a dose-dependent manner (Fig. 1D).

To determine the spectrum of anti-HIV-1 activity of ProTα, we tested the antiviral effect of recombinant ProTα in primary macrophages infected with HIV-1 primary isolates. ProTα (200 ng/ml) inhibited replication of an R5 as well as an X4/R5 HIV-1 primary isolate (Fig. 1E) in primary macrophages. In contrast, ProTα did not exhibit anti-HIV-1_JRFL activity in HeLa CD4⁺/CCR5⁺ cells (Fig. 1F) or Hos CD4⁺/CCR5⁺ cells challenged with the same virus (data not presented). However, the unfractionated CD8⁺ T-cell supernatant of K#1 50K did inhibit HIV-1_JRFL in HeLa CD4⁺/CCR5⁺ cells (Fig. 1F), indicating the presence of additional antiviral factors in the whole unfractionated supernatant.

In contrast to macrophages, recombinant ProTα demonstrated little to no anti-HIV-1 activity in transformed and primary CD4⁺ cells even at a dose of 200 ng/ml. Suppression varied from 50% (in primary CD4⁺ cells infected with HIV-1_BaL and primary isolate X4 or X4/R5) to no suppression of HIV-1_IIIB in primary or HIV-1_IIIB in transformed CD4 cells (Fig. 2). These data suggest that the inhibitory activity of ProTα is cell type dependent and does not account for the broad anti-HIV-1 activity of unfractionated CD8⁺ cell supernatants. No cytotoxic effect of ProTα was observed in macrophages or CD4⁺ T cells treated with up to 1 μg/ml of ProTα as measured by the CellTiter nonradioactive cell proliferation assay from Promega (data not shown).

FIG. 2.

Effect of recombinant ProTα on HIV-1 replication in CD4⁺ T cells. The indicated CD4⁺ T cells were infected with the indicated virus. Unbound virus was washed out, and cells were incubated with either medium alone (black bars) or 200 ng/ml of ProTα (gray bars). The amounts of extracellular HIV-1 Gag p24 protein expressed as means ± standard deviations for triplicate wells are from a single experiment at 7 days postinfection. H9 and CEM cells were infected with HIV-1_IIIB (MOI of 0.1) and treated with recombinant ProTα. PHA-activated primary CD4⁺ cells were treated with recombinant ProTα after infection with HIV-1_IIIB and HIV-1_BaL at an MOI of 0.001 or the HIV-1 X4- or X4/R5-tropic primary isolate at an MOI of 0.001.

Depletion of ProTα from the chromatographic fraction leads to partial removal of the anti-HIV-1 activity.

To ensure that ProTα contributed to the anti-HIV-1 activity of CD8⁺ cell supernatant, a sample of the chromatographic fraction from which ProTα was identified was applied to an anti-ProTα antibody affinity column. The eluate had anti-HIV-1 activity in a single-cycle infection assay similar to that of the parent fraction while the nonbinding flowthrough had lost 80% of the activity (Fig. 3A). Using a similar approach, suppression of HIV-1_BaL in primary macrophages was seen only with the bound eluate of recombinant ProTα (Fig. 3B). These results further indicate that ProTα is a major component of this active fraction derived from CD8⁺ cell supernatants and contributes to the anti-HIV activity.

FIG. 3.

Depletion of ProTα from the chromatographic fraction and from the recombinant preparation of ProTα results in the loss of HIV-1-inhibitory activity, and thymosin alpha 1 is not responsible for the anti-HIV-1 activity of ProTα. (A) Primary macrophages were infected with HIV-1_JRFL for 2 h and treated with the active chromatographic fraction at a concentration of 20 ng/ml of protein or eluate at a concentration of 20 ng/ml of protein from the affinity column or with the nonbinding fraction (flow) at a protein concentration of 20 ng/ml after residual virus had been washed out. Luciferase activity was measured at 72 h postinfection. The results (means ± standard deviations) in relative light units (RLU) are from triplicate determinations (*, P < 0.02 compared to medium). (B) Primary macrophages were infected with HIV-1_BaL at a multiplicity of infection of 0.1 for 2 h and treated with ProTα (20 ng/ml) or eluate from the affinity column or nonbinding fraction (flow) after residual virus had been washed out. HIV-1 p24 antigen was measured on day 7. (C) Primary macrophages were infected with HIV-1_JRFL for 2 h. Residual virus was washed out, and recombinant ProTα at indicated concentrations or chemically synthesized Tα1 was added. Luciferase activity (in RLU) was measured 72 h postinfection. The results (means ± standard deviations) of HIV-1 p24 or RLU are from triplicate determinations. (*, P < 0.02 compared to medium).

Tα1 has no anti-HIV-1 activity.

Tα1 is a peptide derived from the natural proteolytic cleavage of the 28 N-terminal amino acids of ProTα and can be detected in human plasma (43). Circulating levels of this peptide have been reported to be increased in HIV-1-positive individuals (40). Furthermore, it activates dendritic cells through toll-like receptor signaling (39). To determine whether this peptide is responsible for the anti-HIV-1 activity of ProTα, we used chemically synthesized Tα1 to assay for the ability to suppress HIV-1_JRFL-infected macrophages. Tα1 up to 1 μg/ml had no inhibitory activity while ProTα suppressed luciferase gene expression >80% at a concentration of 20 ng/ml (Fig. 3C).

ProTα inhibits HIV-1 LTR promoter-mediated gene expression.

In each of the above assays, ProTα was added following viral infection, suggesting that the effect was on steps in the virus life cycle following entry. To determine the step(s) of the viral life cycle inhibited in macrophages by ProTα, timed experiments were performed comparing ProTα to azidothymidine (AZT; HIV-1 reverse transcription inhibitor) and the integrase inhibitor L-731988. Macrophages were infected with HIV-1_VSV for 2, 24, and 48 h. At the end of each incubation period the virus was washed out and the cells were treated with ProTα, AZT, or integrase inhibitor. ProTα inhibited HIV-1 gene expression when added 2, 24, and 48 h postinfection while AZT had no effect when added at the later time points of 24 and 48 h (Fig. 4A). This suggested that ProTα inhibited HIV-1 gene expression after reverse transcription, which in macrophages is completed by 26 to 48 h (34). Moreover ProTα inhibited HIV-1 gene expression up to 48 h postinfection following completion of integration of HIV-1 in macrophages as indicated by the loss of activity of the integrase inhibitor (Fig. 4B).

FIG. 4.

Effect of ProTα on HIV-1 LTR-driven gene expression. Primary macrophages were infected with HIV-1_VSV for 2, 24, and 48 h. Residual virus was washed out, and ProTα at a concentration of 200 ng/ml or AZT at a concentration of 100 μM (A) or ProTα at a concentration of 200 ng/ml or integrase inhibitor L-731988 at a concentration of 5 μM (B) was added at each time point. Luciferase activity (in relative light units [RLU]) was measured 72 h postinfection, and the results were expressed as means ± standard deviations for triplicate wells.

To determine the effect of ProTα on HIV-1 RNA production, total RNA was extracted from primary macrophages infected with HIV-1_VSV and treated 72 h postinfection with ProTα (200 ng/ml) or medium for 24 h and analyzed by reverse transcription-PCR. Expression of a luciferase reporter gene mRNA was reduced by 82% compared to the medium alone in primary macrophages (data not shown).

ProTα does not induce phosphorylation of STAT1 in primary macrophages and HeLa CD4⁺ cells.

We have shown that activation of the signal transducer and activator of transcription 1 protein (STAT1) is necessary for inhibition of LTR activation and HIV-1 gene expression by unfractionated conditioned media of the HVS/CD8⁺ T-cell line K#1 50K (4) in certain cell lines. To establish whether ProTα also induces STAT1 activation, primary macrophages and HeLa CD4⁺ cells were briefly (15 min) incubated with the protein at 1 μg/ml or with 10% K#1 50K cell-conditioned medium after incubation for 2 h in serum-free medium to reduce baseline phosphorylation. A Western blot of the electrophoresed total lysate of the treated and untreated cells was probed with an antibody to the phosphorylated tyrosine 701 of STAT1 (Fig. 5). These results indicate that in contrast to the unfractionated conditioned medium, ProTα suppression of LTR-driven HIV-1 gene expression is not associated with STAT1 phosphorylation in primary macrophages.

FIG. 5.

Activation of STAT1 in primary macrophages and HeLa CD4⁺ cells was induced by CD8⁺ cell-conditioned medium but not by ProTα treatment. Primary macrophages and HeLa CD4 cells were treated with either 10% CD8⁺ T-cell-conditioned medium or 1 μg/ml of ProTα for 15 min, and then total cell lysate was prepared in a sample loading buffer. Western blotting analysis was done using anti-phospho-STAT1 antibody, and the blots were subsequently stripped and reprobed with anti-STAT1 antibody.

Inhibition of HIV-1 gene expression by ProTα is not selective for the HIV-1 LTR promoter.

To determine whether the effect of ProTα on HIV-1 gene expression is specific for the LTR promoter, we used virus generated by two other HIV-derived lentiviral vectors carrying a luciferase reporter gene under the phosphoglycerate kinase (PGK) promoter pVVPW/LBE or the cytomegalovirus (CMV) promoter pVVCW/LBE. Macrophages were infected with HIV-1_VSV carrying the luciferase gene under the control of the LTR, PGK, or CMV promoter (Fig. 6A, B, and C, respectively). At 48 to 72 h postinfection, when viral integration is completed, ProTα at 200 ng/ml was introduced. ProTα inhibited luciferase expression from all three of these promoters (Fig. 6A, B, and C), suggesting that ProTα is not selective for the HIV-1 LTR. However, ProTα had no effect on the level of expression of β-actin, tubulin, or ribosomal protein s11, suggesting that there is selectivity to the inhibition of transcription (data not shown).

FIG. 6.

Effect of the HDAC inhibitor sodium butyrate on repression of gene expression by ProTα. Macrophages were infected with HIV-1_VSV carrying a luciferase reporter gene under the LTR (A), PGK (B), or CMV (C) promoter for 48 h. After the virus was washed out, 200 ng/ml of ProTα, medium with 4 μM sodium butyrate, or ProTα and medium with 4 μM sodium butyrate was added to the cells. Luciferase activity (relative light units [RLU]) was measured 24 h posttreatment, and the results were expressed as means ± standard deviations for triplicate wells (*, P < 0.05 compared to medium). Similar results were obtained in three independent experiments.

Inhibition of gene expression by ProTα involves HDAC(s) without a direct effect on enzymatic activity of HATs.

Histone acetylation mediated by histone acetyltransferases (HATs) results in accessibility of the promoter region of the integrated HIV-1 provirus to transcriptional factors and is one of the mechanisms by which HIV-1 transcription is regulated (2, 29). ProTα has been shown to associate with the HAT p300 and influence transcription (15, 35, 46). Based on several reports indicating nuclear localization of ProTα and its association with histones and HATs, we tested whether HATs are targeted by ProTα and whether a histone deacetylase (HDAC) inhibitor can rescue HIV-1 gene expression in the presence of ProTα (15, 23, 50).

Primary macrophages were infected with vesicular stomatitis virus (VSV) envelope-pseudotyped HIV-1, and 48 h postinfection the cells were treated with ProTα alone or ProTα with sodium butyrate (a known inhibitor of histone deacetylases) (Fig. 6A). The same experiments were done in macrophages infected with VSV envelope-pseudotyped lentiviruses carrying the reporter gene luciferase under different promoters (PGK and CMV) as well (Fig. 6B and C). ProTα inhibited LTR-, PGK-, and CMV-driven gene expression in primary macrophages, and this inhibition was completely reversed by the histone deacetylase inhibitor sodium butyrate in the case of LTR and PGK promoter-driven transcription (Fig. 6A and B) and partially reversed in the case of CMV promoter-driven transcription (Fig. 6C).

To further investigate the mechanism of ProTα activity, we tested whether binding of ProTα to p300 reduced the histone acetylation function of this enzyme. We used a nonradioactive HAT assay kit (Upstate Biotechnology, Lake Placid, NY). Our results indicated no reduction of the in vitro enzymatic activity of p300 in the presence of different concentrations of recombinant ProTα (data not shown).

DISCUSSION

We identified ProTα as an active anti-HIV-1 component of one of the fractions derived from the supernatant of HSV-transformed CD8⁺ T cells. ProTα is a small acidic protein (12 kDa, 109 amino acids) that has been implicated in a number of diverse biologic activities (17). Known mostly for its association with cell proliferation, it has also been reported to have immunomodulatory activities (6, 44, 47), including chemotaxis. Several of the immunomodulatory activities have been reported with ProTα applied extracellularly; however, it is unclear whether the biologic activity is secondary to uptake or mediated via cell surface signaling. Regulation of ProTα has been reported to occur at the level of translation and intracellular localization (25). ProTα has a well-defined nuclear localization signal sequence, and it has been identified in the nucleus of a number of cells (11, 50, 51). While this protein has no secretory signal sequence, it has, nonetheless, been detected in human serum and the supernatants of cultured lymphocytes (10, 37), and in this report we have demonstrated its presence in supernatants of the HVS-transformed CD8⁺ cell line K#1 50K. Furthermore, a number of cell surface binding proteins have been identified as candidate receptors for extracellular ProTα. Regarding the anti-HIV effect of ProTα that we described here for macrophages, it remains to be determined if this is mediated via uptake of the exogenously added protein or via the induction of a signaling pathway (42).

Our screening assay was set up to specifically focus on postentry events of the HIV-1 life cycle in macrophages. While the HIV-1-suppressive activity of ProTα was significant in macrophages, there was little activity against HIV in T cells. Interestingly, infection of human T cells with HIV-1 is associated with a reduction in detectable amounts of ProTα mRNA (14, 41). Understanding the mechanism of the inhibitory activity of ProTα in the macrophage as well as control of expression in T cells may provide insights into cell-specific control of viral gene expression.

Several reports demonstrate that CD8⁺ T-cell supernatants inhibit HIV-1 LTR-mediated gene transcription (4, 7, 28). While ProTα appears to inhibit LTR-driven luciferase gene expression predominantly in macrophages, the transcriptional inhibition previously ascribed to unfractionated CD8⁺ cell supernatants was observed in T cells and other cell lines as well (4, 13, 31). The latter is associated with STAT-1 activation; however, ProTα did not induce STAT1 activation in macrophages and HeLa CD4⁺ cells. Differential requirements for transcriptional regulators of HIV in cells of monocytic compared to lymphocytic origin have been described elsewhere (20). Several reports demonstrate that CCAAT binding proteins (C/EBP) are necessary for HIV replication in primary macrophages but not in lymphocytes (20, 21, 26). We examined whether CCAAT box binding transcriptional factors are involved in ProTα suppression. Wild-type HIV BaL and an HIV BaL mutant, mC2,C3-BaL, which contains point mutations in two high-affinity C/EBP binding sites of the CCAAT box of LTR, were used for these experiments. Inhibition of wild-type HIV-1 replication by ProTα was observed as expected (97% ± 3%); however, mutated virus replication was suppressed to a lesser extent (63% ± 10%) (data not presented). These results suggest that CCAAT box binding transcriptional factors may be involved in but not completely responsible for ProTα-mediated suppression of HIV LTR.

In HIV-1-infected cells the integrated proviral genome is tightly packaged by chromatin and is silent in the absence of stimulation. Acetylation of the associated histone proteins can stimulate transcription from the integrated HIV-1 proviral genome. Another role ascribed to ProTα is involvement in chromatin remodeling, and in this context it has been reported to interact with a number of HDACs (24) while under other conditions it has been found in association with the HAT p300/CBP (15, 45). While we were not able to demonstrate direct inhibition of p300-associated HAT in an in vitro assay system, ProTα cell-associated suppression of gene expression was reversed by the HDAC inhibitor sodium butyrate (22, 45). This is consistent with, but not definitive proof of, the idea that HDACs are involved in the ProTα-mediated transcriptional inhibition in macrophages.

ProTα is a nonspecific inhibitor of HIV-1 gene expression in macrophages; however, it is not the only postintegration inhibitory factor secreted by the CD8⁺ cell line K#1 50K, as our data demonstrate. CD8⁺ cell supernatants make a number of factors that inhibit HIV-1 specifically, as in the case of β-chemokines, and nonspecifically, as in the case of ProTα. There are additional molecules yet to be identified that contribute to the overall broad HIV suppression seen with unfractionated CD8⁺ T-cell supernatants. Additional questions remain regarding the role of ProTα in in vivo control of replication and the mechanism of postintegration suppression of gene expression as well as whether it can be exploited for the development of novel therapeutics.