Chronic inappropriate (relative to dietary Na+ intake) elevations in circulating aldosterone (ALDO), termed aldosteronism, are associated with remodeling of intramural arteries of the right and left heart. Lesions appear at week 4 of treatment with ALDO and 1% dietary NaCl in uninephrectomized rats (ALDOST) and include invading monocytes, macrophages and lymphocytes with intracellular evidence of oxidative and nitrosative stress, myofibroblasts, and perivascular fibrosis. In this study, we tested the hypothesis that an immunostimulatory state with activated circulating peripheral blood mononuclear cells (PBMCs) precedes this proinflammatory and profibrogenic cardiac phenotype and is initiated by reduction in the cytosolic free Mg2+ concentration ([Mg2+]i). At 1 and 4 wk of ALDOST (preclinical and clinical stages, respectively), we monitored serum Mg2+, PBMC [Mg2+]i and cytosolic free [Ca2+] (via fluorimetry), and expressed genes (via microchip array) as well as markers of oxidative and nitrosative stress in plasma [α1-antiproteinase activity (α1-AP)] and cardiac tissue (immunohistochemical detection of gp91phox subunit of NADPH oxidase and 3-nitrotyrosine). Age- and gender-matched unoperated and untreated (UO) rats and uninephrectomized salt-treated (UN) rats served as controls. Serum [Mg2+] was unchanged by ALDOST. In contrast with UO and UN, [Mg2+]i and plasma α1-AP were each reduced (P < 0.05) at weeks 1 and 4. The decline in PBMC [Mg2+]i was accompanied by Ca2+ loading. Differential (twofold and higher) expression (up- and downregulation) in PBMC transcriptomes was present at week 1 and progressed at week 4. Involved were genes for the α1-isoform of Na+-K+-ATPase, the ATP-dependent Ca2+ pump, antioxidant reserves, inducible nitric oxide synthase, and PBMC activation with autoimmune responses. Expression of 3-nitrotyrosine and activation of gp91phox were seen in inflammatory cells that invaded intramural arteries. Thus early in aldosteronism (preclinical stage), an immunostimulatory state featuring activated circulating PBMCs with reduced ionized [Mg2+]i and oxidative and nitrosative stress precedes and may even predispose to coronary vascular lesions that first appear at week 4.
- peripheral blood mononuclear cells
- ionized magnesium
- oxidative and nitrosative stress
in both humans and rats and whether derived from endogenous or exogenous sources, chronic inappropriate (relative to dietary Na+ intake) elevations in circulating aldosterone (ALDO) are associated with structural remodeling of intramural arteries of the right and left heart and systemic organs (9, 14, 15, 27, 39, 45, 52, 53, 63, 66, 75). Coronary vascular lesions in uninephrectomized rats first appear at 4 wk of ALDO-and-salt treatment (ALDOST) and involve progressively more vascular sites with continued treatment (64). Lesions consist of circulating monocytes, macrophages and lymphocytes that have invaded the perivascular space; fibroblastlike cells or myofibroblasts that express fibrillar type I and III collagens; and ultimately a perivascular fibrosis (15, 27, 39, 45, 52, 53, 64). Sun et al. (65) addressed cellular and molecular events that are involved in the appearance of this proinflammatory and profibrogenic cardiac phenotype. When studied using immunohistochemistry, oxidative and nitrosative stress appeared within inflammatory cells that had invaded the coronary vasculature together with activation of the redox-sensitive nuclear transcription factor NF-κB and upregulated mRNA expression of a proinflammatory mediator cascade that it regulates that includes intercellular adhesion molecule (ICAM)-1, monocyte chemoattractant protein (MCP)-1, and tumor necrosis factor (TNF)-α. Such neurohormonal activation occurs in chronic cardiac failure and is accompanied by increased expression of TNF-family ligands and oxidative stress in peripheral blood mononuclear cells (PBMCs; Refs. 18, 74). The mechanisms that are responsible for the induction of oxidative and nitrosative stress and activation of PBMCs in aldosteronism are unclear.
Mg2+ is involved in myriad reactions and functions in various cells including PBMCs. These include enzymatic reactions, operation of channels, receptors, and intracellular signaling molecules, and conformation of nucleic acids and proteins (22, 50, 54, 55). ALDO influences Mg2+ homeostasis, including increased urinary Mg2+ excretion and reduced PBMC ionized Mg2+ levels (20, 30). In this study, we investigated the hypothesis that an immunostimulatory state with activated circulating PBMCs precedes the proinflammatory and profibrogenic cardiac phenotype in aldosteronism and is induced by a reduction in the cytosolic Mg2+ concentration ([Mg2+]i). Toward this end, we monitored the PBMC transcriptome or the genes that it expresses together with markers of oxidative and nitrosative stress in plasma and invading inflammatory cells, and levels of PBMC [Mg2+]i and cytosolic Ca2+ concentration ([Ca2+]i) at 1 and 4 wk of ALDOST; these represent the preclinical and clinical stages of aldosteronism, respectively.
MATERIALS AND METHODS
Animals. Eight-week-old male Sprague-Dawley rats (Harlan) were used. The study was approved by the institution's Animal Care and Use Committee. Unoperated and untreated (UO) age-matched rats served as one control group (n = 5). Uninephrectomized rats that received 1% NaCl/0.4% KCl in drinking water and standard laboratory feed served as a second control group (n = 10). Separate groups of uninephrectomized salt-treated (UN) rats received ALDO (0.75 μg/h) via implanted minipump for 1 or 4 wk (n = 10 at each time point). This dose of ALDO promptly raises plasma levels to those seen in humans with congestive heart failure or primary aldosteronism (PAL); it also rapidly suppresses plasma renin activity and circulating angiotensin II levels (9, 24). Animals were observed daily to monitor physical activity and food and water consumption. Systolic blood pressure was recorded as previously reported (65). At the conclusion of wk 1 or 4 of ALDOST, animals were weighed and anesthetized, blood was obtained by cardiac puncture, and hearts were harvested.
Plasma Mg2+ concentrations. Total Mg2+ concentrations in 1:20-fold diluted plasma were determined in 100-μl specimens using a Varian model 220 FS double-beam fast sequential atomic absorption spectrophotometer (Varian Techtron; Melbourne, Australia) and were expressed in milligrams per deciliter (5).
[Mg2+]i and [Ca2+]i in PBMCs. A modification of the method of Delva et al. (20) was used for the isolation of PBMCs and the measurement of [Mg2+]i and [Ca2+]i. Briefly, PBMCs were isolated by Histopaque 1083 centrifugation for 30 min at 400 g. Cells were carefully aspirated, washed twice, suspended in PBS, and counted with a hemocytometer.
For the measurement of [Mg2+]i, three separate aliquots of PBMCs (1 × 106 cells each) were suspended in RPMI 1640 and 0.1% BSA (vol/vol) with the cell-permeant dye mag-fura 2-acetoxymethyl ester (mag-fura 2-AM, 10 μmol/l; Molecular Probes; Eugene, OR) and incubated for 1 h at 37°C. After centrifugation, cells were washed twice with RPMI 1640 and 0.1% BSA for removal of extracellular dye and were resuspended in the same medium. Cells were incubated for an additional 45 min at room temperature to allow complete deesterification of the intracellular dye. The cells were then centrifuged and suspended in 2 ml of buffer [that contained (in mmol/l) 140 NaCl, 5 KCl, 1.8 CaCl2, 0.8 MgSO4, 15 HEPES, and 5 d-glucose (pH 7.4)], transferred to a quartz cuvette, and measured fluorimetrically using a spectrofluorometer (model LS-50B, Perkin-Elmer). Fluorescence emission at 510 nm (slit width, 5 nm) was measured at alternate excitations of 335 and 370 nm (slit width, 10 nm). After initial fluorescence was measured, 5 mmol/l each of EDTA and EGTA were added to the cuvette to chelate extracellular Mg2+, and fluorescence values at 335 and 370 nm were read again for calculation of the free resting [Mg2+]i ratio (R). Triton X-100 was then added at a final concentration of 0.1% to lyse the cells, and fluorescence values at 335 and 370 nm were read again for calculation of the minimum fluorescence ratio (Rmin). Subsequently MgSO4 (100 mmol/l) was added to the cuvette, and the fluorescence values at 335 and 370 nm were read for calculation of the maximum fluorescence ratio (Rmax). [Mg2+]i was then calculated as follows: [Mg2+]i = Kd × [(R – Rmin)Sμ] ÷ [(Rmax – R)Sb], where Kd for the Mg2+-mag-fura 2 complex is 1.5 mmol/l (49), R is the free resting fluorescence ratio at 335:370 after addition of EDTA and EGTA, Rmin is the fluorescence ratio at 335:370 after cell lysis with Triton X-100, Rmax is the fluorescence ratio at 335:370 after addition of MgSO4, and Sμ and Sb are the fluorescence intensities at 370 nm with zero Mg2+ and excess Mg2+, respectively.
The [Ca2+]i of PBMCs was measured with the fluorescent dye fura 2-AM. Three separate aliquots of cells (1 × 106 cells each) were incubated in RPMI 1640 medium with fura 2-AM dye (5 μmol/l) for 30 min at 37°C. The cells were centrifuged, washed to remove extracellular dye, and incubated in the same medium for 45 min at room temperature as for the [Mg2+]i assay. For the fluorimetric [Ca2+]i measurements, cells were suspended in the same buffer for [Mg2+]i as described above but without CaCl2. Fluorescence emission values at 510 nm (slit width, 10 nm) with alternating excitations at 340 and 380 nm (slit width, 10 nm) were read at baseline, after addition of EGTA (10 mmol/l) to chelate Ca2+, after addition of Triton X-100 to lyse the cells, and after addition of excess Ca2+ (10 mmol/l CaCl2). [Ca2+]i was calculated with the same equation as that for Mg2+, where Kd for the Ca2+-fura 2 complex was 225 nmol/l (26).
Measurement of plasma α1-antiproteinase. Reactive oxygen and nitrogen species generated from oxidative stress inactivate α1-antiproteinase (α1-AP) by oxidizing an essential methionine at positions 1, 8, or 358 of the active-site loop to methionine sulfoxide (8). Thus the measurement of active α1-AP in body fluids is a biomarker used to assess the presence of oxidative and nitrosative stress (19, 25, 46). Plasma concentrations of α1-AP were measured using a commercially available kit (Oxis Research; Portland, OR). Briefly, plasma was diluted 1:50 with Tris·HCl-phosphate buffer (pH 8.0). Diluted plasma (50 μl) or assay buffer (for controls) was added to a test tube that contained buffer (500 μl) and 2 μM elastase (50 μl) in assay buffer. The tubes were incubated at 37°C for 5 min and subsequently incubated at room temperature for 5 min. This method is based on the measurement of elastase activity and cleavage of N-succinyl-(Ala)3-p-nitroanilide (NASAN), which results in the production of the chromagen p-nitroaniline. Thus the substrate solution NASAN (400 μl) was added to the tube, mixed by vortexing, and immediately transferred to a cuvette. Absorbance values at 410 nm were recorded for at least 1 min to determine the rate of change in absorbance from the slope of the line vs. time (ΔA410). The concentration of α1-AP (in μM) was then calculated as follows: α1-AP = (ΔA410 control – ΔA410 sample ÷ ΔA410 control) × 2d, where d is the dilution factor of the plasma sample before addition to the assay tube, and the factor 2 represents the elastase concentration (in μM). Because α1-AP is an irreversible equimolar inhibitor of elastase, the concentration of active α1-AP in a sample is equivalent to the concentration of elastase that is inhibited. The intra- and interassay percent coefficients of variability for this assay were ≤6%. The detection limit for this assay was 0.5 μM.
PBMC transcriptomes. Total RNA was isolated from purified PBMCs using a tri-reagent (TRIzol, Invitrogen; Carlsbad, CA). Gene-expression analysis was conducted on the Affymetrix rat genome U34A chip (Affymetrix; San Diego, CA) probing ∼7,000 known genes and 1,000 expressed sequence tags (EST). The gene-expression analysis was conducted by a company that provides this core service for our institution on contract. Detailed protocols and descriptions of analyses are available on their web site (http://www.genomeexplorations.com). Briefly, the quality of total RNA was ensured by analysis using an Agilent Bioanalyzer 2100 “Lab on a Chip” system. RNA was then used to synthesize first- and second-strand cDNA. Double-stranded cDNA was used in an in vitro transcription step to synthesize biotin-labeled cRNA. Transcript quality was assessed again at this step by comparing the 5′-to-3′ ratios of a collection of housekeeping genes. Samples that passed both quality-control steps were hybridized to the expression-array chip and relative expression levels for each probe set (gene or EST) were collected and analyzed using Microarray Suite 5.0 software (Affymetrix). This software was used to normalize the data, evaluate the quality of the data sets, and conduct basic comparisons between data from two samples. Comparative analyses of samples produced a list of differentially expressed genes that are defined as genes with expression levels that are significantly different and at least different by a factor of 1 × log2. A total of six UO controls, six ALDOST samples obtained at 1 wk, and six ALDOST samples obtained at 4 wk went into the characterization of transcriptomes. Each sample that was analyzed on expression-array chips consisted of pooled RNA from three animals. We compared transcriptomes from untreated controls to samples obtained at weeks 1 and 4 of ALDOST to produce a list of genes that are affected by the treatment. The experiment was repeated, and only genes that show differential expression in response to treatment in both of these independent experiments are reported in the lists of specific differentially expressed genes (see Tables 2 and 3).
Cardiac pathology. Expressions of oxidative and nitrosative stress were studied by immunohistochemical localization of gp91phox and 3-nitrotyrosine, respectively. Coronal cryostat sections (6 μm) were prepared, air dried, fixed in 10% buffered formalin for 5 min, and washed in PBS for 10 min. Sections were then incubated with primary antibody against gp91phox at a dilution of 1:100 (12) or 3-nitrotyrosine at a dilution of 1:100 (Upstate Biotech; Waltham, MA) in PBS that contained 1% BSA for 60 min. Sections were then washed in PBS for 10 min and incubated with IgG-peroxidase-conjugated secondary antibody (Sigma; St. Louis, MO) at a dilution of 1:150, washed in PBS for 10 min, incubated with 0.5 mg/ml diaminobenzidine tetrahydrochloride 2-hydrate + 0.05% H2O2 for 10 min, and again washed in PBS. Negative control sections were incubated with secondary antibody alone, stained with hematoxylin, dehydrated, mounted, and examined using light microscopy.
Statistical analysis. Plasma Mg2+, [Mg2+]i, [Ca2+]i, and α1-AP results are expressed as means ± SE. Data were analyzed by ANOVA, and significant differences between groups were determined using the Student-Newman-Keuls multiple-comparisons test. Differences were considered statistically significant when P < 0.05.
Animals. During week 1 of ALDOST, animals appeared healthy: they were active, eating, drinking, and gaining body wt (200 ± 5 g, which was not different from body wt of UO and UN control groups, 198 ± 5 and 203 ± 16 g, respectively). After this preclinical stage and at week 4 of ALDOST (clinical stage), animals were lethargic and anorectic and had gained significantly less weight (265 ± 8 g; P < 0.05) compared with UO and UN controls (345 ± 6 and 340 ± 17 g, respectively). We did not observe aural hyperemia, tetany, or convulsions in ALDOST animals. Systolic blood pressure at week 1 ALDOST was no different from UO or UN controls (114 ± 11 vs. 109 ± 12 and 110 ± 7 mmHg, respectively). At week 4, systolic blood pressure was elevated (173 ± 21 mmHg; P < 0.05) compared with controls and week 1 values.
Plasma Mg2+ concentrations. Plasma concentrations of Mg2+ at week 1 of ALDOST (1.52 ± 0.08 mg/dl) were no different than UO controls (1.40 ± 0.06 mg/dl) and UN controls (1.54 ± 0.06 mg/dl). Plasma values for this divalent cation remained unchanged from controls at 4 wk of ALDOST (1.51 ± 0.05 mg/dl).
[Mg2+]i and [Ca2+]i in PBMCs. As seen in Figs. 1 and 2, values for [Mg2+]i and [Ca2+]i did not differ between UO and UN controls. One week of ALDOST was accompanied by a significant (P < 0.05) decline in [Mg2+]i and [Ca2+]i. At week 4 of ALDOST, [Mg2+]i remained reduced compared with controls (P < 0.05), whereas [Ca2+]i had increased (P < 0.05) compared with controls and week 1 values.
Plasma α1-AP. Plasma levels of α1-AP correlate inversely with oxidative and nitrosative stress (19, 25, 46) and with the generation of oxygen metabolites and peroxynitrite, a product of the reaction of superoxide with nitric oxide (34). Levels of α1-AP (Fig. 3) were no different between UO- and UN-control groups. During preclinical and clinical stages that corresponded to wks 1 and 4 of ALDOST, plasma α1-AP values were significantly (P < 0.05) reduced in keeping with an early and sustained induction of oxidative and nitrosative stress.
PBMC transcriptomes. An up- or downregulation in PBMC gene expression was progressively influenced by ALDOST. Relative to controls, we found a differential (twofold and greater) change in expression at weeks 1 and 4 that involved 205 and 431 genes or ESTs, respectively (Table 1). A total of 564 genes or ESTs were differentially expressed at one or both times. Of the 72 genes or ESTs that were differentially expressed at weeks 1 and 4, 37 were increased at both times, 30 were decreased at both times (relative to controls), and 5 were differentially expressed in opposite directions at weeks 1 and 4. Whereas 133 of 205 genes and ESTs were uniquely expressed during the preclinical stage at week 1, 359 of 431 genes and ESTs were uniquely expressed at week 4, which suggests that new pathological processes had been initiated.
Within the lists of specific genes that are affected by ALDOST, we found a number of genes with products that are affected by or dependent on cations (Table 2). For example, the documented decrease in [Mg2+]i appears to affect a downregulation of the α1-isoform of Na+-K+-ATPase and the corresponding increase in intracellular Ca2+ to affect an upregulation of an ATP-dependent Ca2+ pump as well as a number of Ca2+-dependent genes. Furthermore, we found an upregulation of genes that is associated with producing and counteracting oxidative stress such as inducible nitric oxide synthase (iNOS) and Mn2+-SOD (Table 3). Other indications of inflammatory response activation included increased expression of ICAM-1 as well as chemokines, cytokines, and the associated receptors. Finally, activation of lymphocytes and production of specific immune responses including autoimmune responses were evident (Table 3).
Cardiac pathology. Microscopic evidence of cardiac pathology involving the invasion of intramural coronary arteries of both ventricles by monocytes, macrophages, and lymphocytes was first seen at week 4 of ALDOST. At these vascular sites (which involve right and left ventricles), immunohistochemical evidence of gp91phox expression (Fig. 4, A and B) and the presence of 3-nitrotyrosine (Fig. 4, C and D) was found. Cardiac morphology at week 1 ALDOST was no different from controls.
Sun et al. (65) have shown that the proinflammatory and profibrogenic cardiac phenotype that involves intramural arteries of the right and left heart in chronic ALDOST rats is related to an induction of oxidative and nitrosative stress and is prevented by antioxidant cotreatment. In this study, we investigated the hypothesis that activation of circulating monocytes, macrophages, and lymphocytes or PBMCs (including the altered redox potential) is related to reduction in [Mg2+]i composition. Our study led to several major findings.
First, a reduction in [Mg2+]i occurred at week 1 of ALDOST, which is a preclinical stage, before the appearance of systemic illness with anorexia and lethargy, elevated blood pressure, or immunohistochemical evidence of cardiac pathology at week 4. Values for [Mg2+]i observed in PBMCs at week 1 ALDOST were significantly reduced compared with age- and gender-matched UO and UN controls. Our control values conform with those reported by others for lymphocytes obtained from rodents on a normal diet (59). Factors that could have contributed to this decrease in [Mg2+]i were considered. We did not monitor urinary or fecal Mg2+ excretion in our rats that received ALDOST; both are increased in patients with PAL and in adrenalectomized rats treated with ALDO (28, 30) and can ultimately lead to hypomagnesemia, which has been reported in several patients with long-standing adrenal adenoma and where normal Mg2+ homeostasis is restored following surgical removal of the adrenal tumor (30, 36, 43). In the present study, plasma Mg2+ levels at 1 and 4 wks of ALDOST remained within the normal range and were statistically indistinguishable from our two control groups. Dietary Mg2+ deficiency can lead to reduced Mg2+ levels in PBMCs (58, 70, 71). The Mg2+ content of the standard dietary feed we used is 20–40 mmol/kg in keeping with daily rodent requirements and is quite unlike the diets that are used to induce Mg2+ deficiency (which have <2 mmol/kg Mg2+; Refs. 70, 71). We therefore conclude that dietary deficiency was not involved in the observed reduction in PBMC [Mg2+]i. Lymphocytes have ALDO receptors (2), and Delva et al. (20) found a Na+-dependent decline in cultured human lymphocyte [Mg2+]i that reached a maximum within 120 min when these cells were incubated with ALDO in physiological concentrations. Na+/Mg2+ exchange sites that may be operative in this response are illustrated in Fig. 5 (56). This efflux of [Mg2+]i is abrogated by blocking the receptor-ligand binding with canrenoic acid and also by inhibiting transcription and protein synthesis with actinomycin D and cycloheximide, respectively (20). These investigators also observed a reduction in lymphocyte [Mg2+]i in patients with PAL. We therefore conclude that the reductions in PBMC [Mg2+]i that we observed at weeks 1 and 4 in our rat model of aldosteronism are related to an ALDO-mediated Na+-dependent response that likely involves a Na+/Mg2+ exchanger. Previous studies from this laboratory have shown coronary lesions do not appear when ALDO administration is combined with a diet that is deficient in NaCl or with a 1% NaCl diet without ALDO treatment in uninephrectomized controls (10). Hence, we believe that both ALDO and Na+ are needed to drive Mg2+ efflux from PBMCs and initiate the pathogenic inflammatory process. Future in vivo studies that use either an ALDO receptor antagonist or an inhibitor of the Na+/Mg2+ exchanger (56, 61) as cotreatment with ALDOST are planned.
A second major finding of the present study and concordant with the early reduction in [Mg2+]i is the activation of the PBMC transcriptome that appears at week 1 of ALDOST, where 205 genes were either up- or downregulated. Mg2+ is an important intracellular divalent cation that is involved in hundreds of enzymatic reactions, many of which are related to ATPases and GTPases. Most Mg2+ is inactive, bound to ATP within the cytosol and within such organelles as endoplasmic reticulum and mitochondria; its relatively small ionized fraction is biologically active. In response to the decrease in [Mg2+]i that appears in PBMCs at week 1 ALDOST, we found reduced expression of genes for the α1-isoform of Na+-K+-ATPase, ATP-dependent RNA helicase, and cGMP-stimulated phosphodiesterase. Reduced tissue Na+-K+-ATPase activity, which is associated with a Mg2+-deficient diet in rats and hamsters, is accompanied by an increase in Na+ and Ca2+ concentrations in skeletal and cardiac muscle and bone and where the elevation in tissue Ca2+ occurs via a Na+/Ca2+ exchanger (17, 35). Our data indicate that a similar process may occur in PBMCs. Another factor that could account for a reduction in Na+-K+-ATPase activity in our rat model is an increase in endogenous circulating ouabain, which is a Na+-K+-ATPase inhibitor that is known to accompany PAL and is normalized after surgical removal of adrenal adenoma (57).
A third major finding of this study was the responses in PBMC [Ca2+]i and transcriptome events associated with Ca2+ overload. An accompaniment to reduced Na+-K+-ATPase activity would be an increase in Na+ entry followed by the stoichiometric exchange of 3 Na+ for1Ca2+ via a Na+/Ca2+ exchanger (48). The influx of Ca2+ contributes to a Ca2+ overload, which initially is shunted to intracellular stores such as mitochondria where an induction of oxidative and nitrosative stress occurs (6, 23, 51, 61). We found upregulated expression of Mn2+-SOD, which is specific to mitochondria-based antioxidant reserves (60). After the saturation of organelles is a subsequent increase in [Ca2+]i, which may be facilitated by the decline in [Mg2+]i, its physiological antagonist, and nitric oxide-mediated release of Ca2+ from mitochondrial stores (29). Our analysis of PBMC transcriptomes revealed increased expression of iNOS. Furthermore, Wehling et al. (72) found that ALDO reduces the efflux of [Ca2+]i from cultured PBMCs in a Na+-dependent manner inhibited by amiloride. Shifts in intracellular divalent cations are a key component associated with transcriptional activation of many different cell types including lymphocytes (11, 22, 54, 55, 73). We observed induction of a number of transcripts associated with defenses against free radical production, with the presence of oxidative and nitrosative stress indicated by reduced plasma α1-AP and immunohistochemical evidence of gp91phox expression [an NADPH oxidase subunit that is specific to leukocytes and endothelial cells (60)] and the presence of 3-nitrotyrosine in PBMCs that invaded the coronary vasculature. For the PBMC transcriptomes, these included expression of antioxidant reserves such as Mn2+-SOD, glutathione reductase, NADPH oxireductase, and iNOS. Each of these reserves are activated during inflammatory states (33). In humans and experimental animals, neurohormonal activation is accompanied by induction of oxidative and nitrosative stress in plasma, skeletal muscle, heart, and PMBCs (3, 16, 18, 62, 69, 74).
Finally, the early activation of specific immune responses in the PBMC transcriptomes that appeared at week 1 of ALDOST was sustained and progressive at week 4, where 230 genes were upregulated, whereas 201 were downregulated. This included induction of inflammatory responses such as ICAM-1, chemokine receptor protein and receptor, IL-1β and its receptor, and substance P receptor. Upregulation of ICAM-1 and chemokine receptor genes can stimulate leukocyte movement and migration across the endothelium and media into the perivascular space, where we found immunohistochemical evidence of oxidative and nitrosative stress in inflammatory cells and previously have localized (by in situ hybridization) upregulated mRNA expression of ICAM-1, MCP-1, and TNF-α in these cells at sites of cardiac lesions (65). This immunostimulatory state is further accompanied by upregulation of immunoglobulins, in keeping with B lymphocyte activation, and activation of specific, perhaps autoreactive, immune responses that include upregulation of anti-acetylcholine receptor antibody expression and downregulation of RT6 gene, which is depleted in autoimmunity (76). If specific immune responses were involved in attracting activated PBMCs to the coronary vasculature, this could explain why heart tissue is not affected until week 4 of ALDOST. Autoimmune-mediated injury could contribute to a progressive structural remodeling of the heart and impair its pump function.
Weglicki and coworkers (70, 71) presented evidence of oxidative and nitrosative stress within plasma, reduced antioxidant reserves in PBMCs, elevated PBMC proinflammatory cytokine production, and substance P receptor expression at week 1 of a Mg2+-deficient diet in rats; these effects preceded cardiac lesions that appeared at week 3, and a substance P receptor antagonist proved to be cardioprotective. Additional evidence linking PBMC behavior and altered immunity with Mg2+ deficiency in adult animals has been reported including exaggerated superoxide production and Ca2+-mobilizing capacity in response to live bacteria or platelet-activating factors (38) and hyperplasia of bone marrow and thymus (4, 7, 31, 40, 41). In young weanling rats, cDNA array studies demonstrate that Mg2+ deficiency is accompanied by upregulation of stress protein expression in neutrophils and thymocytes; apoptosis and heat shock proteins in neutrophils (13); and cytochrome oxidase, glutathione transferase, SOD, and heat shock proteins in thymocytes (47).
Heart failure is a major health problem of epidemic proportions. Irrespective of its etiologic origins, a dysfunction of this normally efficient muscular pump during either its ejection or filling phases (i.e., systolic and diastolic dysfunction, respectively) is associated with systemic consequences that lead to a progressive downhill clinical course. Why? Neurohormonal system activation. Sustained release of effector hormones of the renin-angiotensin-ALDO system are maladapative. They not only account for salt and water retention and the congestive heart failure syndrome, they become stressors that invoke a progressive systemic illness. Featured are oxidative and nitrosative stress in PBMCs and plasma (18, 21, 32); a “cytokine storm” with elevated plasma levels of proinflammatory cytokines (44, 67, 68); and anorexia, lethargy, and tissue wasting (1). Herein we used a rodent model to address one aspect of the neurohormonal profile of human heart failure, namely, that related to aldosteronism, which is defined as chronic elevation in plasma ALDO that is inappropriate for dietary Na+ intake and its association with immune cell activation. During week 1 of ALDOST (a preclinical stage), we found upregulation of the PBMC transcriptomes that featured inflammatory responses and lymphocyte activation. The triggering mechanism associated with this activation was a reduction in PBMC [Mg2+]i and was accompanied by upregulation of oxidative and nitrosative stress-related genes and those representing antioxidant reserves. This immunostimulatory state preceded (by weeks) the appearance of clinical illness with lethargy and anorexia and the proinflammatory, profibrogenic cardiac phenotype with coronary vascular lesions. Moreover, autoreactive immune responses were observed that may explain the delayed nature of cardiac lesions and raise the prospect of an autoimmune-type response that could contribute to the progressive remodeling and adverse function of myocardium that is seen in humans with chronic cardiac failure. ALDO-mediated, Na+-dependent efflux of Mg2+ has been observed in cultured human lymphocytes, and lymphocyte [Mg2+]i is reduced in patients with PAL (20). Yndestad et al. (74) found upregulated expression of a TNF family of ligands in the transcriptome of PBMCs obtained from patients with heart failure. Our findings not only broaden the concept of an immunoneuroendocrine interface (42), but they raise the intriguing prospect that information derived from the PBMC transcriptomes (as may also be the case for their proteomes) may serve as noninvasive biomarkers of disease risk, onset, and progression in human heart failure. This remains to be elucidated.
In summary, we conducted a detailed molecular characterization of PBMCs at 1 and 4 wks in a rat model of aldosteronism. On the basis of our findings, we propose a paradigm that is depicted in Fig. 6 in which an immunostimulatory state that is already present at week 1 (preclinical stage) precedes systemic illness and coronary vascular remodeling that appears at week 4. Initial shifts in divalent cation concentrations ([Mg2+]i and [Ca2+]i) lead to oxidative and nitrosative stress and activation of PBMCs, which may then target the vasculature of intramural coronary arteries to initiate the proinflammatory, profibrogenic cardiac phenotype.
This work was supported in part by National Institutes of Health Grants R01 DK-62403, R24 RR-15373, and R21 DK-55263 (to I. C. Gerling); National Heart, Lung, and Blood Institute Grants R01 HL-67888 (to Y. Sun) and R01 HL-62229 (to K. T. Weber); and grants from the Center of Excellence in Connective Tissue Diseases (to Y. Sun and K. T. Weber).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society