点击显示 收起
1 From the Centre for Clinical Science & Measurement (MG-M, DJL, BJS, CL, TW, and GAF) and the Centre for Nutrition and Food Safety (SAN), School of Biomedical and Molecular Sciences, University of Surrey, Guildford, Surrey, United Kingdom; the Department of Clinical Biochemistry, The Royal Surrey County Hospital, Guildford, United Kingdom (BJS, CL, TW, NV, and GAF); and the Faculty of Medicine, Mashad University of Medical Sciences, Mashad, Iran (MG-M)
2 Supported by the Iranian Government (MG-M) and the British Heart Foundation (DJL). 3 Address reprint requests and correspondence to GA Ferns, Center for Clinical Science & Measurement, School of Biomedical and Molecular Sciences, University of Surrey, Guildford, Surrey GU2 5XH, UK. E-mail: g.ferns{at}surrey.ac.uk.
ABSTRACT
Background: The heat shock proteins (HSPs) are protein chaperones. Higher titers of antibody to HSPs (anti-HSPs) have been reported in atherosclerosis, which may contribute to immunoactivation in this process.
Objective: We investigated whether dietary antioxidants and fat intake are associated with changes in anti-HSP titers in dyslipidemic subjects.
Design: Patients (n = 238) were recruited from hospital lipid clinics. Control subjects (n = 188) were recruited from university and hospital employees. Food-frequency questionnaires were used to estimate dietary antioxidants and fat.
Results: Dyslipidemic patients had significantly higher titers of anti-HSPs than did control subjects; expressed in medians and interquartile ranges of absorbance units, anti-HSP-60 titers were 0.27 (0.18–0.37) and 0.22 (0.16–0.30), anti-HSP-65 titers were 0.45 (0.28–0.79) and 0.31 (0.22–0.50), and anti-HSP-70 titers were 0.22 (0.17–0.30) and 0.19 (0.13–0.27), respectively. Median and interquartile ranges of serum concentrations of C-reactive protein [1.25 (0.42–3.26) and 0.58 (0.17–1.42)] and mean (±SEM) concentrations of vitamin E (16.36 ± 0.31 and 14.08 ± 0.38) were also significantly higher in patients than in control subjects, respectively. In dyslipidemic patients, the major dietary predictors of the variability in anti-HSP-60 titers were vitamin C (P = 0.005), vitamin E (P = 0.04), and total fat (P = 0.009) intakes; for anti-HSP-65 titers, vitamin C was the major predictor (P = 0.002). These findings remained significant after adjustment for confounding factors.
Conclusions: Anti-HSP-60, -65, and -70 titers are significantly higher in dyslipidemic patients with or without established coronary disease. Our data indicate an association between dietary constituents and the immune response to HSPs in dyslipidemic subjects.
Key Words: Heat shock proteins • HSP-60 • -65 • and -70 • antibody titers • dietary intake • antioxidants • fat
INTRODUCTION
The heat shock proteins (HSPs) are a family of 20–25 molecules expressed by cells in response to stresses such as high temperature, free radicals, sheer stress, and toxins, including oxidized LDL cholesterol (1). They are involved in the renaturation of damaged proteins, allowing them to refold into their native conformation. It has been proposed that, because the structure of HSPs is highly conserved across species, the immune response mounted against bacterial HSPs may result in an immune response (2) that has the potential to target endogenous HSPs and cause complement-mediated endothelial injury (3) and thus to accelerate atherogenesis (1, 3, 4). Antibody titers to HSP-60, –65 and –70 have been reported to be associated with coronary risk factors (5, 6), increased risk of cardiovascular disease (7), the severity of cardiovascular (8), and vascular endpoints in patients with established disease (9, 10). HSP expression has been shown to be increased in atherosclerotic plaque (11) and in experimental animal models; immunization with human HSP-60, or BCG vaccine (that contains high concentrations of HSP-65) enhances the atherogenic process (12–14).
There is substantial evidence that protein-calorie malnutrition and some specific nutrient deficiencies have adverse effects on the immune system. Alterations in the quality and quantity of dietary fat and abnormalities in lipid metabolism have been reported to influence immune responses (15, 16). In vivo and in vitro studies showed that fatty acids can modulate the immune system (17, 18), and animal and human studies showed the importance of vitamin E and zinc in maintaining immune function (19).
The effects of dietary constituents on HSPs have been studied in animal models. Kelly et al (20) showed that vitamin E deprivation for 16 wk combined with exercise for 8 wk induced HSP-72 expression in female rats. Andres et al (21) and Andres and Cascales (22) showed that cyclosporine A–induced expression of HSP-70 by rat hepatocytes, which is mediated by the release of reactive oxygen species, is significantly reduced in the presence of vitamin E in vitro. It is interesting, however, that there were previous reports that vitamin C induced an increase in baseline expression of HSP-60 and -70 in vitro, but its mechanisms of action are unclear (23).
Romano et al (15) reported that a diet rich in saturated fatty acids induces the expression of HSP-25, -60, and -70 in mice splenic lymphocytes. We also showed that serum anti-HSP-65 titers rise in rabbits fed a high-cholesterol diet (24). To our knowledge, no studies have yet examined the association between dietary intake and HSP antibody titers in humans. Thus, the principal aim of our study was to investigate whether dietary factors are associated with HSP antibody titers in patients with dyslipidemia.
SUBJECTS AND METHODS
Study design and subject selection
Two hundred thirty-eight patients were recruited from the lipid clinics at the local hospital (the Royal Surrey County Hospital, Guildford, United Kingdom). One hundred eighty-eight control subjects were employees at the University of Surrey or the Royal Surrey County Hospital. Eighty-two patients were obese [body mass index (in kg/m2) >30]; 42 were diabetic (fasting plasma glucose concentration >7 mmol/L); 55 had established coronary artery disease (CAD), and 186 were hypertensive. Of the latter group, 76 had systolic blood pressure (SBP) 160 mm Hg or diastolic blood pressure (DBP) 100 mm Hg. One hundred ten patients had SBP between 130 and 160 mm Hg or DBP between 85 and 100 mm Hg. One hundred seventy-six patients were hypertriglyceridemic (serum triacylglycerol concentration >1.8 mmol/L), and 216 patients were hypercholesterolemic (serum total cholesterol concentration >5.2 mmol/L). Forty-two and 54 of the patients had a calculated 10-y coronary risk of >30% and 20–30%, respectively (risk calculated with the use of the PROCAM algorithm; 25), and 142 patients had metabolic syndrome by National Cholesterol Education Program Adult Treatment Panel III criteria (26). Patients with a history of established CAD included 9 with unstable angina, 15 with previous myocardial infarction, 10 with a history of coronary artery bypass grafting, and 13 with a history of angioplasty. Fifteen patients had undergone a coronary artery bypass graft after a myocardial infarction, and 5 had undergone angioplasty after a myocardial infarction. Fifty-four subjects were excluded from the control group because they were obese (n = 33), had metabolic syndrome (n = 9), or were on medication (n = 11; Table 1).
View this table:
TABLE 1. Characteristics and medication use of dyslipidemic patients and control subjects1
Each patient gave written informed consent to participate in the study. The study protocol was approved by the South-West Surrey Research Ethics Committee and the Advisory Committee of Surrey University.
Current dietary intakes and estimation of antioxidants and fat
Dietary intakes over the previous 12 mo were assessed by using a food-frequency questionnaire (FFQ) as previously detailed (27, 28). In brief, the FFQ was developed and validated against 7-d weighed records (29) and biochemical markers of antioxidant status (30), and its short- and long-term reproducibility was tested (31). During the initial interview, subjects were instructed in filling out the FFQ, and completed FFQs were checked for inaccuracies and inconsistencies at subsequent interviews.
Anthropometric and other measurements
All subjects were measured for height (in cm) and were weighed (in kg) with the use of a stand-on bioelectrical impedance analyzer (Tanita-305 body fat analyzer; Tanita Corp, Tokyo). The latter was also used to estimate percentage body fat. Body mass index was calculated as described above.
Waist and hip measurements were taken to the nearest millimeter as described previously (32). Blood pressure measurements were made with the use of an automated device (DINAMAP compact monitor, model TS; Critikon, Tampa, FL).
Blood sampling
Blood samples were collected between 0830 and 1030 after a 12-h fast by venipuncture of the antecubital vein. For the anti-HSP assays, blood samples were drawn into EDTA-containing tubes (Vacutainer; Becton Dickinson, Cowley, United Kingdom), which were centrifuged at 1000 x g for 15 min at 4 °C, and plasma aliquots were stored at –80 °C until the day on which HSP antibody titers were measured. Samples for lipid profile and measurement of serum vitamin E and high-sensitivity C-reactive protein (hs-CRP) were taken into plain Vacutainer tubes, and those for measurement of glucose were taken into Vacutainer tubes containing fluoride-oxalate. All chemicals were obtained from Sigma Chemical Co (Poole, United Kingdom) unless stated otherwise.
Analytic methods
Lipid profiles, high-sensitivity C-reactive protein, and blood glucose
A fasting lipid profile, comprising total cholesterol, triacylglycerols, and HDL cholesterol, was obtained for each patient. With the use of the formula of Friedewald et al (33), LDL cholesterol was calculated for all subjects except the patients with serum triacylglycerol concentrations >4.0 mmol/L. Lipids, hs-CRP, and glucose were measured by routine methods with a Bayer Advia 1650 analyzer (Bayer, Newbury, United Kingdom).
Serum vitamin E
Serum vitamin E was measured by HPLC. Briefly, 200 µL of an internal standard (10 µg -tocopherol/mL isopropyl alcohol) was added to 200 µL serum and mixed by vortex (34). Aqueous ammonium sulfate (3.9 mol/L) was added (200 µL), and the solution was again mixed by vortex. After centrifugation (1000 x g for 5 min), 50 µL of supernatant was used for analysis using a 150 x 4.6–mm Prodigy 50-µm ODS2 column (Phenomenex Ltd, Macclesfield, United Kingdom) with methanol as mobile phase and detection at 294 nm. At a flow rate of 1.4 mL/min, the retention time for internal standard and vitamin E was 5.2 and 6.6 min, the respectively. Vitamin E standard and quality-control material were obtained from BioRad Laboratories Ltd (Hemel Hempstead, United Kingdom).
Heat shock protein antibody titers
Plasma HSP antibody titers were measured by using in-house enzyme-linked immunosorbent assays. In brief, a 96-well microtiter plate (Nunc Immunoplate Maxisorp; Scientific Laboratory Supplies Ltd, Nottingham, United Kingdom) was coated with human recombinant HSP-60, -65, or -70 by adding 10 ng recombinant HSP in phosphate-buffered saline (PBS) to the wells of a microtiter plate and incubating overnight at 4 °C. Plates were washed with PBS, blocked with Superblock (Pierce & Wariner, Chester, United Kingdom), and washed 3 times with PBS containing 0.05% (by vol) Tween-20. Plasma samples were diluted 1:15 with PBS containing 0.1% Tween-20 and 1% bovine serum albumin (PBT; Sigma-Aldrich Inc, Poole, United Kingdom), and 100 µL/well in quadruplicate was incubated for 30 min at 37 °C. After washing, bound HSP antibodies were detected by the addition of peroxidase-conjugated goat anti-human immunoglobulin G, which was diluted 1:100 with PBT. After washing with PBS/Tween-20, o-phenylenediamine (0.04%) [dissolved in 0.05 mol citrate/L with 0.1 mol phosphate buffer/L (pH 5) and containing 10 µL of a 30% solution of H2O2/25 mL] was added and incubated for 5 min at room temperature. The reaction was terminated by the addition of 3 mol hydrochloric acid/L. The absorbance was read at 492 nm by using a plate reader with GENESIS 2 software (version 2; Life Sciences, Basingstoke, United Kingdom) (35).
Statistical analysis
Statistical analysis was undertaken with the use of MINITAB software (release 13; Minitab Inc, State College, PA), with determination of descriptive statistics (ie, means, medians, SEMs, and interquartile ranges) for all variables. Data were assessed for normality by using the Kolmogorov-Smirnov test. Between-group comparisons of biochemical variables were assessed by analysis of variance. Categorical data were compared by using Fisher's exact test or chi-square test. Values were expressed as means ± SEMs or medians and interquartile ranges (for nonnormally distributed data). Analysis of covariance was used to assess differences after adjustment for important confounding factors, such as age and physical activity. The hs-CRP concentrations and plasma titers of HSP antibody were found to be nonnormally distributed and were therefore logarithmically transformed before parametric analysis.
Stepwise multiple regression analysis was used to predict whether the anti-HSP titers were related to dietary antioxidants or fat. To enable adjustment for potential confounding factors, we entered into the equation the factors age, sex, obesity, metabolic syndrome or accumulating features of metabolic syndrome, diabetes mellitus, smoking, hypertriglyceridemia, blood pressure, established CAD, calculated 10-y coronary risk factors, and drug treatment. A P value of < 0.05 was considered significant.
RESULTS
Descriptive data
There was a high frequency of obesity (35%), type 2 diabetes (18%), hypertension (79%), and positive smoking habit (18%) in the patient group, findings that are typical of a lipid clinic population. As expected among persons attending such a clinic, serum fasting triacylglycerol, plasma glucose, and total- and LDL-cholesterol concentrations were significantly higher for patients than for control subjects. Anthropometric indexes including waist size, waist-hip ratio, body mass index, and percentage body fat were also significantly higher in the patients than in the control subjects (Table 2). Although current smoking habits did not differ significantly between patients and control subjects, a significantly greater proportion of the patients were former smokers (Table 2). The patients were also significantly older than the control subjects.
View this table:
TABLE 2. Clinical and biochemical measurements in dyslipidemic patients and control subjects1
Heat shock protein antibody titers
Dyslipidemic patients had significantly higher anti-HSP-60, -65, and -70 titers than did the control subjects (Table 3). The results remained the same after adjustment for age. A strong correlation between titers of anti-HSP-60, -65, and -70 (P = 0.001) was observed in patient and control groups.
View this table:
TABLE 3. Comparison of anti-heat shock protein (anti-HSP) titers and serum vitamin E and high-sensitivity C-reactive protein (hs-CRP) concentrations between dyslipidemic patients and control subjects
Serum high-sensitivity C-reactive protein concentrations
Serum hs-CRP concentrations were significantly higher in the dyslipidemic patients than in the control subjects (Table 3). The results remained the same after adjustment for age. No significant association between HSP antibody titers and serum hs-CRP concentrations was observed in the patient or control groups (P > 0.05).
Serum vitamin E concentrations
The dyslipidemic patients had significantly higher serum vitamin E concentrations than did the control subjects (P < 0.001; Table 3). However, the ratio of vitamin E to total cholesterol did not differ significantly between these groups (P > 0.05). Again, the results remained the same after adjustment of the serum vitamin E data for age. There was a strong inverse correlation between serum vitamin E concentrations and HSP-70 antibody titers (r = –0.32, P < 0.001) in the control group.
Dietary intakes of macronutrients and micronutrients
The dyslipidemic patients had significantly higher dietary intakes of protein and total fat (both: P < 0.05) than did the control subjects (Table 4). No significant difference was observed in dietary intakes of carbohydrate, sugar, and energy between the patients and the control subjects (Table 4).
View this table:
TABLE 4. Comparison of macro- and micronutrient dietary intakes between dyslipidemic patients and control subjects1
The dyslipidemic patients had significantly (P < 0.05) higher intakes of monounsaturated fat than did the control subjects (Table 4). There was no significant difference between the control subjects and the dyslipidemic patients with respect to dietary intakes of cholesterol, saturated fat, polyunsaturated fat, or antioxidants. In addition, dietary intakes of antioxidants did not differ significantly between the dyslipidemic patients and the control subjects (P > 0.05; Table 4). Adjustment of dietary intake data for age had no effect on the statistical significance of the results.
Multivariate analysis
For anti-HSP-60 titers, the best-fitting models explained 11% of the variation in the dyslipidemic patients and 3.5% of that in the control subjects (Table 5). The results of a similar analysis for anti-HSP-65 titers are shown in Table 6. In the dyslipidemic subjects, 7.3% of the variation in titers was attributable to dietary vitamin C, obesity, and smoking habits. In the control subjects, 4.4% of the variation in antibody titers to HSP-65 was attributed to dietary vitamin E. For anti-HSP-70 titers, the best-fitting model explained 3.7% of the variation in the dyslipidemic patients and 21% of that in the control subjects (Table 7).
View this table:
TABLE 5. Multifactorial analysis of anti-heat shock protein 60 (absorbance units)1
View this table:
TABLE 6. Multifactorial analysis of anti-heat shock protein 65 (absorbance units)1
View this table:
TABLE 7. Multifactorial analysis of anti-heat shock protein 70 (absorbance units)1
DISCUSSION
This is the first reported investigation of the relation between plasma anti-HSP titers and dietary factors in humans. Our results indicate an association between dietary constituents and the immune response to HSPs in patients with coronary risk factors.
Heat shock protein antibody titers and coronary artery disease risk factors
Pockley et al (36) reported that HSP-65 and -70 antibody titers were elevated in subjects with hypertension, whereas anti-HSP-60 and -70 antigen concentrations and HSP-60 antibody titers were similar to those in normotensive controls. Titers of HSP-60, -65, and -70 antibody are positively related to the risk of vascular disease and to cardiovascular disease endpoints (1). Although high immunoglobulin A–class HSP-60 antibody titers were predictive of CAD risk, the effects were reported to be modest in the absence of other classic risk factors (37). High titers of human HSP-70 are associated with a low CAD risk, probably through HSP-70's protective effects on the cellular response to stress (38). Kocsis et al (39) reported that anti-HSP-70 titers are not elevated in subjects with severe coronary atherosclerosis, whereas anti-HSP-60 and –65 titers were significantly higher in these patients, but there was no association between titers of HSP-70 antibody and either anti-HSP-60 or -65 titers.
Heat shock protein antibody titers and dietary intake of antioxidants
Vitamin E is a potent lipid-soluble antioxidant, and its dietary supplementation has been reported to prevent oxidative injury (40). Vitamin E depletion in rats caused a significant increase in HSP-32 and -70 expression by alveolar and liver cells, which returned almost to normal after vitamin resupplementation (41). The possible modulation of the heat shock signal transduction pathway by vitamin E was previously reported in human skin fibroblasts (42). In the current study, we found an inverse relation between dietary vitamin E intake and anti-HSP-60 titers in patients and anti-HSP-65 titers in the control group. This association may be related to the redox mechanisms affected by vitamin E.
Vitamin C is an important water-soluble antioxidant that is potentially beneficial in reducing oxidative tissue damage by chemical reduction of oxidant species (43). Epidemiologic data suggest that a high intake of vitamin C may protect against oxidative damage in vivo (23, 44). A vitamin C–induced increase in baseline expression of HSP-60 and HSP-70 was previously reported, but there is controversy about the mechanism of this increase, and it has been suggested that vitamin C may exert prooxidant effects in some situations (23). In the current study, the positive association between dietary vitamin C intake and anti-HSP-60 and -65 in dyslipidemic patients may be related to an increased expression of HSPs associated with a prooxidant effect of vitamin C.
Heat shock protein antibody titers and dietary fat intake
Several studies showed that both saturated and unsaturated fatty acids can affect immune responses (17). Romano et al (15) showed that a diet rich in saturated fatty acids induces the expression of HSP-25, -60, and -70 in mice splenic lymphocytes, and we showed that serum anti-HSP-65 titers rise in rabbits fed a high-cholesterol diet (24). In the current study, we found positive associations between anti-HSP titers and dietary total fat, saturated fat, and cholesterol. These associations may be due to a combination of increased expression of the HSPs and an enhanced immune response, both of which are associated with a high-saturated-fat diet. No significant relations were found between anti-HSP titers and unsaturated fat in the current study.
Conclusions
We showed that antibody titers to HSP-60, 65, and -70 are significantly higher in patients with classical coronary risk factors than in those without such factors and are associated with dietary factors. A limitation of our study is its cross-sectional design. It is not possible to be certain that the associations between dietary constituents and HSP antibody titers are due to direct causation. In addition, our study group was a small heterogeneous sample of dyslipidemic patients. It would be important to confirm these findings in a larger sample and perhaps to examine in greater depth the relation in nondyslipidemic patients with metabolic syndrome. Nevertheless, our findings are consistent with previous studies in which a direct effect of antioxidants and fatty acids on HSP antibody titers was shown.
ACKNOWLEDGMENTS
The contributions of authors to this paper are as follows: design of the experiment (MGM, GAF, and SAN); recruitment of subjects (MGM, GAF, CL, and TW); measurement of high-sensitivity C-reactive protein (MGM and NV); measurement of antibody titers to heat shock proteins (MGM, DJL, and GAF); measurement of vitamin E (MGM and BJS); analysis of food-frequency questionnaires (MGM and SAN); statistical analysis (MGM, SAN, and GAF); and manuscript preparation (all). MGM is a scholar of the Iranian Government. DJL was supported by the British Heart Foundation and the University of Surrey. None of the authors had any conflict of interest.
REFERENCES