HIV-1 infection of human PBMC has been shown to elicit secretion of several different cytokines. TNF-α secretion induced by this virus has been of particular interest because it has been associated with the development of HIV-1 dementia and because TNF-α increases viral replication by enhancing NF-κB interaction with the viral promoter, the HIV-1 long terminal repeat. Thus, an autocrine pathway is potentially created in which HIV-1 stimulates its own replication. Conflicting reports exist, however, on the ability of HIV-1 to induce TNF-α secretion in vitro or in vivo. Using experimental protocols that controlled for potential bacterial endotoxin-induced TNF-α secretion, the current study demonstrates significant differences in TNF-α-eliciting properties among primary and laboratory obtained HIV-1. The relative TNF-α-inducing ability of different variants is conserved when tested using PBMC from different individuals. Elicitation of TNF-α secretion was not blocked by exposure of cells to zidovudine, indicating that viral integration was not required to induce secretion. Rather, the interaction between the virus and cell surface is critical for TNF-α induction, as Abs against CD4 or CCR5 blocked the induction of TNF-α synthesis by PBMC when added before virus exposure. Furthermore, the ability to induce TNF-α secretion mapped to a region of the HIV-1 env gene that includes the third hypervariable domain. Differences in the ability of different HIV-1 variants to elicit TNF-α may account for individual differences in HIV-1 disease course.

Cytokines produced by the host in response to ongoing HIV-1 infection serve as important mediators of virus-host interactions. Which cytokines are produced can influence the antiviral efficacy of the host immune response (1, 2, 3, 4, 5, 6, 7), the rate and/or extent of viral replication (8, 9, 10, 11, 12), and the phenotype of the virus that preferentially survives in the host environment (13, 14, 15, 16, 17). Given the demonstrated impact of proinflammatory cytokines on HIV-1 replication in vitro, differences among viral variants in their ability to elicit secretion of different cytokines could significantly influence the course of HIV-1 infection in vivo.

Among cytokines elicited by HIV-1, TNF-α, in particular, has assumed a prominent role in the effort to understand host-virus interactions (1, 3, 8, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29). Binding of TNF-α to one of its two receptors triggers a phospholipase activation pathway that promotes the dissociation in the cytosol of the nuclear transcription factor NF-κB from its inhibitor, I-κB (inhibitory protein that dissociates from NF-κB), thereby permitting NF-κB to move to the nucleus of the cell. In the nucleus, NF-κB is capable of binding to the HIV-1 long terminal repeat, thereby enhancing viral transcription (8, 24, 30).

It has also been demonstrated, however, that HIV-1 infection can enhance secretion of TNF-α by T cells (31), B cells (32), and monocytes or macrophages (25, 33, 34, 35, 36). Thus, TNF-α can participate in an autocrine or paracrine pathway in which HIV-1 enhances its own replication. Because TNF-α secretion can be elicited from a variety of immune cells, it is likely that it is not elicited by HIV-1 binding to the same receptor on all of these cells. It has been demonstrated, for example, that cross-linking of CD4 on monocyte/macrophages results in TNF-α secretion (33), but such binding could not account for the high level of TNF-α mRNA expression observed in B cells (32).

Because of the variability in both HIV-1 proteins and the cellular proteins with which HIV-1 might interact, we hypothesized that different variants of HIV-1 might differ in their ability to elicit TNF-α secretion. We therefore explored the ability of different clinical and laboratory HIV-1 isolates to elicit TNF-α secretion from human PBMC.

Freshly obtained unstimulated PBMC from healthy HIV-seronegative volunteers were placed in a 96-well plate (Corning Glass, Cambridge, MA) at 4 × 105/well in 100 μl of RPMI 1640 tissue culture medium (Life Sciences, Grand Island, NY) containing 10% human AB serum. Different HIV-1 variants (500 pg of p24 Ag/well) or bacterial LPS (6.0 endotoxin units Salmonella typhimurium LPS; Sigma, St. Louis, MO) were added to the cultures in 100 μl of the same medium. Each virus was tested with and without 10 μg/ml bactericidal permeability-increasing protein (37, 38) (BPI;3 Xoma Pharmaceuticals, Berkeley, CA) present in the media to ensure that observed TNF-α production was not attributable to contaminating bacterial endotoxin (39, 40). Sixteen hours later, 100 μl of the cell-free culture supernatant was assayed by ELISA (PerSeptive Diagnostics, Framingham, MA) for the presence of TNF-α. Each test condition was assayed in either duplicate or triplicate.

PBMC from HIV-1-seropositive individuals were cocultured with PBMC from healthy seronegative donors that had been placed in RPMI 1640 medium containing 10% human AB serum and stimulated with PHA (5 μg/ml; Life Sciences) for 2 days. Cocultures were subsequently maintained in media supplemented with IL-2 (2 U/ml; Boehringer Mannheim, Indianapolis, IN). Growth of virus from the cultures was determined by assay for viral p24 Ag (DuPont/NEN, Cambridge, MA) 10 days after the initiation of cultures. Culture supernatant fluids and control supernatant fluids from cultures containing no virus (referred to in the figures as “media controls”) were centrifuged at 800 × g for 10 min to remove cellular debris and then centrifuged at 100,000 × g for 2 h over a 20% sucrose cushion. The sedimented material was resuspended in RPMI 1640 medium and passed through a 0.22-μm filter, tested for p24 concentration, and used at an appropriate dilution in the TNF-α assay.

The HIV-1 molecular clone p120 was originally derived from HIVNL4-3, but contains an envelope derived from the molecular clone HIVHXB2, which was inserted between the SalI and BamHI sites, corresponding to positions 5785–8474 based on the nucleotide sequence of the prototype HIV-1 clone HXB2. The molecular clone p125 is identical to p120, but contains an insert from HIVADA extending from positions 7039–7619 (based on HXB2), which includes the third hypervariable region of the virus envelope. This region is defined by two BglII sites at either end of the insert. The molecular clone p125MNenv contains the envelope gene from HIVMN, amplified by nested PCR reactions, inserted between the SalI and BamHI sites of p125, extending between bases 5785–8474 of the HIV-1 envelope (based on HXB2). The external primers used for this amplification were 5′-AAACTGACAGGATCCATGGAACAGCC-3′ and 5′-TAAGTCATTGGTCTAGAAGGTACCTGAGGT-3′. The internal primers used were 5′-CAGAATTGGGTGTCGACATAGCAGAATAGGCATTATTCGACAGAG-3′ and 5′-TTGCTAAGGATCCATGCACTAATCGACCGGATGTGTCTCTGTCTC-3′. The underlined portions of the primers contain the SalI and BamHI sites used for insertion into p125.

Conditions used for amplification for both first- and second-round PCR were 95°C for 2 min, followed by 10 cycles of 94°C for 15 s, 60°C for 30 s, and 72°C for 2 min, followed by 20 cycles of 94°C for 15 s, 60°C for 30 s, and 72°C for 2 min (including an additional 20 s for each cycle) and one final cycle at 72°C for 7 min.

To create the chimeric viruses, the amplified fragment from HIVMN was cloned into the cloning vector pGEM11 (Promega, Madison, WI) by ligating each amplified SalI-BamHI fragment with T4 DNA ligase (Life Sciences) to the plasmid digested with the corresponding restriction enzymes. The DNA was transformed into Escherichia coli DH5α cells (Invitrogen, Carlsbad, CA) and then purified using standard procedures (41). The pGEM11 construct was then digested with SalI and BamHI; p125 was also digested with the same enzymes independently. Upon agarose gel electrophoresis and purification), the complementary fragments were ligated and transformed as described above. Transfection of virus into COS-7 cells was performed using the DEAE-dextran method, as previously described (42), with virus amplified through replication on PHA and IL-2-stimulated PBMC, which were added to the transfected COS-7 cells.

PBMC were obtained by centrifugation of leukopheresed blood from healthy HIV-seronegative donors on Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) according to the manufacturer’s instructions. The RPMI 1640 medium, human AB serum, and all other culture supplements were purchased as minimal endotoxin reagents.

Human anti-CCR5 Abs were obtained from hybridoma 242.21.6 supernatant fluid and were used either without dilution or at 10-fold dilutions in RPMI 1640 medium with 10% human AB serum. Anti-CD4 Abs (RPA-T4; PharMingen, San Diego, CA) were used at a concentration of 20 μg/ml. Isotype controls for the respective Abs were used at 20 μg/ml (murine IgG and murine IgG1; PharMingen).

The significance of differences in TNF-α secretion between and among different groups was determined by a one-way ANOVA using the Stata statistical package (Stata, College Station, TX).

PBMC from three subjects were stimulated with low passage HIV-1 isolates, LPS, or tissue culture media, and TNF-α secretion was measured 16 h later (Fig. 1). Adherent cells, predominantly monocytes, secreted between 80 and 90% of the total TNF-α produced in response to LPS or viral stimuli (data not shown). Both LPS and virus 9 stimulated secretion of TNF-α, achieving concentrations of this cytokine that were significantly above those elicited by media alone (p < 0.001 for all subjects). None of the other viral isolates stimulated levels of TNF-α that were significantly greater than their respective media controls. The response of subject 1 to virus 11 compared with that subject’s response to tissue culture media approached, but did not achieve, significance (p = 0.054). Individual subjects differed in their TNF-α response to either LPS or virus 9, but the relative relationship of the responsiveness pattern to the different viruses was maintained among the subjects independent of the stimulus. This finding suggests that some subjects tend to secrete more TNF-α in response to an appropriate stimulus than others.

FIGURE 1.

TNF-α secretion by PBMC from three healthy HIV-seronegative individuals in response to a panel of HIV-1 primary clinical isolates (500 pg of p24 of virus/200 μl of culture medium) and to LPS (6 endotoxin units/200 μl of culture medium). A companion culture for each virus or for LPS was established in which 10 μg/ml of BPI was added (see Fig. 2). BPI did not reduce TNF-α secretion in the wells containing PBMC and primary HIV-1 isolates (see Fig. 2). TNF-α levels reflect the mean ± SD of three microwells for each experimental group. Media control wells contained PBMC and neither virus nor LPS.

FIGURE 1.

TNF-α secretion by PBMC from three healthy HIV-seronegative individuals in response to a panel of HIV-1 primary clinical isolates (500 pg of p24 of virus/200 μl of culture medium) and to LPS (6 endotoxin units/200 μl of culture medium). A companion culture for each virus or for LPS was established in which 10 μg/ml of BPI was added (see Fig. 2). BPI did not reduce TNF-α secretion in the wells containing PBMC and primary HIV-1 isolates (see Fig. 2). TNF-α levels reflect the mean ± SD of three microwells for each experimental group. Media control wells contained PBMC and neither virus nor LPS.

Close modal

The virus 9 specimen used in this experiment was a low passage isolate. When passaged multiple times in tissue culture, it subsequently lost its ability to elicit TNF-α secretion, as did many of the primary isolates that initially elicited TNF-α secretion (data not shown).

The ability of LPS to stimulate TNF-α secretion is well established (39, 40) and endotoxin contamination must be considered when evaluating the ability of any agent to elicit TNF-α secretion. Therefore, the effect of BPI, an inhibitor of endotoxin activity, on the response to LPS and to all viral isolates was evaluated in each experiment and is shown for subject 2 from Fig. 1 (Fig. 2). As shown in Fig. 1, BPI effectively abrogated the ability of LPS to elicit TNF-α secretion (p < 0.001). It had, however, no significant effect on the ability of virus 9 to elicit TNF-α secretion (p = 1.0) from subject 2, indicating that contaminating LPS does not account for the ability of virus 9 to elicit this cytokine. For all subjects, the LPS response was significantly reduced by treatment with BPI (p < 0.001 for LPS response compared with LPS plus BPI response for all subjects, shown only for subject 2). However, none of the responses of the subjects to virus 9 was reduced by BPI treatment.

FIGURE 2.

TNF-α secretion with and without BPI by PBMC. BPI was added at the time that PBMC from subject 2 in Fig. 1 were exposed to virus or LPS. Each value represents the mean ± SD of three microwells.

FIGURE 2.

TNF-α secretion with and without BPI by PBMC. BPI was added at the time that PBMC from subject 2 in Fig. 1 were exposed to virus or LPS. Each value represents the mean ± SD of three microwells.

Close modal

Because of the demonstrated ability of HIV-1 to incorporate host cellular Ags into the virus particle during the budding process (43, 44, 45, 46, 47), the possibility was considered that the differential TNF-α response was attributable to reactivity against HLA or other host Ags which might stimulate a MLR. To evaluate this possibility, TNF-α responsiveness was assayed against virus 8, a primary clinical isolate, grown on either autologous or heterologous PBMC (Fig. 3). The mean response to virus grown on autologous cells was 3120 pg/ml of TNF-α compared with 2060 pg/ml of TNF-α produced in response to the same virus grown on heterologous cells. This difference was not statistically significant (p > 0.5).

FIGURE 3.

TNF-α response to virus 8 grown in autologous and heterologous PBMC. BPI was added at the time that PBMC were exposed to virus. Control wells contained PBMC and BPI, but no virus. Each value represents the mean ± SD of two or three microwells.

FIGURE 3.

TNF-α response to virus 8 grown in autologous and heterologous PBMC. BPI was added at the time that PBMC were exposed to virus. Control wells contained PBMC and BPI, but no virus. Each value represents the mean ± SD of two or three microwells.

Close modal

PBMC from a single subject were challenged with the same panel of viruses at two points separated by 5 wk (Fig. 4). The virus that elicited the greatest responses in the first experiment with the cells of this subject (virus 8) also did so in the second experiment. None of the responses to the same virus differed significantly (p > 0.05) at the two time points.

FIGURE 4.

TNF-α response of a single individual at two time points to a panel of HIV-1 isolates. Each group represents the mean ± SD of TNF-α concentrations obtained in two wells. Not shown are the results of companion cultures for each virus and for LPS, which contained 10 mg/ml of BPI. BPI did not significantly reduce the TNF-α secretion elicited by each virus but significantly reduced the LPS-induced TNF-α secretion.

FIGURE 4.

TNF-α response of a single individual at two time points to a panel of HIV-1 isolates. Each group represents the mean ± SD of TNF-α concentrations obtained in two wells. Not shown are the results of companion cultures for each virus and for LPS, which contained 10 mg/ml of BPI. BPI did not significantly reduce the TNF-α secretion elicited by each virus but significantly reduced the LPS-induced TNF-α secretion.

Close modal

To determine whether viral integration and/or protein synthesis were required for TNF-α secretion, TNF-α levels were measured in cultures to which a primary viral isolate and zidovudine (AZT, 50 μg/ml; Burroughs-Wellcome, Research Triangle Park, NC), an inhibitor of RT, were added. The TNF-α secreted in response to viral isolate 8203 was significantly greater than that secreted by PBMC alone (p = 0.008, Fig. 5). A dose of AZT which completely inhibited replication of this virus, as measured by p24 levels in culture supernatant fluid over a period of 2 wk (data not shown), had no significant effect on the ability of this virus to elicit TNF-α secretion (p = 1.0 when compared with virus alone, p = 0.03 when compared with cells and AZT alone). In the absence of virus, AZT also had no significant effect on TNF-α secretion (p = 1.0) compared with the level observed in cultures not exposed to virus.

FIGURE 5.

Effect of AZT on the ability of HIV-1 variants to elicit TNF-α secretion. AZT at a concentration of 50 μm was added at the time of culture initiation. In a parallel culture using the same PBMC source and virus, AZT at a concentration of 50 μm completely inhibited HIV-1 replication, as measured by the absence of increasing p24 concentrations over time (data not shown). Each group represents the mean ± SD of TNF-α concentrations obtained in two wells.

FIGURE 5.

Effect of AZT on the ability of HIV-1 variants to elicit TNF-α secretion. AZT at a concentration of 50 μm was added at the time of culture initiation. In a parallel culture using the same PBMC source and virus, AZT at a concentration of 50 μm completely inhibited HIV-1 replication, as measured by the absence of increasing p24 concentrations over time (data not shown). Each group represents the mean ± SD of TNF-α concentrations obtained in two wells.

Close modal

Because unintegrated HIV-1 was able to elicit TNF-α secretion, we hypothesized that secretion was stimulated by viral interactions at the cell surface. Binding of HIV-1 to the cell surface requires the interaction of the viral envelope with the host cell receptors CD4 and CCR5 or CXCR4 (48). To determine whether this binding event is critical for subsequent TNF-α secretion, PBMC from several subjects were cultured with anti-CD4 or anti-CCR5 Abs before the addition of p125, a CCR5-utilizing molecular clone of HIV-1 that routinely elicited high levels of TNF-α secretion. Both anti-CD4 and anti-CCR5 Abs significantly inhibited TNF-α induction by p125 (Fig. 6). Neither anti-CD4 nor anti-CCR5 Abs inhibited TNF-α synthesis induced by LPS. Isotype controls for the respective Abs also did not significantly alter TNF-α secretion.

FIGURE 6.

Anti-CD4 and anti-CCR5 Abs block TNF-α induction by HIV-1. Anti-CD4 (20 μg/ml) and anti-CCR5 Abs (undiluted or 10-fold dilution) and isotype controls (20 μg/ml) were added to human PBMC in cRPMI-10% human serum. Immediately following the addition of Abs, 500 pg of p24 virus was added for a total culture volume of 200 μl. A companion culture for each virus, Ab, or for LPS was established in which 10 mg/ml of BPI was added (data not shown). TNF-α levels reflect the mean ± SD of at least three microwells for each experimental group. Media control wells contained PBMC alone (no virus, Ab, or LPS).

FIGURE 6.

Anti-CD4 and anti-CCR5 Abs block TNF-α induction by HIV-1. Anti-CD4 (20 μg/ml) and anti-CCR5 Abs (undiluted or 10-fold dilution) and isotype controls (20 μg/ml) were added to human PBMC in cRPMI-10% human serum. Immediately following the addition of Abs, 500 pg of p24 virus was added for a total culture volume of 200 μl. A companion culture for each virus, Ab, or for LPS was established in which 10 mg/ml of BPI was added (data not shown). TNF-α levels reflect the mean ± SD of at least three microwells for each experimental group. Media control wells contained PBMC alone (no virus, Ab, or LPS).

Close modal

Since binding of HIV-1 with the coreceptors was necessary to elicit TNF-α from PBMC, we sought to further define the interaction of HIV-1 envelope with the cell surface that resulted in TNF-α secretion. We therefore examined the ability to elicit TNF-α secretion of chimeric viruses created by substituting parts of the envelope from different viruses into a molecular clone, p125. HIVMN failed to elicit TNF-α secretion (Fig. 7). A viral clone, p125MNenv, which contains the env region of HIVMN also failed to elicit TNF-α secretion. Clone p120, which is identical to p125, but contains the V3 loop region from the HIV-1 clone HXB2, also failed to elicit TNF-α secretion. Only p125 from this group of chimeric viruses was able to elicit TNF-α secretion at levels that were significantly above those observed with the PBL controls (p < 0.001). Levels of TNF-α secreted after cultivation with p125MNenv were not significantly different from those observed with HIVMN (p > 0.5).

FIGURE 7.

Mapping of TNF-α-eliciting ability of different viral variants. The env region from HIVMN, a poor elicitor of TNF-α, was inserted into the HIV-1 plasmid p125, a strong elicitor of TNF-α, and the resultant chimeric virus was evaluated for its ability to elicit TNF-α compared with the parental viruses. Virus clone p120 is identical to p125, but contains the third hypervariable region of the env from HIV-1 clone HXB2. In clone p125 that region is derived from clone HIVADA. BPI did not reduce TNF-α secretion in the wells containing PBMC and the HIV-1 isolates, but did significantly inhibit LPS-induced TNF-α secretion (data not shown). TNF-α levels reflect the mean ± SD of two wells for each experimental group. Media control wells contained PBMC, ultracentrifuged supernatant fluid from cultures of PHA- and IL-2-stimulated PBMC, and neither virus nor LPS.

FIGURE 7.

Mapping of TNF-α-eliciting ability of different viral variants. The env region from HIVMN, a poor elicitor of TNF-α, was inserted into the HIV-1 plasmid p125, a strong elicitor of TNF-α, and the resultant chimeric virus was evaluated for its ability to elicit TNF-α compared with the parental viruses. Virus clone p120 is identical to p125, but contains the third hypervariable region of the env from HIV-1 clone HXB2. In clone p125 that region is derived from clone HIVADA. BPI did not reduce TNF-α secretion in the wells containing PBMC and the HIV-1 isolates, but did significantly inhibit LPS-induced TNF-α secretion (data not shown). TNF-α levels reflect the mean ± SD of two wells for each experimental group. Media control wells contained PBMC, ultracentrifuged supernatant fluid from cultures of PHA- and IL-2-stimulated PBMC, and neither virus nor LPS.

Close modal

These studies demonstrate that different HIV-1 variants have different capabilities to elicit secretion of TNF-α. The ability to elicit TNF-α appears to be intrinsic to certain viral variants and is maintained when those viruses are exposed to cells from different normal individuals. The response pattern of a given individual to a panel of viral variants is generally conserved when observations are made at different time points. Although different individuals show variation in the amount of TNF-α they produce in response to the same virus, the relative stimulatory ability of different viral variants is maintained among different individuals.

Since secretion of TNF-α was observed even when viruses were exposed to PBMC in the presence of inhibitory concentrations of AZT, secretion was not dependent on viral integration.

In fact, the differences among viral variants in their ability to elicit TNF-α mapped in these studies to the viral envelope, specifically to the area around and including the V3 region. Previous studies eliciting TNF-α with purified HIV-1 envelope demonstrated that secretion could be elicited with truncated envelope proteins from many different regions within the viral envelope (49). In the current studies with intact virus, analysis of the ability of chimeric viruses to elicit TNF-α suggested that the region around the V3 loop of the envelope was a critical determinant of the TNF-α-eliciting ability. These studies do not exclude the possibility that other regions of the virus may also influence its ability to elicit TNF-α secretion.

The critical role of envelope in eliciting TNF-α secretion renders it likely that secretion results from interactions of virus and receptors on the cell surface. The current studies demonstrate that TNF-α induction is blocked by both anti-CD4 and anti-CCR5 Abs, which suggests that interactions with these receptors may influence the ability of different viral variants to elicit TNF-α secretion. That being the case, the induction of TNF-α by HIV-1 is not simply due to its interaction with CCR5 because not all viruses that bind CCR5 induce TNF-α. Previous studies have shown that binding of virus to CD4 results in TNF-α secretion (33). Because induction of TNF-α secretion in our studies mapped to the V3 envelope region, binding to the second HIV-1 receptor likely plays a role in this cytokine induction (13, 50), which it may do simply by altering the affinity with which the virus binds to CD4 (51, 52). Although large numbers of syncytia-inducing (SI) viruses were not studied in these experiments, no clinical or laboratory isolates known to contain only SI variants were able to elicit TNF-α secretion. Furthermore, the chimeric viruses p120 and p125MNenv that lost TNF-α-eliciting ability had been switched from nonsyncytia inducing (NSI) to the SI phenotype in creating the chimera. Apart from the envelope change, the viruses were identical to the parental p125.

The role of TNF-α in the pathogenesis of HIV-1 infection has been controversial. The early demonstration that TNF-α could promote expression of virus from T cell lines carrying integrated proviral DNA (8) first suggested that virally induced inflammatory processes could indirectly accelerate the disease course by promoting increased viral expression.

A central role of TNF-α in AIDS pathogenesis has been proposed by others as well (53, 54). Careful analysis of the relationship between viral load and TNF-α levels in asymptomatic HIV-1-infected individuals showed a direct relationship between plasma TNF-α levels and viral load, although, in the same study, levels of TNF-α mRNA in PBMC did not correlate with viral load. These studies suggested that TNF-α in serum originated from other sites within the body, such as lymph nodes, where virus was actively replicating.

Other studies have called into question the role that TNF-α plays in disease progression. Li et al. (55) found no difference in TNF-α mRNA levels in tonsilar tissue from five normal individuals and similar tissue obtained from seven asymptomatic HIV-1-infected individuals. Furthermore, no correlation was found between the level of HIV-1 gene expression in these tissues and the level of TNF-α gene expression. These authors concluded that TNF-α is not critical for HIV-1 replication in lymph nodes. Recently, Lane et al. (56) reported that nanogram concentrations of TNF-α suppressed HIV-1 replication by inducing the CCR5 agonist RANTES. Macrophage inflammatory protein-1α and macrophage inflammatory protein-1β were also induced by TNF-α, but did not inhibit HIV-1 replication.

Foli et al. (57) challenged the entire notion that HIV-1 is capable of eliciting TNF-α secretion. Although acknowledging that TNF-α levels are elevated in HIV-1-infected individuals, these investigators demonstrated that high concentrations of two HIV-1 strains, HIVBa-L and HIVLAI, did not elicit TNF-α secretion when care was taken to ensure that the virus preparations had minimal concentrations of contaminating endotoxin. These investigators concluded that the elevated levels of TNF-α observed in HIV-1-infected individuals was elicited by pathogens or processes other than HIV-1.

In the current studies, effects of potential endotoxin contamination were abrogated by inclusion of BPI, which inhibits LPS-induced TNF-α secretion (38). Most HIV-1 variants were incapable of eliciting TNF-α secretion. Neither HIVBa-L nor HIVIIIB, which is closely related to HIVLAI, could elicit TNF-α secretion (data not shown), as found by Foli et al. (57). However, several clinical isolates were capable of eliciting this cytokine, even in the presence of an LPS inhibitor.

The differences among HIV-1 strains in their ability to elicit TNF-α secretion introduce one more variable that might be associated with differences in disease course. These differences may be particularly relevant to understanding why some individuals with AIDS develop dementia whereas others with advanced disease do not. In a postmortem analysis of cytokine expression in the brains of demented and nondemented AIDS patients, Wesselingh et al. (58, 59) demonstrated significantly increased levels of TNF-α mRNA produced by cells of the macrophage lineage in demented patients compared with nondemented controls. With the exception of IL-4 mRNA levels, which were diminished in demented individuals, TNF-α was the only cytokine for which levels differed significantly from nondemented control subjects. Since the development of dementia does not correlate simply with viral load (60), explaining differences among infected individuals in the development of the dementia syndrome may require phenotypic characterization of the dementia-associated viruses, particularly in terms of their TNF-α-eliciting ability.

Whether the appearance of TNF-α-eliciting HIV-1 variants affects other aspects of disease pathogenesis, including the rate of disease progression, remains to be determined. The knowledge that TNF-α secretion is viral-variant dependent should permit a more systematic approach to this question, with efforts to correlate disease progression with the TNF-α secretion associated with the viral variants present at a given point in the disease course.

We thank Dr. Chris Karp for insightful comments on this manuscript.

1

These studies were supported by Grant DA09973 from the National Institute on Drug Abuse. K.V.K. supported by Grant AI07417 from the National Institute of Allergy and Infectious Diseases.

3

Abbreviations used in this paper: BPI, bactericidal permeability-increasing protein; AZT, zidovudine; SI, syncytia inducing.

1
Estcourt, C., Y. Rousseau, H. M. Sadeghi, N. Thieblemont, M. P. Carreno, L. Weiss, N. Haeffner-Cavaillon.
1997
. Flow-cytometric assessment of in vivo cytokine-producing monocytes in HIV-infected patients.
Clin. Immunol. Immunopathol.
83
:
60
2
Clerici, M., T. A. Wynn, J. A. Berzofsky, S. P. Blatt, C. W. Hendrix, A. Sher, R. L. Coffman, G. M. Shearer.
1994
. Role of interleukin-10 in T helper cell dysfunction in asymptomatic individuals infected with the human immunodeficiency virus.
J. Clin. Invest.
93
:
768
3
Barcellini, W., G. P. Rizzardi, G. Poli, G. Tambussi, C. Velati, P. L. Meroni, A. G. Dalgleish, A. Lazzarin.
1996
. Cytokines and soluble receptor changes in the transition from primary to early chronic HIV type 1 infection.
AIDS Res. Hum. Retroviruses
12
:
325
4
Clerici, M., C. Balotta, L. Meroni, E. Ferrario, C. Riva, D. Trabattoni, A. Ridolfo, M. Villa, G. M. Shearer, M. Moroni, M. Galli.
1996
. Type 1 cytokine production and low prevalence of viral isolation correlate with long-term nonprogression in HIV infection.
AIDS Res. Hum. Retroviruses
12
:
1053
5
Shearer, G. M., M. Clerici, A. Sarin, J. A. Berzofsky, P. A. Henkart.
1995
. Cytokines in immune regulation/pathogenesis in HIV infection.
Ciba Found. Symp.
195
:
142
6
Clerici, M., G. M. Shearer.
1994
. Cellular immunity and a type 1 cytokine profile in protection against HIV infection and progression to AIDS.
Res. Immunol.
145
:
635
7
Clerici, M., G. M. Shearer.
1994
. The Th1-Th2 hypothesis of HIV infection: new insights.
Immunol. Today
15
:
575
8
Folks, T. M., K. A. Clouse, J. Justement, et al
1989
. Tumor necrosis factor-α induces the expression of the human immunodeficiency virus from a chronically infected T cell clone.
Proc. Natl. Acad. Sci. USA
86
:
2365
9
Finnegan, A., K. A. Roebuck, B. E. Nakai, D. S. Gu, M. F. Rabbi, S. Song, A. L. Landay.
1996
. IL-10 cooperates with TNF-α to activate HIV-1 from latently and acutely infected cells of monocyte/macrophage lineage.
J. Immunol.
156
:
841
10
Weissman, D., G. Poli, A. S. Fauci.
1994
. Interleukin 10 blocks HIV replication in macrophages by inhibiting the autocrine loop of tumor necrosis factor α and interleukin 6 induction of virus.
AIDS Res. Hum. Retroviruses
10
:
1199
11
Saville, M. W., K. Taga, A. Foli, S. Broder, G. Tosato, R. Yarchoan.
1994
. Interleukin-10 suppresses human immunodeficiency virus-1 replication in vitro in cells of the monocyte/macrophage lineage.
Blood
83
:
3591
12
Masood, R., Y. Lunardi-Iskandar, T. Moudgil, Y. Zhang, R. E. Law, C. L. Huang, R. K. Puri, A. M. Levine, P. S. Gill.
1994
. IL-10 inhibits HIV-1 replication and is induced by Tat.
Biochem. Biophys. Res. Commun.
202
:
374
13
Bjorndal, A., H. Deng, M. Jansson, J. R. Fiore, C. Colognesi, A. Karlsson, J. Albert, G. Scarlatti, D. R. Littman, E. M. Fenyo.
1997
. Coreceptor usage of primary human immunodeficiency virus type 1 isolates varies according to biological phenotype.
J. Virol.
71
:
7478
14
Lu, Z., J. F. Berson, Y. Chen, J. D. Turner, T. Zhang, M. Sharron, M. H. Jenks, Z. Wang, J. Kim, J. Rucker, et al
1997
. Evolution of HIV-1 coreceptor usage through interactions with distinct CCR5 and CXCR4 domains.
Proc. Natl. Acad. Sci. USA
94
:
6426
15
Sozzani, S., S. Ghezzi, G. Iannolo, W. Luini, A. Borsatti, N. Polentarutti, A. Sica, M. Locati, C. Mackay, T. N. Wells, et al
1998
. Interleukin 10 increases CCR5 expression and HIV infection in human monocytes.
J. Exp. Med.
187
:
439
16
Di Marzio, P., J. Tse, N. R. Landau.
1998
. Chemokine receptor regulation and HIV type 1 tropism in monocyte-macrophages.
AIDS Res. Hum. Retroviruses
14
:
129
17
Cohen, O. J., A. Kinter, A. S. Fauci.
1997
. Host factors in the pathogenesis of HIV disease.
Immunol. Rev.
159
:
31
18
Butera, S. T., B. D. Roberts, T. M. Folks.
1996
. Ligand passing by the p75 tumour necrosis factor receptor enhances HIV-1 activation.
Cytokine
8
:
745
19
Biglino, A., A. Sinicco, B. Forno, A. M. Pollono, M. Sciandra, C. Martini, P. Pich, P. Gioannini.
1996
. Serum cytokine profiles in acute primary HIV-1 infection and in infectious mononucleosis.
Clin. Immunol. Immunopathol.
78
:
61
20
Esser, R., W. Glienke, H. von Briesen, H. Rubsamen-Waigmann, R. Andreesen.
1996
. Differential regulation of proinflammatory and hematopoietic cytokines in human macrophages after infection with human immunodeficiency virus.
Blood
88
:
3474
21
Herbein, G., L. J. Montaner, S. Gordon.
1996
. Tumor necrosis factor α inhibits entry of human immunodeficiency virus type 1 into primary human macrophages: a selective role for the 75-kilodalton receptor.
J. Virol.
70
:
7388
22
Khatissian, E., L. Chakrabarti, B. Hurtrel.
1996
. Cytokine patterns and viral load in lymph nodes during the early stages of SIV infection.
Res. Virol.
147
:
181
23
Lee, B. N., N. Ordonez, E. J. Popek, J. G. Lu, A. Helfgott, N. Eriksen, H. Hammill, C. Kozinetz, M. Doyle, M. Kline, et al
1997
. Inflammatory cytokine expression is correlated with the level of human immunodeficiency virus (HIV) transcripts in HIV-infected placental trophoblastic cells.
J. Virol.
71
:
3628
24
Mellors, J. W., B. P. Griffith, M. A. Ortiz, M. L. Landry, J. L. Ryan.
1991
. Tumor necrosis factor α/cachectin enhances human immunodeficiency virus type 1 replication in primary macrophages.
J. Infect. Dis.
163
:
78
25
Moriuchi, H., M. Moriuchi, A. S. Fauci.
1997
. Nuclear factor-κB potently up-regulates the promoter activity of RANTES, a chemokine that blocks HIV infection.
J. Immunol.
158
:
3483
26
Poli, G., A. Kinter, J. S. Justement, et al
1990
. Tumor necrosis factor α functions in an autocrine manner in the induction of human immunodeficiency virus expression.
Proc. Natl. Acad. Sci. USA
87
:
782
27
Popik, W., P. M. Pitha.
1993
. Role of tumor necrosis factor α in activation and replication of the tat-defective human immunodeficiency virus type 1.
J. Virol.
67
:
1094
28
Ramazzotti, E., M. Vignoli, M. C. Re, G. Furlini, M. La Placa.
1996
. Enhanced nuclear factor-κB activation induced by tumor necrosis factor-α in stably tat-transfected cells is associated with the presence of cell-surface-bound Tat protein.
AIDS
10
:
455
29
Salazar-Gonzalez, J. F., O. Martinez-Maza, N. Aziz, J. A. Kolberg, T. Yeghiazarian, L. P. Shen, J. L. Fahey.
1997
. Relationship of plasma HIV-RNA levels and levels of TNF-α and immune activation products in HIV infection.
Clin. Immunol. Immunopathol.
84
:
36
30
Nabel, G., D. Baltimore.
1987
. An inducible transcription factor activates expression of human immunodeficiency virus in T-cells.
Nature
326
:
711
31
Shatrov, V. A., F. Ratter, A. Gruber, W. Droge, V. Lehmann.
1996
. HIV type 1 glycoprotein 120 amplifies tumor necrosis factor-induced NF-κB activation in Jurkat cells.
AIDS Res. Hum. Retroviruses
12
:
1209
32
Rieckmann, P., G. Poli, C. H. Fox, J. H. Kehrl, A. S. Fauci.
1991
. Recombinant gp120 specifically enhances tumor necrosis factor-α production and Ig secretion in B lymphocytes from HIV-infected individuals but not from seronegative donors.
J. Immunol.
147
:
2922
33
Merrill, J. E., Y. Koyanagi, I. S. Chen.
1989
. Interleukin-1 and tumor necrosis factor α can be induced from mononuclear phagocytes by human immunodeficiency virus type 1 binding to the CD4 receptor.
J. Virol.
63
:
4404
34
Wahl, L. M., M. L. Corcoran, S. W. Pyle, L. O. Arthur, A. Harel-Bellan, W. L. Farrar.
1989
. Human immunodeficiency virus glycoprotein (gp120) induction of monocyte arachidonic acid metabolites and interleukin 1.
Proc. Natl. Acad. Sci. USA
86
:
621
35
Zembala, M., J. Pryjma, A. Plucienniczak, A. Szczepanek, M. Jasinski, I. Ruggiero, P. Piselli, V. Colizzi.
1995
. Interaction of HIV-1 gp120 molecule fragments with human monocytes: different requirements for tumor necrosis factor-α and IL-6 production.
Clin. Immunol. Immunopathol.
75
:
131
36
Rimaniol, A. C., F. D. Boussin, D. Dormont, J. F. Bach, F. Zavala.
1997
. Mechanisms of downmodulation and release of tumour necrosis factor receptor induced by human immunodeficiency virus type 1 in human monocytes.
Cytokine
9
:
9
37
Marra, M. N., M. B. Thornton, J. L. Snable, C. G. Wilde, R. W. Scott.
1994
. Endotoxin-binding and -neutralizing properties of recombinant bactericidal/permeability-increasing protein and monoclonal antibodies HA-1A and E5.
Crit. Care Med.
22
:
559
38
Wilde, C. G., J. J. Seilhamer, M. McGrogan, N. Ashton, J. L. Snable, J. C. Lane, S. R. Leong, M. B. Thornton, K. L. Miller, R. W. Scott, et al
1994
. Bactericidal/permeability-increasing protein and lipopolysaccharide (LPS)-binding protein: LPS binding properties and effects on LPS-mediated cell activation.
J. Biol. Chem.
269
:
17411
39
Chen, A. R., K. P. McKinnon, H. S. Koren.
1985
. Lipopolysaccharide (LPS) stimulates fresh human monocytes to lyse actinomycin D-treated WEHI-164 target cells via increased secretion of a monokine similar to tumor necrosis factor.
J. Immunol.
135
:
3978
40
Beutler, B. A., I. W. Milsark, A. Cerami.
1985
. Cachectin/tumor necrosis factor: production, distribution, and metabolic fate in vivo.
J. Immunol.
135
:
3972
41
Maniatis, T., E. F. Fritsch, J. Sambrook.
1982
.
Molecular Cloning: A Laboratory Manual
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
42
Lee, Y. M., X. B. Tang, L. M. Cimakasky, J. E. Hildreth, X. F. Yu.
1997
. Mutations in the matrix protein of human immunodeficiency virus type 1 inhibit surface expression and virion incorporation of viral envelope glycoproteins in CD4+ T lymphocytes.
J. Virol.
71
:
1443
43
Gelderblom, H., H. Reupke, T. Winkel, R. Kunze, G. Pauli.
1987
. MHC-antigens: constituents of the envelopes of human and simian immunodeficiency viruses.
Z. Naturforsch. C: Biosci.
42
:
1328
44
Fortin, J. F., R. Cantin, G. Lamontagne, M. Tremblay.
1997
. Host-derived ICAM-1 glycoproteins incorporated on human immunodeficiency virus type 1 are biologically active and enhance viral infectivity.
J. Virol.
71
:
3588
45
Rizzuto, C. D., J. G. Sodroski.
1997
. Contribution of virion ICAM-1 to human immunodeficiency virus infectivity and sensitivity to neutralization.
J. Virol.
71
:
4847
46
Arthur, L. O., J. W. Bess, R. C. Sowder, R. E. Beneveniste, D. L. Mann, J.-C. Chermann, L. E. Henderson.
1992
. Cellular proteins bound to immunodeficiency viruses: implications for pathogenesis and vaccines.
Science
258
:
1935
47
Arthur, L. O., J. W. Bess, Jr, R. G. Urban, J. L. Strominger, W. R. Morton, D. L. Mann, L. E. Henderson, R. E. Benveniste.
1995
. Macaques immunized with HLA-DR are protected from challenge with simian immunodeficiency virus.
J. Virol.
69
:
3117
48
Deng, H., R. Liu, W. Ellmeier, S. Choe, D. Unutmaz, M. Burkhart, P. Di Marzio, S. Marmon, R. E. Sutton, C. M. Hill, et al
1996
. Identification of a major co-receptor for primary isolates of HIV-1.
Nature
381
:
661
49
Koka, P., K. He, D. Camerini, T. Tran, S. S. Yashar, J. E. Merrill.
1995
. The mapping of HIV-1 gp160 epitopes required for interleukin-1 and tumor necrosis factor α production in glial cells.
J. Neuroimmunol.
57
:
179
50
de Jong, J. J., A. de Ronde, W. Keulen, M. Tersmette, J. Goudsmit.
1992
. Minimal requirements for the human immunodeficiency virus type 1 V3 domain to support the syncytium-inducing phenotype: analysis by single amino acid substitution.
J. Virol.
66
:
6777
51
Hill, C. M., H. Deng, D. Unutmaz, V. N. Kewalramani, L. Bastiani, M. K. Gorny, S. Zolla-Pazner, D. R. Littman.
1997
. Envelope glycoproteins from human immunodeficiency virus types 1 and 2 and simian immunodeficiency virus can use human CCR5 as a coreceptor for viral entry and make direct CD4-dependent interactions with this chemokine receptor.
J. Virol.
71
:
6296
52
Broder, C. C., R. G. Collman.
1997
. Chemokine receptors and HIV.
J. Leukocyte Biol.
62
:
20
53
Matsuyama, T., Y. Hamamoto, G. Soma, D. Mizuno, N. Yamamoto, N. Kobayashi.
1989
. Cytocidal effect of tumor necrosis factor on cells chronically infected with human immunodeficiency virus (HIV): enhancement of HIV replication.
J. Virol.
63
:
2504
54
Okamoto, T., T. Matsuyama, S. Mori, Y. Hamamoto, N. Kobayashi, N. Yamamoto, S. F. Josephs, F. Wong-Staal, K. Shimotohno.
1989
. Augmentation of human immunodeficiency virus type 1 gene expression by tumor necrosis factor α.
AIDS Res. Hum. Retroviruses
5
:
131
55
Li, Q., K. Gebhard, T. Schacker, K. Henry, A. T. Haase.
1997
. The relationship between tumor necrosis factor and human immunodeficiency virus gene expression in lymphoid tissue.
J. Virol.
71
:
7080
56
Lane, B. R., D. M. Markovitz, N. L. Woodford., R. Rochford, R. M. Streiter, M. J. Coffey.
1999
. TNF-α inhibits HIV-1 replication in peripheral blood monocytes and alveolar macrophages by inducing the production of RANTES and decreasing C-C chemokine receptor 5 (CCR5) expression.
J. Immunol.
163
:
3653
57
Foli, A., M. W. Saville, L. T. May, D. S. Webb, R. Yarchoan.
1997
. Effects of human immunodeficiency virus and colony-stimulating factors on the production of interleukin 6 and tumor necrosis factor α by monocyte/macrophages.
AIDS Res. Hum. Retroviruses
13
:
829
58
Wesselingh, S. L., C. Power, J. D. Glass, W. R. Tyor, J. C. McArthur, J. M. Farber, J. W. Griffin, D. E. Griffin.
1993
. Intracerebral cytokine messenger RNA expression in acquired immunodeficiency syndrome dementia.
Ann. Neurol.
33
:
576
59
Wesselingh, S. L., K. Takahashi, J. D. Glass, J. C. McArthur, J. W. Griffin, D. E. Griffin.
1997
. Cellular localization of tumor necrosis factor mRNA in neurological tissue from HIV-infected patients by combined reverse transcriptase/polymerase chain reaction in situ hybridization and immunohistochemistry.
J. Neuroimmunol.
74
:
1
60
Glass, J. D., H. Fedor, S. L. Wesselingh, J. C. McArthur.
1995
. Immunocytochemical quantitation of human immunodeficiency virus in the brain: correlations with dementia.
Ann. Neurol.
38
:
755