Macrophage Inflammatory Protein 1α Inhibits Postentry Steps of Human Immunodeficiency Virus Type 1 Infection via Suppression of Intracellular Cyclic AMP

ABSTRACT Primary isolates of human immunodeficiency virus type 1 (HIV-1) predominantly use chemokine receptor CCR5 to enter target cells. The natural ligands of CCR5, the β-chemokines macrophage inflammatory protein 1α (MIP-1α), MIP-1β, and RANTES, interfere with HIV-1 binding to CCR5 receptors and decrease the amount of virions entering cells. Although the inhibition of HIV-1 entry by β-chemokines is well documented, their effects on postentry steps of the viral life cycle and on host cell components that control the outcome of infection after viral entry are not well defined. Here, we show that all three β-chemokines, and MIP-1α in particular, inhibit postentry steps of the HIV-1 life cycle in primary lymphocytes, presumably via suppression of intracellular levels of cyclic AMP (cAMP). Productive HIV-1 infection of primary lymphocytes requires cellular activation. Cell activation increases intracellular cAMP, which is required for efficient synthesis of proviral DNA during early steps of viral infection. Binding of MIP-1α to cognate receptors decreases activation-induced intracellular cAMP levels through the activation of inhibitory G proteins. Furthermore, inhibition of one of the downstream targets of cAMP, cAMP-dependent PKA, significantly inhibits synthesis of HIV-1-specific DNA without affecting virus entry. These data reveal that β-chemokine-mediated inhibition of virus replication in primary lymphocytes combines inhibitory effects at the entry and postentry levels and imply the involvement of β-chemokine-induced signaling in postentry inhibition of HIV-1 infection.

The major cellular targets of human immunodeficiency virus type 1 (HIV-1) are CD4 ϩ T lymphocytes and macrophages. Infection of these cells is initiated by interactions between viral envelope proteins and specific cellular receptors. In addition to CD4 glycoprotein, which is a major HIV-1 receptor, several members of the chemokine receptor family have been identified as coreceptors for HIV-1 (6,(11)(12)(13). Most primary strains of HIV-1 have an R5 phenotype and use CCR5, a member of the ␤-chemokine receptor family, to enter target cells. The natural ligands of CCR5, the ␤-chemokines macrophage inflammatory protein 1␣ (MIP-1␣), MIP-1␤, and RANTES, represent host factors with potential anti-HIV-1 activity (7). In vitro studies clearly demonstrate that these molecules suppress HIV-1 entry into target cells by interfering with interactions between the virus and CCR5 receptors (27,38,39). However, the consequences of downstream signaling events induced by ␤-chemokine binding to their cognate receptors, CCR5 and/or CCR1, on the replication of HIV-1 R5 strains are not well defined, although several studies show an effect of ␤-chemokine-induced signaling on the replication of CXCR4-using viruses (15,24).
Chemokine receptors belong to the superfamily of seven transmembrane-domain, G protein-coupled receptors. Several major intracellular signaling pathways are triggered by G protein-coupled receptors upon ligand binding. These pathways include the cyclic AMP (cAMP)/protein kinase A (PKA) pathway, the phosphatidylinositol/calcium/protein kinase C pathway, and the mitogen-activated protein kinase pathway (3,14,33). In eukaryotic cells, the cAMP/PKA pathway is one of the most common and versatile signaling pathways regulated by G protein-coupled receptors. Two G proteins, G s and G i , regulate intracellular levels of cAMP through the direct modulation of the activity of adenylyl cyclase, an enzyme that catalyzes the conversion of ATP to cAMP. G s activates adenylyl cyclase, which results in increased cellular levels of cAMP, whereas G i inhibits adenylyl cyclase.
Intracellular cAMP is a second messenger that influences metabolism, cell shape, chemokine receptor expression, and gene transcription via reversible protein phosphorylation (8,34). Although cAMP has been shown to activate ion channels and Rap guanine exchange factors Epac 1 and 2 (reviewed in reference 26), the principal cAMP target is cAMP-dependent PKA. PKA is a serine/threonine kinase that regulates a number of cellular processes important for immune activation (37). Hyperactivation of the cAMP/PKA pathway has been implicated in the T-cell dysfunction associated with a common variable immunodeficiency (2) and HIV-1 infection (28). HIV-1 infection results in a functional impairment of CD4 ϩ T cells before a quantitative decline becomes evident. The inability of T cells to generate a vigorous response to HIV-1 antigens continues to persist in patients on highly active antiretroviral therapy despite markedly reduced or undetectable levels of HIV-1 RNA. It has been shown that HIV-1 infection is asso-ciated with increased intracellular levels of cAMP and constitutive activation of PKA (19). Inhibition of the cAMP/PKA pathway, either by reducing intracellular cAMP levels with adenosine analogues (19) or PKA antagonists (1), can restore immune responses in T cells isolated from HIV-1-infected patients. Besides affecting T-cell responses, the cAMP/PKA pathway can accelerate HIV-1 replication directly (29,31) as well as indirectly through modulation of cytokine production (9,23). Furthermore, it has been shown recently that the active catalytic subunit of PKA (C-PKA) incorporates within highly purified HIV-1 particles and that impairment of host cell C-PKA activity by a synthetic inhibitor at the time of virus release results in the production of virions with reduced infectivity (5).
In this study, we demonstrate that MIP-1␣, in addition to inhibiting virus entry, inhibits postentry steps of the HIV-1 life cycle in primary lymphocytes via suppression of intracellular levels of cAMP. The magnitude of MIP-1␣-mediated postentry inhibition depends on activation-induced levels of intracellular cAMP, which are highly donor dependent. We further demonstrate that activation-induced levels of intracellular cAMP and active cAMP-dependent PKA contribute to the efficient synthesis of proviral DNA during early steps of the HIV-1 life cycle.

MATERIALS AND METHODS
Primary lymphocyte cultures. Peripheral blood mononuclear cells from healthy donors were obtained by Ficoll-Hypaque gradient centrifugation. To eliminate monocytes, cells were incubated for 2 h in PRIMARIA flasks (Falcon). Lymphocytes (nonadherent cells) were collected and subjected to the second round of purification by overnight adherence. Nonadherent cells recovered after the second round of adherence were greater than 80% CD3 ϩ , as determined by fluorescence-activated cell sorter analysis. Cells were then resuspended in RPMI 1640 medium supplemented with 10% heat-inactivated fetal calf serum and stimulated either with phytohemagglutinin (PHA) (5 g/ml PHA) or with anti-CD3 antibody (1 g/ml) coated onto plates. PHA was present in the culture medium for 3 days. Afterwards, cells were washed and resuspended in medium containing 20 U/ml of interleukin-2 (Roche). Cell proliferation was determined using an [ 3 H]thymidine incorporation assay.
Viruses and infection. HIV-1 R5 strains HIV-1 ADA and HIV-1 92US660 were used in this study. Immediately before infection, an aliquot of the viral stock was treated with 200 U/ml of RNase-free DNase (Roche Molecular Biochemicals) to eliminate any viral DNA contamination. Cells were infected for 2 h at 37°C with an amount of virus corresponding to 5 ϫ 10 4 cpm of reverse transcriptase activity per million cells. The amount of the virus was determined by a standard reverse transcriptase assay.
Cell assays. To analyze the effects of ␤-chemokines on HIV infection, cells were treated with 200 ng/ml of MIP-1␣, MIP-1␤, or RANTES, obtained from Serono Pharmaceuticals, at the times indicated. To determine the effects of MIP-1␣ on intracellular levels of cAMP in HIV-1-infected lymphocytes, PHAstimulated cells were incubated with the virus for 1 h at 4°C to allow binding of the virus to cellular receptors. Following the addition of MIP-1␣, cells were incubated for two additional hours before preparing cell lysates for cAMP analysis. PKA inhibitors PKA 14-22 (a cell-permeable peptide inhibitor) and H89 (both from Calbiochem) were added to the cultures at the indicated concentrations 1 h before infection. Since H89 is solubilized in dimethyl sulfoxide (DMSO), parallel cultures were treated with DMSO alone to rule out the possibility that the observed effects on HIV-1 infection were due to solvent.
Analysis of intracellular cAMP. Cell lysates prepared from primary lymphocyte cultures were analyzed for levels of intracellular cAMP using a commercially available enzyme immunoassay system (Amersham Pharmacia Biotech) according to the manufacturer's instructions.
Detection of HIV-1-specific DNA by PCR. Samples from infected cultures were prepared and subjected to PCR analysis using HIV-1-specific primers amplifying pol and long terminal repeat (LTR) RU5 transcripts as described previously (35). Amplified DNA was analyzed by Southern blot hybridization using 32 P-labeled probes and quantified using an Instant Imager System (Packard). Results are expressed in counts per minute. Amplification of the ␣-tubulin gene was used to control for the amount of DNA in each sample. Serial dilutions of 8E5/LAI cells, containing one HIV-1 genome per cell, were included in each amplification reaction to standardize the results.
Statistical analysis. Statistical and curve fit analysis of the correlation between cAMP levels and MIP-1␣-mediated postentry inhibition and the dependence of MIP-1␣-mediated decreases in intracellular cAMP on initial cAMP levels were performed using GraphPad Prism and Sigma Plot software packages, respectively. P values are considered significant when they are Ͻ0.05.

␤-Chemokines inhibit postentry steps of HIV-1 infection in primary lymphocytes.
Although in vitro studies clearly demonstrate that ␤-chemokines inhibit HIV-1 entry into target cells, the consequences of chemokine-induced signaling events on viral replication are not well defined. In the current study, we examined whether ␤-chemokine-induced signaling affects infection of primary cells with R5 strains of HIV-1 by focusing on the postentry steps of the viral life cycle. We treated primary lymphocytes with MIP-1␣, MIP-1␤, or RANTES either before and after, before, or only after infection with HIV-1 ADA (R5 strain). In pretreated samples, ␤-chemokines were also present during infection in order to prevent the reappearance of CCR5 receptors down-regulated during chemokine pretreatment and to further allow competition with the virus for receptor binding. Afterwards, we analyzed synthesis of HIV-1-specific DNA (pol transcripts) during one viral life cycle (24 h) by PCR. All three ␤-chemokines inhibited HIV-1 DNA synthesis not only when they were present before and after or before infection but also when they were added only after infection with HIV-1 ADA (Fig. 1A). It has been shown previously that CCR5 receptors are resistant to trypsin treatment (41). Thus, in the next experiments we treated cells with trypsin immediately after incubation with the virus and prior to ␤-chemokine treatment in order to verify that the inhibition is not due to the interference of ␤-chemokines with the entry of residual noninternalized virus. The results obtained with trypsin-treated cells were similar to those presented in Fig. 1A (data not shown). These data indicate that ␤-chemokines, in addition to inhibiting virus entry, inhibit postentry steps of the HIV-1 viral life cycle in primary lymphocytes. To further confirm this observation, we performed an additional set of experiments. It has been shown that HIV-1 can enter and initiate reverse transcription in quiescent lymphocytes; however, the reverse transcription process is not completed. The virus can be rescued from these cells by subsequent stimulation, which reinitiates DNA synthesis (40). Thus, we infected lymphocytes with the primary R5 isolate HIV-1 92US660 without prior activation. Cells were stimulated with anti-CD3 antibody and treated with ␤-chemokines 72 h after infection as shown in the schematic (Fig. 1B, upper panel). All three ␤-chemokines decreased synthesis of viral DNA under these experimental settings (Fig. 1B, bottom panel), thus confirming the inhibitory effect of ␤-chemokines on postentry steps of the HIV-1 life cycle. Although we observed postentry inhibition in lymphocytes isolated from almost all blood donors, the magnitude of the inhibition varied among donors. Postentry inhibition ranged from 10 to 90%, independently of whether cells were activated with anti-CD3 or PHA. Furthermore, the relative efficacy of each chemokine varied from donor to donor. Since MIP-1␣ treatment gave the most consistent results (Fig. 2), with dose-dependent effects (Fig. 3), we decided to focus our study on MIP-1␣-mediated effects on HIV-1 infection.
MIP-1␣-mediated postentry inhibition is abolished by PTX and by exogenously supplied cAMP. MIP-1␣ can bind and signal through both CCR5 and CCR1. Previous studies demonstrated that chemokine binding induces coupling of both CCR5 and CCR1 to inhibitory G (G i ) proteins (25,42). To determine whether MIP-1␣-mediated postentry inhibition of HIV-1 involves activation of G i proteins, we incubated parallel cultures of lymphocytes in the presence or absence of 5 nM pertussis toxin (PTX) prior to HIV-1 infection and MIP-1␣ treatment. Pretreatment with PTX, a specific inhibitor of G i protein signaling, abolished MIP-1␣-mediated inhibition without significantly affecting the levels of pol transcripts in infected cells that were not treated with MIP-1␣ (Fig. 4). Activated G i proteins inhibit adenylyl cyclase, resulting in decreased intracellular levels of cAMP. Since the involvement of cAMP in the enhancement of HIV-1 replication has been demonstrated previously (29,31), we reasoned that MIP-1␣ might inhibit synthesis of HIV-1-specific DNA via suppression of intracellular cAMP. We predicted that in such a case exogenously supplied cAMP would override the postentry inhibitory effect of MIP-1␣. As expected, treatment with a cell- be linked to differences in activation-induced intracellular levels of cAMP. We analyzed intracellular cAMP levels in parallel cultures of lymphocytes (quiescent and PHA-stimulated) isolated from five blood donors. Our results (Fig. 6A)  high degree of donor-dependent variability in activation-induced levels of cAMP. Next we examined whether high cAMP levels might restrain MIP-1␣-mediated postentry inhibition. We analyzed lymphocytes isolated from 10 blood donors for an association between activation-induced intracellular cAMP levels and magnitude of MIP-1␣-mediated inhibition of HIV-1-specific DNA synthesis. Regression analysis of the data yielded a significant inverse correlation between activationinduced cAMP levels and postentry inhibition (Fig. 6B). These results suggest that the efficacy of MIP-1␣-mediated suppression of intracellular cAMP might be limited in cells isolated from donors that respond to cell activation with large increases in intracellular cAMP. Indeed, analysis of intracellular cAMP in HIV-1-infected lymphocytes treated with MIP-1␣ and the comparison of detected levels to those in untreated cells revealed that MIP-1␣-mediated suppression of intracellular cAMP highly depends on activation-induced levels of cAMP (Fig. 6C).
Inhibition of cAMP-dependent PKA suppresses synthesis of HIV-1-specific DNA. One well-defined downstream target of cAMP is cAMP-dependent PKA, but other cAMP effectors have been identified as well (26). Therefore, we wanted to determine whether PKA is involved in the synthesis of HIV-1-specific DNA in primary lymphocytes. We treated cells with two different inhibitors of PKA, inhibitory peptide PKA 14-22 and chemical inhibitor H89, 1 h prior to infection with HIV-1 92US660 and analyzed HIV-1-specific pol transcripts 24 h later by PCR. Both inhibitors markedly suppressed synthesis of HIV-1-specific DNA in lymphocyte cultures (Fig. 7A). To verify that PKA regulates postentry steps of HIV-1 infection without affecting virus entry, we treated lymphocytes with PKA inhibitors as described above and infected them for 2 h. Afterwards, we analyzed the synthesis of HIV-1-specific strongstop DNA (LTR RU5), the early product of reverse transcription synthesized shortly after viral entry. Cultures pretreated with MIP-1␣ served as a control. As can be seen in Fig. 7B, inhibition of PKA did not abolish HIV-1 entry into primary lymphocytes. Levels of strong-stop DNA were similar in treated and untreated samples, while pretreatment with MIP-1␣ markedly decreased these levels. These data suggest that PKA, a downstream target of cAMP, is involved in the regulation of HIV-1-specific DNA synthesis during early steps after virus entry.

DISCUSSION
In this study, we investigated the effects of ␤-chemokines on postentry steps of HIV-1 infection in primary lymphocytes. The presented data indicate that ␤-chemokine-induced signaling inhibits synthesis of HIV-1-specific DNA during early steps of the viral life cycle in this cellular target of the virus (Fig. 1). Focusing on the MIP-1␣-mediated inhibition, we show that MIP-1␣ suppresses activation-induced intracellular cAMP levels ( Fig. 5 and 6C) through the activation of inhibitory G proteins (Fig. 4). The outcome of MIP-1␣-induced signaling depends on the state of activation of adenylyl cyclase, as reflected by intracellular levels of cAMP, and is highly donor dependent ( Fig. 6A and B). Furthermore, our data demonstrate that inhibition of a downstream target of cAMP, cAMPdependent PKA, results in the suppression of synthesis of HIV-1-specific DNA without affecting HIV-1 entry (Fig. 7). Altogether, our results support a model (Fig. 8) in which lymphocyte stimulation activates adenylyl cyclase, leading to increased cAMP levels and PKA activation. This cascade, most likely in combination with other cellular factors, contributes to the enhancement of synthesis of proviral DNA. Several studies have suggested that viral proteins might also increase intracellular cAMP (16,28). Several early studies have shown that increased intracellular levels of cAMP enhance HIV-1 replication, while inhibition of a downstream target of cAMP, cAMP-dependent PKA, has the opposite effect (18,(29)(30)(31). In their recent paper, Cartier et al. (5) demonstrate that active cAMP-dependent PKA is incorporated in HIV-1 particles where it interacts and possibly catalyzes phosphorylation of the viral CAp24 gag protein, thus suggesting one mechanism by which cAMP/PKA modulates HIV-1 infection. Decreased amounts of virions with reduced infectivity are produced from cells transfected with pNL4-3 and cultivated in the presence of PKA inhibitors. Cartier et al. suggest that C-PKA plays a role in viral particle release as well as in either uncoating events or transcription or integration steps of the HIV-1 life cycle. Our results indicate that the cell-associated cAMP/PKA pathway might also modulate HIV-1 infection through the enhancement of synthesis of HIV-1-specific DNA during early postentry reverse transcription steps and is one of the targets of MIP-1␣-induced signaling. It is likely that the inhibition of HIV-1 replication during longterm cultivation in cells deprived of PKA activity might combine several mechanisms, since cAMP/PKA is a versatile pathway that can affect several cellular functions (reviewed in reference 36) required for efficient replication of the virus.
While several studies, including data presented here, show that the cAMP/PKA pathway enhances HIV-1 infection in lymphocytes, the studies in macrophages appear to yield quite different data. Hayes et al. (17) recently showed inhibition of HIV-1 replication in macrophages by prostaglandin E 2 through the activation of PKA. Therefore, we analyzed the effects of ␤-chemokines and PKA inhibitors on HIV-1 replication in primary macrophages. Interestingly, contrary to what we have shown for primary lymphocytes, in macrophages isolated from the same donors, neither ␤-chemokines nor PKA inhibitors showed the inhibitory effects on postentry steps of the HIV-1 life cycle (data not shown). This might suggest that chemokines trigger signaling events in macrophages different from those induced in primary lymphocytes. However, consid-ering the data published by Hayes et al. (17), a more likely explanation would be that cAMP/PKA might exert differential effects on HIV-1 replication in primary lymphocytes and macrophages.
The results presented in this report reveal that ␤-chemokine-mediated inhibition of HIV-1 replication in primary lymphocyte cultures is complex and can be mediated through several mechanisms. Although it is difficult to extrapolate in vitro data to the in vivo setting, especially the relative importance of chemokine-mediated inhibition of entry versus postentry steps, it is plausible to envision a hypothetical implication of our data. It has been suggested that elevated ␤-chemokine levels play an important role in preventing HIV-1 infection in uninfected individuals with multiple high-risk sexual exposures (21). However, different results have been reported when ␤-chemokine levels were analyzed during established HIV-1 infection (22,32) where other factors might counterbalance their inhibitory effects. Taking into consideration that increased intracellular cAMP levels and constitutively active PKA have been reported for HIV-1-infected patients (1,10,19), our studies raise the possibility that intracellular cAMP might be one of those host cellular factors counterbalancing the antiviral effects of ␤-chemokines.