Antiretroviral treatment induces a shift to type-2 cytokine responses in HIV-1 infected pregnant women.

ARTICLE

INTRODUCTION

An increasing number of women are being infected with human immunodeficiency virus type 1 (HIV-1) and the majority of them are of reproductive age. Treatment of HIV-1-infected pregnant women with zidovudine (ZDV) correlated strongly with a reduced risk of vertical transmission of HIV-1 [1, 2]. Recent studies suggest that a decrease in maternal viral load (VL) is insufficient to fully explain the reduced viral transmission, suggesting that the protective effect of ZDV results, at least in part, from a mechanism other than the reduction of maternal VL [3, 4].

On the other hand, pregnancy is known to alter the profile of cytokines produced by T cells [5], which are known to profoundly affect HIV-1 replication [6, 7]. Thus, inflammatory cytokines are able to induce HIV replication not only in T cells and macrophages [6, 7] but also in placental cells [8]. Moreover, a switch from a Th1-type to a Th2-type cytokine response has been correlated to disease progression [9], although this view has been challenged [10].

Although it is widely assumed that in utero transmission of HIV-1 has little effect on the fetus and that HIV-1-related disease is a postnatal phenomenon, several studies of pregnancy outcome in HIV-1-seropositive mothers have shown an increase in foetal loss [11-13]. It has been hypothesized that in some cases of recurrent abortion of unknown cause, immunological factors play a role in foetus rejection [14]. It has been suggested that Th1-type cytokines may lead to adverse effects on the early embryo either by direct embryotoxicity or by damaging the placenta directly or indirectly via the activation of cytotoxic cells [15]. Therefore, it is tempting to speculate that an increased Th1-type/Th2-type cytokine profile in the placental environment could be responsible for the increased pregnancy loss in HIV-1-infected women as has been suggested [16, 17] or be involved in favoring vertical transmission.

To our knowledge, no study has attempted to investigate the relationship between the cytokine profile produced by maternal peripheral blood mononuclear cells (PBMC) during pregnancy in HIV-1 seropositive women and mother-to-infant viral transmission. Thus we have assessed the cytokine profile in HIV-1-infected pregnant women, and we studied the effect of ZDV treatment during pregnancy on this cytokine profile. Our results indicate that both HIV-1 infection and pregnancy favoured a Th2-type response by T cells. Strikingly, ZDV-treated mothers had a significantly higher Th2-type response than untreated HIV-infected ones.

METHODS

Patients

This was a prospective study involving 82 Caucasian women and conducted between January 1991 and June 1996, in the Hospital General Universitario Gregorio Marañon, in Madrid, Spain. We studied the maternal PBMC cytokine profile during their pregnancy and the effect of ZDV treatment. Thirty-five HIV-1-infected pregnant women (IP) were analyzed and subdivided into two groups according to antiretroviral treatment: 26 who only received ZDV therapy during pregnancy, labor and delivery (mean treatment 4.11 + 1.00 months; range 1-9 months) (IP-T), and 9 who were not treated with any antiviral drugs (IP-NT). Of those, 15 women were included in clinical category A (CD4 mm³ >= 500: 6; CD4 mm³ 200-500: 9) 3 in B (CD4 mm³ 200-500: 2; CD4 mm³ < 200: 1), and 1 in category C (CD4 mm³ < 200: 1) of the Center for Disease Control and Prevention Classification (18). They were compared with several control groups that included: 16 HIV-1-uninfected non-pregnant women, as healthy women controls (C); 18 HIV-1-uninfected pregnant women, as pregnant controls (CP) and 13 non-pregnant HIV-1 infected as HIV-1 infection controls (I). In all groups of pregnant women, the duration of membrane rupture was of less than 4 hours and none had preterm labour. None of them suffered from sexually transmitted diseases during pregnancy or underwent elective caesarean sections. Moreover, none were subjected to procedures such amniocentesis. Since protocol ACTG 076 was published, all mothers were informed of this protocol and were given the possibility to be enrolled in it. The 9 HIV-1-infected mothers that did not receive ZDV were IDUs (injection drug users) during pregnancy who where not controlled in our Hospital until delivery. The study was conducted according to the declaration of Helsinki, and approved by the Ethical Committee of the Hospital.

Detection of HIV-1 infection

In all HIV-1 infected women, blood samples were collected in EDTA tubes, separated within 4 hours and plasma stored at - 70° C. Serum samples were tested for HIV specific antibodies by Western blot (Pasteur-Sanofi) [19]. Viral load was measured in 200 µl plasma using a quantitative reverse polymerase chain reaction (PCR) assay (Amplicor Monitor, Roche Diagnostic System).

Quantification of T cell subsets in peripheral blood

T lymphocyte subsets in peripheral blood were quantified by direct immunofluorescence using monoclonal antibodies of the T series and flow cytometry (FACScan, Becton-Dickinson, Immunocytometry Systems, San José, CA, USA) as previously described [19].

Mononuclear cell preparation

PBMC were isolated from pregnant women the days around delivery, by Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) density-gradient separation (gradient centrifugation). Cells were washed and resuspended in RPMI-1640 (Seromed, Biochrom) medium supplemented with 10% foetal calf serum, 2 mM L-glutamine (Flow), 100 IU/ml penicillin and 100 µg/ml streptomycin (Gibco, Paisley, Scotland, UK) at a concentration of 10⁶ cells/ml.

Cytokine production and proliferation assays

The cells were either stimulated or not with phytohaemagglutinin (PHA, Welcome, 1 µg/ml). PBMC were seeded in 96-well U-bottom microtiter plates (10⁵/100 µl per well). After 3 days of culture at 37° C in 5% carbon dioxide and 95% air, culture supernatants were harvested and their IL-12, IFN-gamma, TNF-alpha and IL-5 content, quantified by specific ELISA assays. Commercially available kits were used according to the manufacturer's instruction (IL-12, Medgenix Diagnostics, Belgium; IFN-gamma, CLB; Amsterdam, Holland, TNF-alpha, Bender MedSystems, Vienna, Austria and IL-5 Endogen Inc., Cambridge, MA, USA). Cytokine concentration was assayed in duplicate cultures.

Cell proliferation was evaluated by incorporation of (³H) thymidine in parallel cultures during the last 16 hours of culture. Cells were harvested in glass fiber filters using an automatic cell harvester, and radioactive incorporation was evaluated in a liquid scintillation spectrometer. The assay was carried out in triplicate cultures.

Statistical analysis

To analyse the differences among groups we used the Mann-Whitney U test, a non-parametric statistical test. This type of analysis is as potent as parametric analysis when applied to small groups, and also, it guarantees the normalization of data. The analyses were based on the ranges of the values of the different variables as well as on the median values.

RESULTS

Characteristics of the patients at entry to the study

Clinical characteristics, CD4⁺ and CD8⁺ T cell percentages, CD4/CD8 ratios and VL from the 82 women included in the study are summarised in Table 1. There were no significant differences in any of these markers between infected women pregnant (IP) or not (I), or between pregnant women, infected treated or not (IP-T and IP-NT). As expected, all HIV-1-infected women either pregnant or not, had mean CD4 values lower and CD8 mean values higher than uninfected ones (Table 1). Furthermore, treated and untreated HIV-1-infected pregnant women did not differ statistically as regards the CD4, CD8 and CD4/CD8 values, nor in the mean value of the VL at delivery (Table 1). Treatment with ZDV reduced vertical transmission from 11.1% in the untreated women to 0% in ZDV treated mothers.

Cytokine secretion and T cell proliferation by PBMC

We analyzed the proliferation of and the production of cytokines (IL-12, IFN-gamma, IL-5 and TNF-alpha) by resting and PHA-activated PBMC from HIV-infected pregnant women treated or not with ZDV, and compared them with several control groups: healthy women either pregnant or not, and HIV-1-infected non-pregnant women. Several of the most relevant results obtained are shown from each individual in Figures 1 and 2 and the summary of all the results in Tables 2 and 3.

The basal production of cytokines by PBMC varied in the different groups. The most significant results in unstimulated PBMC were as follows: a) IL-12 production was significantly higher in IP than in I or C groups, although was not significant with respect to CP. More interestingly, treatment with ZDV during pregnancy greatly reduced this basal IL-12 production by PBMC from IP (Table 3); b) IL-5 production was significantly increased in the IP group with respect to the other groups. This production was higher in IP-T than in the untreated IP-NT-group (Table 3); c) TNF-alpha production was slightly up-regulated in IP with respect to CP. Control and I PBMC did not spontaneously produce TNF-alpha; d) IFN-gamma production was barely detectable in any group.

By comparing the mean values of cytokine production in PHA-stimulated PBMC cultures (Table 2), the most statistically significant results were as follows: a) IL-12 production was similar in all groups; b) IFN-gamma production by IP, I and CP-groups was significantly reduced with respect to C group; c) IL-5 production was higher in the IP group than in the CP group. Moreover, production of this cytokine was even higher in PBMC from the IP-T group than in PBMC from the IP-NT group (Table 3); d) TNF-alpha production was significantly higher in IP group than in the remaining groups.

Furthermore, the mean of the individual Th1-type/Th2-type cytokine ratios by PHA-stimulated cultures was compared using IFN-gamma (a Th1-type cytokine) and IL-5 (a Th2-type cytokine) (Tables 2 and 3). In agreement with some previous reports [5], we found that non-pregnant HIV-infected women had a lower IFN-gamma/IL-5 ratio than controls. HIV-infected pregnant women also had a lower IFN-gamma/IL-5 ratio than the uninfected CP-group. Interestingly, the IP-T women showed an IFN-gamma/IL-5 ratio lower than the IP-NT group. However, these two groups had similar levels of T cell proliferation in response to PHA (Table 3), ruling out a non-specific effect on T cell activation as responsible for the observed variations.

DISCUSSION

We and others have clearly established that the use of ZDV during pregnancy, labour, and the neonatal period decreases the risk of HIV-1 transmission [4, 9, 18-20]. It has been proposed that maternal VL were highly predictive of perinatal transmission risk and that ZDV exerts a major protective effect by reducing maternal VL prior to delivery [21]. However, recent studies challenged that view, since a low maternal VL, although correlating with a decreased risk of infection in the new-born, is insufficient to fully explain the effect of ZDV on vertical transmission of HIV-1 [3, 4]. Thus, only a small part of the beneficial effect of the ZDV treatment during pregnancy can be explained by its effect on decreasing VL. Therefore, this protection must result, at least in part, from a mechanism other than the reduction of the maternal VL [3, 4]. In addition, the putative implication of genetic factors in the ZDV-induced resistance to viral transmission, i.e. the deletion in the CCR5 HIV co-receptor, has also been discarded [3, 4, 9, 18-22].

On the other hand, pregnancy modifies the profile of cytokines produced by T cells [20]. Many cytokines are known to either increase or decrease HIV-1 replication [6, 7]. Moreover, the regulation of Th1-type and Th2-type responses is mutually exclusive [9, 23]. Induction of Th1-type cytokines downregulates Th2-type cytokines and vice versa. This prompted us to study the cytokine profile from T cells of ZDV-treated and untreated HIV-infected pregnant women. Normal pregnancy is associated with a lack of strong maternal cell-mediated anti-foetal immunity and a dominant humoral immune response suggesting that pregnancy is a Th2-type dominant situation (24). In agreement with that, we have found that baseline IFN-gamma production was significantly lower in the CP than in the C group (p < 0.01), although no significant differences in the IL-5 production, non-stimulated or in response to PHA were observed. More interestingly, the IFN-gamma /IL-5 ratio after PHA stimulation was lower in the CP group than in group C, indicating a switch to a Th2-type cytokine profile production by T cells of pregnant women.

In addition, it has been proposed that a switch from a Th1-type to a Th2-type cytokine profile in HIV infection is related to progression of disease [9], although this view has been challenged by other authors [10]. We have found that the production of IFN-gamma in PHA-stimulated PBMC was significantly lower in infected than in healthy control women. Moreover, the IFN-gamma/IL-5 ratio in stimulated PBMC was significantly lower in the I than in the C group, suggesting a switch to a Th2 cytokine profile.

Thus, both HIV-1 infection and pregnancy seem to independently favour the production of Th2-type cytokines. In infected pregnant women, basal IL-5 production was higher than in CP or I groups and the IFN-gamma/IL-5 ratio (Th1-type/Th2-type) in PHA stimulated PBMC was lower than in the CP group. However, TNF-alpha production by non-stimulated or PHA-stimulated PBMC was higher in the IP than in the CP group. These results suggest that the IP-group has a more pronounced Th2-type response than the CP or I group, but higher production of TNF-alpha in response to mitogens. This could explain the increase in foetal loss observed in HIV-infected women [11-13] since TNF-alpha is thought to be harmful to the maintenance of pregnancy and may damage the placenta [5, 24].

More interestingly, when we compared the cytokine production of HIV-1-infected women, treated or not with ZDV, we found that PBMC from the treated-group produced higher levels of IL-5, and much lower levels of IL-12, either spontaneously or after PHA stimulation, than those from untreated women. In addition, IFN-gamma production by PHA-stimulated PBMC was also lower. Therefore, the Th1-type/Th2-type cytokine ratio was lower in the treated than in the untreated group. IL-12 is a good inducer of HIV replication and is required for the generation of Th1 responses [6]. This effect of ZDV treatment could explain, at least in part, its protective effects in reducing vertical transmission, apart from its ability to decrease the maternal VL. Reduced IL-2 together with higher production of Th2-type cytokines induced by ZDV may downregulate IFN-gamma production and other Th1 cytokines, which are thought to facilitate HIV infection [9, 23]. An increase in inflammatory cytokines in the placenta has been correlated with an increased HIV replication of placental trophoblast cells [8, 17]. So, a reversal of the Th1/Th2 ratio may decrease viral load levels in the placenta, thus decreasing the risk of vertical transmission. On the other hand, some reports have indicated that IFN-gamma may be responsible for the damage of the placental trophoblast [5, 24]. This damage may eventually break the placental barrier and favour the virus spreading to the foetus. ZDV by increasing Th2/Th1 balance may prevent transmission by either restricting HIV replication or by decreasing Th1-type-associated placental damage or by a combination of both mechanisms. However, we cannot discard other effects of ZDV in maternal and obstretic risk factors.

The mechanism by which ZDV alters cytokine regulation in vivo is unknown. However, ZDV, apart from its well-known antiretroviral activities, has been described to exert other effects in hematopoeitic cells. Thus, ZDV is able to affect erythropoeisis [25] and suppresses myeloid cell differentiation in vivo [26, 27]. Moreover, we have found that ZDV treatment also alters normal B lymphocyte development and immunoglobulin production [28]. Thus, ZDV in vivo seems to act of cells of the lymphoid system, altering their function.

CONCLUSION

In conclusion, it seems evident that there is a switch from a predominant Th1 to Th2-type cytokine response during pregnancy as well as during HIV-1 infection. This switch is further favored by ZDV treatment of HIV-1-infected mothers during pregnancy and is associated with reduced vertical HIV-1 transmission.

Accepted for publication: 13/07/00

Acknowledgements. We thank Dolores Gurbindo, MD (Sección de Inmunopediatría); Paloma Segovia, MD (Servicio de Obstetricia y Ginecología) and Pilar Miralles, MD (Servicio de Microbiología) Hospital General Universitario Gregorio Marañón, Madrid, Spain for their participation in the selection and control of patients, and Consuelo Muñoz for her excellent technical assistance.

This work was funded by grants from the Fondos de Investigación Sanitaria (FISS 00/0207), Comunidad de Madrid and Fundación para la investigación y la prevención del SIDA en España (FIPSE 3008/99), the "Programa Nacional de Salud" (SAF 99-0022), and Bristol-Myers, S.A. Grupo Bristol-Myers Squibb.

REFERENCES

1. Simonds R J, Steketee R, Nesheim S, Matheson P, Palumbo P, Alger L, et al. 1998. Impact of zidovudine use on risk and risk factors for perinatal transmission of HIV. Perinatal AIDS Collaborative Transmission Studies. AIDS 12: 301.

2. Sperling R S, Shapiro D E, Coombs R W, Todd J A, Herman S A, McSherry G D, et al. 1996. Maternal viral load, zidovudine treatment, and the risk of transmission of human immunodeficiency virus type 1 from mother to infant. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. N. Engl. J. Med. 335: 1621.

3. Alonso Arias R, Gurbindo M D, Miralles P, Segovia P, Fernandez-Cruz E, Munoz-Fernandez M A. 1998. Zidovudine treatment prevents vertical transmission of human immunodeficiency virus-1 independently of viral load. Acta Paediatr. 87: 1208.

4. The European Collaborative Study. 1999. Maternal viral load and vertical transmission of HIV-1: an important factor but not the only one. AIDS 13: 1377.

5. Raghupathy R. 1997. Th1-type immunity is incompatible with successful pregnancy. (Review) Immunol. Today 18: 478.

6. Fauci AS. 1996. Host factors and the pathogenesis of HIV-induced disease. (Review) Nature 384: 529.

7. Munoz-Fernandez M A, Navarro J, Garcia A, Punzon C, Fernandez-Cruz E, Fresno M. 1997. Replication of human immunodeficiency virus-1 in primary human T cells is dependent on the autocrine secretion of tumor necrosis factor through the control of nuclear factor-kappa B activation. J. Allergy Clin. Immunol.100: 838.

8. Lee B N, Ordonez N, Popek E J, Lu J G, Helfgott A, Eriksen N, 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.

9. Clerici M, Shearer G M. 1993. A Th1 -> Th2 switch is a critical step in the etiology of HIV infection. (Review) Immunol. Today 14: 107.

10. Graziosi C, Pantaleo G, Gantt K R, Fortin J P, Demarest J F, Cohen O J, et al. 1994. Lack of evidence for the dichotomy of Th1 and Th2 predominance in HIV-infected individuals. Science 265: 248.

11. Langston C, Lewis D E, Hammill H A, Popek E J, Kozinetz C A, Kline M W, et al. 1995. Excess intrauterine fetal demise associated with maternal human immunodeficiency virus infection. J. Infect. Dis. 172: 1451-60.

12. Temmerman M, Chomba E N, Ndinya-Achola J, Plummer F A, Coppens M, Piot P. 1994. Maternal human immunodeficiency virus-1 infection and pregnancy outcome. Obstet. Gynecol. 83: 495.

13. Semprini A E, Ravizza M, Bucceri A, Vucetich A, Pardi G. 1990. Perinatal outcome in HIV-infected pregnant women. Gynecol. Obstet. Invest. 30: 15.

14. Hill J A. 1992. Immunological contributions to recurrent pregnancy loss. Baillieres Clin. Obstet. Gynaecol. 6: 489.

15. Clark D A, Chaouat G. 1989. What do we know about spontaneous abortion mechanisms? (Review) Am. J. Reprod. Immunol. 19: 28-37.

16. Hill J A, Polgar K, Anderson D J. 1995. T helper 1-type immunity to trophoblast in women with recurrent spontaneous abortion. JAMA 273: 1933.

17. Shearer W T, Reuben J, Lee B N, Popek E J, Lewis D E, Hammill H H, et al. 1997. Role of placental cytokines and inflammation in vertical transmission of HIV infection. (Review) Acta Paediatr. (Suppl) 421: 39.

18. CDCP: Center for Diseases Control Prevention. 1992. Revised classification system for HIV-1 infection and expanded surveillance case definition for AIDS among adolescent and adults. MMWR 41: RR-17.

19. Munoz-Fernandez M A, Obregon E, Navarro J, Borner C, Gurbindo M D, Sampelayo T H, et al. 1996. Relationship of virologic, immunologic, and clinical parameters in infants with vertically acquired human immunodeficiency virus type 1 infection. Pediatr. Res. 40: 597.

20. Connor E M, Sperling R S, Gelber R, Kiselev P, Scott G, O'Sullivan M J, et al. 1994. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076 Study Group. N. Engl. J. Med. 331: 1173.

21. Dickover R E, Garratty E M, Herman S A, Sim M S, Plaeger S, Boyer P J, et al. 1996. Identification of levels of maternal HIV-1 RNA associated with risk of perinatal transmission. Effect of maternal zidovudine treatment on viral load. JAMA 275: 606.

22. Edelstein R E, Arcuino L A, Hughes J P, Melvin A J, Mohan K M, King P D, et al. 1997. Risk of mother-to-infant transmission of HIV-1 is not reduced in CCR5/delta32ccr5 heterozygotes. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol 16: 243.

23. Poli G, Fauci A S. 1993. Cytokine modulation of HIV expression. (Review) Semin. Immunol. 1993; 5: 165.

24. Wegmann T G, Lin H, Guilbert L, Mosmann T R. 1993. Bidirectional cytokine interactions in the maternal-foetal relationship: is successful pregnancy a Th2 phenomenon? (Review) Immunol. Today 14: 353.

25. Shah M M, Li Y, Christensen R D. 1996. Effects of perinatal zidovudine on hematopoiesis: a comparison of effects on progenitors from human fetuses versus mothers. AIDS 10: 1239.

26. Sperling R S, Roboz J, Dische R, Silides D, Holzman I, Jew E. 1992. Zidovudine pharmacokinetics during pregnancy. Am. J. Perinatol. 9: 247.

27. Schibler K R, Liechty K W, White W L, Christensen R D. 1993. Production of granulocyte colony-stimulating factor in vitro by monocytes from preterm and term neonates. Blood 82: 2478.

28. Resino S, Gurbindo M D, Bellón J M, Jiménez J L, León J A, Navarro J, et al. 1999. Dinámica de producción de inmunoglobulinas en niños no infectados nacidos de madres infectadas por el VIH-1: efecto de la zidovudina. Med. Clin. (Barc) 112: 568.