INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Alcohol suppresses many components of the immune response, thereby increasing the morbidity and mortality of a wide spectrum of bacterial infections, particularly pneumonia (1, 2). Similarly, human immunodeficiency virus (HIV)-infected individuals develop a number of secondary infections which are often caused by bacterial pathogens. As with alcohol abuse, a major site of these infectious episodes is the lung (3). Several studies have shown that a large number of HIV-infected individuals, as well as those at a high risk for HIV infection, abuse alcohol (4, 5). Indeed, as many as 41% of HIV-infected individuals meet the criteria for classification as alcoholics (4). Because excess alcohol consumption often coexists with HIV disease, it is important to understand the interaction of these two immunocompromising conditions on host defense. Presently, there are limited data and no consensus concerning the effects of excessive alcohol consumption on susceptibility to HIV infection and/or disease progression following primary infection (6).

Tumor necrosis factor-alpha (TNF-) functions as an important proximal mediator (or "alarm" cytokine) in the proinflammatory cascade that initiates and reinforces the inflammatory response of the host to invading pathogens (10). Whereas the systemic release of TNF- may have deleterious effects, numerous studies have demonstrated that in vivo neutralization of TNF- impairs the clearance of a variety of microorganisms in several animal models of infection, including Pneumocystis carinii (11), Mycobacterium tuberculosis (12), Streptococcus pneumoniae (13), and Listeria monocytogenes (14). Acute alcohol intoxication has been shown to suppress TNF- production and neutrophil recruitment into the lung in response to both lipopolysaccharide (LPS) and bacteria, resulting in a greater lung bacterial burden in intoxicated animals (15, 16).

If alcohol abuse in the setting of HIV/acquired immunodeficiency syndrome (AIDS) results in an impaired pulmonary TNF- response, possible sequelae include a greater number of and/or more severe secondary infectious complications. Thus, using simian immunodeficiency virus (SIV)-infected rhesus macaques as a model of HIV/AIDS, the aims of this study were to determine: (1) the effects of SIV infection on alveolar macrophage (AM) TNF- production at the asymptomatic and terminal stages of infection, and (2) the effects of in vitro alcohol exposure on the AM TNF- response in samples from both uninfected and SIV-infected animals, at the asymptomatic and terminal stages of infection. This animal model was chosen because it most closely resembles human HIV infection (17) and allows for a well controlled investigation to be performed. The present study was performed using in vitro alcohol exposure of AMs, before conducting longitudinal studies examining the in vivo effects of acute and chronic alcohol intoxication on these responses.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animals

Fifteen rhesus monkeys (Macaca mulatta) of either sex born and maintained at the Tulane Regional Primate Research Center (TRPRC) (Covington, LA) were used in this study. Seven animals were inoculated with SIV and eight served as uninfected controls. All animals were verified to be free of both simian T-cell lymphotropic virus-I and simian type D retrovirus prior to study enrollment. The animals were housed individually in a Biosafety Level 2 (BSL-2) containment building. Monkeys were fed a commercial primate chow, supplemented with fruit, and provided water ad libitum. Physical examination, SIV inoculation, and blood withdrawal were performed on ketamine HCl (10 mg/kg, intramuscularly; Ketaset; Fort Dodge Animal Health, Fort Dodge, IA) anesthetized animals after a 12-h fast.

For SIV infection, a 1-ml inoculum of SIV_DeltaB670 containing approximately 10,000 ID₅₀ (50% infective dose) of virus was used to infect seven of the 15 animals. The inoculum consisted of 1 ml of filtered cell culture supernatant from human phytohemagglutinin-stimulated lymphoblast cells infected with SIV_DeltaB670. Animals were inoculated with SIV_DeltaB670 via the saphenous vein. All animals were monitored twice daily for overall appearance, physical activity, stool consistency, food consumption, and hydration. Physical examinations were performed on the animals weekly for the first month after SIV inoculation and monthly thereafter. All control animals received periodic physical examinations.

To study AM TNF- production, bronchoalveolar lavage fluid (BALF) samples were obtained from uninfected and SIV-infected macaques. The SIV-infected animals used in this study originally constituted an unvaccinated, virus-challenged control group in a vaccine study and were subsequently transferred to this project. At the time of transfer, all animals were past the acute phase of their SIV disease, but none had progressed to AIDS. For SIV-infected animals, samples were obtained at both the asymptomatic and terminal stage of SIV infection. The asymptomatic stage was defined as the period of infection following the acute stage in which the animals had no clinical signs of secondary infection and appeared healthy. The terminal stage was defined as that point in time during the AIDS phase of the disease during which the animal demonstrated signs of secondary infection and was no longer responsive to supportive therapy and care. Thus, these samples were obtained immediately prior to killing. Euthanasia of SIV-infected macaques was accomplished by an intravenous barbiturate overdose (100 mg/kg, intravenously, Beuthanasia D; Schering-Plough, Omaha, NE).

All animal experiments performed were approved by the Animal Care and Use Committee of the Louisiana State University Medical Center and the TRPRC.

Hematology and Lymphocyte Subset Analysis

Complete blood counts and differentials were performed in the Clinical Laboratory at the TRPRC using a Coulter counter for total leukocyte counts and Wright-Giemsa staining of blood smears for leukocyte differentials. Blood lymphocyte subsets were determined, as previously described (18), using fluorochrome-conjugated monoclonal antibodies against human phenotypic cell surface antigens and analyzed on a Becton Dickinson FACSCalibur flow cytometer. The following antibodies were used: phycoerythrin (PE)-T11 for total CD2⁺ T lymphocytes (mouse IgG₁, clone T11; Coulter, Hialeah, FL), fluorescein isothiocyanate (FITC)-Leu-2a for CD8⁺ cytotoxic T lymphocytes (mouse IgG₁, clone Leu-2a; Becton Dickinson, San Jose, CA), and FITC-OKT4 for CD4⁺ helper T lymphocytes (mouse IgG_2b, clone OKT4; Ortho Diagnostics, Raritan, NJ).

Bronchoalveolar Lavage

For bronchoalveolar lavage (BAL), animals were preanesthetized with glycopyrrolate (0.05 mg/kg, intramuscularly, Robinul-V; Fort Dodge Animal Health, Fort Dodge, IA) and anesthetized with tiletamine HCl/ zolazepam (5 mg/kg of each, intramuscularly, Telazol; Fort Dodge Animal Health). All blood collections were performed while the animal was anethetized but before BAL. A pediatric bronchofiberscope (BF Type-3C20; Olympus, Lake Success, NY) was inserted via the trachea and into a bronchus of the right lower lobe of the lung until the endoscope was in a wedge position. Lavage was performed by instilling, through the bronchoscope, 30 ml of warm phosphate-buffered saline (PBS) (GIBCO BRL, Gaithersburg, MD) containing 0.1% dextrose (Sigma Chemical Co., St. Louis, MO), followed by gentle aspiration and collection of the fluid with a syringe. This procedure was repeated four times for a total instilled volume of 120 ml. Using this procedure, the average percentage of BALF recovered in uninfected, asymptomatic SIV-infected animals, and terminal SIV-infected animals was 85.9 ± 1.3%, 83.4 ± 2.9%, and 75.3 ± 6.6% of the instilled volume, respectively. After BAL, the animal was maintained on 100% oxygen for 5 to 10 min, returned to its cage, and then closely monitored for the next 4 h.

In Vitro Culture of AMs

Contaminating red blood cells were removed from the recovered BALF cells using E-LYSE (Cardinal Associates, Santa Fe, NM) according to the manufacturer's protocol. Subsequently, the cells were washed twice in RPMI-1640 media (GIBCO BRL). The total number of cells obtained was quantified using a hemacytometer and morphologic differentiation of cells was performed on a cytospin preparation stained with Diff-Quik Stain Kit (Baxter, Miami, FL). Recovered BALF cells were resuspended in RPMI-1640 media supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 5 mM glutamine (GIBCO BRL), and 10% fetal calf serum (FCS) (Hyclone Laboratories, Inc., Logan, UT) and aliquoted at 50,000 AMs per well in a 96-well flat-bottom tissue culture plate (Corning Costar Corp., Cambridge, MA). After 30 min at 37° C in 5% CO₂, the wells were washed twice to remove the nonadherent cells (primarily lymphocytes and neutrophils). Subsequently, media alone or media containing ethanol (McCormick Distilling Co., Inc., Weston, MO) was added to the wells to achieve a final ethanol concentration of 0, 25, or 100 mM in 200 µl of media. Four hours later, LPS (Escherichia coli 0111:B4; List Biological Laboratories, Inc., Campbell, CA) was added to the wells to achieve a final concentration of 0, 0.01, or 10 µg/ml. The cells were cultured for an additional 15 h and then the supernatant harvested and stored at 80° C for later analysis.

In experiments designed to examine the effects of ethanol on the LPS-induced TNF- messenger RNA (mRNA) concentrations, the ethanol and LPS incubation periods were changed to optimize detection of TNF- mRNA. AMs were isolated and cultured as previously described. Subsequently, ethanol was added to achieve a final concentration of 0 or 100 mM. Thirty minutes later, LPS was added to achieve a final concentration of 0 or 0.1 µg/ml and the incubation was continued for either 0, 0.5, 2, or 8 h after LPS addition.

Measurement of TNF- Protein

TNF- protein levels in supernatant samples were determined, according to the manufacturer's protocol, with Genzyme's (Cambridge, MA) enzyme-linked immunosorbent assay (ELISA) for human TNF- known to cross-react with rhesus macaque TNF-.

Assay of TNF- mRNA

At 0, 0.5, 2, and 8 h after LPS addition to AM cultures, the supernatants were harvested to quantitate TNF- protein levels, while RNA was obtained from cells to measure TNF- mRNA by complementary DNA (cDNA)-equalized reverse transcription-polymerase chain reaction (cERT-PCR) (19, 20). Total RNA was isolated and combined from duplicate wells using the TRIZOL RNA isolation reagent (GIBCO BRL) according to the manufacturer's protocol. Reverse transcription (60 min at 37° C) was performed in 40-µl reactions in the presence of 1 mM deoxyadenosine triphosphate-deoxyguanosine triphosphate- deoxythymidine triphosphate (dATP-dGTP-dTTP) (Pharmacia Biotech Inc., Piscataway, NJ), 0.03 mM deoxycytidine triphosphate (dCTP) (Pharmacia Biotech), 0.125 mg/ml random hexamer primer (GIBCO BRL), 1 mM dithiothreitol (GIBCO BRL), 10 U/µl Moloney murine leukemia virus reverse transcriptase (GIBCO BRL), 1× GeneAmp PCR Buffer II (in mM: 10 Tris-HCl, 50 KCl, and 2.5 MgCl₂, pH 8.3; Perkin-Elmer, Branchburg, NJ), 0.5 U/µl rRNasin ribonuclease inhibitor (Promega Corp., Madison, WI), and 1 µCi/reaction [-³²P]dCTP (NEN Research Products, Boston, MA). The resulting cDNA was quantitated by electrophoresing an aliquot of the reaction on a denaturing polyacrylamide gel and scanning the gel on a PhosphorImager (Molecular Dynamics, Mountain View, CA). Image Quant (Molecular Dynamics) analysis software was used to determine the resulting mass of cDNA.

After this quantitation, an equal amount of cDNA from each sample was analyzed and quantitated, using competitive PCR (19, 20). Primers and the competitive template were both designed for human TNF- (CLONTECH Laboratories, Inc., Palo Alto, CA) and performed well with rhesus macaque samples owing to the close TNF- sequence homology. The TNF- primer sequences were: TNF-A: 5'-GAG TGA CAA GCC TGT AGC CCA TGT TGT AGC A-3', and TNF-B: 5'-GCA ATG ATC CCA AAG TAG ACC TGC CCA GAC T-3' and amplified a 444-bp fragment. The TNF- competitor DNA template amplified a 617-bp product. PCR was performed in 50-µl reactions using 0.25 ng of cDNA with a final concentration of: 1× GeneAmp PCR Buffer II, 0.2 mM deoxyribonucleoside triphosphates (dNTPs), 0.2 µM each of TNF- primer A and B, 0.01 U/ml AmpliTaq DNA Polymerase (Perkin-Elmer), 0.02 attomoles TNF- competitor template, and 1 µCi/reaction [-³²P]dCTP. The PCR reaction was conducted in a PTC-100 Programmable Thermal Controller (MJ Research, Inc., Watertown, MA) for 30 cycles of amplification (1 min at 94° C, 1 min at 60° C, and 1 min at 72° C). A final extension step was performed for 7 min at 72° C. An aliquot of the PCR reaction was separated on a denaturing polyacrylamide gel, exposed overnight to a storage Phosphor screen, and scanned on a PhosphorImager. Analysis of the PCR products was performed by separately integrating the TNF- and competitor product counts. The ratio of TNF- product to competitor product and the concentration of competitor were used to quantitate TNF- mRNA levels between treatment groups. We have previously shown that this method is quantitative over four logs of TNF- mRNA concentrations (19).

Statistical Analysis

Statistical analysis was performed with a commercially available statistical software program (SigmaStat, SPSS, Inc., San Rafael, CA). All data are presented as mean ± SEM. One-way analysis of variance (ANOVA) was used to determine differences in the responses observed between samples obtained from SIV-negative animals, asymptomatic SIV-positive animals, and terminal SIV-positive animals. Statistically significant differences between in vitro treatment groups were determined using a one-way repeated measures ANOVA (one-way RM ANOVA). This allowed for individual differences among animals to be accounted for, thus removing animal variability from the analysis. After ANOVA, the Student-Newman-Keuls test was performed to determine group differences. Statistical significance was set at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Clinical Findings

Several studies were conducted simultaneously using this cohort of animals during the course of SIV infection. Results from the first study have been published and the clinical course of these animals has been previously described (21). Briefly, BAL was performed on asymptomatic SIV-infected macaques at a mean of 237 ± 53 d postinoculation and on terminal SIV-infected macaques at a mean of 463 ± 92 d postinoculation. At the time of BAL, blood CD4⁺ lymphocyte counts and the CD4⁺/CD8⁺ ratio were 1,388 ± 177 cells/mm³ and 0.95 ± 0.07, respectively, in uninfected animals; 760 ± 148 cells/mm³ and 0.47 ± 0.04, respectively, in asymptomatic SIV-positive animals; and 205 ± 28 cells/ mm³ and 0.42 ± 0.09, respectively, in terminal SIV-positive animals. Among the three groups examined, the CD4⁺ lymphocyte counts were statistically different (p < 0.05), whereas the CD4⁺/ CD8⁺ ratio was statistically different in SIV-negative animals versus SIV-positive animals. Several opportunistic pathogens were observed in animals at the terminal stage of SIV infection, but the presence and/or the type of pathogen involved did not correlate with observed changes in AM TNF- production.

BAL Cell Recovery

The total number of cells recovered by BAL was 8.7 ± 1.0 × 10⁶ in SIV-negative animals, 41.9 ± 17.6 × 10⁶ in SIV-positive animals at the asymptomatic stage of infection, and 8.0 ± 2.9 × 10⁶ in SIV-positive animals at the terminal stage of disease. Although no statistically significant differences existed between the groups, BAL cell counts tended to be greater in asymptomatic SIV-positive animals compared with the other groups examined. Compared with the percentage of AMs recovered in the BALF from SIV-negative animals (92.3 ± 1.2%), SIV-positive animals had a decreased percentage of AMs at both the asymptomatic (66.7 ± 5.9%, p < 0.05) and terminal stage of SIV infection (69.2 ± 5.1%, p < 0.05). A greater percentage of lymphocytes was present in BALF samples from SIV-positive animals at both the asymptomatic (22.9% ± 3.6%) and terminal stage (15.2 ± 3.8%), compared with those in SIV-negative animals (6.3 ± 0.9%, p < 0.05). The percentage of neutrophils present in the BALF of SIV-negative animals (1.3 ± 0.3%) and at the terminal stage of SIV infection (15.6 ± 7.3%) did not differ statistically from that in asymptomatic SIV-positive animals (10.3 ± 3.1%). A statistically significant difference (p < 0.05) was observed in BALF percentage of neutrophils in SIV-negative animals versus SIV-positive animals at the terminal stage of infection.

Effect of In Vivo SIV Infection on AM TNF- Production

There was no difference in spontaneous TNF- production by AMs collected from SIV-negative animals (109 ± 44 pg/ml), and SIV-positive animals at the asymptomatic (65 ± 51 pg/ml) or terminal stages of infection (23 ± 8 pg/ml). LPS, at a concentration of 0.01 and 10 µg/ml, stimulated TNF- production by AMs from both SIV-negative and -positive animals (Figure 1). AMs from SIV-negative animals demonstrated a greater production of TNF- in the presence of 10 µg/ml LPS compared with 0.01 µg/ ml LPS (p < 0.05). However, no consistent LPS dose-response effect was observed in samples from SIV-positive animals. Interestingly, LPS-induced (0.01 and 10 µg/ml) TNF- production was significantly lower in AMs obtained at the asymptomatic stage of SIV infection compared with cells from uninfected animals (Figure 1, p < 0.05). At the terminal stage of disease, TNF- production in response to LPS returned to normal and did not differ from samples obtained from uninfected animals (Figure 1).

View larger version (23K): [in this window] [in a new window]   Figure 1.   LPS-induced (0.01 or 10 µg/ml) TNF- production by AMs from uninfected and SIV-infected macaques, at the asymptomatic and terminal stages of infection. Values are mean ± SEM; n = 5-8; *p < 0.05 compared with 0.01 µg/ml LPS in the same animal group. Bars with different letters are statistically different (p < 0.05 by one-way ANOVA).

In Vitro Ethanol Suppresses AM TNF- Production

Within specific experiments, suppression of the AM LPS- induced TNF- response was observed after exposure to 25 mM ethanol. However, statistical analysis of all 25-mM ethanol experimental data did not reveal a statistically significant suppressive effect, with either dose of LPS used or in any of the groups of animals examined (data not shown). In contrast, treatment of AMs with 100 mM ethanol suppressed the LPS-induced TNF- response (p < 0.05, Figure 2). A similar degree of ethanol-induced suppression was demonstrated using either 0.01 or 10 µg/ml LPS. Finally, as reflected in Figure 2, the suppressive effect of ethanol was observed in AMs from both SIV-negative and -positive animals, regardless of the stage of infection examined. Because the AM TNF- response in SIV-positive asymptomatic animals was already suppressed, further suppression by ethanol resulted in a markedly diminished response to LPS compared with AMs from uninfected animals that were not exposed to in vitro ethanol.

View larger version (26K): [in this window] [in a new window]   Figure 2.   Suppression by ethanol of the LPS-induced (0.01 or 10 µg/ml) TNF- response by AMs from uninfected and SIV-infected macaques, at the asymptomatic and terminal stages of infection. Values are mean ± SEM; n = 5-8.

Time Course of Ethanol-induced Suppression and Effects on TNF- mRNA Levels

Conflicting findings have been reported in the literature as to the effects of ethanol on LPS-induced TNF- mRNA transcription (22, 23). To further examine this issue, TNF- protein and mRNA production by AMs from uninfected control animals were measured after in vitro exposure to ethanol and LPS. Whereas no TNF- protein was measured in supernatant samples at 0.5 h after LPS addition, at 2 h TNF- protein levels were significantly elevated and a further increase was observed at the 8-h time point (Figure 3). Ethanol (100 mM) exposure of AMs for as short as 30 min before LPS addition suppressed AM TNF- protein production by 84% at 2 h and by 70% at 8 h after LPS addition (Figure 3). In contrast, similar concentrations of TNF- mRNA were induced by LPS in the absence and presence of 100 mM ethanol at 0.5, 2, and 8 h after LPS addition (Figure 3). These data indicate that the ethanol-induced suppression of the TNF- response may result from as brief as a 30-min exposure to ethanol, occurs soon after the addition of LPS, and does not result from decreased TNF- transcript production.

View larger version (34K): [in this window] [in a new window]   Figure 3.   Determination of ethanol's effects on LPS-induced TNF- protein and mRNA production by macaque AMs. Cells were incubated with ethanol (0 or 100 mM) for 30 min followed by stimulation with LPS (0 or 0.01 µg/ml). Supernatants and cells were harvested for measurement of TNF- protein and mRNA concentrations, respectively, at 0.5, 2, and 8 h after LPS addition. (Top panel ) TNF- protein levels. (Middle panel ) A representative polyacrylamide gel of cERT-PCR of macaque AM TNF- mRNA (lanes are equivalent to the groups shown directly below in the graph). (Bottom panel ) TNF- mRNA levels. COMP = competitor template; values are mean ± SEM; n = 4; Within a time point, bars with different letters are statistically different (p < 0.05 by one-way RM ANOVA).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Alcohol exposure suppresses TNF- production in several cell types (22) and animal models of infection (15, 16, 25), which suggests a possible mechanism underlying the observed increased susceptibility of these hosts to infection. Because alcohol abuse causes immunosuppression and frequently coexists with HIV infection (4), it may function as a potential cofactor in HIV disease progression. The aim of this study was to examine this interaction by investigating the effects of in vitro alcohol exposure on TNF- release by AMs from SIV-infected rhesus macaques. Ethanol at a concentration of 25 mM failed to consistently suppress the LPS-induced TNF- response by AMs, isolated from uninfected or SIV-infected animals at both stages of infection examined, but incubation with 100 mM ethanol for as short as 30 min suppressed this response. Because of the vital role of TNF- in generating an effective immune response to many of the secondary pathogens observed in HIV-infected individuals including M. tuberculosis (12), P. carinii (11), S. pneumoniae (13), and Mycobacterium avium-intracellulare complex (26), these findings suggest that excessive alcohol consumption by HIV-positive individuals may further compromise host defense and enhance susceptibility to secondary infections.

The most significant effect of ethanol on the AM TNF- response was observed with an in vitro ethanol concentration of 100 mM. This corresponds to a blood alcohol level of approximately 0.46%. These experimental observations are clinically significant because similar blood alcohol levels are obtained under in vivo conditions and are observed in chronic alcohol abusers after an episode of acute alcohol intoxication. Furthermore, ethanol concentrations below 100 mM would also be expected to suppress the TNF- response because some suppressive effects were observed with as little as 25 mM ethanol.

If in vivo alcohol intoxication were to similarly suppress the AM TNF- response in HIV/AIDS patients, host resistance to secondary pathogens might be depressed and thereby increase infectious complications associated with HIV infection. This secondary infection may now enhance HIV replication. Several recent studies have reported that plasma HIV viral load increases after the development of secondary infections in HIV-infected individuals. Goletti and coworkers (27) measured greater levels of plasma HIV RNA during M. tuberculosis infection in the lung compared with values obtained both before and after mycobacterial infection. In another study, plasma HIV RNA levels were measured in HIV-infected patients before, during, and after the development of bacterial pneumonia. Before the onset of pneumonia, the median plasma HIV RNA concentration was 60,000 copies per ml and rose to a median value of 245,000 copies per ml during the clinical course of pneumonia, with the level decreasing toward normal upon resolution of the secondary infection (28). Alcohol abuse could similarly increase secondary infectious complications in HIV-infected individuals by compromising the TNF- response. Unresolved, secondary infections would then ensue, ultimately driving HIV replication and possibly accelerating HIV disease progression.

The role of TNF- mRNA expression in the ethanol-induced suppression of the TNF- response by AMs was also investigated in this study using samples from uninfected control animals. In support of previously published findings, LPS induced a similar rise in TNF- mRNA levels both in the absence and presence of ethanol (100 mM) as determined by cERT-PCR. Using a rodent model, both Kolls and coworkers (23) and Xie and coworkers (29) have reported that after intratracheal injection of LPS, BALF TNF- protein levels are markedly diminished by acute alcohol intoxication, while TNF- mRNA levels are similar in AMs from control and intoxicated animals. In addition, the concentration of the 26-kD membrane bound form of TNF- is higher on AMs obtained from the intoxicated rats, suggesting that the ethanol-induced decrease in TNF- production is occurring at a post-transcriptional level (23). Szabo and coworkers (22), using human monocytes, have also reported that ethanol decreases stimulated TNF- protein production. In contrast to the study of Kolls and coworkers (23) and the present findings, they observed that decreased TNF- mRNA levels accompanied the decreased TNF- protein production, indicating that under their experimental conditions the ethanol-induced suppression of TNF- release resulted from decreased mRNA transcription or decreased TNF- mRNA half-life. Possible explanations for these discrepant findings in the ability of ethanol to alter TNF- mRNA expression include differences in the method used for TNF- mRNA analysis (RT-PCR versus Northern blot analysis) and the model systems used in the experiments (rodent and nonhuman primate AMs versus human peripheral blood monocytes). In this study, cERT-PCR showed no differences in LPS-induced TNF- transcripts in the presence of ethanol, indicating that ethanol's effects are occurring at a post-transcriptional level in this experimental model.

Despite the fact that the lung is the most common site of secondary infections in HIV infection (3), only a limited number of studies have been conducted to determine AM TNF- production during HIV infection. Many of the studies report both enhanced constitutive and LPS-stimulated TNF- mRNA and protein production by AMs obtained during HIV infection (30, 31). However, Cox and coworkers (32) observed decreased TNF- production by AMs from patients with AIDS.

To the best of our knowledge, there is only a single study, by Horvath and coworkers (33), examining AM TNF- production in the SIV model. These investigators found no constitutive secretion and normal LPS-induced production of TNF- after in vitro infection of AMs with SIV. In contrast to our findings they reported that AMs from SIV-infected macaques at the asymptomatic stage of disease progression produce greater amounts of TNF- in response to LPS compared with cells obtained from uninfected animals or during AIDS. The in vitro experimental protocol used in their study to assess AM TNF- production was quite comparable to the present study, but their study did differ in a number of ways. A different strain of SIV, SIV_mac, was used for their in vivo infection of macaques. Although no information was presented on the clinical history of the macaques used in the study, the times postinoculation when the experiments were performed were somewhat different from those of the current study. AMs were obtained from two of the three asymptomatic animals over a year following infection and from two of the five AIDS animals at 1,088 and 1,477 d postinfection. These numbers indicate that the clinical course of infection induced by their viral strain tended to be more protracted compared with that of SIV_DeltaB670, the strain used in the present study, in which the median survival time is 207 d postinfection. Thus, one explanation for the observed differences includes a different strain of SIV for infection and the clinical disease induced by the virus. A strength of the present study is that the response of samples from a greater number of animals was examined at the early asymptomatic stage and the fact that the animals were serially sampled so measurements at the two stages of disease were made in the same animals.

The mechanism responsible for decreased TNF- release by AMs isolated at the asymptomatic stage of disease was not examined in this study. Possible mechanisms include the presence of a secondary infection which induced tolerance of the AMs to a subsequent LPS challenge (34). However, no secondary infections were clinically evident when BAL was performed on these animals at this stage of SIV infection. Alternatively, while similar percentages of neutrophils and lymphocytes were recovered at the asymptomatic and terminal stages of SIV infection, no determination of these cells' phenotype, functional state, or degree of activation was made. Additionally, the percentage of cells in the lung, specifically lymphocytes and AMs, infected with SIV at the time of BAL was unknown. Differences in any of these parameters between stages of infection and their in vivo consequences on AM function could account for the diminished LPS-induced TNF- response by AMs observed at the asymptomatic stage of infection. Although in vitro studies have failed to show a direct effect of HIV/SIV on monocyte/ macrophage TNF- release (33, 35), the in vivo consequences of SIV exposure or infection on AM functional responses may be quite different, and responsible for the observed changes in TNF- production during the asymptomatic period.

In summary, LPS-induced AM TNF- production is suppressed during the asymptomatic stage of SIV infection and ethanol suppresses TNF- production by AMs obtained from both SIV-negative and -positive rhesus macaques, regardless of the stage of SIV examined. If similar events occur during in vivo alcohol intoxication in SIV-infected macaques or HIV- infected humans, an increased number of and/or severity of secondary infections may result, leading to increased viral replication and disease progression.