点击显示 收起
Abstract |
---|
Top Abstract Introduction Methods Results Discussion References |
---|
Methods and Results— We studied 35 nondiabetic adult FPLD subjects (of whom 24 were women) with either the LMNA R482Q or R482W missense mutations and 51 matched normal first-degree relatives (of whom 27 were women). Compared with normal controls, LMNA mutation carriers had significantly higher plasma insulin and more dyslipidemia, higher mean triglycerides and lower HDL cholesterol, significantly higher nonesterified free fatty acids and CRP, and significantly lower leptin and adiponectin than controls. Subgroup analyses showed that these differences were more pronounced in women. Other biomarkers such as resistin, fibrinogen, and plasminogen activator inhibitor-1 were not different between groups.
Conclusions— LMNA mutations in nondiabetic patients with FPLD are associated with several metabolic and biochemical changes, particularly in women. The unfavorable profile might contribute to the increased susceptibility to CHD seen in LMNA mutation carriers.
Key Words: insulin resistance diabetes DNA analysis cytokines risk factor
Introduction |
---|
Top Abstract Introduction Methods Results Discussion References |
---|
Methods |
---|
Top Abstract Introduction Methods esults Discussion References |
---|
Biochemical and Genetic Determinations
Blood was centrifuged at 2000 rpm for 30 minutes, and the plasma was stored at -70°C. Assays of fasting plasma concentrations of lipids, lipoproteins (including direct measurements of LDL cholesterol after ultracentrifugation), apolipoprotein (apo) A1 and B, glucose, glycosylated hemoglobin, insulin, C-peptide, and leptin were performed by using established procedures.10,11 Fasting plasma insulin concentration was determined by radioimmunoassay (RIA, Pharmacia), which had a sensitivity of 0.4 ng/mL and intra-assay and interassay coefficients of variation of 5.2% and 8.7%, respectively. C-peptide concentration was determined by RIA (Diagnostic Products) with a minimal detection limit of 43 pmol/L and 0% cross-reactivity with insulin. Serum nonesterified FFA concentration was measured by using the ACS-ACOD kit (Wako), which had a detection limit of 0.01 mmol/L and a maximum coefficient of variation of 2.7%. Fasting plasma leptin concentration was determined by quantitative ELISA (Quantikine Human Leptin, R&D Systems), which had a minimal detectable concentration of 0.5 ng/mL, a limit of linearity of 100 ng/mL, and an interassay coefficient of variation of 8.3%. Fasting plasma adiponectin concentration was determined by RIA (Linco), which had a detection limit of 1 ng/mL and an interassay coefficient of variation of 5%. Plasma acylation-stimulating protein (ASP) concentration was determined by ELISA (Quidel), which had a detection limit of 200 ng/mL and an interassay coefficient of variation of 5%. Fasting plasma resistin concentration was determined by ELISA (BioVendor), which had a detection limit of 2 ng/mL and an interassay coefficient of variation of 7.2%. Serum tumor necrosis factor (TNF)- concentration was determined by ELISA (Biosource International), which had a maximum coefficient of variation of 8.2%. Serum CRP concentration was measured using a high-sensitivity automated nephelometric method (Beckman Coulter), which had a detection limit of 0.05 mg/L and a maximum coefficient of variation of 6%. Serum plasminogen activator inhibitor-1 (PAI-1) concentration was measured with the Stachrom PAI kit (Diagnostica Stago) that had a detection limit of 5 IU/mL and a maximum coefficient of variation of 4%. Serum FBG was analyzed according to the von Clauss method,27 which had a maximum coefficient of variation of 9.5%. DNA extraction and LMNA genotyping were performed as described,10,11 using sequence-proven DNA standards.
Statistical Analyses
Clinical and biochemical traits for LMNA mutation carriers were compared with matched family controls. Differences between genotypes in qualitative traits were compared with Fisher’s exact test in SAS version 6.12 (SAS Institute). All continuous variables assumed a distribution that was not significantly different from normal after logarithmic transformation. Differences between genotypes in logarithmically transformed quantitative traits were compared by Student’s t tests from the general linear models procedure in SAS. For the purpose of presentation, untransformed variables are shown in the tables. For all quantitative variables, the entire sample size afforded >75% power to detect a 1.6-fold difference and >90% power to detect a 2.0-fold difference (=0.05). Post hoc subgroup analyses according to sex were also performed. A value of P<0.05 was taken as the nominal level of significance.
Results |
---|
Top Abstract Introduction Methods Results Discussion References |
---|
|
Biochemical Phenotypes
Biochemical traits related to atherosclerosis (unadjusted mean±SE) are shown in In the entire study sample, there were numerous significant differences between FPLD subjects and normal family controls. For instance, among the plasma lipoproteins, mean triglyceride was significantly higher and mean HDL cholesterol was significantly lower in the FPLD subjects than in family controls. These significant differences were also seen in subgroup analysis according to sex and were consistent with the findings in Twenty-four of 24 dyslipidemic subjects defined by age- and sex-matched cut points had high triglyceride and low HDL cholesterol levels, whereas only 4 of 24 had elevated LDL cholesterol values (data not shown). A post hoc subgroup analysis showed that dyslipidemic FPLD subjects also had significantly lower leptin but higher FFA and insulin levels than did normolipidemic subjects (data not shown). Plasma concentrations of lipoprotein(a) and apo B were not different between FPLD subjects and controls
|
Among metabolic variables, mean insulin and C-peptide were significantly higher in FPLD subjects than in family controls In the subgroup analysis according to sex, insulin and C-peptide were significantly higher in female FPLD subjects than in controls, but there was no difference in males. These findings suggest that insulin resistance is more pronounced in nondiabetic FPLD women than men. In the overall sample, serum FFA concentrations were significantly higher in FPLD subjects than in family controls. In the subgroup analysis according to sex, this difference was significant in both men and women.
As previously observed, mean leptin28 and adiponectin29 concentrations were significantly lower in FPLD subjects than in family controls. In the subgroup analyses according to sex, these analytes were both significantly lower in female FPLD subjects than in controls, but there was no difference in men. Again, this was consistent with a more severe biochemical phenotype related to insulin resistance in nondiabetic FPLD women than men. No differences were seen for ASP, resistin, or TNF- between FPLD subjects and controls, both in the entire sample and in the subgroups divided by sex. Among inflammatory and coagulation markers, only mean CRP concentration was significantly higher in FPLD subjects than in family controls. Trends to higher mean CRP concentrations were seen in FPLD subjects of both sexes, but again the magnitude of the difference appeared to be more pronounced in nondiabetic FPLD women. Post hoc pairwise correlation analysis showed significant correlations between insulin or C-peptide and both FFA and CRP. For instance, in the overall sample, C-peptide was correlated with both FFA and CRP, with Pearson coefficients of 0.21 (P=0.038) and 0.50 (P<0.0001), respectively. Similar correlations were seen between insulin and these analytes, both overall and in subgroups divided by sex (data not shown). There were no differences in mean concentrations of PAI-1 or FBG in the entire sample or in the subgroups.
Discussion |
---|
Top Abstract Introduction Methods Results Discussion References |
---|
The association with CHD in FPLD is consistent with the epidemiological association of CHD with common hyperinsulinemia.30 The basis for the increased risk of CHD in the common insulin resistance syndrome is complex, with contributions from several different component metabolic abnormalities.30 The cardinal plasma lipoprotein abnormality in both common insulin resistance and FPLD is elevated triglycerides and depressed HDL cholesterol, with little or no increase in plasma LDL cholesterol or apo B.10,11 Absence of an association of FPLD with lipoprotein(a) was consistent with the absence of an association of lipoprotein(a) with common insulin resistance. Alternatively, because lipoprotein(a) is so strongly genetically determined and widely variable in concentration, the sample size might have been too small to detect a difference. Future analyses of lipoprotein subfractions are planned for these kindreds. The presence of the high triglyceride/low HDL cholesterol profile in a monogenic syndrome of high CHD risk is relevant, because fibrate treatment to normalize this profile does not affect LDL cholesterol and still reduces CHD risk.31
Because an elevated FFA level is a feature of common insulin resistance,32,33 it was important to document that FFA was elevated in FPLD, both in the overall sample and the subgroups divided by sex. FFA was correlated with insulin and C-peptide concentrations in this study. Elevated FFA underlies the insulin resistance syndrome through various mechanisms,32,33 although the basis for increased FFA in FPLD is probably different from that in common insulin resistance. In this regard, it may be of interest that FFA was increased in male FPLD subjects, whose insulin and C-peptide concentrations were not significantly increased. Nonesterified FFA acts as both a metabolic substrate and a signal controlling glucose utilization in muscle and liver and insulin metabolism in pancreatic beta cells.32,33 In FPLD, the absence of peripheral subcutaneous fat results in lipids being targeted for deposition in central fat stores. FFA is inextricably associated with the insulin resistance associated with central obesity, because central adipocytes are resistant to insulin action and portal FFA delivery to the liver increases when visceral triglyceride stores increase.32,33 In addition, the increased muscle triglyceride stores in FPLD could interfere with insulin action or glucose utilization, although such a causal relationship has not been demonstrated.32,33 Furthermore, increased FFA and prolonged dyslipidemia might be toxic to pancreatic beta cells, and this could predispose to pancreatic failure, leading to diabetes. In the subgroup analysis of men, the presence of significantly increased FFA in FPLD, with no significant difference in insulin or C-peptide, suggests that FFA elevation precedes the onset of hyperinsulinemia in affected subjects.
The adipocyte plays a key role in the development of insulin resistance.34 The finding of decreases in both leptin and adiponectin in FPLD subjects, particularly women, confirms previous observations.28,29 The significant decrease in these 2 adipocyte-derived hormones in FPLD is consistent with evidence that links leptin and adiponectin to the metabolic consequences of lipodystrophy. For instance, leptin and adiponectin administered together corrected insulin resistance in a mouse model of lipodystrophy.35 Also, a decrease in adiponectin often precedes the development of insulin resistance.36 A differentiating feature between FPLD and common insulin resistance is that although both are characterized by low adiponectin, FPLD is characterized by low leptin due to an absolute deficiency of subcutaneous adipose stores, whereas common insulin resistance is characterized by increased leptin due to increased subcutaneous adipose stores.37 The absence of differences in ASP and resistin between FPLD and controls suggests that these analytes might be less strongly related to human insulin resistance.23
The other novel finding was the presence of a significantly increased serum CRP concentration in nondiabetic FPLD subjects in the overall sample. These trends were also seen in the subgroup analyses according to sex, particularly in women. CRP was positively correlated with insulin and C-peptide concentrations in this study. Prospective epidemiological studies identified high-normal or increased serum CRP as an independent predictor of CHD.12–20 Furthermore, among healthy, nondiabetic subjects, elevated CRP was associated with measures of insulin sensitivity, suggesting that chronic subclinical inflammation was part of the insulin resistance syndrome.22 However, until now there had been little information about serum inflammatory markers, particularly CRP, in monogenic syndromes of insulin resistance. The present findings suggest that increased serum CRP can be associated with central fat exclusively, because peripheral subcutaneous fat is absent in FPLD subjects and total body fat content was reduced, as indicated by decreased serum leptin. It is also possible that CRP is primarily elevated in FPLD owing to defective LMNA and its subsequent effects on adipocyte biology and chronic inflammation.
The specific association between increased CRP and FPLD is highlighted by the absence of significant differences for other acute-phase response markers. For instance, TNF-, a cytokine produced by adipose tissue that may also play an important role in insulin resistance,38 was not increased in FPLD subjects. Also, interleukin-6, which plays a key role in adipocyte biology,39 was not different in FPLD subjects compared with controls in a preliminary analysis (data not shown). Furthermore, PAI-1 and FBG, procoagulant markers whose concentrations were previously shown to be elevated in common insulin resistance,21,40,41 were not elevated in FPLD. However, in FPLD, only CRP was significantly increased, to levels consistent with those associated with increased CHD risk in populations,42 with no differences in the measured cytokines, such as TNF-, or procoagulants, such as PAI-1 and FBG. Studies in populations have indicated that abdominal obesity with insulin resistance is associated with increases in both CRP43,44 and PAI-1,40 which contrasts with the increase in CRP only in FPLD. Although elevated CRP is also associated with specific clinical components of the common insulin resistance syndrome, such as blood pressure,45 any independent association of elevated CRP with particular biochemical and/or clinical characteristics of FPLD would be difficult to ascertain in this relatively small study sample.
In summary, nondiabetic FPLD resulting from mutant LMNA was associated with hyperinsulinemia, high triglycerides, low HDL, hypoleptinemia, hypoadiponectinemia, elevated FFAs, and elevated CRP but not with changes in LDL cholesterol, apo B, lipoprotein(a), resistin, ASP, TNF-, PAI-1, or FBG. The metabolic disturbances in nondiabetic FPLD subjects, particularly in women, were all deleterious, proatherosclerotic, and consistent with the increased CHD risk seen in Canadian FPLD families with mutant LMNA,10,11 particularly in women. Our findings indicate that elevations in specific analytes, such as FFA and CRP, can occur relatively early in the clinical course of FPLD. Furthermore, the elevated FFA in FPLD was consistent with the elevated FFA in common insulin resistance. However, the specific elevation in CRP, without changes in the other cytokine and procoagulant markers typically associated with common insulin resistance, suggests that alterations in such variables are not always correlated and interdependent. Perturbations in each variable might follow a unique time course under the influence of distinct determinants, at least in monogenic syndromes such as FPLD.
Acknowledgments |
---|
Received September 19, 2002; accepted November 4, 2002.
References |
---|
Top Abstract Introduction Methods Rsults Discussion References |
---|