| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
-Defensin Is Associated with Cardiovascular Morbidity and Mortality in Type 1 Diabetic PatientsMedical Research Laboratories (G.J., T.K.H., A.F., J.F.), Clinical Institute and Medical Department M (Diabetes and Endocrinology), Aarhus University Hospital, and Faculty of Health Sciences (H.-H.P.), Aarhus University, DK-8000 Aarhus, Denmark; and Steno Diabetes Center (L.T., A.S.A., H.-H.P.), Gentofte Hospital, DK-2820 Copenhagen, Denmark
Address all correspondence and requests for reprints to: Dr. Jan Frystyk, Medical Research Laboratories, Aarhus University Hospital, Nørrebrogade 44, DK-8000 Aarhus C, Denmark. E-mail: jan{at}frystyk.dk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-Defensins are antimicrobial peptides of the innate immune system. In addition, experimental evidence suggests that -defensins are proatherogenic.
Objective: The objective of the study was to examine the predictive value of plasma -defensin as a clinical marker of cardiovascular disease (CVD) in patients with type 1 diabetes.
Methods: In an observational, prospective design, 389 patients with long-lasting type 1 diabetes were examined for CVD at study start (1993; baseline) and followed up through the Danish National Register for a median of 10.1 yr (range 0.2–10.4 yr). Plasma was collected in 1993 and stored at –80 C until analysis of plasma -defensin using an in-house RIA.
Results: At baseline, plasma -defensin was significantly higher in patients with than without nephropathy [median and interquartile ranges: 305 (205–321) vs. 223 (182–263) µg/liter; P < 0.0001]. During follow-up, 98 patients reached the primary end point (fatal and nonfatal events of CVD). Prospectively a baseline -defensin within the upper vs. the lower tertile significantly increased the covariate-adjusted risk for CVD-related morbidity and mortality to a hazard ratio of 2.8 (1.3–5.9) (median and 95% confidence intervals, P = 0.006).
Conclusion: This study suggests that plasma -defensin may serve as a clinical risk marker for CVD-related morbidity and mortality in type 1 diabetes. However, future studies are needed to clarify whether plasma -defensin is causally linked to the development of CVD or an innocent bystander.
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-Defensins are small, cystein-rich peptides of approximately 30 amino acids. Together with β- and 
-defensins, they constitute a family of evolutionary-related antimicrobial peptides of the innate immune system. The group of -defensins constitutes six distinct peptides, of which -defensin-1, -2, and -3 share many similarities. Their primary structure differs with one amino acid only, and all three peptides are synthesized constitutively and stored intracellularly in azurophil granules in bone marrow precursors of neutrophil leukocytes. Due to their origin, the -defensins are often referred to as human neutrophil peptides. Although -defensins are capable of stimulating the immune system at different levels, one of their main functions is to interact with cell membranes of invading organisms, causing disruption of ion fluxes and eventually cell lysis (1, 2, 3).
Chronic low-grade inflammation with elevated circulating levels of various cytokines and acute-phase reactants is a hallmark of obesity, insulin resistance, and type 2 diabetes, and this proinflammatory condition has been causally linked to the development of premature cardiovascular disease (CVD) (4, 5). In addition, an increased leukocyte count is a well-established risk factor for coronary heart disease (6), and recent data link this observation to the neutrophil-derived -defensins (7). In vitro experiments have shown that -defensins stimulate the binding of low-density lipoprotein cholesterol and lipoprotein(a) to cultured human umbilical vascular endothelial cells and vascular smooth muscle cells (8) as well as the cell matrix (9), and importantly this occurs without a concomitant increase in the degradation of the lipoprotein particles (8, 9). Furthermore, in human cerebral vessels obtained during autopsy, -defensins and lipoprotein(a) colocalized primarily within atherosclerotic areas with an intensity that correlated with the severity of the atherosclerotic changes (8). Finally, -defensins inhibit the fibrinolytic activity of tissue plasminogen activator in vitro, a finding that indicates that -defensins may also be important participants in thrombotic microvascular occlusions (10). In vivo data are sparse, but a recent study found a significant correlation between the deposition of -defensin in skin biopsies and the occurrence of coronary artery disease (CAD) as determined by elective coronary artery catheterization (7).
Although -defensins are kept primarily within phagocytic vacuoles, activation of neutrophils results in release of -defensins into the circulation and during severe infection levels may approach 200 mg/liter (11). However, even in the noninfected state circulating -defensins are easily detectable, with levels averaging approximately 200 µg/liter in healthy subjects (12). Still, there are no data on the possible relationship between the circulating baseline (i.e. noninfected) levels of the -defensins and the presence of CVD. In this context it is interesting that a recent study in severe obese, normoglycemic women found that massive weight loss caused by gastric bypass was associated with reduced plasma levels of -defensins as well as C-reactive protein and IL-6 (13).
Due to the potential involvement of -defensin in atherosclerosis and thrombosis, we found it of interest to measure baseline plasma concentrations of -defensins-1 to -3 in type 1 diabetic patients with and without vascular complications. We hypothesized that there was a correlation between the concentration of -defensins and the risk for CVD, and to test this hypothesis, we examined 389 type 1 diabetic subjects, who were prospectively followed up for an average of approximately 10 yr.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The study cohort included 389 type 1 diabetic patients, who during the period January 1, to December 31, 1993, were recruited from the outpatient clinic at Steno Diabetes Center (Gentofte, Denmark). Clinical data on the study population at baseline and during the approximately 10 yr of follow-up have been described previously (14, 15). Originally the study was designed as a case-control study, aiming to compare long-lasting type 1 diabetic patients with and without diabetic nephropathy. Therefore, at study start in 1993, the patients were selected so that about half of the cohort (n = 199) had diabetic nephropathy, defined as a persistent urinary albumin excretion (UAE) greater than 300 mg per 24 h in at least two of three consecutive 24-h urine collections, the presence of retinopathy, and absence of other kidney or renal tract disease. The remaining 190 patients had normoalbuminuria (UAE 
30 mg per 24 h) and were matched for age, sex, body mass index (BMI), and diabetes duration; patient characteristics are given in Table 1
. The estimated glomerular filtration rate (GFR) was calculated by the formula of Levey et al. (16) in all patients, whereas the measured GFR was determined by the 51Cr-EDTA-method but only in patients with nephropathy (17). To avoid the use of two differently based GFR values within the same statistical analysis, all correlations were based on estimated GFR data. The medical examination and history performed at baseline revealed that 25 of the participants (four with normoalbuminuria and 21 with nephropathy) had previously been diagnosed with either myocardial infarction (MI) or stroke, and accordingly, these patients were considered to suffer from manifest CVD.
|
Definition of end points during follow-up
The study cohort was followed up until December 31, 2003, or until death (n = 76) or emigration (n = 3). In January 2004 all patients were traced through the Danish National Register. The median time span until death, emigration, or registration in January 2004 was 10.1 yr (range 0.2–10.4 yr). The outcome parameter was a composite cardiovascular end point of cardiovascular death, a history of nonfatal MI including coronary reperfusion and coronary artery bypass grafting, and nonfatal stroke as well as amputation and vascular surgery caused by peripheral atherosclerotic disease as recently described (14). The study was approved the local ethics committee, and all patients gave their informed consent.
Defensin RIA
The plasma level of -defensin was determined by a novel, validated, in-house, solid-phase RIA based on a monoclonal antibody, which recognizes -defensin isoforms 1–3. Ninety-six-well breakable microtiter plates (catalog no. 473768; Nunc, Roskilde, Denmark) were incubated overnight at 5 C with 4 mg/liter antimouse IgG (catalog no. M8890; Sigma-Aldrich, Copenhagen, Denmark) dissolved in 15 mmol/liter sodium carbonate and 35 mmol/liter sodium hydrogen carbonate (pH 9.6). After washing [wash buffer: 50 mmol/liter Tris-HCl (pH 8.0), 0.9% (wt/vol) NaCl, 0.5% (vol/vol) Tween 20, and 0.05% (wt/vol) NaN3], all wells were blocked with 1% (wt/vol) BSA (Sigma-Aldrich) dissolved in 40 mmol/liter phosphate buffer and 0.05% (wt/vol) NaN3 (pH 8.0) and incubated for 3 h at room temperature. After washing, all wells were added 50 µl of standard (purified -defensin-1, catalog no. D2040; Sigma-Aldrich) or diluted plasma (1 in 25), 50 µl of 125I-labeled -defensin (
10.000 cpm) and 100 µl (200 µg/liter) of a specific monoclonal antibody, which recognizes -defensin-1, -2, and -3 (clone DEF 3, catalog no. T-1034; BMA, Augst, Switzerland). Iodinated -defensin-1 was prepared in-house using the chloramine-T method. All reagents were dissolved in 40 mmol/liter phosphate buffer containing 5.0% (wt/vol) BSA, 9% (wt/vol) NaCl, and 0.5% (wt/vol) Tween 20. All samples (standards and unknown samples) were analyzed in duplicates with the exception of nonspecific binding and the 0 standard, which were analyzed in quadruplicates. After addition of antigen, tracer, and antibody, the microtiter plates were incubated for 2 d at 5 C, washed three times, and counted in a 
-counter for 5 min.
-Defensin standards were made by serial dilution and ranged from 1 to 250 µg/liter and serum as well as plasma diluted in parallel with the standard curve (data not shown). The nonspecific binding averaged approximately 80 cpm, whereas the signals of the lowest and highest standards averaged approximately 2400 and approximately 400 cpm, respectively. Half-maximal binding occurred at a standard concentration of approximately 100 µg/liter. The lower detection limit was estimated to less than 1.95 µg/liter. Mean within-assay coefficient of variation of standards and samples was less than 6%. The in-between assay coefficient of variation was estimated by repetitive analysis of a control sample and averaged 9% (seven assays).
To study the antigen recovery, a preparation of -defensin (100 µL, 500 µg/liter) dissolved in assay buffer was incubated with an equal volume of EDTA plasma (n = 10) and measured against EDTA plasma added buffer without -defensin. Before assay, all samples were incubated at 5 C overnight to allow binding of -defensin standards to plasma proteins. All samples were assayed in two dilutions: 1:10 and 1:20. The recovery averaged 109 ± 4% (range 90–128%) and 106 ± 2% (range 98–119%) after a dilution of 1:20 and 1:10, respectively. Finally, we studied the effect of repetitive freezing and thawing (nine cycles) on serum and plasma levels of -defensin. However, repeated freezing and thawing did no affect the -defensin levels (data not shown).
Other measurements
Total cholesterol, highly sensitive C-reactive protein (hsCRP) and glycosylated hemoglobin (HbA1c) were determined by routine laboratory assays.
Statistics
For nonnormally distributed variables, values are given as median and interquartile ranges (IQR). All other values are given as means ± SD. Comparisons between groups were performed by Student’s unpaired t test, Mann-Whitney U test, or Kruskall Wallis test as appropriate. Spearman correlation with two-tailed probability values was used to estimate the strength of association between variables. Multiple linear regression and forward stepwise regression analyses were used to determine the strongest predictors of circulating -defensin, and a general linear model univariate procedure was used to investigate the association of -defensin with nephropathy.
The cumulative risk of death was calculated based on the entire follow-up period ending in 2004. Groups were compared using the log rank test and displayed on Kaplan-Meier plots. To evaluate the effect of baseline -defensin levels on the combined CVD end point and mortality, multivariate Cox proportional hazards regression analyses with manual stepwise backward removal of nonsignificant variables were performed. Included baseline variables were the presence of CVD at baseline, age (per 10 yr interval), gender, nephropathy status, smoking status, HbA1c (per 1% increase), systolic BP (per 10 mm Hg increase), BMI (per 2 kg/m2 increase), total cholesterol (per 10 mg/dl increase), estimated GFR (per 10 ml/min increase), and hsCRP tertiles.
Statistical significance was assumed for P < 0.05. All statistical calculations were performed with SPSS for Windows version 13.0 (SPSS, Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
In the whole study cohort (n = 389), the median baseline concentration of plasma -defensin was 258 (IQR 205–321) µg/liter. Patients with nephropathy had significantly higher levels of plasma -defensin, compared with those with normoalbuminuria: 305 (IQR 205–321) vs. 223 (IQR 182–263) µg/liter; P < 0.0001. This difference remained significant after adjustment for known risk factors for nephropathy (sex, diabetes duration, daily insulin dose, HbA1c, BMI, systolic BP, cholesterol, smoking status, hsCRP) as well as differences in GFR and UAE [estimated means (95% confidence interval [CI]): 289 (95% CI 275–304) vs. 263 (95% CI 247–278) µg/liter in nephropathy vs. normoalbuminuria; P = 0.03]. There was no gender-related difference in plasma -defensin: males 258 (IQR 206–333) vs. females 257 (IQR 202–306) µg/liter, P = 0.35, whereas smokers had slightly higher levels than nonsmokers: 262 (IQR 215–328) vs. 251 (IQR 200–307) µg/liter, P < 0.05.
A total of 25 patients (hereof 21 with nephropathy) were diagnosed with CVD at baseline, and their median plasma -defensin exceeded the levels observed in the remaining patients: 310 (IQR 243–442) vs. 258 (IQR 202–313) µg/liter; P < 0.005. The predictors of plasma -defensin were found by multiple linear regression analyses among the following variables: CVD at baseline, estimated GFR, systolic BP, diastolic BP, UAE, total cholesterol, BMI, age, diabetes duration, daily insulin dose, HbA1c, gender, and smoking status (Table 2). In the whole study cohort, plasma -defensin was strongest correlated to estimated GFR, and the same was true in patients with nephropathy. In normoalbuminuric patients, hsCRP appeared to be the strongest predictor.
|
During follow-up, 98 patients reached the primary end point (composite CVD), hereof 79 patients with nephropathy and 19 with normoalbuminuria at baseline. When compared with the remaining cohort, patients who reached the composite CVD end point had significantly higher baseline levels of -defensin: 311 (259–391) vs. 242 (193–300) µg/liter, P < 0.0001. A similar result was obtained after exclusion of the 25 subjects, who suffered from CVD at baseline: levels of -defensin averaged 315 (268–372) µg/liter in patients reaching the primary end point, compared with 239 (192–298) µg/liter in those who remained event free (P < 0.0001).
We also observed significant differences if patients with normoalbuminuria or nephropathy at baseline were analyzed separately (event vs. event free normoalbuminuric patients): 236 (214–298) vs. 220 (180–261) µg/liter, P < 0.05; (event vs. event free patients with nephropathy): 337 (281–417) vs. 295 (232–344) µg/liter, P < 0.0001.
When the entire cohort was divided according to tertiles of baseline -defensin, 43% of patients in the upper tertile reached the primary combined end point during follow-up, compared with 25% in the middle tertile and 8% in the lowest tertile (P < 0.0001, Fig. 1A).
|
-defensin tertiles yielded a covariate adjusted hazard ratio (HR) of 2.8 (95% CI 1.3–5.9; P = 0.006) for reaching the composite CVD end point (Table 3).
|
-defensin within the middle and upper tertile increased the risk for reaching the composite CVD end point significantly (Table 4). This was in contrast to what was observed in patients with normoalbuminuria, in whom the significant predictors included gender, systolic BP, and CVD at baseline (Table 5). The Kaplan-Meier plots showing the risk for the primary composite end point are shown in Fig. 1A
(all patients), Fig. 1B (patients with normoalbuminuria), and Fig. 1C (patients with nephropathy).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-defensin in the development of CVD (7, 8, 9, 10), and this prompted us to perform the first prospective investigation of the potential use of plasma -defensin as a new biomarker of CVD-related morbidity and mortality in a clinical setting. The study cohort comprised 389 well-characterized patients with long-lasting type 1 diabetes, who were examined with blood samples during 1993 (i.e. at baseline) and subsequently followed up for more than 10 yr (14). The study showed that elevated baseline levels of -defensin were associated with an increased risk for cardiovascular morbidity and mortality, even after adjustment for known confounders. Hence, our study supports the hypothesis that -defensins are involved in the processes leading to CVD.
The regression analyses indicated that a deteriorated kidney function was an important determinant of plasma -defensin levels. Firstly, GFR was found to be the most important predictor of plasma -defensin in patients with nephropathy as well as in the total cohort (Table 2). Second, -defensin levels were elevated in patients with nephropathy compared with those with normoalbuminuria. It has previously been reported that kidneys suffering from diabetic nephropathy have a reduced ability to degrade peptide hormones such as insulin and GH (18, 19), and this may therefore offer a possible explanation. Alternatively, it could be speculated that the urinary excretion of -defensins is reduced in patients suffering from nephropathy. However, it was recently reported (albeit in preliminary form) that -defensins are cleared from the circulation without being excreted in the urine (7).
The GFR-adjusted plasma levels of -defensin were higher in patients with nephropathy than patients with normoalbuminuria, indicating that the difference is not simply explained by variations in renal function. However, adjustment for GFR may not necessarily eliminate all confounders caused by differences in kidney function, and therefore, we cannot fully exclude that the elevation of plasma -defensin in patients with nephropathy is caused by other renal confounders than GFR.
The observation that -defensin is elevated in patients with nephropathy, compared with patients with normoalbuminuria, is in agreement with studies of other proinflammatory markers, i.e. hsCRP (20), mannose-binding lectin (21), soluble intracellular adhesion molecule-1, and plasminogen activator inhibitor type 1 (22). Having the proatherosclerotic properties of -defensin in mind, it is tempting to speculate that the elevated plasma levels of -defensin, in concert with other proinflammatory and proatherogenic factors may contribute to the markedly increased incidence of premature CVD in patients suffering from nephropathy (23).
In the present cross-sectional part of the study, we found elevated levels of plasma -defensins in patients diagnosed with CVD at study start. However, even more important, in the prospective part, an elevated plasma -defensin appeared to be an independent predictor of cardiovascular morbidity and mortality: during the 10 yr of follow-up, the risk for CVD and death increased approximately 3-fold when comparing subjects in the upper and lower tertile of plasma -defensin at baseline. Importantly, the significant relationship between -defensin and the composite end point was maintained after exclusion of the 25 patients who suffered from CVD at baseline (i.e. when they were included in the study). Thus, the present study supports previous experimental in vitro data showing that -defensins exert proatherogenic effects (7, 8, 9, 10).
Subgroup examination using the Cox proportional hazards regression analysis indicated that an elevated plasma -defensin was an independent risk marker in patients with nephropathy only. On the other hand, it should be stressed that less than 20% of the primary end points came from patients with normoalbuminuria, and therefore, in all probability, the few numbers of cases have yielded an insufficient statistical power.
Our data are in agreement with recent findings by Nassar et al. (7), who investigated the skin density of -defensin in 53 middle-aged men examined immediately before coronary artery catheterization. In that cohort, the prevalence of diabetes (type not specified), hypertension, previous MI, and smoking was about 50%, whereas there was no information on kidney function, which we found to be of importance. Nevertheless, the authors were able to demonstrate that the skin density of -defensin independently predicted the likelihood for CAD and was an even better independent predictor for CAD severity than age (7).
We fully acknowledge that our study has some limitations. First, the intraindividual variation over time in plasma -defensin is unknown and remains to be investigated. Second, the effect of long-term storage on plasma levels of -defensin is not known. However, all samples were collected and stored at –80 C within a 4-month period in 1993, and this eliminates the confounding effect of variable storage times. Furthermore, the peptide appears to be relatively stable because repeated thawing-freezing for up to nine cycles had no affect on -defensin levels, in neither serum nor plasma (data not shown). Third, the apparent role of -defensin as an independent risk marker in type 1 diabetic patients (and in particular in those suffering from nephropathy) cannot be extrapolated to other study cohorts without reservations, and therefore, there is a call for similar studies in for instance healthy subjects and patients with type 2 diabetes. Finally, due to the nature of the study design, our data do not allow us to conclude that the relationship between plasma levels of -defensin and the presence of CVD is casually linked; thus, future studies have to decide whether circulating -defensins are pathogenically linked to the development of CVD in vivo or merely an innocent bystander.
In conclusion, we show that the plasma levels of the proinflammatory and proatherogenic peptide -defensin are elevated in type 1 diabetic patients suffering from CVD and predict the future risk for cardiovascular morbidity and mortality.
|
Acknowledgments |
|---|
-defensin. |
Footnotes |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online January 22, 2008
Abbreviations: BMI, Body mass index; BP, blood pressure; CAD, coronary artery disease; CI, confidence interval; CVD, cardiovascular disease; GFR, glomerular filtration rate; HbA1c, glycosylated hemoglobin; HR, hazard ratio; hsCRP, highly sensitive C-reactive protein; IQR, interquartile range; MI, myocardial infarction; UAE, urinary albumin excretion.
Received August 27, 2007.
Accepted January 16, 2008.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
-Defensin: link between inflammation and atherosclerosis. Atherosclerosis 194:452–457[CrossRef][Medline]-defensins in plasma and BAL fluid of patients with active pulmonary tuberculosis. Chest 121:519–526[CrossRef][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
