UCSNP-43 has allelic frequencies of G = 0.62 and A = 0.38 in Pima Indians. Since the evidence for linkage in the NIDDM1 region in Mexican-Americans was attributable to the affected sib pairs homozygous for the high-frequency G allele (3
), the data for Pima Indians were analyzed similarly. Individuals homozygous for the UCSNP-43 G allele (GG) were compared with individuals not homozygous for the UCSNP-43 G allele (i.e., A/A and G/A). The two groups had similar ages (mean ± SE = 26 ± 1, 27 ± 1 years) and percentage of body fat (mean ± SE = 31 ± 1 for both groups). Metabolic characteristics of the two groups were compared after adjusting for age, sex, percentage of body fat, and nuclear family membership, since many of the subjects were siblings (Table ). People homozygous for the G allele had a higher mean fasting plasma glucose (P
= 0.01). In contrast, the mean fasting rate of glucose appearance, or endogenous glucose production, was lower in the G/G group (P
= 0.0004). The mean plasma insulin concentration 2 hours after ingesting 75 g glucose was higher in the G/G group (P
= 0.05), indicative of insulin resistance. The results of the euglycemic clamp studies confirmed this with lower mean glucose-disposal rates in the G/G group during the low-dose (physiologic) insulin infusion (P
= 0.006) and the high-dose (maximal) insulin infusion (P
= 0.05). The lower rate of glucose disposal appeared to be the result of a lower glucose-oxidation rate, since differences in carbohydrate oxidation were more significant (P
= 0.04 and P
= 0.009 for the low- and high-dose insulin infusion, respectively) than differences in nonoxidative glucose disposal (P
= 0.26 and P
= 0.23 for the low- and high-dose insulin infusion, respectively).
Differences in substrate oxidation rates between the genotype groups were also observed in data from the respiratory chamber (Table ). People homozygous for the UCSNP-43 G allele oxidized more lipid (P = 0.03) and less protein (P = 0.01) over 24 hours than people with the G/A or A/A allelic combinations (after adjusting for age, sex, percentage of body fat, and energy balance). The G/G group also oxidized less carbohydrate, but this difference (P = 0.81) did not reach statistical significance. The differences in substrate oxidation rates occurred despite all individuals having eaten a diet of similar energy and substrate content during the indirect calorimetric measurements. These results indicate significant differences in nutrient partitioning between the groups. In addition, the mean sleeping metabolic rate was lower in the G/G group (P = 0.01, after adjusting for age, sex, fat-free mass, fat mass, and nuclear family membership), most likely as a result of the lower rate of endogenous glucose production, which is an energy-costly process to the extent that it is due to gluconeogenesis. Note that whenever multiple phenotypes are analyzed, some may differ significantly by chance. The P values in this study were not adjusted for multiple testing, although three main effects were tested (glucose turnover, insulin secretion, and energy/substrate metabolism). To examine whether the effects of the UCSNP-43 G/G genotype on glucose turnover and nutrient partitioning might lead to an increased risk of type 2 diabetes, a random sample of 720 full-heritage Pima Indians was genotyped. Overall, there was no statistically significant association between the UCSNP-43 genotype and prevalence of diabetes (Table ).
In vitro evidence suggests that the nucleotide sequence encompassing UCSNP-43 may be involved in regulating CAPN10
). Therefore, we investigated the in vivo expression of CAPN10
mRNA in skeletal muscle, a major target for insulin action, in 18 male nondiabetic Pima subjects. Individuals homozygous for the G allele had 53% lower mean CAPN10
transcript levels compared with heterozygotes, whereas A/A homozygotes had the highest mean levels (Figure a). As predicted from the larger data set, individuals with lower CAPN10
transcription (G/G genotype) had reduced basal and insulin-stimulated carbohydrate oxidation rates, but perhaps due to small sample size (n
= 18), this association did not reach statistical significance. However, a significant correlation (r
= 0.79; P
= 0.003) was observed between the level of CAPN10
expression in skeletal muscle and the rate of carbohydrate oxidation measured in the respiratory chamber over 24 hours (Figure b).
Figure 1 Association between CAPN10 mRNA expression and UCSNP-43 genotype and 24-hour carbohydrate oxidation rates. (a) CAPN10 mRNA expression, as measured by real-time PCR, in skeletal muscle of nondiabetic males representing G/G (n = 8), G/A (n = (more ...)