Elevated substance P levels in HIV-infected men

The neuropeptide, substance P, is a potent modulator of neuroimmunoregulation. Substance P and its receptor modulate HIV infection. HIV-seropositive men had significantly higher plasma substance P levels compared with uninfected controls, which were associated with decreased CD16 and CD56 natural killer (NK) cell populations. The changes in plasma substance P levels and decreases in NK subsets did not correlate with CD4 cell levels, but a diurnal pattern was suggested for substance P. The balance between substance P expression and functions of immune cells may be important in the immunopathogenesis of HIV infection. Substance P, a potent modulator of neuroimmunoregulation [1–3], stimulates human peripheral blood mononuclear cells to produce inflammatory cytokines [4]. Substance P and its receptor (neurokinin-1) receptor are present in human immune cells, including lymphocytes and monocyte-derived macrophages [2,3]. There is an important relationship between substance P and HIV infection of human immune cells [5]. Substance P and its receptor (neurokinin-1), may be involved in the modulation of HIV infection both in vivo and in vitro. Using in-vitro culture systems, we have demonstrated that substance P upregulates HIV replication in human peripheral blood monocyte-derived macrophages [5]. The substance P inhibitor, CP96-345, inhibits HIV R5 strain replication in mononuclear phagocytes, probably through interaction with CCR5 [6]. HIV-seropositive children had higher plasma levels of substance P in comparison to healthy control children [7]. We investigated whether HIV-infected individuals have elevated substance P levels in plasma using samples from a well-documented cohort of gay men. Samples were obtained from subjects who were participating in the Coping in Health and Illness Project, a National Institutes of Health sponsored cohort study [8,9] based at the University of North Carolina, USA. The cohort included 65 HIV-seropositive gay men, who were asymptomatic at study entry, and 51 HIV-seronegative gay men who served as controls. The participants were recruited from rural and urban areas of North Carolina, USA Each subject contributed two samples (a.m. and p.m.). The HIV disease status of the seropositive subjects was: CD4 T cell count 0–199 cells/mm3 (n = 12); 200–499 cells/mm3 (n = 38); and greater than 500 cells/mm3 (n = 15). At the time of blood sampling, of the 65 HIV-seropositive subjects, 29 subjects were receiving monotherapy with zidovudine and 36 subjects were not receiving antiretroviral therapy. Sample collection procedure: Before each visit, subjects took nothing by mouth starting at midnight on the night before the blood sample collection. Blood was drawn from subjects, who remained recumbent for plasma cortisol, lymphocyte phenotyping, and substance P levels at 9.10 a.m. (a.m. sample) and between 3.30 and 4.30 p.m. (p.m. sample). The plasma samples were stored at −70°C. Substance P was measured using a simple chromotographic procedure for partial purification and sensitive enzyme immunoassay [10]. NK cells were enumerated using flow cytometry [8–10]. The two sample t-test was used to compare the differences in substance P levels between HIV-seronegative and seropositive men with a.m. samples. A paired t-test was used to compare the differences in substance P levels for HIV-seropositive men between a.m. and p.m. samples. Distributions of CD4, CD16, CD56, and CD57 cells and baseline cortisol values also were examined. Pearson correlation coefficients were calculated to investigate the association between substance P and lymphocyte subsets. Multiple linear regression was used to examine the relationship between substance P and lymphocyte subsets, while controlling for baseline cortisol levels. The measurement of plasma substance P levels for 65 HIV-seropositive men and 51 HIV-seronegative men revealed higher mean substance P levels in HIV-seropositive men (55.6 ± 13.3 pg/ml, mean ± SD) in comparison with HIV-seronegative men (31.5 ± 9.7 pg/ml, mean ± SD) (Fig. 1, P < 0.001). In HIV-seropositive men, paired a.m. and p.m. samples were compared (n = 49). The mean ± SD values for a.m. samples were 55.7 ± 13.7 pg/ml, and for p.m. samples were 41.9 ± 7.8 pg/ml (Fig. 1, P < 0.001), suggesting a diurnal pattern for substance P.Fig. 1.: Substance P levels in HIV-seronegative and seropositive men. Note: Mean and 95% confidence interval.We further examined the relationship between substance P levels and lymphocyte subsets in HIV-seropositive men, with and without controlling for cortisol values. Overall, there were no strong correlations between substance P levels and the lymphocyte subsets, except for one borderline significant negative correlation with CD56 cells (r = −0.22, P = 0.084). When the cohort was separated into subgroups consisting of subjects with high versus low baseline cortisol values, however, further associations were detected. Although there was no correlation between substance P and lymphocyte subsets in the low cortisol group, in the high cortisol group there was a significant negative correlation with CD56 cells (r = −0.37, P = 0.038) and a borderline significant correlation with CD16 cells (r = −0.32, P = 0.072). Finally, we examined the relationship between substance P and lymphocyte subsets in a regression model, which controlled for the specific baseline cortisol values (beyond grouping into high versus low groups). We found a statistically significant inverse relationship for both CD56 (P = 0.035) and CD16 (P = 0.047) NK cell populations, and a borderline inverse relationship for CD4 cells (P = 0.088). CD57 cells were not associated with substance P with or without accounting for cortisol levels. There was no evident relationship between plasma substance P levels and plasma cortisol levels. When examining the relationship between substance P and lymphocyte subsets, baseline cortisol was added in the regression model (data not shown). These data are the first determination of alterations in plasma substance P levels in HIV-infected adults, and concur with earlier limited findings in HIV-infected young children [7]. We have demonstrated a significant elevation of plasma substance P in HIV-seropositive gay men compared with high-risk HIV-seronegative gay men. The in vivo findings in this HIV-positive cohort of men further support our hypothesis that there is an interaction between substance P and HIV infection. We have previously demonstrated that substance P upregulates HIV in monocytes/macrophages in vitro[5]. We have also shown that HIV infection enhances substance P mRNA and protein expression in human immune cells in vitro (Ho et al., in preparation). Furthermore, we have demonstrated an elevation of substance P levels in a.m. compared with p.m. samples. This pattern may have similarities to the known diurnal cycles for hypothalamopituitary axis hormones (cortisol, dehydroepiandrosterone). Further testing of samples at serial timepoints, however, is required to confirm a consistent diurnal pattern. In this cohort of very well-characterized HIV-infected gay men, there was no relationship between substance P levels and CD4 cell status. The present study also supports observations that a reduction in NK cell numbers occurs early in the course of HIV disease, and may contribute to its immunopathogenesis [11]. Individuals with high cortisol levels may have immune responses that are uniquely sensitive to neuropeptides, such as substance P. In summary, our study demonstrates elevated levels of plasma substance P in HIV-seropositive men, and an association of these elevated substance P levels with decreased circulating CD16 and CD56 NK cell populations. The observed increases in substance P and decreases in NK subsets did not correlate with CD4 cell strata. The changes in substance P levels and NK cell subsets occurred in a sample of subjects in the early phase of HIV disease (primarily Centers for Disease Control and Prevention stages A1 and A2). The possibility that substance P is involved in the pathogenesis of HIV disease and its progression warrants further study. Steven D. Douglasa,b Wen-Zhe Hoa,b David R. Gettesc Avital Cnaana,b Huaqing Zhaoa Jane Lesermand John M. Petittoe Robert N. Goldend Dwight L. Evansc