Peripheral Blood Mitochondrial DNA Content Is Related to Insulin Sensitivity in Offspring of Type 2 Diabetic Patients

Subjects were selected by random cluster sampling at Mokdong, which has a population of ∼22,000, located in the southwestern part of Seoul (12). We selected 13 of 124 apartment complexes. Of a total 1,804 eligible apartment residents, 750 underwent a 75-g oral glucose tolerance test and were classified as having normal glucose tolerance (NGT), impaired glucose tolerance (IGT), or newly diagnosed diabetes based on the World Health Organization criteria (13). Among the 750 subjects examined, 82 offspring of type 2 diabetic patients as well as 52 age-, sex-, and BMI-matched NGT subjects without a family history of diabetes were selected by questionnaire. Informed consent was obtained from all subjects. The protocol used was approved by the ethics committee of Ewha Women’s University Hospital.

In all subjects, blood pressure, height, weight, and waist and hip circumferences were measured. Total body fat content was measured as fat mass (kg) and percent body fat (%) using a bioelectric impedance analyzer. Cross-sectional body fat areas of two sites (subcutaneous abdomen and intra-abdomen) were measured by computed tomography at the umbilical level using an established method (14,15).

To determine the insulin sensitivity index (S_I), glucose-dependent glucose elimination (glucose effectiveness [S_G]), and insulin secretion, the insulin-modified frequently sampled intravenous glucose tolerance test (FSIGT) was used, as described previously (15,16). Glucose (0.3 g glucose/kg body wt) was administered and insulin (0.0125 U/kg body wt) was infused after glucose. S_I and S_G were calculated using a MINMOD computer program (Version 3.0; University of Southern California, Los Angeles, CA). Acute insulin response to glucose (AIR) was calculated as the mean difference between the insulin level at 2, 3, 4, 5, 6, and 8 min and the basal level. Serum glucose concentrations were determined using the glucose oxidase method. Total cholesterol, triglyceride, and HDL cholesterol levels were determined by enzymatic methods using a Hitachi 7150 autoanalyzer (Hitachi, Tokyo). Nonesterified fatty acid levels were measured by the enzymatic method. Commercial radioimmunoassay kits were used to measure serum insulin, C-peptide (Diagnostic Products, Los Angeles, CA), and serum leptin (Linco Research, St. Charles, MO).

The total DNA was extracted from peripheral blood leukocytes, and its concentration was measured with a spectrophotometer. The mtDNA content was measured by a real-time polymerase chain reaction (PCR) method using an ABI Prism 7700 (PE Biosystems, CA) as previously described (17). The mtDNA quantity was corrected by simultaneous measurement of the nuclear DNA, 28S rRNA. The mtDNA- and 28S rRNA-specific fluorescent probes were labeled internally using the fluorescent dyes 5-carboxyfluorescein (FAM) on the 5′ end, and 6-carboxy-tetramethyl-rhodamine (TAMRA) on the 3′ end. The primers for mtDNA were mt2981 (5′ACGACCTCGATGTTGGATC3′) and mt3245 (5′GCTCTGCCATCTTAACAAACC3′) and were used together with the mtDNA probe (5′FAM-TTCAGACCGGAGTAATCCAGGTCG-TAMRA3′ made to nt 3071-3095). The primers of 28S rRNA were 28S-7358 (5′TTAAGGTAGCCAAATGCCTCG3′) and 28S-7460 (5′CCTTGGCTGTGGTTTCGCT3′) and the 28S rRNA probe was 5′-FAM-TGAACGAGATTCCCACTGTCC-CTACCTACTATC-TAMRA3′ made to nt 7408-7440.

The PCR was performed separately for mtDNA and the reference 28S rRNA amplification. The PCR mixture contained the primers (10 pmol each), 200 nmol/l Taqman probe, 200 nmol/l dATP, 200 nmol/l dCTP, 200 nmol/l dGTP, 400 nmol/l dUTP, 4.5 mmol/l MgCl₂, 1.25 U AmpliTaqGold polymerase, 0.5 U AmpErase Uracil N-glycosylase, and 1 × PCR buffer A. Amplification was performed under the following conditions: one cycle of 50°C for 2 min, one cycle of 95°C for 10 min, and 40 cycles of 95°C for 15 s and 60°C for 1 min. The amount of starting target in a particular reaction mixture was measured by interpolation from a standard curve of Ct values generated from known concentrations of the standard DNAs.

All results are expressed as means ± SEM. Statistical analysis was performed using SPSS software for Windows (SPSS, Chicago, IL). Results of the control subjects and the offspring were analyzed using Students’ t test and analysis of variance. Pearson’s correlation and multiple linear regression analysis were used to evaluate the relationships among mtDNA content and the indexes of insulin secretion and sensitivity and the parameters of insulin resistance syndrome. Results were considered significant when the corresponding P value was <0.05.

According to the criteria (13), 52 of the 82 offspring of type 2 diabetic patients had NGT, 21 had IGT, and 9 had newly diagnosed type 2 diabetes. The clinical characteristics of the subjects are shown in Table 1. Age, sex distribution, BMIs, and lipid profiles were not significantly different for the control subjects and the NGT offspring. In diabetic offspring, age, BMI, waist-to-hip ratio, and fasting glucose, fasting insulin, and cholesterol levels were significantly higher than in the control NGT and IGT offspring. Whereas visceral fat areas and visceral-to-subcutaneous fat areas increased, the insulin sensitivity index was significantly lower in diabetic subjects than in the other groups. AIR was significantly lower in NGT and IGT offspring of type 2 diabetic patients than in the control subjects, and it was lowest in the offspring with diabetes (Table 2).

The pb-mtDNA/28S rRNA content was ∼17% lower in the offspring of type 2 diabetic subjects (1,084.7 ± 62.6, n = 82) than in the control subjects (1,304.0 ± 99.2, n = 52) at P = 0.051. When the offspring group was divided into three groups, control and NGT subjects were found to have significantly different mtDNA levels (P < 0.05) (Table 1). Interestingly, pb-mtDNA levels tended to increase in those in whom glucose intolerance or diabetes developed. Because the insulin resistance parameters were not significantly different for IGT and NGT offspring, the two groups were combined and compared with the control subjects in terms of mtDNA content; the difference between the control and NGT+IGT subjects was still statistically significant (P < 0.05) (Fig. 1).

When pb-mtDNA content was plotted against either clinical parameters or indexes of insulin resistance, significant correlations were observed between mtDNA content and either the logarithmically transformed insulin sensitivity index (r = 0.286, P < 0.05) or the fasting C-peptide concentration (r = −0.248, P < 0.05) in all subjects. Correlation between mtDNA content and the visceral fat area showed marginal significance (r = −0.198, P = 0.057). On the other hand, BMI, waist-to-hip ratio, fasting glucose, lipid levels and body fat content, fasting insulin, and AIR did not correlate with pb-mtDNA content. In offspring of type 2 diabetic patients, diastolic blood pressure correlated negatively with mtDNA content (r = −0.209, P = 0.061). Because pb-mtDNA is believed to be influenced by the disease state, data analysis was performed separately on the control subjects and the offspring. The pb-mtDNA content correlated negatively with age (r = −0.316, P < 0.05) in the control subjects, whereas the correlation was blurred in offspring of type 2 diabetic patients.

Pb-mtDNA of the offspring correlated positively with logarithmically transformed insulin sensitivity (r = 0.244, P < 0.05) (Table 3); the linear relationship was maintained in offspring with NGT and IGT (r = 0.253, P < 0.05) (Table 3) but not in offspring with diabetes. Stepwise regression analysis selected visceral fat area and mtDNA content as the main predictors of insulin sensitivity in the combined group of offspring with NGT and IGT (Table 4).

This study shows for the first time that pb-mtDNA content is lower in the offspring of type 2 diabetic patients than in age-, sex-, and BMI-matched control subjects without a family history of diabetes. Along with our previous finding that decreased pb-mtDNA content precedes the onset of diabetes (7), the present data support the notion that quantitative decreases of pb-mtDNA are a determinant of type 2 diabetic development. Although pb-mtDNA content shows a tendency of increasing as glucose metabolism deteriorates, this was without statistical significance. The fact that pb-mtDNA content is lower in the offspring of diabetic patients suggests that mtDNA copy number must be inherited. A previous study showed that cord blood mtDNA correlates positively with maternal pb-mtDNA content (18), which also supports the inheritance of mtDNA copy number. Considered together, these findings suggest that there is a heritable factor controlling mtDNA level.

Aging is an another factor known to influence the mtDNA content of tissues (7,19,20,21). Aging is known to be associated with deletions and mutations of mtDNA resulting from the combined effects of intense oxidative damage (20) and the low efficiency of the mtDNA repair system (21). Age-dependent decline of pb-mtDNA was observed in the present study as well as in our previous study (7).

Although mtDNA content is supposed to be related to age and be under genetic control, the mechanisms regarding regulation of mtDNA content are not entirely clear. Tfam is known to play a role in setting the mtDNA status quantitatively (by interacting with mitochondrial single-stranded binding protein and DNA polymerase γ) and in maintaining the normal glucose level (4). Therefore, the quantitative and qualitative aspects of Tfam must be investigated to clarify the mechanism that controls mtDNA content.

The increase of mtDNA content between offspring with NGT and offspring with diabetes is quite interesting because previous studies have reported that the mtDNA content in muscle (6) and peripheral blood (7) of diabetic patients is reduced. It is possible that the low level of mtDNA causes development of diabetes and that they try to compensate for their lack of mtDNA by enhancing mitochondrial biogenesis. We also have unpublished observations that subjects with diabetic complications show atypical increases in mtDNA copy number. Further studies are needed to clarify the regulation of mtDNA content in response to the state and duration of diabetes.

Although the depletion of mtDNA could impair mitochondrial and pancreatic β-cell function, mitochondrial function can be measured only in the tissues of patients or animals (4,6) and cannot be measured easily in noninvasive samples or in nondiseased subjects. It is believed, therefore, that the pb-mtDNA content could provide an alternative index. An interesting finding of the present study is that the pb-mtDNA content correlates with the insulin sensitivity of the offspring. Fasting serum C-peptide levels correlated negatively with the pb-mtDNA content of the subjects, suggesting that decreased mtDNA content is associated with insulin resistance in offspring. However, mtDNA content did not correlate with either fasting blood glucose or waist-to-hip ratio, and the pb-mtDNA content correlated negatively with diastolic blood pressure, as reported previously in a population-based study (7). The relationship between pb-mtDNA content and the energy utilization pattern of the whole body (22) also suggests that the quantitative status of mtDNA could be an indicator of insulin sensitivity. In summary, we have observed that the pb-mtDNA content decreased in the offspring of type 2 diabetic patients and, furthermore, that the pb-mtDNA content is correlated with insulin sensitivity. These findings suggest that the quantitative mtDNA status might be an early genetic marker for type 2 diabetes and possibly for insulin resistance syndrome.

This study was supported by the Korean National Institute of Health Grant 00-4-5 (to Y.K.P.), a grant from the Good Health 21 Program, Ministry of Health and Welfare Grant HMP-97-M-3-0045 (to Y.S.), and a MOSR/KISTEP grant (to K.S.P. and H.L.)