Cytokines and chemokines play an important role in controlling both innate and adaptive immune responses against infections. The persistent pro-inflammatory cytokine profile seen during HIV-1 infection contributes to the pathogenesis of the disease . Leukemia inhibitor factor (LIF) is a pleotropic cytokine that belongs to the interleukin-6 (IL-6) cytokine family . Like IL-6, LIF plays a key role in inflammation. In addition to its pro-inflammatory activity it also plays an important role in the hypothalamus–pituitary–adrenal-mediated stress response . Thus, by inducing corticosteroid synthesis it also has anti-inflammatory properties. Furthermore, LIF is essential in fertility and placental function and is also a key cytokine in the neuroimmune network . We have previously shown that the expression of placenta-derived LIF is associated with protection from transmission of HIV-1 from mother to child. Furthermore, LIF inhibits HIV-1 replication in vitro in peripheral blood mononuclear cells (PBMC) as well as in placenta and thymus explants models . This occurred with at least a 100-fold lower concentration than the inhibitory concentration needed to induce a similar effect with β-chemokines (RANTES and MIP-1 α and β) . LIF-mediated anti-HIV-1 activity occurs prior to reverse transcription and is dependent upon cell-surface expression of its signalling receptor, gp130 . We have previously shown that LIF and its two receptors, gp130 and LIFR-α, were significantly upregulated in lymphoid tissue (LT) of subjects with primary HIV-1 infection (PHI) . In addition, significantly less HIV-1 core antigen expression was found, in LT, in CD4+gp130+ cells as compared with CD4+gp130− cells . In this study we have assessed the concentration of LIF and of the soluble gp130 (sgp130) in plasma from primary HIV-1 infected subjects included in the QUEST study, the first placebo controlled treatment trial in acute HIV . In brief, subjects were treated for > 18 months before randomization to adding vaccines (ALVAC-HIV vCP1452 or ALVAC-HIV vCP1452 + Remune] or placebo for 6 months. At the end of the immunization period, all subjects stopped antiretroviral therapy (ART) and were followed up for a further 6 months to determine the primary virological end-point (viremia < 1000 HIV-1 RNA copies/ml plasma in the absence of treatment). No statistical difference was found between the pooled vaccine arms and the placebo arm 6 months post-stopping treatment in term of the percentage of subjects achieving the primary virological end-point . In our present sub-study, we selected a group of ‘controllers’ defined as subjects with a viral load < 1000 HIV-1 RNA copies/ml plasma and a group of ‘non-controllers’ who were defined as those with > 9000 copies of HIV-1-RNA copies/ml plasma 6 months post-stopping ART. We assessed whether the HIV-1 infected subjects with low HIV viremia 6 months post-stopping ART showed initial higher LIF levels as compared with those who did not.
Material and methods
Plasma samples were collected from 22 primary HIV-1 infected subjects from the QUEST study (Table 1), 10 acutely Epstein–Barr virus (EBV)-infected individuals with infectious mononucleosis and 12 HIV-1-seronegative healthy controls. Here we analysed the plasma concentrations of LIF and of sgp130 from samples taken during PHI and the post-ART phase. During PHI, samples were analysed 7 days prior to the initiation of ART and at day 1, 7, 14 and week 4, 8, 12, 20, 24 and 28 post-initiation of treatment (day minus 7 = mean duration of 11 days from onset of PHI symptoms). In the post-ART phase, samples were analysed and at week −4, 1, 4, 8, 12, 16, and 20 after cessation of treatment. Ten ‘controllers’ (< 1000 HIV-1 RNA copies/ml plasma at 6 months post-stopping ART) and 12 ‘non-controllers’, (> 9000 HIV-1 RNA copies/ml plasma at 6 months post-stopping ART) were selected for analysis. Among a total of 14 ‘controllers’ we selected 10 that had had samples taken most frequently and 12 ‘non-controllers that had the highest viral load (> 140 000 HIV-1 RNA copies/ml plasma) at the beginning of the PHI phase. There was a significant difference in viral load between the two groups at onset of study.
Plasma samples from 10 subjects with acute primary EBV infection (positive for anti-VCA IgM and negative for anti-EBNA IgG) were collected at the Infectious Unit at Karolinska University Hospital, Stockholm, Sweden. HIV-1-seronegative plasma samples were provided by 12 healthy blood donors. The QUEST study was performed after approval from the Institutional Review Boards and Ethical Committees at each participating site. All subjects included in the study gave their signed informed consent prior to study enrolment.
HIV-1 RNA detection in plasma
Quantitative plasma HIV-1 RNA was performed using Roche Amplicor test (Version 1.0; limit of detection 400 copies/ml plasma; Roche Molecular Systems Inc., Alameda, California, USA) and the Roche Amplicor Monitor test (UltraSensitive Version 1.5; limit of detection 50 copies/ml plasma; Roche Molecular Systems Inc.). A modified version of the UltraSensitive version, with a limit of detection of 3 copies/ml plasma, was used when viremia was below the limit of detection of the commercial UltraSensitive assay [8,9].
CD4 and CD8 T-cell counts
Assessment of CD4 and CD8 T-cell counts in peripheral blood was performed by routine flow cytometric analysis.
To measure LIF plasma concentration, an in-house LIF-ELISA was developed. A polyclonal capture and detection antibody pair was selected that detected both Escherichia coli (E. coli) derived recombinant human LIF (rhLIF) and native forms of LIF. It has previously been shown that monoclonal antibody against E. coli derived LIF may not detect different forms of native glycosylated LIF . Costar high binding 96-well plates (Corning Inc., Corning NY, USA) were coated with 2 μg/ml polyclonal capturing Ab (AF-250-NA, R&D Systems Europe, Abingdon, UK) overnight at 4°C. The following day the plate was blocked for 1 h at 37°C with phosphate buffered saline (PBS) containing 1% bovein serum albumin. Thereafter the standards (rhLIF, R&D Systems Europe), positive controls [native LIF produced from the human stromal cell line HS-5 (ATCC, Manassas, Virginia, USA) and from phytohaemagglutinin stimulated PBMC)] and samples were added and the plate was incubated overnight at 4°C. Biotinylated polyclonal detection antibody (BAF250, R&D Systems Europe) was added at concentration of 200 ng/ml and the plate was incubated overnight at 4°C. The following day the plate was incubated with alkaline phosphatase-conjugated streptavidin for 30 min at room temperature followed by addition of substrate solution; phosphatase substrate tablets (Sigma-Aldrich Sweden AB, Stockholm, Sweden) and diethanolamine with 0.05 mM MgCl2, pH 9.8. The optical density was determined by using a micro plate reader (Labsystem Genesis ELISA plate reader, Thermo Labsystems, Franklin, MA, USA) set to read 405 nm. The plate was washed six times with PBS containing 0.05% Tween-20 between each reaction step. The concentration of LIF was calibrated from a dose response curve with rhLIF (R&D Systems Europe) as standard. The optical density reaction was linear up to 1000 pg/ml rhLIF, with a limit of detection of less than 10 pg/ml. The intra-assay variation was 5.3% and the inter-assay variation was 5.8% CV.
The concentration of sgp130 in plasma from the patients included in this study was measured using a commercially available sgp130-ELISA (R&D Systems Europe). The measurement was performed according to the manufacturer's instructions and the standards (recombinant sgp130) and samples were analysed in duplicate and presented as sgp130 concentration in pg/ml. The optical density reaction was linear up to 800 pg/ml of sgp130 with the minimum detectable dose of less than 80 pg/ml. All samples were run in a 1: 100 dilution.
Area under the viral load-curve was calculated for all HIV-1 infected patients and a non-parametric two-tailed Mann–Whitney U test with 95% confidence interval (CI) distance was used to analyse statistical significance between ‘controllers’ and ‘non-controllers’ with regards to the plasma concentration of HIV-1 RNA and to CD4 and CD8 T-cells counts. Wilcoxon signed Rank test with 95% CI was used to analyse statistical significance between HIV-1 or acutely EBV infected and HIV-1-seronegative healthy controls with regards to the concentration of LIF in plasma. Demanding regression with 95% CI was used to analyse the correlation between plasma viral load and plasma concentration of LIF. A non-parametric two-tailed Mann–Whitney U test with 95% CI was used to analyse statistical significance between HIV-1 or acute EBV infected and HIV-1-seronegative healthy controls with regards to the concentration of sgp130 in plasma. Finally, area under the curve was also used to compare the concentration of HIV-1 RNA, LIF or sgp130 in PHI phase versus post-ART phase followed by non-parametric two-tailed Mann–Whitney U test with 95% CI. P < 0.05 was considered significant. Statistical analyses were performed with Graph Pad Prism 4 (Graph Pad Software, Inc., San Diego, CA, USA).
Patients with sustained control of HIV-1 replication had significantly lower viral load throughout the PHI phase
The concentration of HIV-1 RNA in plasma was measured in order to investigate how the viral load during PHI phase changed over time in the two groups of individuals (‘controllers’ and ‘non-controllers’). Data showed that ‘controllers’ had significantly (P < 0.0004) lower viral load during PHI and the ensuing weeks as compared with ‘non-controllers’ (Fig. 1a and b). However, no significant differences between the groups were found regarding CD4 and CD8 T-cell counts at this early stage of disease (Fig. 1c and d).
Enhanced LIF-plasma levels in acute HIV-1 infection
LIF-ELISA was used to measure LIF plasma concentration during PHI. LIF was significantly increased in PHI individuals at day 1 and 14 (P < 0.002 and P < 0.001, respectively) and also at week 4 (P < 0.0005), 8 (P < 0.002), 12 (P < 0.004), 20 (P < 0.004), 24 (P < 0.03) and 28 (P < 0.0005) after enrolment in the study, respectively, as compared with HIV-1-seronegative healthy controls (Fig. 2a). Plasma samples from acutely EBV-infected individuals did also have significantly (P < 0.001) higher LIF concentration as compared with HIV-1-seronegative healthy controls (Fig. 2a). There were no significant differences in the LIF plasma concentration between acute EBV and PHI at day minus 7, day 1 and 14 post initiation of ART, respectively. However, at week 4, 8, 12, 20, 24 and 28 during PHI, LIF concentration was significantly (P < 0.02, P < 0.008, P < 0.01, P < 0.03, P < 0.02 and P < 0.005, respectively) reduced in the HIV-1 plasma as compared with the acute EBV plasma. There were no significant differences in the LIF plasma concentration between ‘controllers’ and ‘non-controllers’, although the median value showed a trend towards a higher LIF concentration in ‘non-controllers’ as compared with ‘controllers’ from day minus 7 to day 1 post-initiation of ART (Fig. 2b). Overall the kinetic response pattern was comparable between the two groups. A positive significant correlation (P < 0.01 and r = 0.19) between levels of plasma HIV-1 viral load and LIF concentration was found in HIV-1 infected individuals (Fig. 2c). However, no statistical correlation was found when the PHI cohort was divided into ‘controllers’ (Fig. 2d, P < 0.32 and R = 0.12) and ‘non-controllers’ (Fig. 2e, P < 0.07 and R = 0.19).
Enhanced sgp130 concentration in plasma from patients with acute HIV-1 infection
Sgp130 was significantly increased in PHI at day 14 (P < 0.01) and at week 4 (P < 0.003), 8 (P < 0.007), 12 (P < 0.01), 20 (P < 0.006), 24 (P < 0.01) and 28 (P < 0.02) as compared with HIV-1-seronegative healthy controls (Fig. 2f). However, no significant differences between ‘controllers’ and ‘non-controllers’ were found (data not shown). Furthermore, a significant increase in sgp130 concentration was found at week 4 (P < 0.03) in HIV-1 plasma as compared with EBV plasma (Fig. 2f). There were no significant differences in plasma concentration of sgp130 between the acute EBV samples and HIV-1-seronegative healthy controls (Fig. 2f). Furthermore, the median levels of sgp130 increased between day 1 and day 14 during the PHI phase while the detectable levels of LIF were decreasing during the same time period (Fig. 2g).
Patients with sustained HIV-1 replication control had significantly lower viral load throughout the post-ART phase
Plasma HIV-1 RNA was measured during the 24-week period after ART interruption and statistical analysis showed that the ‘controllers’ had significantly (P < 0.0002) lower viral load throughout the whole post-ART phase as compared with ‘non-controllers’ (Fig. 3a). No significant differences between the two groups were found regarding CD4 and CD8 T-cell counts at week 1 after discontinuation of ART (data not shown). However, at week 16 the ‘non-controllers’ had significantly lower CD4 T-cell counts as compared with the ‘controllers’(P < 0.05), but no difference was found regarding CD8 T-cell count levels (data not shown).
Significantly lower plasma levels of both LIF and of sgp130 in the post-ART phase versus the PHI phase
Despite viral rebound in six out of 10 ‘controllers’ and 12 out of 12 ‘non-controllers’ there were no significant differences in the LIF or the sgp130 plasma concentration in the post-ART phase. Both levels were comparable between HIV-1 infected and uninfected controls (Fig. 3b and d). Furthermore no differences were found between ‘controllers’ and ‘non-controllers’ (data not shown). However, both LIF and sgp130 concentrations in HIV-1 infected individuals were significantly (P < 0.04 and P < 0.02, respectively) reduced in the post-ART phase as compared with the PHI phase (Fig. 3c and e).
Many studies have shown that the HIV-1 viral set-point in blood after PHI predicts the rate of progression to AIDS . The magnitude of this set-point is probably determined by the initial interaction between the virus and the host immune system. Studies in both man and macaques have shown that HIV-1/SIV RNA peak viraemia levels correlate with the viral set-point during the natural course of HIV-1/SIV infection [12,13]. Therefore, a high initial HIV-1 RNA level may lead to a higher viral set-point and faster progression to AIDS as compared with a lower initial HIV-1 RNA peak . We found that the individuals defined as virological ‘controllers’ in this study had a significantly lower HIV-1 viral load throughout both the PHI phase, the post-ART phase (and at 6-months post-stopping ART: by definition) as compared with those who were ‘non-controllers’.
Chronic immune activation is a hallmark of HIV-1 infection and is thought to play an independent role from viraemia in AIDS pathogenesis [15,16]. Studies have shown that during chronic infection a high HIV-1 load correlated with a stronger induction of cytokines and chemokines and other inflammatory mediators than those who had lower viral load during chronic infection [17–19]. Studies in sootey mangabeys with SIV infection show normal T-cell regeneration parameters, low level of immune activation and moderate immune response (but a high viral load) whereas SIV-infected macaques show a state of generalized immune activation associated with a high viral load, decreasing CD4 T-cell numbers and rapid clinical progression . In the present study we have investigated the potential role of LIF, a pleotropic cytokine, in controlling HIV-1 in the early stages of the infection and after cessation of ART. We found that the plasma concentrations of LIF and of the soluble form of LIF's signalling receptor, sgp130 were significantly increased in PHI individuals. Peak concentrations of LIF were observed during the first week after inclusion in the study whereas sgp130 reached its maximum 2–4 weeks later. The concentrations of both molecules decreased over time during the PHI phase probably due to successful ART. At this stage, although we could not see a significant difference in levels between ‘controllers’ and ‘non-controllers’, the median value showed a trend towards an initial higher concentration of LIF and sgp 130 in ‘non-controllers’ as compared with ‘controllers’. The fact that the peak of LIF occurred already at day 7 before initiation of ART to day 1 post-initiation of ART indicates that LIF induction was a part of an early, virally induced pro-inflammatory response rather than an adaptive immune response. Furthermore, this induction of LIF did not seem to be HIV-1 specific, since similar LIF levels were also detected in plasma from acutely EBV-infected individuals. As a positive correlation between HIV-1 RNA and LIF plasma concentration was found, one may speculate that it was the initial viral dissemination which drove LIF production and that sgp130 was released into the circulation to block and neutralize LIF, in order to regulate the biological activity of LIF [21,22]. The data show that the median detectable levels of LIF were decreasing between day one and day 14 during the PHI phase whereas the median levels of sgp130 increased between day one and day 14. This indicated that LIF and sgp130 probably became associated into complexes. The LIF-ELISA probably did not detect the LIF/sgp130 complexes, as addition of sgp130 in a huge excess blocked detectable levels of LIF (data not shown) However, as these complexes are not biologically active they would not contribute to LIF-mediated inhibition of HIV-1 replication.
To our surprise we did not detect elevated levels of LIF during the ART cessation phase. Plasma levels of both LIF and of sgp130 were significantly reduced in the post-ART phase as compared with PHI. This implies that the viral reactivation occurring after treatment interruption did not induce the same magnitude of pro-inflammatory response as in PHI. All together, LIF seems to have a dual role in HIV-1 infection. Patterson et al. have shown that recombinant human LIF added prior to HIV-1 exposure resulted in significant anti-HIV-1 activity while Broor et al. have demonstrated that if LIF is added to HIV-1 infected cells, it increased viral replication in latently infected cells [3,23]. Our data support these views since even though LIF was produced early during HIV-I infection we were not able to demonstrate any correlation between LIF plasma levels and viral control. Even if LIF induction occurred at an early time point during the establishment of HIV-1 infection (peak levels detected within the first weeks of PHI), it may still be ‘too late’ to protect CD4gp130-expressing cells from HIV-1 infection. Furthermore, we have demonstrated previously that gp130 is expressed in less than 50% of all the CD4 cells present in LT of PHI individuals. Thus LIF cannot induce significant anti-viral activity in the entire CD4 cell pool. However, cells that express gp130 may be protected against HIV infection as previously shown in LT from PHI individuals where CD4 cells that expressed gp130 on their cell surface had significantly less HIV-1 core antigen as compared with those which did not . As it has recently been shown that > 50% of all CD4 memory T cells, mainly at mucosal sites but also throughout the body, become infected early after both SIV and HIV-1 infection it would be interesting to determine the gp130 cell surface expression in residing CD4CD45RO memory T cells localized in mucosal tissue [24–26]. Furthermore, studies are needed to explore whether HIV-1 is able to downregulate the expression of gp130 in order to circumvent LIF's anti-HIV effects in order to know if LIF is a potential endogenous HIV-1 inhibitor in vivo.
The authors thank all patients who participated in this study, the QUEST study group and Dr Ilona Lewensohn-Fuchs at department of Clinical Virology, Karolinska University Hospital. This work was supported by the Swedish Foundation for Strategic Research, the Berth von Kantzow Foundation, National Institutes of Health (grant no. AI 41536), Swedish Cancer Society (grant no.2490), Swedish Medical Research Council (grant no.10850), Swedish Physicians Against AIDS Research Foundation, Swedish Society for Medical Research.
1. Kedzierska K, Crowe SM. Cytokines and HIV
-1: interactions and clinical implications. Antivir Chem Chemother 2001; 12:133–150.
2. Auernhammer CJ, Melmed S. Leukemia-inhibitory factor-neuroimmune modulator of endocrine function. Endocr Rev 2000; 21:313–345.
3. Patterson BK, Behbahani H, Kabat WJ, Sullivan Y, O'Gorman MR, Landay A, et al
. Leukemia inhibitory factor inhibits HIV
-1 replication and is upregulated in placentae from nontransmitting women. J Clin Invest 2001; 107:287–294.
4. Cocchi F, DeVico AL, Garzino-Demo A, Arya SK, Gallo RC, Lusso P. Identification of RANTES, MIP-1 alpha, and MIP-1 beta as the major HIV
-suppressive factors produced by CD8+ T cells. Science 1995; 270:1811–1815.
5. Tjernlund A, Fleener Z, Behbahani H, Connick E, Sonnerborg A, Brostrom C, et al
. Suppression of leukemia inhibitor factor in lymphoid tissue in primary HIV
infection: absence of HIV
replication in gp130-positive cells. AIDS 2003; 17:1303–1310.
6. Perrin L. Therapeutic vaccination in primary HIV
infection, the Quest trial. Vaccine 2002; 20:2004–2006.
7. Kinloch-de Loes S, Hoen B, Smith DE, Autran B, Lampe FC, Phillips AN, et al
. Impact of therapeutic immunisation on HIV-1 viremia after discontinuation of antiretroviral therapy initiated during acute infection.J Infect Dis
8. Yerly S, Rutschmann OT, Opravil M, Marchal F, Hirschel B, Perrin L. Cell-associated HIV
-1 RNA in blood as indicator of virus load in lymph nodes. The Swiss HIV
Cohort Study. J Infect Dis 1999; 180:850–853.
9. Yerly S, Kaiser L, Perneger TV, Cone RW, Opravil M, Chave JP, et al
. Time of initiation of antiretroviral therapy: impact on HIV
-1 viraemia. The Swiss HIV
Cohort Study. AIDS 2000; 14:243–249.
10. Taupin JL, Gualde N, Moreau JF. A monoclonal antibody based ELISA
for quantitation of human leukaemia inhibitory factor. Cytokine 1997; 9:112–118.
11. Mellors JW, Kingsley LA, Rinaldo CR Jr, Todd JA, Hoo BS, Kokka RP, et al
. Quantitation of HIV
-1 RNA in plasma
predicts outcome after seroconversion. Ann Intern Med 1995; 122:573–579.
12. Lifson JD, Nowak MA, Goldstein S, Rossio JL, Kinter A, Vasquez G, et al
. The extent of early viral replication is a critical determinant of the natural history of simian immunodeficiency virus infection. J Virol 1997; 71:9508–9514.
13. Lindback S, Karlsson AC, Mittler J, Blaxhult A, Carlsson M, Briheim G, et al
. Viral dynamics in primary HIV
-1 infection. Karolinska Institutet Primary HIV
Infection Study Group. AIDS 2000; 14:2283–2291.
14. Blattner WA, Ann Oursler K, Cleghorn F, Charurat M, Sill A, Bartholomew C, et al
. Rapid clearance of virus after acute HIV
-1 infection: correlates of risk of AIDS. J Infect Dis 2004; 189:1793–1801.
15. Giorgi JV, Hultin LE, McKeating JA, Johnson TD, Owens B, Jacobson LP, et al
. Shorter survival in advanced human immunodeficiency virus type 1 infection is more closely associated with T lymphocyte activation than with plasma
virus burden or virus chemokine coreceptor usage. J Infect Dis 1999; 179:859–870.
16. Giorgi JV, Lyles RH, Matud JL, Yamashita TE, Mellors JW, Hultin LE, et al
. Predictive value of immunologic and virologic markers after long or short duration of HIV
-1 infection. J Acquir Immune Defic Syndr 2002; 29:346–355.
17. Fahey JL, Taylor JM, Manna B, Nishanian P, Aziz N, Giorgi JV, et al
. Prognostic significance of plasma
markers of immune activation, HIV
viral load and CD4 T-cell measurements. AIDS 1998; 12:1581–1590.
18. Sabin CA, Devereux H, Phillips AN, Janossy G, Loveday C, Lee CA. Immune markers and viral load after HIV
-1 seroconversion as predictors of disease progression in a cohort of haemophilic men. AIDS 1998; 12:1347–1352.
19. Salazar-Gonzalez JF, Martinez-Maza O, Nishanian P, Aziz N, Shen LP, Grosser S, et al
. Increased immune activation precedes the inflection point of CD4 T cells and the increased serum virus load in human immunodeficiency virus infection. J Infect Dis 1998; 178:423–430.
20. Chakrabarti LA. The paradox of simian immunodeficiency virus infection in sooty mangabeys: active viral replication without disease progression. Front Biosci 2004; 9:521–539.
21. Modrell B, Liu J, Miller H, Shoyab M. LIF
and OM directly interact with a soluble form of gp130, the IL-6 receptor signal transducing subunit. Growth Factors 1994; 11:81–91.
22. Narazaki M, Yasukawa K, Saito T, Ohsugi Y, Fukui H, Koishihara Y, et al
. Soluble forms of the interleukin-6 signal-transducing receptor component gp130 in human serum possessing a potential to inhibit signals through membrane-anchored gp130. Blood 1993; 82:1120–1126.
23. Broor S, Kusari AB, Zhang B, Seth P, Richman DD, Carson DA, et al
. Stimulation of HIV
replication in mononuclear phagocytes by leukemia inhibitory factor. J Acquir Immune Defic Syndr 1994; 7:647–654.
24. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, Beilman GJ, et al
. CD4+ T cell depletion during all stages of HIV
disease occurs predominantly in the gastrointestinal tract. J Exp Med 2004; 200:749–759.
25. Li Q, Duan L, Estes JD, Ma ZM, Rourke T, Wang Y, et al
. Peak SIV replication in resting memory CD4(+) T cells depletes gut lamina propria CD4(+) T cells.Nature
26. Mattapallil JJ, Douek DC, Hill B, Nishimura Y, Martin M, Roederer M. Massive infection and loss of memory CD4(+) T cells in multiple tissues during acute SIV infection.Nature