Haplotype combination 112/121 and its intrinsic variants (UCSNP43, -19, and -63) identified within the calpain 10 gene are associated with increased risk of type 2 diabetes in Mexican-Americans. We evaluated whether this haplotype combination and its constituent haplotypes and variants contribute to increased susceptibility to impaired fasting glucose (IFG)/impaired glucose tolerance (IGT) and type 2 diabetes in a South Indian population. Two study groups were used: 95 families ascertained through a proband with type 2 diabetes and 468 subjects recruited as part of an urban survey (69.1% with normal glucose tolerance, 12.8% with IFG/IGT, and 18.2% with type 2 diabetes). The four-locus haplotype combination 1112/1121 (UCSNP44, -43, -19, and -63) in South Indians conferred both a 10.7-fold increased risk for IFG/IGT (P = 0.001) and a 5.78- to 6.52-fold increased risk for type 2 diabetes in the two study groups (families P = 0.025, urban survey P = 0.015). A combination of the 1112 haplotype with the 1221 haplotype also appeared to increase risk for both IFG/IGT and type 2 diabetes. Contrary to what might be expected, quantitative trait analysis in the families found that transmission of the disease-related 1121 and 1112 haplotypes was associated with a reduced hip size and lower waist-to-hip ratio, respectively. This study supports the paradigm that specific haplotype combinations of calpain 10 variants increase risk of both IFG/IGT and type 2 diabetes. However, the relative infrequency of the “at-risk” combinations in the South Indian population suggests that calpain 10 is not a common determinant of susceptibility to type 2 diabetes.

The identification of susceptibility genes responsible for the genetic component of type 2 diabetes could greatly assist in the elucidation of the underlying pathophysiological mechanisms leading to disease and is central to the development of more effective preventative and therapeutic strategies for this condition. In the first reported genome-wide scan for type 2 diabetes genes, Hanis et al. (1) identified a 12-cM interval in Mexican-Americans at the distal telomeric region of chromosome 2q37 that potentially concealed a type 2 diabetes susceptibility gene, NIDDM1. An epistatic interaction between chromosome 15 and NIDDM1 subsequently reduced the interval (2) and enabled the successful mapping of a putative type 2 diabetes susceptibility gene, calpain10 (CAPN10) to the NIDDM1 region (3). The CAPN10 gene encodes a protein that is a member of the calpain-like cysteine protease family and is ubiquitously expressed in humans. Single-nucleotide polymorphisms (SNPs) in the CAPN10 gene were shown to be associated with an increased risk of type 2 diabetes in both Mexican-Americans and Botnian Finns (3).

The strongest association, with evidence of linkage, was with the common G allele (also denoted allele 1) of an SNP, UCSNP43 (G/A), in intron 3. Differences in the DNA binding affinity of nuclear proteins and in the transcription activity of an intronic fragment encompassing allelic variants of both UCSNP43 and another SNP 11 bp upstream, UCSNP44 (T/C), suggest these polymorphisms may play a role in the regulation of expression of CAPN10 or another gene located nearby. The uncommon T allele of another intronic variant, UCSNP63, was also found to be significantly associated with type 2 diabetes in the Botnian Finns but not in the Mexican-Americans or Germans. Although no association with Pima Indians was found between type 2 diabetes and the CAPN10 UCSNP43 polymorphism, normoglycemic subjects homozygous for the G allele of UCSNP43 were found to have reduced glucose oxidation rates and reduced skeletal muscle CAPN10 mRNA levels (4). The greatest risk of type 2 diabetes in Mexican-Americans was defined by a combination of haplotypes constructed from three of the identified SNPs, UCSNP43, -19, and -63, with the heterozygote haplotype 112/121 combination conferring an overall 3.0-fold increase in risk for type 2 diabetes (3). Furthermore, the same haplotype combination, though not statistically significant, was associated with a 2.6- and 5.0-fold increase in risk in Botnian Finns and Germans, respectively (3). The 112/121 haplotype combination was estimated to account for 14% of the population-attributable risk for type 2 diabetes in Mexican-Americans but only 4% in Europeans, due to the relative infrequency of the 112 haplotype in Europeans.

The aim of this study was to investigate whether the calpain10 gene polymorphisms UCSNP44, -43, -19, and -63, either individually or as haplotypes/haplotype combinations, contribute to the susceptibility of type 2 diabetes or related intermediate traits in a South Indian population.

South Indian families.

A total of 95 South Indian families were recruited from a diabetic clinic in Chennai, India. Ascertainment was via an offspring with type 2 diabetes, as defined by recent World Health Organization (WHO) criteria and as previously described (5,6). Of the probands, 63.5% were male, with a mean age of onset for type 2 diabetes of 34 years (range 29–38), a mean BMI of 26.5 ± 4.5 kg/m2, and a mean waist-to-hip ratio (WHR) of 0.94 ± 0.05. Clinical details for female probands were: a mean age of onset of type 2 diabetes of 32 years (range 23–37), a mean BMI of 27.8 ± 4.6 kg/m2, and mean WHR of 0.86 ± 0.07. A total of 85.2% of the male and 71% of the female probands were taking oral hypoglycemic agents (OHAs), whereas 7.4% of the male probands and 22.6% of the female probands were on insulin therapy. The mean age of the fathers (n = 95) was 65.4 ± 7.6 years, with a mean BMI of 25.3 ± 9.6 kg/m2 and a mean WHR of 0.96 ± 0.05; the corresponding figures in the mothers (n = 95) were 58.5 ± 7.1 years, 25.7 ± 4.3 kg/m2, and 0.88 ± 0.07, respectively. In the parental group, 69.6% had type 2 diabetes, as did 21% of co-siblings of the probands. Mature-onset diabetes of the young was excluded in families if the proband had an age of onset of diabetes before 25 years and an autosomal-dominant history of diabetes in two generations. Type 1 diabetes was excluded on clinical grounds. None of the diabetic subjects in the family had a history of ketoacidosis or ketonuria, an acute onset of symptoms, or weight loss before diagnosis, nor was there a family history of type 1 diabetes. Furthermore, if the proband was insulin treated, then this did not commence within a year of diagnosis. None of the probands carried the mitochondrial mt3243 mutation (7).

South Indian cross-sectional urban survey.

DNA samples were available from 468 South Indian subjects initially recruited as part of an urban population-based survey of the prevalence of type 2 diabetes and associated risk factors, using a cluster analysis design across all socioeconomic groups (8). Of these subjects, 48.6% were female, with mean age of 42 years (range 34–53), a mean BMI of 23.1 ± 4.6 kg/m2, and a mean WHR of 0.85 ± 0.06. The other 51.4% were male, with a mean age of 47 years (40–55), a mean BMI of 22.0 ± 3.7 kg/m2, and a mean WHR of 0.91 ± 0.06. In the total study population, 323 (69.1%) subjects had normal glucose tolerance, 60 (12.8%) had impaired glucose tolerance (IGT) and impaired fasting glucose (IFG), and 85 (18.1%) had type 2 diabetes. In both the family study and the urban survey, glucose tolerance was defined by the most recent WHO criteria (9). Male IFG/IGT subjects had a mean age of recruitment of 52 ± 15.3 years, a mean BMI of 22.5 ± 3.3, and a WHR of 0.92 ± 0.05; the corresponding figures in female subjects were 44.9 ± 12.5, 24.4 ± 4.9, and 0.86 ± 0.06, respectively. Male type 2 diabetic subjects were recruited at a mean age of 55 ± 10.8 years, with a mean BMI of 23.4 ± 3.3 kg m2 and a WHR of 0.94 ± 0.05; the corresponding figures in females were 50.7 ± 10.4, 24.1 ± 4.1, and 0.87 ± 0.04. A total of 38% of the type 2 diabetic subjects were newly diagnosed, and 62% had established diabetes at the time of the survey. In the latter group, 15% were on diet therapy, 83% on OHAs, and 2% on insulin.

Genotyping.

The genotyping was essentially carried out as previously described (1). UCSNP43 and UCSNP44 were both genotyped by mutagenically separated PCR, UCSNP19 (insertion/deletion) by size separation, and both UCSNP63 and UCSNP110 by PCR-restriction fragment-length polymorphism using the restriction enzyme HhaI. All fragments were analyzed on Metaphor agarose (Flowgen, U.K.) and visualized using ethidium bromide, and an Alpha Innotec digital camera system (Flowgen, U.K.).

Paternity testing.

Families were previously typed with 5 multiallelic markers from two chromosomes and 11 biallelic loci (including this study) from 4 different chromosomes. Families suspected of having an ex-paternity member were tested with the paternity/forensic testing kit PowerPlex 16 (Promega, Madison, WI), which uses a panel of 15 short tandem repeat markers and amelogenin for sex specification, and then typed using an ABI 310 genetic analyzer (ABI, Foster City, CA).

Statistical analysis.

In the families, individual SNPs were tested for evidence of association with diabetes using the transmission disequilibrium test (TDT) as implemented with the ETDT program (11). Transmission of haplotypes by TDT was investigated using the TRANSMIT program (available online from www.hgmp.mrc.ac.uk) (12). In the unrelated urban cohort, standard contingency table analyses were performed; if an SNP genotype homozygous frequency was <5% in any of the polymorphisms, then the low-frequency homozygotes were pooled with the heterozygotes. Global P values were calculated for haplotypes and haplotype combinations using Fisher’s exact test. Relationships between genotypes, haplotypes, and quantitative traits in the urban survey were examined by ANOVA in SPSS for Windows (version 10).

Quantitative TDT analyses were undertaken in the families for both individual polymorphisms and haplotypes using the orthogonal method of Abecasis, as implemented in the QTDT program (available online from www.well.ox.ac.uk/asthma/qtdt) (13). The “total association” option in QTDT was also used to implement a non-TDT association test that incorporates multiple members of the same pedigree, while allowing for shared polygenic and environmental variances.

Haplotypes unambiguously generated by the TRANSMIT program in the families allowed unambiguous assignment of haplotypes to the unrelated urban survey subjects. Haplotype combination frequencies were derived and odds ratios (ORs) and 95% CIs computed using a two-way contingency table method (14), with significance levels calculated using a Yates-corrected χ2 (1df) where appropriate. To detect transmission distortion of a haplotype combination, we counted the number of transmissions of a haplotype combination to affected offspring as a proportion of the total number of parents who could have potentially passed the combination to their offspring (i.e., where one parent had at least one haplotype and the other parent had the other haplotype). Deviations from Hardy-Weinberg equilibrium were tested using the χ2 goodness-of-fit test. Linkage disequilibrium (LD) between four calpain 10 loci was assessed using the Estimated Haplotypes program (15).

Allele/haplotype frequencies.

Allele frequencies in the urban survey for the four bialleleic variants were (allele1/2), 0.81/0.19 for UCSNP44, 0.86/0.14 for UCSNP43, 0.44/0.56 for UCSNP19, and 0.97/0.03 for UCSNP63. Similarly, in the parents from the families, the allele frequencies, as assessed by gene counting, were 0.85/0.15 for UCSNP44, 0.89/0.11 for UCSNP43, 0.41/0.59 for UCSNP19, and 0.97/0.03 for UCSNP63.

Three loci, UCSNP44 (P = 0.96), UCSNP43 (P = 0.84), and UCSNP63 (P = 0.36), were all in Hardy-Weinberg equilibrium, whereas UCSNP19 was not (P = 0.009). This deviation of UCSNP19 was not caused by inbreeding, since it resulted from increased heterozygosity rather than homozygosity. Pairwise LD statistics are presented in Table 1. UCSNP43, -44, and -19 are all in tight LD with each other. UCSNP43 and -44 are in weak LD with UCSNP63, whereas UCSNP19 is in tight LD with UCSNP63; this is consistent with their respective map positions. We detected 8 four-locus haplotypes of a possible 16; 5 were found in the South Indians at frequency >1% (Table 2). In agreement with the British/Irish study (10), all South Indian subjects positive for the UCSNP44 variant were also positive for the UCSNP110 variant; therefore, the latter was not included in the haplotype.

Association studies.

Family-based association studies in South Indians did not reveal any excess transmission from heterozygous parents for any of the four individual SNP alleles to probands with type 2 diabetes (Table 3). Analysis by TRANSMIT also found no evidence of excess transmission of any haplotype to type 2 diabetic offspring (global P = 1.0), including both the 1121 (P = 0.18) and 1112 (P = 0.56) haplotypes (full details are shown in an online Appendix at http://diabetes.diabetesjournals.org).

In the urban survey, there was no difference between the genotype frequencies of normoglycemic control subjects, IFG/IGT, and type 2 diabetes for the UCSNP44, -43, and -19 variants. However, the presence of the uncommon allele 2 (T) of UCSNP63 was significantly increased in IFG/IGT subjects, with a frequency of 17% compared with control subjects 4% (P = 0.001). Although there was only a slight nonsignificant increase in frequency of allele 2 in the unrelated type 2 diabetic subjects (5.9%), there was a significant increase (11%) in the type 2 diabetic probands (P = 0.02). The genotype frequency of the other three variants in the probands was similar to that in normoglycemic subjects.

Analysis of the five common haplotypes found the frequency of the 1112 haplotype also significantly increased in both the urban IFG/IGT subjects (global χ2P = 0.001, 4 df, 99% CI 0.0001–0.001) and the probands (global P = 0.004, 4 df, 99% CI 0.002–0.005). None of the other observed haplotype frequencies in type 2 diabetic subjects were significantly different from those of control subjects (Table 4).

Haplotype combinations.

Global χ2 comparisons of the 14 identified haplotype combinations between the nondiabetic subjects and the various study groups revealed the following P values (99% CI): urban IFG/IGT 0.02 (0.019–0.023), urban type 2 diabetic subjects 0.27 (0.26–0.28), and type 2 diabetic probands 0.015 (0.012–0.018) (Table 5 and online Appendix). Further analysis found that the 1112/1121 heterozygous haplotype combination was associated with an increased risk of IFG/IGT in urban subjects (OR 10.74, P = 0.001) and urban type 2 diabetic subjects (6.52, P = 0.015) and probands (5.78, P = 0.025). In addition, the haplotype combination 1112/1221 was absent in the normoglycemic group (n = 312) and in the urban type 2 diabetic subjects, but it was present in low numbers in the IFG/IGT subjects (P = 0.018) and probands (P = 0.003). These results suggest that a heterozygous haplotype combination of the 1112 haplotype with either 1121 or 1221 appears to influence susceptibility to glucose tolerance.

In assessing increased transmission of the 1112/1121 haplotype combination to affected offspring, we identified four sets of parents capable of generating the 1112/1121 haplotype in offspring, where each parent carried only one copy of each haplotype. Three type 2 diabetic probands received the “at risk” haplotype combination, and one did not, compared with the one-quarter expected proportion with no excess transmission (P = 0.05). Although the number of families was small, the observed frequency of the simultaneous transmission of both haplotypes to the type 2 diabetic probands was greater than the expected one-quarter probability for transmissions under the null hypothesis that affection is not associated with the haplotype combination. In addition, there was a transmission of the combination from a 5th set of parents, where one parent was homozygous for 1121 and the other parent for 1112.

Quantitative trait analysis

Quantitative TDT analysis.

Analysis of the transmission of the five haplotypes with a frequency >1% in the families, using sex as a covariant, found that the transmission of the 1121 haplotype to offspring was associated with a decrease in BMI (P = 0.019) and a narrower hip size (P = 0.007) and was weakly associated with a decreased fasting blood glucose (P = 0.03) (Table 6). In contrast, transmission of the 1221 haplotype was associated with larger hips P = 0.008. After correction for the number of haplotypes analyzed, only the hip remained significantly associated. Tests for parent-of-origin effect found the 1121 and 1221 haplotype associations with hip were both paternal in origin (P = 0.02 and P = 0.016, respectively). Transmission of the 1112 haplotype to offspring was associated with a decrease in WHR (P = 0.03), further confirmed by the “total association” analysis option within the QTDT program (P = 0.003). Analysis of the individual CAPN10 polymorphisms by QTDT, also with sex as a covariant, found that the transmission of allele 1 of UCSNP43 to type 2 diabetic offspring was associated with a decrease in BMI (P = 0.04) and hip size (P = 0.0056). A “total association” option analysis of the individual variants found that UCSNP63 was associated with variance in both waist circumference (P = 0.02) and WHR (P = 0.004). After correction for the four loci analyzed, only associations with hip and WHR remained significant, confirming the haplotype data.

Quantitative trait analysis in the unrelated urban survey samples.

Analysis of traits in the urban survey found no association between either individual SNPs or haplotypes of CAPN10 and quantitative traits (weight, height, BMI, waist circumference, hip, WHR, fasting blood glucose, and 2-h glucose) (data not shown). Association analyses of haplotype combinations were also all negative.

We have found that in the South Indian subjects, the presence of the 1112/1121 heterozygous haplotype combination confers both a 10.7-fold increase in risk of IFG/IGT and a 6.3- and 5.8-fold increased risk of type 2 diabetes in unrelated subjects and probands, respectively. These haplotype associations therefore both replicate and extend those previously published in the seminal publication of Horikawa et al. (3). The haplotypes constructed in our study differ only from the Mexican-Americans by the inclusion of a fourth SNP, UCSNP44. The Mexican-American 112 haplotype is equivalent to our 1112 haplotype, since the 2112 haplotype was not found in either the South Indian population or in U.K. Europeans (10). The 121 haplotype is also almost wholly represented by the 1121 haplotype in our study; the 2121 haplotype had a population frequency of 0.3–0.5% in South Indian subjects. Therefore, our 1112/1121 combination is equivalent to the Mexican-American “at risk” 112/121 combination. In the Mexican-Americans, this haplotype combination conferred a threefold increase (OR 2.8 and 3.58, from two data sets) in risk of type 2 diabetes. In the Horikawa et al. (3) study, it was suggested that there was also a concomitant 2.5- and 5.0-fold increase in risk of type 2 diabetes with the 112/121 haplotype combination in both Finnish and German populations, respectively, even though the ORs achieved were not statistically significant. In South Indians, a second haplotype combination, 1112/1221, also appeared to be associated with IFG/IGT and type 2 diabetes, although the numbers involved are very small and the haplotype combination was absent in the normoglycemic subjects. Homozygosity for the 1121 and 1221 haplotypes was not associated with an increased risk for disease, although only two subjects in the study were 1112 homozygous. These findings are in agreement with the Mexican-American data, emphasizing the importance of heterozygous combinations of calpain10 gene haplotypes in determining susceptibility to disease.

We did not find an association between polymorphisms of UCSNP44, -43, or -19 and either IFG/IGT or type 2 diabetes. We are therefore unable to confirm in the South Indians any of the other previously observed associations with either UCSNP43 or -44 (3,4,10,1618). However, the presence of the uncommon allele 2 (T) of UCSNP63 was significantly increased in both IFG/IGT subjects and type 2 diabetic probands. There are only two haplotypes detected in South Indian and Mexican-Americans that carry allele 2 (T) of UCSNP63; these are 1112 and 1122. The 1122 haplotype has a frequency of only 0.5% in the South Indians and 1% in Mexican-Americans (3), and it may be absent in U.K. Europeans (10). Therefore, not surprisingly, the 1112 haplotype was also significantly increased in both IFG/IGT subjects (OR 6.5) and type 2 diabetic probands (OR 4.1) (Table 4). A similar significant increase of UCSNP63 allele 2 (T) was also observed in Botnian Finns (3). UCSNP63 allele 2 is in LD with allele 1 of the other three SNPs, USNP44, -43, and -19. The relationship is much weaker between UCSNP63 and UCSNP44/-43. Therefore, the uncommon allele 2 (T) of UCSNP63 (associated with type 2 diabetes in both the South Indians and Botnian Finns) may be marking a separate disease susceptibility determinant to UCSNP44 and UCSNP43 (3), and thus it may be contributing to the risk attributed to the disease-associated haplotype combinations.

The failure to detect evidence of linkage with type 2 diabetes using TDT-based methods for either the 1112 or 1121 haplotypes individually or UCSNP63 allele 2 could have occurred for several reasons. First, if the allele frequencies among the affected and nonaffected subjects in the South Indian families were the same as those in the population study group, then there would only be a 39% power to detect the association (at P = 0.05) observed with UCSNP63 and type 2 diabetes in the case control studies. UCSNP63, and hence the 1112 haplotype, is found at a substantially lower frequency in both South Indian and European populations than in Mexican-Americans, and this would also further impede TDT analysis.

Second, family association methods have reduced the power to detect disease association under certain circumstances. The at-risk haplotype combination 112/121 is attributed to the greatest risk for type 2 diabetes in both Mexican-Americans and South Indians. Homozygosity for the individual haplotypes was not associated with an increased risk of type 2 diabetes in either ethnic group. Furthermore, in Mexican-Americans (3) other haplotype combinations including either of the at-risk 112 or 121 haplotypes were found to be neutral or protective to disease susceptibility. Consequently, unless the independent transmission of the two individual haplotypes can be assessed collectively, then observations of association in the offspring are less likely unless the individual haplotype association is strong.

The association between CAPN10 and type 2 diabetes has similarities to that between type 1 diabetes and HLA (19,20), although the strength of the overall association is far greater in type 1 diabetes. There are several HLA haplotypes that determine susceptibility to type 1 diabetes. These include the DR4/DQ8 and DR3/DQ2 haplotypes, which determine susceptibility, and the DR2/DQ6 haplotype, which determines dominant protection even in the presence of DR4/DQ8 or DR3/DQ2. Individuals homozygous for DR3/DQ2 or DR4/DQ8 are not at increased risk of disease, whereas subjects with a combination of the DR3 and DR4 haplotypes do have an increased risk of disease. The analogy can be further extended because there are likely to be several type 1 diabetes-associated susceptibility determinants carried on multiple HLA haplotypes over a 3500-kb distance and covering a large number of genes in the HLA class I, II, III, and IV regions. Furthermore, in different ethnic groups, different susceptibility haplotypes are found. Although there is some evidence that UCSNP43/-44 polymorphisms might have an effect on gene expression, it does not exclude other susceptibility determinants on disease-associated CAPN10 haplotypes nor, indeed, a closely linked locus as yet unidentified. The analogy of the importance of the haplotype combination, rather than an individual haplotypes or SNP, might also explain why preliminary data would suggest varying associations with different SNPs. Thus, for instance, UCSNP63 associates with type 2 diabetes in Botnian Finns and South Indians but not in Mexican-Americans. UCSNP44 associates with type 2 diabetes in British/Irish subjects (10) but has not been confirmed in other ethnic groups, and the UCSNP43 association with type 2 diabetes appears so far to be unique to Mexican-Americans.

Type 2 diabetes is likely to be a polygenic disease in which susceptibility determinants are determined by associations between intermediate traits and diabetes susceptibility genes. Quantitative trait analysis of haplotypes in the South Indian families found significant associations with three haplotypes, coincidentally the same three haplotypes that in combination conferred the greatest risks to both IFG/IGT and type 2 diabetes in the South Indians. The transmission of the 1121 haplotype was associated with a decrease in BMI and narrower hip size. Analysis of the individual variants with traits provided evidence that UCSNP43 may contribute in part to this association, since allele 1 was also associated with decreased BMI and hip measurement. The transmission of the 1112 haplotype was associated with a reduction in WHR. The reduction in WHR associated with the 1112 haplotype also appears to be partly due to the presence of the uncommon allele 2 of UCSNP63, which individually was also significantly associated with variance in WHR and waist circumference. Contrary to our findings in South Indians, homozygosity for the UCSNP43 G allele has been reported to be to be associated with an increased BMI/WHR and lower sleeping metabolic rate in Chinese and Pima Indians, respectively (4,21). However, other reported associations (mainly from published abstracts) of the UCSNP43 variant with both disease and intermediate traits have been inconsistent within and between ethnic groups (17,18,2225). Furthermore, in the original Mexican-American study (3), homozygosity for both the UCSNP43 and risk-associated haplotypes was not associated with an increased risk of disease. Finally, in only one of these studies was the impact of extended haplotypes in determining phenotypic outcome assessed (25).

From our studies, one might predict that South Indian subjects at risk of type 2 diabetes with the 1112/1121 combination haplotype would have a low BMI and WHR, with a smaller hip measurement. Although increased BMI, waist circumference, and WHR are widely recognized as predisposing factors toward the development of type 2 diabetes, this is not the case for hip measurements. Interestingly, it has been suggested that narrow hips reflect a decrease in muscle mass that might predispose to type 2 diabetes, especially in male individuals (26). Furthermore, in a second study comparing body tomographic cardiovascular risk factors between Swedish and Asian Indian subjects (27), the higher frequency of IGT in the Indian subjects was not related to the preponderance of visceral fat but to a lower leg muscle-to-total muscle ratio.

Despite the evidence of linkage and association between CAPN10 and diabetes, the biology of CAPN10 and its role in the pathogenesis in diabetes is not well understood. Studies using cysteine protease inhibitors have implicated members of the calpain family as being involved in either the promotion (28) or inhibition of protein secretion (29), depending on the cell type. A role of calpains in insulin secretion from mouse pancreatic islets has been demonstrated using protease inhibitors affecting exocytosis of insulin (30,31). Inhibition of calpain has also been found to modify insulin-mediated glucose transport in muscle and adipocytes (31). However, these studies have not specifically targeted calpain 10. Finally, calpains have been implicated in influencing signaling pathways that control differentiation of myoblasts (32), osteoblasts (33), chondrocytes (34), and preadipocytes to adipocytes (35). It is therefore clear that work should now be directed at understanding the biology of calpain10.

TABLE 1

Linkage disequilibrium between CAPN10 polymorphisms in the South Indian population

CAPN10 SNPsUCSNP44UCSNP43UCSNP19UCSNP63
UCSNP44 — <0.0001 <0.0001 0.034 
UCSNP43 <0.0001 — <0.0001 0.015 
UCSNP19 <0.0001 <0.0001 — <0.0001 
UCSNP63 0.034 0.015 <0.0001 — 
CAPN10 SNPsUCSNP44UCSNP43UCSNP19UCSNP63
UCSNP44 — <0.0001 <0.0001 0.034 
UCSNP43 <0.0001 — <0.0001 0.015 
UCSNP19 <0.0001 <0.0001 — <0.0001 
UCSNP63 0.034 0.015 <0.0001 — 

Data are P values. UCSNP44, -43, -19, and -63 refer to individual SNPs of the calpain 10 (CAPN10) gene. P values refer to linkage disequilibrium statistics calculated using the Estimated Haplotypes program.

TABLE 2

Comparison of haplotype frequencies of both South Indian families and urban survey with other ethnic groups

Haplotypes for SNPs 44/43/19/63Estimated haplotype frequencies
South India family parents (n = 172)South India survey (n = 452)*U.K. EuropeansMexican-Americans
1121 0.37 0.42 0.36 0.32 
2121 0.00 0.005 0.00 NK 
1221 0.18 0.14 0.26 0.27 
1111 0.19 0.21 0.16 0.17 
2111 0.19 0.19 0.16 0.006 
2211 0.00 0.00 0.00 0.00 
1122 0.007 0.00 0.00 0.01 
1112 0.03 0.03 0.07 0.23 
Haplotypes for SNPs 44/43/19/63Estimated haplotype frequencies
South India family parents (n = 172)South India survey (n = 452)*U.K. EuropeansMexican-Americans
1121 0.37 0.42 0.36 0.32 
2121 0.00 0.005 0.00 NK 
1221 0.18 0.14 0.26 0.27 
1111 0.19 0.21 0.16 0.17 
2111 0.19 0.19 0.16 0.006 
2211 0.00 0.00 0.00 0.00 
1122 0.007 0.00 0.00 0.01 
1112 0.03 0.03 0.07 0.23 
*

Frequencies derived from a control population group (10);

frequencies derived from a random sample (3). NK, not known.

TABLE 3

ETDT program analysis in South Indian families for association/linkage with type 2 diabetes

Transmitted alleles (n)Nontransmitted alleles (n)P
UCSNP44 21 31 0.17 
UCSNP43 25 26 0.89 
UCSNP19 31 42 0.2 
UCSNP63 11 0.49 
Transmitted alleles (n)Nontransmitted alleles (n)P
UCSNP44 21 31 0.17 
UCSNP43 25 26 0.89 
UCSNP19 31 42 0.2 
UCSNP63 11 0.49 
TABLE 4

Haplotype frequencies for case control studies

Haplotypes with >1% frequencyUrban subjects with normal glucose toleranceUrban subjects with IGT/IFGUrban subjects with type 2 diabetesType 2 diabetic probands
1121 43.3 (257) 42 (47) 41.6 (70) 35.3 (65) 
2111 19.7 (117) 13 (15) 18.4 (31) 22.8 (42) 
1221 13 (77) 15.8 (18) 12.5 (21) 18.9 (33) 
1111 22.4 (133) 19.3 (22) 24.4 (41) 18 (33) 
1112 1.7 (10) 9.6 (11)* 3.0 (5) 6.0 (11) 
Haplotypes with >1% frequencyUrban subjects with normal glucose toleranceUrban subjects with IGT/IFGUrban subjects with type 2 diabetesType 2 diabetic probands
1121 43.3 (257) 42 (47) 41.6 (70) 35.3 (65) 
2111 19.7 (117) 13 (15) 18.4 (31) 22.8 (42) 
1221 13 (77) 15.8 (18) 12.5 (21) 18.9 (33) 
1111 22.4 (133) 19.3 (22) 24.4 (41) 18 (33) 
1112 1.7 (10) 9.6 (11)* 3.0 (5) 6.0 (11) 

Data are % (n). Statistically significant differences in haplotype frequencies compared with subjects with normal glucose tolerance. The four locus haplotypes refer to CAPN10 allele numbers for UCSNP44, -43, -19, and -63, respectively.

*

Fisher exact test global P values: *P = 0.001

Fisher exact test global P values: †P = 0.004

TABLE 5

Haplotype combinations that influence risk to IFG/IGT and type 2 diabetes

Haplotype combinationControl subjects (n = 312)
IGT/IFG subjects (n = 56)
Unrelated diabetic subjects (n = 84)
Diabetic probands (n = 94)
FrequencyFrequencyOR (95% CI)FrequencyOR (95% CI)FrequencyOR (95% CI)
1112/1121 0.009 (3) 0.09 (5) 10.74 (2.0–55.2)* 0.06 (5) 6.52 (1.32–35.3) 0.05 (5) 5.78 (1.18–31.2) 
1112/1221 0.0 (0) 0.04 (2) INF (1.38–INF) 0.0 (0) 0.0 (0) 0.04 (4) INF (2.22–INF) 
Haplotype combinationControl subjects (n = 312)
IGT/IFG subjects (n = 56)
Unrelated diabetic subjects (n = 84)
Diabetic probands (n = 94)
FrequencyFrequencyOR (95% CI)FrequencyOR (95% CI)FrequencyOR (95% CI)
1112/1121 0.009 (3) 0.09 (5) 10.74 (2.0–55.2)* 0.06 (5) 6.52 (1.32–35.3) 0.05 (5) 5.78 (1.18–31.2) 
1112/1221 0.0 (0) 0.04 (2) INF (1.38–INF) 0.0 (0) 0.0 (0) 0.04 (4) INF (2.22–INF) 

Data are % (n) for frequency. ORs and 95% CIs were computed using a two-way contingency table. Significance values were calculated using a Yates-corrected χ2 (1 df).

*

P = 0.001;

P = 0.015;

P = 0.025. The four-figure haplotypes refer to CAPN10 allele numbers for UCSNP44, -43, -19, and -63, respectively. INF, infinity.

TABLE 6

QTDT analysis (Abecasis orthogonal model) of four-loci haplotypes with a frequency of >1% in a South Indian population

TRAITHaplotypes with >1% population frequency
11212111122111111112
BMI 0.019 NS NS NS NS 
Waist NS NS NS NS NS 
Hip 0.007 NS 0.008 NS NS 
WHR NS NS NS NS 0.03* 
Age at diagnosis of type 2 diabetes NS NS NS NS NT 
Height NS NS NS NS NS 
Fasting blood glucose 0.03 NS NS NS NS 
TRAITHaplotypes with >1% population frequency
11212111122111111112
BMI 0.019 NS NS NS NS 
Waist NS NS NS NS NS 
Hip 0.007 NS 0.008 NS NS 
WHR NS NS NS NS 0.03* 
Age at diagnosis of type 2 diabetes NS NS NS NS NT 
Height NS NS NS NS NS 
Fasting blood glucose 0.03 NS NS NS NS 

NT, not tested due to low haplotype frequency/numbers. The four-figure haplotypes refer to CAPN10 allele numbers for UCSNP44, -43, -19, and -63, respectively.

*

P = 0.003 by total association analysis.

This study was supported by funds from Diabetes U.K. (formerly known as the British Diabetic Association).

We gratefully acknowledge the release of prepublication data from Nancy Cox and Graeme I. Bell (Chicago, IL).

1.
Hanis CL, Boerwinkle E, Chakraborty R, Ellsworth DL, Concannon P, Stirling B, Morrison VA, Wapelhorst B, Spielman RS, Gogolin-Ewens KJ, Shepard JM, Williams SR, Risch N, Hinds D, Iwasaki N, Ogata M, Omori Y, Petzold C, Rietzch H, Schroder HE, Schulze J, Cox NJ, Menzel S, Boriraj VV, Chen X, et al.: A genome-wide search for human non-insulin-dependent (type 2) diabetes genes reveals a major susceptibility locus on chromosome 2.
Nat Genet
13
:
161
–166
1996
2.
Cox NJ, Frigge M, Nicolae DL, Concannon P, Hanis CL, Bell GI, Kong A: Loci on chromosomes 2 (NIDDM1) and 15 interact to increase susceptibility to diabetes in Mexican Americans.
Nat Genet
21
:
213
–215
1999
3.
Horikawa Y, Oda N, Cox NJ, Li X, Orho-Melander M, Hara M, Hinokio Y, Lindner TH, Mashima H, Schwarz PE, del Bosque-Plata L, Horikawa Y, Oda Y, Yoshiuchi I, Colilla S, Polonsky KS, Wei S, Concannon P, Iwasaki N, Schulze J, Baier LJ, Bogardus C, Groop L, Boerwinkle E, Hanis CL, Bell GI: Genetic variation in the gene encoding calpain-10 is associated with type 2 diabetes mellitus.
Nat Genet
26
:
163
–175
2000
4.
Baier LJ, Permana PA, Yang X, Pratley RE, Hanson RL, Shen GQ, Mott D, Knowler WC, Cox NJ, Horikawa Y, Oda N, Bell GI, Bogardus C: A calpain-10 gene polymorphism is associated with reduced muscle mRNA levels and insulin resistance.
J Clin Invest
106
:
R69
–R73
2000
5.
Cassell PG, Neverova M, Janmohamed S, Uwakwe N, Qureshi A, McCarthy MI, Saker PJ, Albon L, Kopelman P, Noonan K, Easlick J, Ramachandran A, Snehalatha C, Pecqueur C, Ricquier D, Warden C, Hitman GA: An uncoupling protein 2 gene variant is associated with a raised body mass index but not type II diabetes.
Diabetologia
42
:
688
–692
1999
6.
McCarthy MI, Hitman GA, Shields DC, Morton NE, Snehalatha C, Mohan V, Ramachandran A, Viswanathan M: Family studies of non-insulin-dependent diabetes mellitus in South Indians.
Diabetologia
37
:
1221
–1230
1994
7.
McCarthy M, Cassell P, Tran T, Mathias L, ’t Hart LM, Maassen JA, Snehalatha C, Ramachandran A, Viswanathan M, Hitman GA: Evaluation of the importance of maternal history of diabetes and of mitochondrial variation in the development of NIDDM.
Diabet Med
13
:
420
–428
1996
8.
Ramachandran A, Snehalatha C, Dharmaraj D, Viswanathan M: Prevalence of glucose intolerance in Asian Indians: urban-rural difference and significance of upper body adiposity.
Diabetes Care
15
:
1348
–1355
1992
9.
Alberti KG, Zimmet PZ: Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1. Diagnosis and classification of diabetes mellitus provisional report of a WHO consultation.
Diabet Med
15
:
539
–553
1998
10.
Evans JC, Frayling TM, Cassell PG, Saker PJ, Hitman GA, Walker M, Levy JC, O’Rahilly S, Subba Rao PV, Bennett AJ, Jones EC, Menzel S, Prestwich P, Simecek N, Wishart M, Dhillon R, Fletcher C, Millward A, Demaine A, Wilkin T, Horikawa Y, Cox NJ, Bell GI, Ellard S, McCarthy MI, Hattersley AT: Variation in the calpain 10 gene and type 2 diabetes: studies of association between the gene for calpain-10 and type 2 diabetes mellitus in the United Kingdom.
Am J Hum Genet
69
:
544
–552
2001
11.
Sham PC, Curtis D: An extended transmission/disequilibrium test (TDT) for multi-allele marker loci.
Annals Hum Genet
59
:
323
–336
1995
12.
Clayton D: A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission.
Am J Hum Genet
65
:
1170
–1177
1999
13.
Abecasis GR, Cardon LR, Cookson WOC: A general test of association for quantitative traits in nuclear families.
Am J Hum Genet
66
:
279
–292
2000
14.
Fleiss JL:
Statistical Methods for Rates and Proportions.
2nd ed. New York, New York, John Wiley & Sons,
1981
15.
Xie X, Ott J: Testing linkage disequilibrium between a disease gene and marker loci (Abstract).
Am J Hum Genet
53
:
1107
1993
16.
Ren Q, Hasstedt S, Hanis C, Elbein SC: Increased transmission of NIDDM1 variant in Caucasian familial type 2 Diabetes (Abstract).
Diabetes
49 (Suppl. 1)
:
A200
2000
17.
Ehrmann DA, Hara M, Polonsky KS, Cox NJ: Polymorphisms in PPP1R3 and calpain 10 (CAPN10) influence oral glucose tolerance in women with polycystic ovary syndrome and their first-degree relatives (Abstract).
Diabetes
49 (Suppl. 1)
:
A198
2000
18.
Lynn S, Evans JC, White C, Frayling TM, Hattersley AT, Turnbull DM, Horikawa Y, Cox NJ, Bell GI, Walker M: Variation in the calpain-10 gene affects blood glucose levels in the British population.
Diabetes
51
:
247
–250
2002
19.
Dorman JS, Bunker CH: HLA DQ locus of the human leukocyte antigen complex and type 1 diabetes mellitus: a HuGE review.
Epidemiol Rev
22
:
218
–227
2000
20.
Undlien DE, Lie BA, Thorsby E: HLA complex genes in type 1 diabetes and other autoimmune diseases. Which genes are involved?
Trends Genet
17
:
93
–100
2001
21.
Ji L, Chen L, Han X: The role of calpain10 gene polymorphisms in genetic susceptibility to type 2 diabetes in a Chinese population (Abstract).
Diabetes
50 (Suppl. 2)
:
A206
–A207
2001
22.
Stumvoll M, Fritsche A, Madaus A, Stefan N, Weisser M, Machicao F, Haring H: Functional significance of the UCSNP-43 polymorphism in the CAPN10 gene for proinsulin processing and insulin secretion in nondiabetic Germans.
Diabetes
50
:
2161
–2163
2001
23.
Xiang K, Zheng T, Fang Q, Jia W, Wang Y, Zhang R, Sheng K, Jie L, Lu J, Lu H: Calpain-10 SNP43 polymorphism is associated with insulin sensitivity and insulin levels during glucose load in Chinese non-diabetic subjects (Abstract).
Diabetes
50 (Suppl. 2)
:
A232
2001
24.
Ng MCY, So W-Y, Critchley JAJH, Cockram CS, Chan JCN: Association of calpain 10 genetic polymorphisms with type 2 diabetes and insulin response in non-diabetic Chinese (Abstract).
Diabetes
50 (Suppl. 2)
:
A234
2001
25.
Furuta M, Furuta H, Ueda K, Kanamori M, Horikawa Y, Kawashima H, Tamai M, Nishi M, Sanke T, Nanjo K: Relationship between genetic variation in the calpain 10 gene and insulin sensitivity in Japanese (Abstract).
Diabetes
50 (Suppl. 2)
:
A490
2001
26.
Seidell JC, Han TS, Feskens EJM, Lean MEJ: Narrow hips and broad waist circumferences independently contribute to increased risk of non-insulin-dependent diabetes mellitus.
J Int Med
242
:
401
–406
1997
27.
Chowdhury B, Lantz H, Sjostrom L: Computerised tomography-determined body composition in relation to cardiovascular risk factors in Indian and matched Swedish males.
Metabolism
45
:
634
–644
1996
28.
Croce K, Flaumenhaft R, Rivers M, Furie B, Furie BC, Herman IM, Potter DA: Inhibition of calpain blocks platelet secretion, aggregation, and spreading.
J Biol Chem
274
:
36321
–36327
1999
29.
Yamazaki T, Haass C, Saido TC, Omura S, Ihara Y: Specific increase in amyloid-protein 42 secretion ratio by calpain inhibition.
Biochemistry
36
:
8377
–8383
1997
30.
Sreenan SK, Zhou YP, Otani K, Hansen PA, Currie KP, Pan CY, Lee JP, Ostrega DM, Pugh W, Horikawa Y, Cox NJ, Hanis CL, Burant CF, Fox AP, Bell GI, Polonsky KS: Calpains play a role in insulin secretion and action.
Diabetes
50
:
2013
–2020
2001
31.
Zhou, Y-P, Sreenan S, Bindokas VP, Pan C-Y, Currie KPM, Lee JP, Fox AP, Miller RJ, Cox NJ, Polonsky KS: Calpain inhibitors impair insulin secretion after 48-hours: a model for beta-cell dysfunction in type 2 diabetes?
Diabetes
49 (Suppl. 1)
:
A80
2000
32.
Ueda Y, Wang MC, Ou BR, Huang J, Elce J, Tanaka K, Ichihara A, Forsberg NE: Evidence for the participation of the proteasome and calpain in early phases of muscle cell differentiation: J Biochem Cell Bio
30
:
679
–694
1998
33.
Murray SS, Grisanti MS, Bentley GV, Kahn AJ, Urist MR, Murray EJ: The calpain-calpastatin system and cellular proliferation and differentiation in rodent osteoblastic cells.
Exp Cell Res
233
:
297
–309
1997
34.
Yasuda T, Shimizu K, Nakagawa Y, Yamamoti S, Niibayashi H, Yamamuro T: m-Calpain in rat growth plate chondrocyte cultures: its involvement in the matrix mineralization process.
Develop Biol
170
:
159
–168
1995
35.
Patel YM, Lane DM: Role of calpain in adipocyte differentiation.
Proc Natl Acad Sci U S A
96
:
1279
–1285
1999

Address correspondence and reprint requests to Professor Graham A. Hitman, Department of Diabetes and Metabolic Medicine, the Royal London Hospital, Whitechapel Road, London E1 1BB, U.K. E-mail: g.a.hitman@mds.qmw.ac.uk.

Received for publication 18 June 2001 and accepted in revised form 6 February 2002.

Additional information for this article can be found in an online appendix at http://diabetes.diabetesjournals.org.

IFG, impaired fasting glucose; IGT, impaired glucose tolerance; LD, linkage disequilibrium; OHA, oral hypoglycemic agent; OR, odds ratio; SNP, single-nucleotide polymorphism; TDT, transmission disequilibrium test; WHO, World Health Organization; WHR, waist-to-hip ratio.

Supplementary data