Abstract
Background. Human immunodeficiency virus (HIV)–infected patients demonstrate increased activation of the renin-angiotensin-aldosterone system (RAAS). We evaluated changes in immune markers with physiological RAAS activation.
Methods. Immune activation markers were assessed serially in 18 HIV-infected and 7 non–HIV-infected subjects consuming an ad libitum diet followed by a standardized low-sodium diet.
Results. Levels of CCL-2 (P = .0004) and soluble CD163 (P = .0001) significantly increased with sodium restriction and RAAS activation, compared with levels in individuals with ad libitum sodium intake, among chronically treated HIV-infected subjects (mean duration of ART [±SEM], 11 ± 1 years), but not among non–HIV-infected subjects of similar age and sex.
Conclusions. Dietary sodium restriction, which activates RAAS, uniquely stimulates critical indices of immune activation during HIV infection.
Clinical Trials Registration. NCT01407237.
Immune activation persists even among human immunodeficiency virus (HIV)–infected patients with well-controlled viremia and may contribute to the development of metabolic dysfunction and cardiovascular risk in the HIV-infected population. Indeed, prior studies show that circulating markers of monocyte and macrophage activation are associated with arterial inflammation [1], high-risk morphology plaque [2], and insulin resistance [3] in the antiretroviral therapy (ART)–experienced HIV-infected population. In this regard, greater mechanistic insight into immune activation and metabolic dysregulation during HIV infection could inform the field with regard to potential immunomodulatory strategies.
Emerging data suggest that the renin-angiotensin-aldosterone system (RAAS) may be a key driver of inflammation [4], aside from its traditional role in regulating blood volume and sodium homeostasis through the mineralocorticoid receptor. Recent physiologic studies from our group show for the first time that HIV-infected patients have increased aldosterone levels in association with excess visceral adiposity, low adiponectin levels, and insulin resistance, compared with non–HIV-infected individuals, during dietary sodium restriction—a RAAS-activated state [5]. Therefore, we hypothesized that RAAS activation and increased aldosterone levels among the HIV-infected population may contribute to metabolic sequelae mechanistically through stimulation of immune activation.
METHODS
Study Participants
Twenty HIV-infected subjects were previously recruited from the Boston area through HIV-focused community health centers and clinics and local media advertisements for a study investigating RAAS physiology (high-sodium vs low-sodium diet) in relation to visceral fat. In addition, 10 non–HIV-infected subjects were recruited from the same communities to serve as controls. Data from these subjects were included in a prior publication [5]. Inclusion and exclusion criteria were identical for both groups, with the exception of HIV serostatus, as detailed elsewhere [5]. In brief, groups were selected to be of comparable age (excluded for ages <18 or >65 years), sex, and waist circumference. HIV-infected subjects were receiving stable ART regimens for ≥3 months. Subjects who currently were using tobacco, receiving estrogen-replacement therapy, or had hypertension were excluded, to limit the confounding effects of these conditions on RAAS activation. All participants provided informed consent to participate. This study received institutional review board approval from the Partners Human Research Committee.
Standardization of Dietary Sodium Intake to Stimulate RAAS Activity
All subjects were initially assessed for sodium balance at baseline during their daily ad libitum diet by analysis of 24-hour urinary sodium excretion. Only those subjects with a urine sodium excretion rate of >150 mmol/24 hours, conditions associated with RAAS suppression, were included in the final analysis. Following the baseline visit, subjects were asked to consume a standardized low-sodium diet (mean [±standard error of the mean {SEM}], 10 ± 2 mEq Na+, 100 ± 2 mEq K+, and 1000 ± 50 mg Ca2+) for 6 days to induce RAAS activation. Participants were considered to be in low sodium balance if the 24-hour urinary sodium excretion level was estimated during the visit to be <50 mmol. Twenty-five of 30 subjects in whom RAAS physiology was previously examined [5] met criteria for inclusion in the current study to evaluate changes in markers of immune activation under conditions of the ad libitum versus low-sodium diet, unique from the prior investigative focus. The sequence of events is depicted in Supplementary Figure 1. The aim of the study was to assess endogenous activation of the RAAS through sodium restriction, and we therefore did not assess effects of exogenous angiotensin II, a known proinflammatory mediator, on immune markers.
Biochemical and Metabolic Parameters
Serum aldosterone, plasma renin activity, urine sodium, urine creatinine, and insulin were assayed as described in the Supplementary Methods.
Markers of Immune Activation
CCL-2 (R&D Systems) and soluble CD163 (sCD163; Trillium Diagnostics) levels were measured according to each manufacturers' protocol. Data on indices of immune activation have not previously been assessed. Blood samples were obtained for analysis of RAAS components and immune activation indices after subjects were asked to fast for 12 hours and lie supine overnight to control environmental, circadian, and posture effects.
Statistical Analysis
Data are presented as means ± SEM if normally distributed or median [interquartile range (IQR)] if not normally distributed. Categorical variables are reported as proportions. Within-group comparisons (HIV infected or non–HIV infected) of RAAS, metabolic, and immune activation parameters were made between the low-sodium (RAAS-activated) and ad libitum sodium (RAAS-suppressed) conditions, using the paired Student t test, for normally distributed variables, or the Wilcoxon signed rank test, for nonnormally distributed variables. Between-group comparisons (HIV infected vs non–HIV infected) of the absolute change in the same parameters between the low-sodium and ad libitum sodium conditions were performed with the Student t test or the Wilcoxon rank sum test for normally or nonnormally distributed variables, respectively. Statistical significance was defined as a P value of < .05. All statistical analyses were performed using SAS JMP (version 12.0).
RESULTS
Baseline Demographic and Clinical Characteristics
As per study design, HIV-infected and non–HIV-infected subjects did not differ by age or sex. There were fewer white subjects among HIV-infected subjects versus non–HIV-infected subjects (45% vs 80%). Metabolic profiles were overall similar, but the HIV-infected population tended to have higher HOMA-IR levels. The mean duration of HIV infection (±SEM) was 18 ± 1 years. Infection in HIV-infected subjects was well controlled by ART, with a mean ART duration (±SEM) of 11 ± 1 years, a mean CD4+ T-cell count (±SEM) of 571 ± 73 cells/µL, and a mean nadir CD4+ T-cell count (±SEM) 165 ± 51 cells/µL (Supplementary Table 1).
Comparison of RAAS Parameters Between Ad Libitum and Low-Sodium Conditions
Manipulation of dietary sodium intake successfully produced expected changes in RAAS parameters, and these changes were similar in the HIV-infected and non–HIV-infected groups.
Urinary Sodium Excretion
Twenty-four hour urinary sodium concentrations were significantly lower under the sodium-restricted diets as compared to the ad libitum diets in both groups and did not differ by HIV status, suggesting adequately controlled and contrasting states to compare RAAS physiology in the groups. Median urinary sodium concentrations were 15 mmol/24 hours (IQR, 11–38 mmol/24 hours) and 14 mmol/24 hours (IQR, 12–24 mmol/24 hours) in the HIV-infected and non–HIV-infected groups, respectively (P = .83) under the sodium-restricted conditions and 224 mmol/24 hours (IQR, 200–276 mmol/24 hours) and 259 mmol/24 hours (IQR, 221–271 mmol/24 hours), respectively (P = .41), under ad libitum conditions (Supplementary Table 2).
RAAS Parameters
Among the HIV-infected subjects, median plasma renin activity (2.45 ng/mL/hour [IQR, 1.40–3.50 ng/mL/hour] and 0.70 ng/mL/hour [IQR, 0.28–1.43 ng/mL/hour]; P = .0002) and serum aldosterone levels (13.9 ng/dL [IQR, 9.4–31.9 ng/dL] and 4.0 ng/dL [IQR, 2.8–7.0 ng/dL]; P ≤ .0001) were significantly increased during low-sodium and ad libitum sodium conditions, respectively. Plasma renin activity and aldosterone levels increased significantly and to a comparable degree with sodium restriction among non–HIV-infected subjects (Supplementary Table 2).
In further analyses, there was no significant relationship between the change in urinary sodium levels and change in serum aldosterone levels between low-sodium and ad libitum sodium diets among either the HIV-infected or non–HIV-infected groups.
Comparison of Immune Activation Markers in Response to Dietary Sodium Restriction and RAAS Activation
Immune Activation
Median levels of both markers of immune activation, CCL-2 (173 pg/mL [IQR, 147–231 pg/mL] and 123 pg/mL [IQR, 101–195 pg/mL]; P = .0004) and sCD163 (1255 ng/mL [IQR, 695–1447 ng/mL] and 665 ng/mL [IQR, 471–996 ng/mL]; P = .0001), were significantly increased during low-sodium conditions, relative to ad libitum sodium conditions, in the HIV-infected group (Table 1 and Figure 1). In contrast to the HIV-infected subjects, significant changes in immune activation markers were not seen among the non–HIV-infected group during low-sodium versus ad libitum sodium conditions (median, 113 pg/mL [IQR, 107–125 pg/mL] vs 113 pg/mL [IQR, 97–136 pg/mL] for CCL-2 [P = .94] and 495 ng/mL [IQR, 461–759 ng/mL] vs 506 ng/mL [IQR, 413–516 ng/mL] for sCD163 [P = .22]; Table 1 and Figure 1).
Comparison of Immune Markers and Metabolic Indices Between Ad Libitum and Low Sodium Conditions Within and Between Groups of Subjects With and Those Without Human Immunodeficiency Virus (HIV) Infection
| Marker, Group | Sodium Condition | Change Between Conditions | P Value | ||
|---|---|---|---|---|---|
| Ad Libitum | Low | Within-Group Changea | Between-Group Changeb | ||
| CCL-2 level, pg/mL | |||||
| Non–HIV infected (n = 7) | 113 (97–136) | 113 (107–125) | 6 (−11–13) | .94 | .01 |
| HIV infected (n = 18) | 123 (101–195) | 173 (147–231) | 28 (19–76) | .0004 | |
| sCD163 level, ng/mL | |||||
| Non–HIV infected (n = 7) | 506 (413–516) | 495 (461–759) | 127 (−55–307) | .22 | .07 |
| HIV infected (n = 18) | 665 (471–996) | 1255 (695–1447) | 346 (106–539) | .0001 | |
| HOMA-IR index | |||||
| Non–HIV infected (n = 7) | 0.6 (0.4–1.4) | 0.9 (0.7–2.3) | 0.4 (−0.3–0.6) | .22 | .45 |
| HIV infected (n = 18) | 1.0 (0.6–2.0) | 1.8 (0.9–2.6) | 0.5 (0.1–0.9) | .02 | |
| Marker, Group | Sodium Condition | Change Between Conditions | P Value | ||
|---|---|---|---|---|---|
| Ad Libitum | Low | Within-Group Changea | Between-Group Changeb | ||
| CCL-2 level, pg/mL | |||||
| Non–HIV infected (n = 7) | 113 (97–136) | 113 (107–125) | 6 (−11–13) | .94 | .01 |
| HIV infected (n = 18) | 123 (101–195) | 173 (147–231) | 28 (19–76) | .0004 | |
| sCD163 level, ng/mL | |||||
| Non–HIV infected (n = 7) | 506 (413–516) | 495 (461–759) | 127 (−55–307) | .22 | .07 |
| HIV infected (n = 18) | 665 (471–996) | 1255 (695–1447) | 346 (106–539) | .0001 | |
| HOMA-IR index | |||||
| Non–HIV infected (n = 7) | 0.6 (0.4–1.4) | 0.9 (0.7–2.3) | 0.4 (−0.3–0.6) | .22 | .45 |
| HIV infected (n = 18) | 1.0 (0.6–2.0) | 1.8 (0.9–2.6) | 0.5 (0.1–0.9) | .02 | |
Data are median (interquartile range).
Abbreviations: HOMA-IR, homeostatic model assessment of insulin resistance; sCD163, soluble CD163.
a By the Wilcoxon signed rank test.
b By the Wilcoxon rank sum test.
Comparison of Immune Markers and Metabolic Indices Between Ad Libitum and Low Sodium Conditions Within and Between Groups of Subjects With and Those Without Human Immunodeficiency Virus (HIV) Infection
| Marker, Group | Sodium Condition | Change Between Conditions | P Value | ||
|---|---|---|---|---|---|
| Ad Libitum | Low | Within-Group Changea | Between-Group Changeb | ||
| CCL-2 level, pg/mL | |||||
| Non–HIV infected (n = 7) | 113 (97–136) | 113 (107–125) | 6 (−11–13) | .94 | .01 |
| HIV infected (n = 18) | 123 (101–195) | 173 (147–231) | 28 (19–76) | .0004 | |
| sCD163 level, ng/mL | |||||
| Non–HIV infected (n = 7) | 506 (413–516) | 495 (461–759) | 127 (−55–307) | .22 | .07 |
| HIV infected (n = 18) | 665 (471–996) | 1255 (695–1447) | 346 (106–539) | .0001 | |
| HOMA-IR index | |||||
| Non–HIV infected (n = 7) | 0.6 (0.4–1.4) | 0.9 (0.7–2.3) | 0.4 (−0.3–0.6) | .22 | .45 |
| HIV infected (n = 18) | 1.0 (0.6–2.0) | 1.8 (0.9–2.6) | 0.5 (0.1–0.9) | .02 | |
| Marker, Group | Sodium Condition | Change Between Conditions | P Value | ||
|---|---|---|---|---|---|
| Ad Libitum | Low | Within-Group Changea | Between-Group Changeb | ||
| CCL-2 level, pg/mL | |||||
| Non–HIV infected (n = 7) | 113 (97–136) | 113 (107–125) | 6 (−11–13) | .94 | .01 |
| HIV infected (n = 18) | 123 (101–195) | 173 (147–231) | 28 (19–76) | .0004 | |
| sCD163 level, ng/mL | |||||
| Non–HIV infected (n = 7) | 506 (413–516) | 495 (461–759) | 127 (−55–307) | .22 | .07 |
| HIV infected (n = 18) | 665 (471–996) | 1255 (695–1447) | 346 (106–539) | .0001 | |
| HOMA-IR index | |||||
| Non–HIV infected (n = 7) | 0.6 (0.4–1.4) | 0.9 (0.7–2.3) | 0.4 (−0.3–0.6) | .22 | .45 |
| HIV infected (n = 18) | 1.0 (0.6–2.0) | 1.8 (0.9–2.6) | 0.5 (0.1–0.9) | .02 | |
Data are median (interquartile range).
Abbreviations: HOMA-IR, homeostatic model assessment of insulin resistance; sCD163, soluble CD163.
a By the Wilcoxon signed rank test.
b By the Wilcoxon rank sum test.
Comparison of the markers of immune activation CCL-2 and soluble CD163 during ad libitum and low-sodium diets in subjects with (A and C) and those without (B and D) human immunodeficiency virus (HIV) infection. Box plot represents the 25th and 75th percentiles, and lines within the boxes represents medians.
Metabolic Function
The median HOMA-IR index increased significantly during the sodium-restricted condition, compared with the ad libitum condition, in the HIV-infected group (1.8 [IQR, 0.9–2.6] vs 1.0 [IQR, 0.6–2.0]; P = .02) but not in the non–HIV-infected group (0.9 [IQR, 0.7–2.3] vs 0.6 [IQR, 0.4–1.4]; P = .22; Table 1).
DISCUSSION
In the current study, we demonstrate stimulation of immune activation markers, specifically CCL-2 and sCD163, through carefully controlled studies of sodium manipulation to activate the RAAS among HIV-infected ART recipients, changes not seen in the control group. Sodium restriction also worsened insulin resistance in the HIV-infected group but not in the non–HIV-infected group. To our knowledge, this is the first study to investigate the effects of dietary sodium manipulation to activate RAAS on immune indices and provides novel insight into a potential hormone-related mechanism of immune activation.
The RAAS may interact with and stimulate the immune system via a number of different mechanisms. Data suggest that the mineralocorticoid receptor and other components of the RAAS are expressed in immune cells, including monocytes and T lymphocytes [6]. In this regard, an increased aldosterone level could potentiate inflammation by direct mineralocorticoid receptor activation in immune cells. Mineralocorticoid receptor activation in macrophages stimulates release of tumor necrosis factor α, CCL-2, and interleukin 12, while mineralocorticoid receptor activation in T cells promotes cellular differentiation to T-helper type 1 (Th1) and Th17 cells, which secrete interferon γ, interleukin 17, interleukin 21, and interleukin 22. Based on our new data, we postulate that immune activation may be modulated by RAAS activation during HIV infection.
In this regard, immune subsets with particular relevance to HIV have been related to mineralocorticoid receptor signaling and may be influenced by excess aldosterone. A Th17 population was identified as contributing to HIV persistence despite ART and is differentiated from other subsets by production of increased interleukin 17F [7]. In the context of RAAS activation, SGK1, whose genomic expression is rapidly upregulated by mineralocorticoids, enhances Th17 cell differentiation [8], whereas mineralocorticoid receptor blockade downregulates Th17 cells [9]. In the non–HIV-infected population, studies suggest that CD4+ Th17 cells have a role in metabolic-related complications, including insulin resistance [10].
Characterization of the systemic inflammatory and immune response in conditions of RAAS activation in humans is limited. One potential clinical implication of our findings may be that RAAS activation contributes to increased arterial inflammation and cardiovascular disease risk in HIV-infected patients [1]. We and others have established a significant link between vulnerable plaque morphology and monocyte activation, including sCD163, in HIV-infected individuals, as opposed to traditional risk factors [2], and have shown that CCL-2 is independently related to the number of coronary segments with plaque in HIV-infected patients [11]. To that end, RAAS-mediated immune activation may be an important mechanism and anti-inflammatory treatment target for cardiovascular disease in patients with HIV infection.
Aldosterone-mediated vascular injury is linked to an increase in CCL-2 in preclinical studies [12]. Animal studies have also suggested that CD163 expression on macrophages [13] is decreased after treatment with mineralocorticoid receptor blockade and is associated with reduced inflammation. In addition, data in mice show that aldosterone promotes foam cell formation, T cells, and monocyte infiltration and is associated with an unstable lipid-laden and inflammatory plaque phenotype prone to rupture [14], effects which can be attenuated with mineralocorticoid receptor blockade.
No prior studies have specifically evaluated effects of mineralocorticoid receptor blockade on immune indices or cardiovascular disease in the absence of left ventricular dysfunction. Among HIV-infected patients, telmisartan, an angiotensin-receptor blocker, did not demonstrate a metabolic benefit on endothelial dysfunction during HIV infection [15] and has not been studied formally studied for its effects on immune activation in HIV-infected patients. Mineralocorticoid receptor antagonists such as eplerenone avoid the aldosterone escape phenomena that would otherwise occur with an angiotensin-receptor blocker and may provide additional immunomodulatory benefit.
In the current study, we used dietary manipulation as a maneuver to provide critical physiologic insight into the effects of RAAS stimulation and suppression on immune activation. In the non–HIV-infected population, low-sodium diets are generally recommended to reduce cardiovascular injury. It may be that, in HIV-infected individuals, the protective mechanisms associated with sodium restriction are diminished or that endogenous aldosterone levels are too high relative to the dietary sodium intake. Further studies on these differential effects in HIV-infected individuals are needed. The mechanisms by which insulin sensitivity worsens with sodium restriction also merit further evaluation and could relate to RAAS activation or changes in adiponectin or inflammatory pathways with sodium restriction [5].
Limitations of this study include a small sample size. Nonetheless, we detected a significant change in markers of monocyte and macrophage activation during a RAAS-activated state in the HIV-infected group but not the non–HIV-infected group, using rigorous techniques under standardized conditions. More-detailed immunophenotyping investigating the relationship between aldosterone and specific T-cell and monocyte subpopulations in HIV-infected individuals are necessary. This study suggests an acute effect of RAAS activation on immune indices, but this result may also be due to other unknown factors resulting from sodium restriction, such as increased sympathetic activity, which could not be measured.
In conclusion, these studies suggest a potential link between RAAS activation and immune activation during HIV infection and that blockade of RAAS activation may be a potentially useful strategy to reduce immune activation during HIV infection.
Notes
Acknowledgments. We thank the nursing staff at the Massachusetts General Hospital Clinical Research Center and Brigham and Women's Hospital Center for Clinical Investigation, for their dedicated patient care; and the volunteers who participated in this study.
Financial support. This work was supported by the National Institutes of Health (NIH; grants R01DK49302 to S. K. G., K24 HL103845 to G. K. A., and M01RR01066, UL1 RR025758, and UL1 TR001102 to the Harvard Catalyst/Harvard Clinical and Translational Science Center, from the National Center for Research Resources and National Center for Advancing Translational Sciences); Harvard cMeRIT (to S. S.); and the National Institute of Diabetes and Digestive and Kidney Diseases, NIH (pilot and feasibility grant P30DK040561 to the Harvard Nutrition and Obesity Research Center).
Potential conflicts of interest. T. S. has received research funding from Kowa Pharmaceuticals and has served on advisory boards for Theratechnologies. G. K. A. has been a consultant for Pfizer. S. K. G. has received research funding from Bristol-Myers Squibb, Immunex, Gilead, Kowa, Navidea, and Theratechnologies and served as a consultant for Navidea, Merck, Bristol-Myers Squibb, Gilead, Theratechnologies, AstraZeneca, and NovoNordisk, all unrelated to this manuscript. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.
References
