Strong evidence for the direct modulation of the immune system by opioids is well documented. μ-Opioids have been shown to alter the release of cytokines important for both host defense and the inflammatory response. Proinflammatory chemokines monocyte chemoattractant protein-1 (MCP-1), RANTES, and IFN-γ-inducible protein-10 (IP-10) play crucial roles in cell-mediated immune responses, proinflammatory reactions, and viral infections. In this report, we show that [d-Ala2,N-Me-Phe4,Gly-ol5]enkephalin (DAMGO), a μ-opioid-selective agonist, augments the expression in human PBMCs of MCP-1, RANTES, and IP-10 at both the mRNA and protein levels. Because of the proposed relationship between opioid abuse and HIV-1 infection, we also examined the impact of DAMGO on chemokine expression in HIV-infected cells. Our results show that DAMGO administration induces a significant increase in RANTES and IP-10 expression, while MCP-1 protein levels remain unaffected in PBMCs infected with the HIV-1 strain. In contrast, we show a dichotomous effect of DAMGO treatment on IP-10 protein levels expressed by T- and M-tropic HIV-infected PBMCs. The differential modulation of chemokine expression in T- and M-tropic HIV-1-infected PBMCs by opioids supports a detrimental role for opioids during HIV-1 infection. Modulation of chemokine expression may enhance trafficking of potential noninfected target cells to the site of active infection, thus directly contributing to HIV-1 replication and disease progression to AIDS.
Endogenous and exogenous opioids exert physiologic effects on the CNS via engagement of seven transmembrane, G protein-coupled opioid receptors expressed by certain neuronal cells. Three major classes of opioid receptors, μ, κ, and δ, have been identified and cloned from neuronal cell lines or brain tissue (1, 2, 3, 4, 5, 6). Since initial studies of opioid receptor expression in the CNS, evidence has accumulated which suggests that there is a neuroimmune circuit involving opioid pathways. For example, radiolabeled agonist- and antagonist-binding analyses have shown that opioid receptors are also expressed on cells of the immune system (7, 8, 9). Moreover, μ-, κ-, and δ-opioid receptors have been cloned by RT-PCR from mRNA isolated from lymphocytes and macrophages and are identical to receptors in the CNS (10, 11, 12). Furthermore, it has been shown that the μ-opioids possess broad immunomodulatory activity, including the inhibition of NK cell activity (13, 14), mitogen responses (14, 15, 16, 17, 18), Ab production (19, 20, 21, 22, 23, 24), delayed-type hypersensitivity reactions (25, 26, 27), thymocyte surface marker expression (28, 29, 30, 31, 32), and macrophage and polymononuclear cell function (33, 34, 35, 36, 37, 38, 39, 40, 41). μ-Opioids have also been shown to alter the release of cytokines important for both host defense and the inflammatory response. On the one hand, studies have suggested that exposure to morphine suppresses IFN-γ, IL-2, and IL-4 production by lymphocytes (18, 42, 43, 44). Moreover, exposure of PBMCs to morphine inhibited production of IFN-γ in response to Con A and varicella zoster virus (45). In contrast, human PBMCs treated with morphine produced increased levels of TGFβ in response to LPS or PHA (46). In subsequent studies, Chao et al. (47) found that morphine inhibited release of TNF from PBMCs in response to LPS or PHA. The morphine-induced suppression of TNF-α release was partially inhibited by Abs to TGFβ, indicating that morphine suppression was mediated through production of TGFβ (47). Paradoxically, Chao et al. (48) found that morphine elevated TNF-α production in murine neonatal microglial cell cultures. The nature of this apparently contradictory activity of morphine remains undefined.
Based on previous studies showing the capacity of μ-opioids to alter proinflammatory cytokine expression, we hypothesized that opioid administration might alter chemokine production. Our data show that the μ-selective agonist [d-Ala2,N-Me-Phe4,Gly-ol5]enkephalin (DAMGO)3 elevates both the mRNA transcripts and protein expression of the CC chemokines, monocyte chemoattractant protein (MCP)-1, and RANTES, as well as the C chemokine IFN-γ-inducible protein (IP)-10 in PBMC cultures. Since MCP-1, RANTES, and IP-10 are potent chemoattractants for both monocytes and certain populations of lymphocytes, this up-regulation of chemokine expression by opioids may influence trafficking of potential noninfected target cells to the site of active infection. Thus, enhanced expression of MCP-1, RANTES, and IP-10 may directly contribute to HIV-1-induced T cell depletion, leading to immunosuppression, pathogenesis, and progression to AIDS.
Materials and Methods
Isolation of PBMCs and culture
PBMCs were obtained from the whole blood of normal donors and isolated by Ficoll-Paque Plus (Pharmacia Biotech, Piscataway, NJ) density gradient centrifugation. Isolated PBMCs were plated to a cell density of 2 × 106 cells/ml in 24-well tissue culture plates. Cell cultures were maintained in RPMI 1640 medium (Life Technologies, Rockville, MD) supplemented with 10% heat-inactivated (56°C, 30 min) endotoxin-free FCS (HyClone, Logan, UT), 10 μg/ml gentamicin reagent solution (Life Technologies), and 1 mM l-glutamine (Life Technologies). PBMCs were cultured in the presence or absence of the T cell mitogen PHA at 5 μg/ml for 24 h.
Opioid agonists and antagonists and treatment of PBMCs
The μ-selective agonist DAMGO and the μ-selective antagonist H-d-Phe-Cys-Tyr-d-Trp-Arg-Thr-Pen-Thr-NH2 (CTAP) were purchased from Multiple Peptide Systems (San Diego, CA). PHA-stimulated and nonstimulated PBMCs were pretreated with medium alone or medium containing CTAP for 1 h before adding the designated concentration of DAMGO. PBMCs were then cultured in the presence of opioid for 24–96 h.
The T-tropic IIIB and M-tropic JRFL strains of HIV were obtained from the National Institute of Allergy and Infectious Diseases AIDS Research and Reference Reagent Program operated by ERC Bioservices Corporation (Rockville, MD). The IIIB strain of HIV-1 was propagated in the human T cell line Molt4. The multiplicity of infection (MOI) of the IIIB strain was determined by quantitating syncytia formed by HIV-infected lymphocytes when cocultured with exponentially growing CD4-bearing SupT1 cells. All M-tropic strains of HIV were propagated in cultures of PBMCs from normal donors. Virus was isolated from culture supernatants and purified by pelleting at 110,000 × g for 90 min. This procedure produces stock virus of between 106 and 107 syncytia-forming units/0.1 ml. The 50% tissue culture infectious dose for the M-tropic virus was determined using PBMCs. Briefly, graded doses of virus were inoculated onto PBMCs and the extent of replication was measured every 3 days in the culture fluid by a p24 Ag ELISA. The final titer was calculated at the peak time of virus production.
HIV-1 infection of PBMCs
PBMCs were cultured in the presence or absence of PHA and, after 24 h in culture, the cells were treated with DAMGO for 1 h. In designated experiments, the cells were treated with CTAP for 1 h before addition of agonist. Titered viral strains were resuspended in serum-free, low-endotoxin RPMI 1640 medium and then added to the PBMCs at an MOI of 0.1. PHA-stimulated and nonstimulated, opioid-treated PBMCs cultures were infected for 2.5 h at 37°C, and excess HIV was removed by washing. Cells and/or supernatant were harvested after the designated culture period to determine chemokine mRNA or protein levels, respectively.
Measurement of chemokine mRNA by RNase protection analysis
The expression of chemokines was measured by RNase protection analysis using the RiboQuant MultiProbe RNase Protection Assay System (PharMingen, San Diego, CA). Briefly, 10 μg of RNA from each sample, isolated using the RNAzol method (Cinna/Biotecx Laboratories International, Friendswood, TX), was allowed to hybridize in solution with the radiolabeled antisense RNA probe generated with the RNA probe set for the human chemokines lymphotactin (Ltn), RANTES, macrophage-inflammatory protein (MIP)-1α, MIP-1β, IP-10, MCP-1, IL-8, and I-309 according to manufacturer’s instructions. The hybridized 32P-labeled probe-transcript duplex was subjected to digestion with RNase, and the protected probes were purified and resolved on 5% denaturing polyacrylamide. The gels were then dried and exposed to a phosphor imaging screen, and protected fragments were visualized and quantitated using a model GS-525 phosphor imager (Bio-Rad, Hercules, CA). Results are expressed as relative units, which are calculated after normalizing the L32 OD values.
Measurement of chemokine protein levels by ELISA
The concentration of chemokines present in culture supernatants was determined by ELISA, using matched mouse mAb capture and detection Abs in a sandwich ELISA (PharMingen). Antichemokine “capture” Abs used in these experiments were rabbit polyclonal anti-human RANTES, murine monoclonal anti-human MCP-1 (clone 10F7), and anti-human IP-10 (clone 4D5). The capture Abs were coated onto plastic microwell plates (Nunc Maxisorb; Nunc, Naperville, IL) and blocked with 1% BSA-containing PBS, and graded dilutions of culture supernatant or recombinant standard were added. After washing, the captured chemokine proteins were detected using biotin-conjugated anti-chemokine “detection” Abs rabbit polyclonal anti-human RANTES, anti-human MCP-1 (clone 5D3-F7), and anti-human IP-10 (clone 6D4), followed by HRP-linked streptavidin. Following the addition of 2,2′-azino-bis-(3-ethylbenzthiazoline-6-sulfonic acid) in buffer, the level of colored product was measured spectrophotometrically (405 nm).
The data are presented as the mean ± SD. Where appropriate, the statistical difference between experimental groups was assessed by using Student’s t test.
Analysis of chemokine mRNA and protein expression in PBMCs
Because of the critical function of chemokines as proinflammatory mediators (18, 35, 47) and based on the established capacity of μ-opioids to alter proinflammatory cytokines, we were interested in evaluating the effect of the μ-opioid agonist DAMGO on the expression of C, CC, and CXC chemokines. To address this question, we first determined the basal level of chemokine mRNA expression in PBMCs. Cultures of PBMCs were maintained for 24, 48, 72, and 96 h in the presence of PHA, and total RNA was isolated at different time points and evaluated by RNase protection assay. The level of expression of the constitutive gene L32 was included to allow quantitative comparisons among the different samples. RNase protection analysis showed detectable levels of CC chemokines MIP-1β, MIP-1α, MCP-1, and I-309, as well as CXC chemokine IL-8, immediately after isolation (time 0) (Fig. 1,A). In contrast, the C chemokine Ltn, the CC chemokine RANTES and the CXC chemokine IP-10 were not detected in nonstimulated cells at time 0 (Fig. 1,A). CC chemokines RANTES, MCP-1, and I-309, as well as CXC chemokine IL-8, mRNA expression peaked at 48 h (Fig. 1, A and B); however, Ltn, MIP-1α, MIP-1β, and IP-10 mRNA expression peaks later, at 72 h in culture. Supernatants from PHA-stimulated PBMCs were evaluated for the levels of MCP-1, RANTES, and IP-10 protein (Fig. 1 C). ELISA results show the peak production of MCP-1 was at 72 h, whereas RANTES and IP-10 reached a peak at 48 h. However, there was no significant difference in the levels of these chemokines between 48 and 72 h; we adopted the 72-h time point for our further studies.
DAMGO treatment of PBMCs results in an increase in chemokine protein expression
In an effort to better understand the influence of μ-opioids on cytokine production, we investigated the effect of DAMGO administration on the expression of chemokines by PBMCs. Cells were cultured with 100 nM DAMGO for 24, 48, and 72 h, and the level of MCP-1, IP-10, and RANTES expression was determined (Fig. 2). The results show that the DAMGO treatment induced the production of each of these chemokines, with maximal production at 48–72 h. We extended these studies to determine the effect of DAMGO on chemokine production in nonactivated and PHA-stimulated PBMCs. Cells were cultured in the presence or absence of PHA for 24 h, followed by DAMGO treatment, and MCP-1, RANTES, and IP-10 protein production was determined. Under these conditions, DAMGO administration to nonstimulated PBMCs resulted in a dose-dependent 3-, 4.5-, and 5-fold increase in the protein levels of MCP-1, RANTES, and IP-10, respectively (Fig. 3), and a 3.5-, 3-, and 4-fold increase in these chemokines in PHA-activated PBMCs. However, only after treatment with higher doses of DAMGO (1 μM) to nonactivated, as well as PHA-stimulated, PBMCs did we observe an increase in MCP-1 levels. In contrast, lower and more physiological concentrations of DAMGO (0.1 nM) induced RANTES expression in nonactivated PBMCs. Interestingly, even subnanomolar concentrations of DAMGO could significantly increase IP-10 protein levels in PHA-activated PBMCs.
DAMGO-induced expression of chemokines occurs at the mRNA level and is mediated by the μ-opioid receptor
In an effort to characterize the mechanism of DAMGO-induced up-regulation of chemokine protein levels by PBMCs, we conducted experiments to quantitate chemokine mRNA levels after DAMGO treatment. RNase protection analysis showed that DAMGO administration induces a significant increase in the levels of RANTES and MCP-1 mRNA in nonstimulated PBMCs (Fig. 4), suggesting that the opioid acts at the level of transcription. In addition, to determine whether DAMGO induction of chemokine expression is mediated through the μ-opioid receptor, we conducted experiments with CTAP, a selective μ-opioid antagonist (Fig. 4). PBMCs were cultured in the presence of 10−6 M CTAP for 1 h before administration of 10−7 M DAMGO, RNA was isolated from cells at 48 h, and the level of chemokine mRNA was determined. We found that CTAP alone had no effect on the expression of either MCP-1 or RANTES; however, CTAP pretreatment abolished the DAMGO-induced increase in both MCP-1 and RANTES production (Fig. 4). Attempts made to detect the induction of IP-10 mRNA and the effect of CTAP administration were unsuccessful due to the low level of expression of this transcript (data not shown). These results suggest that the DAMGO-induced increase in MCP-1 and RANTES expression is mediated through the μ-opioid receptor.
DAMGO induces expression of MCP-1, RANTES, and IP-10 in HIV-infected stimulated and nonactivated PBMCs
Evidence suggests that chemokines differentially regulate HIV replication during HIV disease progression to AIDS. Certain β chemokines have been shown to inhibit viral replication by competing with HIV for binding to the HIV coreceptors (49). Recent studies have also shown that pretreatment of T cells with β chemokines MIP-1α, MIP-1β, and RANTES increased the absorption and replication of some T-tropic HIV strains (50). In view of these findings and the established capacity of μ-opioids to augment the replication of HIV-1 in vitro (51), we tested the effect of DAMGO on chemokine production by HIV-1-infected PBMCs. Activated and nonactivated PBMCs were treated with DAMGO and, after 1 h, cells were infected with either HIVJRFL (M-tropic) or HIVIIIB (T-tropic) strains of HIV-1. MCP-1, RANTES, and IP-10 expression were then determined by ELISA. The results show that both M- and T-tropic HIV infection alone augmented the levels of MCP-1. However, DAMGO pretreatment of T-tropic HIV-1-infected nonactivated cells resulted in a significant increase in RANTES (Fig. 5,C) and IP-10 (Fig. 5,E), but not MCP-1 (Fig. 5,A), levels. Treatment with as little as 0.1 nM DAMGO induced a 7-fold increase in IP-10 protein levels in T-tropic HIV-infected nonactivated PBMCs (Fig. 5,E). Conversely, DAMGO pretreatment failed to substantially alter MCP-1, RANTES, or IP-10 expression in T-tropic HIV-1-infected PHA-activated cells (Fig. 5, B, D, and F). DAMGO treatment exerted differential effects on RANTES and IP-10 protein expression in both nonactivated and activated M-tropic HIV-infected PBMCs. Specifically, in M-tropic HIV-infected activated PBMCs, administration of DAMGO at a high concentration (1 μM) significantly elevated RANTES expression 13-fold (Fig. 5,D), whereas levels as low as 1 nM reduced IP-10 expression 3-fold (Fig. 5,F). We observed the same opposing effects of DAMGO treatment on RANTES and IP-10 in nonactivated cells (Fig. 5, C and E). Our results suggest that DAMGO administration differentially regulates RANTES and IP-10 protein expression in HIV-infected PBMCs, and this appears to be dependent on viral tropism.
In the present studies, we have found that DAMGO pretreatment of either activated or naive PBMCs resulted in a significant increase in MCP-1, RANTES, and IP-10 chemokine mRNA and protein expression. Administration of DAMGO, at concentrations as low as 1 pM, enhanced IP-10 protein levels in PHA-stimulated PBMCs. We were able to block the induction of MCP-1 and RANTES mRNA expression with the μ-selective opioid antagonist CTAP, showing that these effects are mediated through the μ-opioid receptor. Moreover, preliminary studies show that δ- and κ-opioid agonists do not induce significant changes in chemokine expression (data not shown), supporting the notion that the μ-opioid receptor is responsible for mediating the DAMGO effects. These findings provide additional evidence that μ-opioids modulate immune function by altering the production of proinflammatory chemokines by cells of the immune system. Reports from several laboratories have shown that endogenous endorphins and enkephalins increase the production of proinflammatory cytokines, including IL-1, IL-2, and IFN-γ (52, 53, 54, 55). Evaluation of these findings is complicated by the fact that these opioids are not highly selective for a particular opioid receptor class. Moreover, these results contrast with the results of studies on the effect of morphine or highly selective exogenous agonists on cytokine expression. For example, studies reported by Lysle et al. (18) and Peterson et al. (45) show that IL-2 and IFN-γ production is inhibited following morphine administration. Chao et al. (46) demonstrated a significant increase in TGFβ production following morphine treatment of LPS- or PHA-activated PBMCs. It is well documented that TGFβ elicits many opposing cellular effects, depending on the cell type, maturation-differentiation status of the responding cell, and the local physiologic cellular environment (56). However, in cells of the immune system, TGFβ inhibits mitogen-induced synthesis of IFN-γ, IL-2, IL-3, GM-CSF, and TNF-α (57). The documented immunosuppressive activity of TGFβ may explain the inhibition of IL-2 and IFN-γ production following morphine administration. In contrast, recent reports suggest that TGFβ increases MCP-1 and IP-10 expression, but inhibits RANTES production by a variety of cell types (58, 59, 60, 61). Given the contrasting pro- and anti-inflammatory activities reported for the μ-opioids and TGFβ, we wished to examine the effect of a highly selective μ-opioid agonist on the expression of the chemokine family of critical proinflammatory cytokines. Our data support the notion that opioids are immunomodulatory and have the capacity to enhance the production of MCP-1, IP-10, and RANTES by cells of the immune system.
The correlation between drug abuse in general, and heroin abuse specifically, with HIV-1 infection is well established. Morphine, a major breakdown product of heroin, exhibits μ-opioid agonist activity. Studies by Peterson et al. (51) in 1990 showed morphine administration potentiated HIV-1 replication in human PBMCs. These findings suggested that opioids could act as cofactors in the pathogenesis of HIV-1 in i.v. drug users. However, we still do not fully understand the mechanism of immunomodulation by opioids during HIV-1 infection. Indeed, conflicting results have been reported for the effect of morphine on the progression of SIV infection in monkeys (62, 63). The disagreement in the results from these primate studies could be due to differences in the SIV strain and morphine dosages employed by these laboratories. Our data suggest a mechanism by which opioids could enhance HIV-1 replication through modulation of chemokine expression. On the one hand, recent experiments showed that CC chemokines RANTES, MIP-1α, and MIP-1β can act as HIV-1-suppressive molecules in CD4+ T cells by competitively binding HIV-1 coreceptors required for entry into target cells (49). In contrast, several studies show that chemokines may exhibit enhancing rather than inhibitory effects on HIV-1 replication (64, 65, 66). For example, Dolei et al. (50) determined that pretreatment of T cells with RANTES, MIP-1α, and MIP-1β increased the replication of T-tropic HIV-1 strains in a dose-dependent manner. These findings were associated with increased accumulation of CXCR4 transcripts and CC chemokine-induced PBL proliferation (50). Gordon et al. (67) demonstrated enhancement of both M- and T-tropic HIV-1 infection when RANTES was administered before or simultaneously with HIV-1 infection. The mechanism responsible for enhanced viral replication by RANTES may be related to chemokine-induced cellular activation (68, 69, 70). Recent studies have also shown that RANTES may increase attachment of HIV-1 to target cells via glycosaminoglycans and also activate a signal transduction pathway that enhances viral infectivity (71). Clearly, RANTES may exert a complex set of positive and negative influences on HIV-1 replication.
We also analyzed the effect of DAMGO on chemokine expression in HIV-1-infected cells. The results showed that DAMGO augmented expression of RANTES in a dose-dependent manner in both M- and T-tropic HIV-1-infected cells. The ability of DAMGO to induce HIV-1-infected cells to produce RANTES and IP-10 is significant given the activity of this chemokine to attract both T cells and macrophages, bringing potential target cells to the site of active HIV infection.
Our studies revealed an interesting dichotomy in the DAMGO-induced IP-10 response between M- and T-tropic HIV-1-infected cells. A significant increase in IP-10 expression was induced in T-tropic HIV-1-infected (nonactivated) cells. In contrast, DAMGO induced a 3-fold decrease in IP-10 production by M-tropic HIV-1-infected cells. Taub et al. (72) showed that IP-10 chemoattracts activated T lymphocytes and promotes T cell adhesion to endothelial cells. The elevated DAMGO-stimulated IP-10 production by T-tropic HIV-1-infected cells would be expected to promote the infection by attracting noninfected T cells to the site of infection. In contrast, the reduced expression of IP-10 by DAMGO-stimulated M-tropic HIV-1-infected cells would likely result in the attraction of greater numbers of monocytes. In this way, the DAMGO administration is likely to promote the spread of HIV to target cells which bear the ideal HIV-1 coreceptor.
MCP-1 plays a major role in two distinctly different host responses: cellular immune reactions and responses to acute tissue injury (73). MCP-1 can be produced by leukocytes of both lymphocyte and monocyte lineages and is specific for monocytes, macrophages, and activated T cells. This suggests that MCP-1 is a key effector cell mediator of delayed hypersensitivity, immune responses due to tissue injury, bacterial invasion, and viral infection (74). In addition to its function in monocytic infiltration to sites of injury of inflammation, MCP-1 has also been shown to play a role in monocyte recruitment into the CNS during HIV-1 infection mediated by Tat (74, 75, 76). However, in addition to CD4+ T cells, monocytes have been identified as early targets for HIV-1 infection, disseminating the virus to different organs, including the lungs, skin, and lymph nodes, as well as the brain (77). In this study, we find that DAMGO increases the expression of MCP-1 by PBMCs. Studies on the response of astrocytes and microglial cells to DAMGO would be particularly significant, since these cells are a major source of MCP-1, in the brain, and the brain is a natural site for significant levels of endogenous μ-opioids (76).
This work was supported by National Institute on Drug Abuse Grants DA-11130, DA-06650, DA-12113, T32 DA-07237, and F31 DA-05894.
Abbreviations used in this paper: DAMGO, [d-Ala2,N-Me-Phe4,Gly-ol5]enkephalin; IP-10, inducible protein-10; Ltn, lymphotactin; MCP-1, monocyte chemoattractant protein-1; MIP-1, macrophage-inflammatory protein-1; MOI, multiplicity of infection; CTAP, H-d-Phe-Cys-Tyr-d-Trp-Arg-Thr-Pen-Thr-NH2.