A strong genetic component of the β-cell defect of type 2 diabetes is undisputed. We recently developed a modification of the classic hyperglycemic clamp to assess β-cell function in response to various stimuli (10 mmol/l glucose, additional glucagon-like peptide [GLP]-1, and arginine). Subjects at risk for developing type 2 diabetes (impaired glucose-tolerant individuals, women with gestational diabetes, and individuals with a family history of type 2 diabetes) clearly showed a significantly decreased mean secretory response to all secretagogues compared with controls. We also showed that normal glucose-tolerant carriers of the Gly972Arg polymorphism in the insulin receptor substrate 1 have significantly reduced insulin secretion in response to glucose and arginine but not to GLP-1. More remarkably, however, the relative impairment of the different secretory phases varied greatly in the same individual, indicating a substantial heterogeneity of β-cell dysfunction. Specific prominence of this heterogeneity may reflect a specific cellular defect of the β-cell. In subjects sharing this pattern of heterogeneity, any underlying genetic variant may be enriched and thus more likely not only to be identified but also to be related to a pathophysiological mechanism. In conclusion, we believe that careful clinical characterization of β-cell function (and dysfunction) is one way of identifying and understanding the genetic factors leading to the insulin secretory failure of type 2 diabetes.

Type 2 diabetes is a heterogeneous disorder characterized by varying degrees of impaired insulin secretion and insulin resistance. Although the site of the primary genetic lesion (β-cell or insulin target tissues) is still a matter of controversy, it is generally agreed that 1) the disease has strong genetic and environmental (acquired) components (1–4); 2) its mode of inheritance is polygenic (5–7), meaning that several genetic abnormalities are required for disease manifestation; and 3) the variation in β-cell function segregates among family members (4,8,9), indicating that β-cell dysfunction has a relevant genetic component. On the other hand, what is still unclear is the relative impact of genetic versus environmental factors and the precise nature of the genetic lesion(s) responsible for β-cell dysfunction in common type 2 diabetes.

It is worth noting that entirely different molecular defects (for example, in the “glucose sensor” glucokinase and the transcription factor hepatocyte nuclear factor 1α) can result in the clinical phenotype of abnormal insulin secretion (10). This result strongly indicates that abnormalities in many different genes may be involved in β-cell dysfunction. In addition, different polymorphisms in the same gene can have different functional consequences, as illustrated by mutations in the insulin promoting factor (IPF)-1. A homozygote frameshift mutation resulted in agenesis of the pancreas and in maturity-onset diabetes of the young in heterozygotes (11), whereas several missense mutations led only to a late-onset type 2 diabetic phenotype (12–14). It is possible that less detrimental, still unidentified genetic variants of IPF-1 contribute to impaired insulin secretion to a minor degree, which may be difficult to assess in vivo. The impairment in β-cell function by such a mutation alone may not be sufficient to result in overt hyperglycemia. However, in combination with other minor (or major) genetic factors, it may well be the straw that breaks the camel’s back.

Among the experimental strategies used to identify genetic factors of β-cell failure, the candidate gene approach has not been very successful, with the exception of rare monogenetic forms of type 2 diabetes (e.g., detrimental point mutations in the glucokinase gene). On the other hand, positional cloning based on linkage analysis of many pooled pedigrees led to the identification of a mutation in calpain 10 (15), which is now considered a true diabetes gene. However, the functional relevance of this cysteine protease and its connection with glucose metabolism is still unclear (16). There is evidence for a functional involvement of calpain 10 in muscle and β-cells. Moreover, studies both in knockout animals (17) and transfected β-cells (18) have produced intriguing results regarding the role of abnormal signaling through the insulin receptor substrate (IRS) molecules in insulin secretion, which still await translation into the clinical context. This obvious complexity suggests that a more careful clinical characterization of β-cell function may be required as the basis for state-of-the-art genetic analyses.

However, the characterization of potential subphenotypes of type 2 diabetes after the onset of overt hyperglycemia and other metabolic derangements is virtually impossible because hyperglycemia per se impairs both insulin sensitivity and secretion, masking any primary abnormality. It is thus necessary to study cohorts at an increased genetic risk of developing type 2 diabetes before the onset of overt hyperglycemia. These cohorts include subjects with impaired glucose tolerance, normal glucose-tolerant individuals with a family history of type 2 diabetes, and women with a history of gestational diabetes. In Tübingen (southwestern Germany), we clinically characterize such populations by a variety of clinical-experimental methods as the basis for genetic association studies (euglycemic clamp for insulin sensitivity, stable isotopes for lipolysis and glucose production, magnetic resonance spectroscopy for intramyocellular lipids, flow-associated vasodilation for endothelial dysfunction, and hyperglycemic clamp for β-cell function). In addition to classifying insulin secretion (high vs. low) by a standard hyperglycemic clamp, we made an attempt to quantify different aspects of insulin secretion by measuring the secretory response to different secretagogues in the same subjects. The hypothesis is that this might result in patterns of abnormal β-cell function specific for a certain genotype.

We therefore designed a hyperglycemic clamp (10 mmol/l) in which the classic 2-h square-wave hyperglycemia is followed by additional administration of glucagon-like peptide 1 (GLP-1) (1.5 pmol · kg−1 · min−1 over 80 min) and a final arginine bolus (5 g) (19). The test produced a distinct pattern of plasma insulin and C-peptide concentrations in response to the three secretagogues. First and second phases of glucose-stimulated insulin secretion, an early and a late phase of GLP-1–induced insulin secretion, and an acute response to arginine could be calculated. Moreover, the combination of secretagogues used in this clamp resulted in insulin and C-peptide concentrations that exceeded those seen in the test hitherto considered to provoke the maximal insulin response by more than 100% (arginine at >25 mmol/l glucose [20]). The maximal serum insulin and C-peptide concentrations observed during this test may come closest to the assessment of the maximal insulin secretory capacity in humans in vivo. The novel test takes 3.5 h and was shown to be highly reproducible (19). It can be used in various contexts requiring differential characterization of islet function in selected human populations. In addition, insulin sensitivity can be determined by relating the glucose infusion rate during the second phase of the hyperglycemic clamp to the insulin concentration during the same interval.

Our first application of this test was its use in a group of subjects with impaired glucose tolerance (IGT) compared with a normal glucose-tolerant control group that was perfectly matched for BMI, waist-to-hip ratio, sex, and age (21). Most phases of insulin secretion were significantly reduced in the subjects with IGT (Fig. 1). Insulin sensitivity was not significantly different between the two groups. This was a result of the matching process, which excluded the impact of obesity, body fat distribution, age, and sex. The novel finding, however, was that the group with IGT subjects did not behave homogeneously with respect to the relative impairment of the different phases. Displaying the individual data makes it apparent that the pattern of impairment varies greatly among subjects (Fig. 2). For example, one subject (Fig. 2, filled diamonds) reached 55% of the normal first phase but not even 20% of the average maximal secretion. Another subject (Fig. 2, open semicircle, upper half) produced an almost normal maximal secretion, whereas the GLP peak was entirely absent. A particularly surprising finding of the study was the absence of the sharp peak in insulin secretion rate in response to the start of the GLP infusion (after deconvolution of C-peptide concentrations) in the group with IGT. This may indicate that, at least in a subgroup, a β-cell defect specifically in the GLP-1–dependent signaling pathway is involved in the pathogenesis of β-cell failure. In conclusion, it is very obvious that the impairment of insulin secretion varies greatly among subjects with IGT not only quantitatively but also qualitatively. A precise understanding of the individual abnormality may be essential for the identification of the genetic lesion(s) involved.

A total of 50 and 25% of genes are shared among first- and second-degree relatives, respectively. Twin studies showed concordance rates for any abnormality of glucose metabolism (either type 2 diabetes or impaired glucose tolerance) as high as 96% (22), underlining the substantial contribution of genetic factors. Furthermore, 40% of first-degree relatives of patients with type 2 diabetes develop the disease as opposed to ∼6% of the general population (23). Moreover, there is evidence from different laboratories demonstrating subtle but significant abnormalities in β-cell function in first-degree relatives (4,9,24). We applied our modified hyperglycemic clamp to subjects with a family history of type 2 diabetes to determine specific patterns of β-cell response to different secretagogues.

All phases of insulin secretion were reduced to various extents in the subjects with a diabetic relative (n = 37) compared with the group without a known family history of type 2 diabetes (n = 20) (Fig. 3). Similar to the analysis of the group with IGT, first-phase (2,243 ± 168 vs. 3,188 ± 361 pmol/min, P = 0.02) and GLP-1–induced peak insulin secretion rates (1,561 ± 175 vs. 2,390 ± 480 pmol/min, P = 0.02) were significantly lower, whereas second phases and the response to arginine were not significantly different (P > 0.20 for both). A similar heterogeneity of β-cell dysfunction to that in the group with IGT was detectable in this normal glucose-tolerant group with an increased genetic risk for type 2 diabetes.

Women with transient diabetes during pregnancy (gestational diabetes mellitus [GDM]) carry an increased risk of developing type 2 diabetes later in life (25). It has been suggested that GDM and type 2 diabetes are the same disorder (26). Therefore, it is likely that both have similar underlying genetic abnormalities. GDM occurs when the physiological reduction in insulin sensitivity during the second half of pregnancy cannot be countered by an adequate increase in insulin secretion. Thus, GDM represents a natural fast motion model of conditions such as obesity where a (possibly genetic) inability of the β-cell to compensate for reduced insulin sensitivity over the years results in type 2 diabetes.

Again, all phases of insulin secretion were more or less reduced in women with GDM (n = 7) compared with women with no evidence of glucosuria or a large-for-date baby (n = 20) (Fig. 4). First-phase insulin secretion (1,769 ± 239 vs. 2,317 ± 233 pmol/min, P = 0.04) was significantly lower in women with GDM, whereas the second phase, the GLP-1–induced peak, and the response to arginine were not significantly different (P > 0.15 for all). Also in this group of women with GDM, a marked intra-individual heterogeneity in the response to the different secretagogues was observed, indicating that a number of different (possibly genetically determined) factors modulate the response to the different secretagogues.

In addition to analyzing insulin secretion in populations with a well-established reduction in insulin secretion, we assessed whether a specific genetic variant contributes to the interindividual variation of β-cell function. The Gly972Arg polymorphism in IRS-1 originally represented a prime candidate for genetic causes of impaired insulin signaling. Over the years, however, conflicting data accumulated (27–33), and this polymorphism has not held up to the expectations as an insulin resistance gene. Recently, it was shown that insulin-secreting cells overexpressing the IRS-1 Gly972Arg variant had a decreased sulfonylurea- and glucose-stimulated insulin secretion compared with cells overexpressing the wild-type IRS-1 (18). Moreover, cell apoptosis of isolated human islets obtained from organ donors heterozygous for Arg972 was increased twofold compared with wild-type IRS-1 carriers. It was consequently suggested that this polymorphism might represent a genetic variant unifying insulin resistance and β-cell dysfunction. We studied 8 subjects with and 36 subjects without the polymorphism using the modified hyperglycemic clamp method. To provide greater statistical power, we also analyzed oral glucose tolerance tests (OGTTs) of 212 subjects (31 with and 181 without the mutation) using a number of recently validated indexes to estimate β-cell function from insulin (and C-peptide) concentrations obtained during an OGTT.

During the modified hyperglycemic clamp, insulin secretion rates were significantly lower in Gly/Arg compared with Gly/Gly during the first phase (1,711 ± 142 vs. 3,014 ± 328 pmol/min, P = 0.05) and after maximal stimulation with arginine (5,340 ± 639 vs. 9,075 ± 722 pmol/min, P = 0.03). During second-phase insulin secretion, the differences approached statistical significance, whereas during the GLP-1 phases, they were not significant (Fig. 5). Several validated indexes of β-cell function from the OGTT were significantly lower in X/Arg (n = 31) compared with Gly/Gly (n = 181) (P between 0.002 and 0.05). Insulin sensitivity, as determined by the euglycemic-hyperinsulinemic clamp, was not different. In a subanalysis, we compared subjects with a BMI greater or less than 25 kg/m2. The C-peptide area under the curve (AUC)/glucoseAUC (CPAUC/GAUC) was significantly greater in the Gly/Gly obese (n = 69) than in the Gly/Gly lean subjects (n = 112) (461 ± 17 vs. 373 ± 10, P < 0.001). In contrast, the CPAUC/GAUC was virtually identical between the lean (n = 19) and obese (n = 12) subjects in the X/Arg group (336 ± 25 vs. 341 ± 27, P = 0.91; P = 0.078, analysis of variance [ANOVA]) (Fig. 6). In summary, our results suggest that the Gly972Arg polymorphism in IRS-1 is associated with decreased insulin secretion in response to glucose. It is possible that this polymorphism contributes to the variation in insulin secretion in normal glucose-tolerant humans. Although with ANOVA the interaction effect showed only a tendency, it is possible that this contribution may be particularly important in the presence of obesity.

In addition to simply quantifying the insulin secretory response to glucose, the modified hyperglycemic clamp permits the assessment of the differential secretory response to different secretagogues, i.e., the “heterogeneity of insulin secretion.” To quantify this heterogeneity, we performed the following procedures: in a first step, the different phases of insulin secretion (first phase, second phase, peak after GLP-1, peak after arginine) are normalized by the population mean of each phase to account for the huge absolute differences (e.g., between second phase and the peak after arginine); subsequently, the normalized second phase, peak after GLP-1, and peak after arginine are each expressed relative to the first phase to account for overall differences in β-cell function (e.g., between impaired glucose-tolerant and normal glucose-tolerant subjects). We are left with percentages where, for example, a 140% second phase and 60% arginine peak indicate a high second phase and a low arginine phase relative to the individual’s first phase. Thus, the spatial position in the three-dimensional construct (the dimensions are second phase, peak after GLP-1, and peak after arginine, all relative to first phase) is the measure for the heterogeneity of β-cell function.

In a second step, we subjected these spatial points from 80 subjects to a standard cluster analysis arbitrarily set for four clusters. This procedure clusters the 80 points (80 individuals) around four centers of gravity (Fig. 7). The individuals in one cluster have the relative height of the three secretion phases in common. We then assessed the distribution of genotypes within the four clusters. We specifically examined the frequency of polymorphisms reported to be associated with β-cell dysfunction or an increased risk for type 2 diabetes by χ2 testing. Because the groups were small, truly significant differences could not be seen upon multiple comparison. Nevertheless, the frequency of carriers of the IRS-1 972 polymorphism was different in cluster 1 versus cluster 2 and 4 or of the CAPN10 UCSNP-43 polymorphism in cluster 1 versus cluster 4 (Table 1). A relatively low maximal secretion (arginine) appears to be associated with the IRS-1 972 polymorphism. Carriers of the G allele in CAPN10 UCSNP-43 were absent in the cluster with the relative increase in all phases relative to the first phase. The enrichment of a specific genotype may help to identify the genetic contribution of subtypes of β-cell dysfunction. We are well aware that this kind of analysis is preliminary at best and is only intended as a potential perspective of approaching clinical characterization of β-cell dysfunction in the future. With increasing number, the statistical power will increase and more clusters and additional dimensions, for example, the proinsulin-to-insulin ratio in response to an acute stimulus as an index for proinsulin processing (34), can be introduced.

We presented evidence that careful clinical characterization of β-cell function will lead to the identification of secretion defects in selected subgroups prone to develop type 2 diabetes, such as IGT, GDM, and relatives of affected patients. Moreover, as exemplified by the IRS-1 972 polymorphism, quantification of β-cell function can help to identify the contribution of a genetic variant to the biological variation of insulin secretion. In addition to categorizing insulin secretion in response to glucose, however, determining the heterogeneity of the secretory response to different secretagogues may be a novel way to understand in vivo function and dysfunction of genetic variants allegedly contributing to β-cell dysfunction and type 2 diabetes. However, only longitudinal data will provide the final proof of our concept.

FIG. 1.
FIG. 1. Glucose infusion rates (GIR), blood glucose concentration (BG), insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant (NGT) and IGT subjects. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale (35).
View largeDownload slide

Glucose infusion rates (GIR), blood glucose concentration (BG), insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant (NGT) and IGT subjects. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale (35).

FIG. 1.
FIG. 1. Glucose infusion rates (GIR), blood glucose concentration (BG), insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant (NGT) and IGT subjects. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale (35).
View largeDownload slide

Glucose infusion rates (GIR), blood glucose concentration (BG), insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant (NGT) and IGT subjects. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale (35).

Close modal
FIG. 2.
FIG. 2. Relative impairment of insulin secretion in response to different secretagogues in subjects with IGT. Data are expressed as percent of insulin secretion in subjects with normal glucose tolerance (NGT). Horizontal lines denote statistical difference (P < 0.05) using Duncan’s procedure for multiple comparisons (35). Negative values for the peak response to GLP-1 result from expressing the peak as the value above the subsequent plateau. AIR, acute insulin response.
View largeDownload slide

Relative impairment of insulin secretion in response to different secretagogues in subjects with IGT. Data are expressed as percent of insulin secretion in subjects with normal glucose tolerance (NGT). Horizontal lines denote statistical difference (P < 0.05) using Duncan’s procedure for multiple comparisons (35). Negative values for the peak response to GLP-1 result from expressing the peak as the value above the subsequent plateau. AIR, acute insulin response.

FIG. 2.
FIG. 2. Relative impairment of insulin secretion in response to different secretagogues in subjects with IGT. Data are expressed as percent of insulin secretion in subjects with normal glucose tolerance (NGT). Horizontal lines denote statistical difference (P < 0.05) using Duncan’s procedure for multiple comparisons (35). Negative values for the peak response to GLP-1 result from expressing the peak as the value above the subsequent plateau. AIR, acute insulin response.
View largeDownload slide

Relative impairment of insulin secretion in response to different secretagogues in subjects with IGT. Data are expressed as percent of insulin secretion in subjects with normal glucose tolerance (NGT). Horizontal lines denote statistical difference (P < 0.05) using Duncan’s procedure for multiple comparisons (35). Negative values for the peak response to GLP-1 result from expressing the peak as the value above the subsequent plateau. AIR, acute insulin response.

Close modal
FIG. 3.
FIG. 3. Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant subjects with and without a family history (FH) of type 2 diabetes. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.
View largeDownload slide

Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant subjects with and without a family history (FH) of type 2 diabetes. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.

FIG. 3.
FIG. 3. Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant subjects with and without a family history (FH) of type 2 diabetes. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.
View largeDownload slide

Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal glucose-tolerant subjects with and without a family history (FH) of type 2 diabetes. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.

Close modal
FIG. 4.
FIG. 4. Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal control women and women with GDM. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.
View largeDownload slide

Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal control women and women with GDM. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.

FIG. 4.
FIG. 4. Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal control women and women with GDM. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.
View largeDownload slide

Blood glucose concentrations, insulin secretion rates (ISR) (calculated from plasma C-peptide concentrations by deconvolution), and plasma concentrations of insulin at baseline and during 200 min of the hyperglycemic clamp in normal control women and women with GDM. Note that for reasons of clarity, the y-axis for insulin was split and continued on the right with a different scale.

Close modal
FIG. 5.
FIG. 5. Blood glucose and insulin secretion rates during a modified hyperglycemic clamp in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. The indicated P values were calculated at 2.5, 80–120, 125, 160–180, and 182.5 min.
View largeDownload slide

Blood glucose and insulin secretion rates during a modified hyperglycemic clamp in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. The indicated P values were calculated at 2.5, 80–120, 125, 160–180, and 182.5 min.

FIG. 5.
FIG. 5. Blood glucose and insulin secretion rates during a modified hyperglycemic clamp in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. The indicated P values were calculated at 2.5, 80–120, 125, 160–180, and 182.5 min.
View largeDownload slide

Blood glucose and insulin secretion rates during a modified hyperglycemic clamp in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. The indicated P values were calculated at 2.5, 80–120, 125, 160–180, and 182.5 min.

Close modal
FIG. 6.
FIG. 6. Effect of obesity on β-cell function (expressed as CPAUC/GAUC during an OGTT) in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. Subjects within one genotype were divided into an obese and a lean group according to BMI (cutoff 25 kg/m2). P = 0.078 for the interaction effect between the IRS 972 variant and obesity (BMI ≥ 25 kg/m2), ANOVA.
View largeDownload slide

Effect of obesity on β-cell function (expressed as CPAUC/GAUC during an OGTT) in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. Subjects within one genotype were divided into an obese and a lean group according to BMI (cutoff 25 kg/m2). P = 0.078 for the interaction effect between the IRS 972 variant and obesity (BMI ≥ 25 kg/m2), ANOVA.

FIG. 6.
FIG. 6. Effect of obesity on β-cell function (expressed as CPAUC/GAUC during an OGTT) in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. Subjects within one genotype were divided into an obese and a lean group according to BMI (cutoff 25 kg/m2). P = 0.078 for the interaction effect between the IRS 972 variant and obesity (BMI ≥ 25 kg/m2), ANOVA.
View largeDownload slide

Effect of obesity on β-cell function (expressed as CPAUC/GAUC during an OGTT) in subjects without (Gly/Gly) and with (X/Arg) the Gly972Arg polymorphism in IRS-1. Subjects within one genotype were divided into an obese and a lean group according to BMI (cutoff 25 kg/m2). P = 0.078 for the interaction effect between the IRS 972 variant and obesity (BMI ≥ 25 kg/m2), ANOVA.

Close modal
FIG. 7.
FIG. 7. Heterogeneity of insulin secretion. Each spatial point represents one individual. The x, y, and z coordinates indicate the insulin secretory response during the second phase, GLP-1, and arginine relative to the first phase of glucose-stimulated insulin secretion. Thus, the spatial position in the three-dimensional construct (the dimensions are second phase, peak after GLP-1, and peak after arginine relative to first phase) is the measure for the heterogeneity of β-cell function. The plot no longer contains information regarding high versus low insulin secretion. The symbol coding denotes the result of a cluster analysis performed on 80 nondiabetic subjects. The individuals in one cluster have the relative height of the three secretion phases in common (see text for more details).
View largeDownload slide

Heterogeneity of insulin secretion. Each spatial point represents one individual. The x, y, and z coordinates indicate the insulin secretory response during the second phase, GLP-1, and arginine relative to the first phase of glucose-stimulated insulin secretion. Thus, the spatial position in the three-dimensional construct (the dimensions are second phase, peak after GLP-1, and peak after arginine relative to first phase) is the measure for the heterogeneity of β-cell function. The plot no longer contains information regarding high versus low insulin secretion. The symbol coding denotes the result of a cluster analysis performed on 80 nondiabetic subjects. The individuals in one cluster have the relative height of the three secretion phases in common (see text for more details).

FIG. 7.
FIG. 7. Heterogeneity of insulin secretion. Each spatial point represents one individual. The x, y, and z coordinates indicate the insulin secretory response during the second phase, GLP-1, and arginine relative to the first phase of glucose-stimulated insulin secretion. Thus, the spatial position in the three-dimensional construct (the dimensions are second phase, peak after GLP-1, and peak after arginine relative to first phase) is the measure for the heterogeneity of β-cell function. The plot no longer contains information regarding high versus low insulin secretion. The symbol coding denotes the result of a cluster analysis performed on 80 nondiabetic subjects. The individuals in one cluster have the relative height of the three secretion phases in common (see text for more details).
View largeDownload slide

Heterogeneity of insulin secretion. Each spatial point represents one individual. The x, y, and z coordinates indicate the insulin secretory response during the second phase, GLP-1, and arginine relative to the first phase of glucose-stimulated insulin secretion. Thus, the spatial position in the three-dimensional construct (the dimensions are second phase, peak after GLP-1, and peak after arginine relative to first phase) is the measure for the heterogeneity of β-cell function. The plot no longer contains information regarding high versus low insulin secretion. The symbol coding denotes the result of a cluster analysis performed on 80 nondiabetic subjects. The individuals in one cluster have the relative height of the three secretion phases in common (see text for more details).

Close modal
TABLE 1

Genotype distribution among secretion clusters

AlleleCluster 1 (%)Cluster 2 (%)Cluster 3 (%)Cluster 4 (%)P* (four groups)P* (selected comparison)
PPARγ Prol2 67 57 78 67 0.52 — 
IRS-1 Arg972 25 11 18 0 0.39 0.058 (1 vs. 2 + 4) 
CAPN10 UCSNP-43/G 62 33 64 0 0.11 0.02 (1 vs. 4) 
AlleleCluster 1 (%)Cluster 2 (%)Cluster 3 (%)Cluster 4 (%)P* (four groups)P* (selected comparison)
PPARγ Prol2 67 57 78 67 0.52 — 
IRS-1 Arg972 25 11 18 0 0.39 0.058 (1 vs. 2 + 4) 
CAPN10 UCSNP-43/G 62 33 64 0 0.11 0.02 (1 vs. 4) 

PPARγ, peroxisome proliferator–activated receptor γ.

*

χ2 test.

View Large

This study was supported in part by a research grant from the European Community (QLRT-1999–00674).

We thank all the research volunteers for their participation. We gratefully acknowledge the excellent technical assistance of Isolde Riedlinger, Ulrike Schmidt, Sabine Obermüller, and Sabine Wolff. We are also indebted to Elke Hardt, Anna Teigeler, Norbert Stefan, and Alexander Madaus, who helped with many experiments. We gratefully acknowledge the help of Dr. Fausto Machicao, who performed the DNA analysis.

1.
Kahn CR: Insulin action, diabetogenes, and the cause of type II diabetes.
Diabetes
 
43
:
1066
–1084,
1994
2.
Hamman RF: Genetic and environmental determinants of non-insulin-dependent diabetes mellitus (NIDDM).
Diabete Metab Rev
 
8
:
287
–338,
1992
3.
Bouchard C: The genetics of obesity in humans.
Curr Opin Endocrinol Diabetes
 
3
:
29
–35,
1996
4.
Yki-Järvinen H: Pathogenesis of non-insulin-dependent diabetes mellitus.
Lancet
 
343
:
91
–95,
1994
5.
Turner RC, Hattersley AT, Shaw JT, Levy JC: Type II diabetes: clinical aspects of molecular biological studies.
Diabetes
 
44
:
1
–10,
1995
6.
Granner DK, O’Brien RM: Molecular physiology and genetics of NIDDM: importance of metabolic staging.
Diabetes Care
 
15
:
369
–395,
1992
7.
Aitman TJ, Todd JA: Molecular genetics of diabetes mellit.
. Baillieres Clin Endocrinol Metab
 
9
:
631
–656,
1995
8.
Lehtovirta M, Kaprio J, Forsblom C, Eriksson J, Tuomilehto J, Groop L: Insulin sensitivity and insulin secretion in monozygotic and dizygotic twins.
Diabetologia
 
43
:
285
–293,
2000
9.
Van Haeften TW, Dubbeldam S, Zonderland ML, Erkelens DW: Insulin secretion in normal glucose-tolerant relatives of type 2 diabetic subjects: assessments using hyperglycemic glucose clamps and oral glucose tolerance tests.
Diabetes Care
 
21
:
278
–282,
1998
10.
Winter WE, Nakamura M, House DV: Monogenic diabetes mellitus in youth: the MODY syndromes.
Endocrinol Metab Clin North Am
 
28
:
765
–785,
1999
11.
Stoffers DA, Zinkin NT, Stanojevic V, Clarke WL, Habener JF: Pancreatic agenesis attributable to a single nucleotide deletion in the human IPF1 gene coding sequence.
Nat Genet
 
15
:
106
–110,
1997
12.
Hansen L, Urioste S, Petersen HV, Jensen JN, Eiberg H, Barbetti F, Serup P, Hansen T, Pedersen O: Missense mutations in the human insulin promoter factor-1 gene and their relation to maturity-onset diabetes of the young and late-onset type 2 diabetes mellitus in Caucasians.
J Clin Endocrinol Metab
 
85
:
1323
–1326,
2000
13.
Hani EH, Stoffers DA, Chevre JC, Durand E, Stanojevic V, Dina C, Habener JF, Froguel P: Defective mutations in the insulin promoter factor-1 (IPF-1) gene in late-onset type 2 diabetes mellitus.
J Clin Invest
 
104
:
R41
–R48,
1999
14.
Macfarlane WM, Frayling TM, Ellard S, Evans JC, Allen LI, Bulman MP, Ayres S, Shepherd M, Clark P, Millward A, Demaine A, Wilkin T, Docherty K, Hattersley AT: Missense mutations in the insulin promoter factor-1 gene predispose to type 2 diabete
. J Clin Invest
 
104
:
R33
–R39,
1999
15.
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, 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
16.
Permutt AM, Bernal-Mizrachi E, Inoue H: Calpain 10: the first positional cloning of a gene for type 2 diabetes.
J Clin Invest
 
106
:
819
–821,
2000
17.
Withers DJ, Gutierrez JS, Towery H, Burks DJ, Ren JM, Previs S, Zhang Y, Bernal D, Pons S, Shulman GI, Bonner Weir S, White MF: Disruption of IRS-2 causes type 2 diabetes in mice.
Nature
 
391
:
900
–904,
1998
18.
Porzio O, Federici M, Hribal ML, Lauro D, Accili D, Lauro R, Borboni P, Sesti G: The Gly972→Arg amino acid polymorphism in IRS-1 impairs insulin secretion in pancreatic beta cells.
J Clin Invest
 
104
:
357
–364,
1999
19.
Fritsche A, Stefan N, Hardt E, Schützenauer S, Häring H, Stumvoll M: A novel hyperglycemic clamp for characterization of islet function in humans: assessment of three different secretagogues, maximal insulin response and reproducibility.
Eur J Clin Invest
 
30
:
411
–418,
2000
20.
Ward WK, Bolgiano DC, McKnight B, Halter JB, Porte D Jr: Diminished B cell secretory capacity in patients with noninsulin-dependent diabetes mellitus.
J Clin Invest
 
74
:
1318
–1328,
1984
21.
Yki-Järvinen H: Role of insulin resistance in the pathogenesis of NIDDM.
Diabetologia
 
38
:
1378
–1388,
1995
22.
Medici F, Hawa M, Pyke DA, Leslie RD: Concordance rate for type II diabetes mellitus in monozygotic twins: actuarial analysis.
Diabetologia
 
42
:
146
–150,
1999
23.
Köbberling J, Tillil H: Empirical risk figures for first-degree relatives of non-insulin-dependent diabetics. In
The Genetics of Diabetes Mellitus
Köbberling J, Tattersall R, Eds. London, Academic Press,
1982
, p.
202
–209
24.
O’Rahilly S, Turner RC, Matthews DR: Impaired pulsatile secretion of insulin in relatives of patients with non-insulin-dependent diabetes.
N Engl J Med
 
318
:
1225
–1230,
1988
25.
Buchanan T, Catalano P: The pathogenesis of GDM: implications for diabetes after pregnancy.
Diabetes Rev
 
3
:
584
–601,
1995
26.
Pendergrass M, Fazioni E, DeFronzo R: Non-insulin-dependent diabetes mellitus and gestational diabetes mellitus. Same disease, another name?
Diabetes Rev
 
3
:
566
–583,
1995
27.
Zhang Y, Wat N, Stratton IM, Warren Perry MG, Orho M, Groop L, Turner RC: UKPDS 19: heterogeneity in NIDDM: separate contributions of IRS-1 and beta 3-adrenergic-receptor mutations to insulin resistance and obesity respectively with no evidence for glycogen synthase gene mutations: UK Prospective Diabetes Study.
Diabetologia
 
39
:
1505
–1511,
1996
28.
Koch M, Rett K, Volk A, Maerker E, Haist K, Deninger M, Renn W, Haring HU: Amino acid polymorphism Gly 972 Arg in IRS-1 is not associated to lower clamp-derived insulin sensitivity in young healthy first degree relatives of patients with type 2 diabetes.
Exp Clin Endocrinol Diabetes
 
107
:
318
–322,
1999
29.
Hager J, Zouali H, Velho G, Froguel P: Insulin receptor substrate (IRS-1) gene polymorphisms in French NIDDM families (Letter).
Lancet
 
342
:
>1430,
1993
30.
Imai Y, Philippe N, Sesti G, Accili D, Taylor SI: Expression of variant forms of insulin receptor substrate-1 identified in patients with noninsulin-dependent diabetes mellitus.
J Clin Endocrinol Metab
 
82
:
4201
–4207,
1997
31.
Ura S, Araki E, Kishikawa H, Shirotani T, Todaka M, Isami S, Shimoda S, Yoshimura R, Matsuda K, Motoyoshi S, Miyamura N, Kahn CR, Shichiri M: Molecular scanning of the insulin receptor substrate-1 (IRS-1) gene in Japanese patients with NIDDM: identification of five novel polymorphisms.
Diabetologia
 
39
:
600
–608,
1996
32.
Almind K, Bjorbaek C, Vestergaard H, Hansen T, Echwald S, Pedersen O: Aminoacid polymorphisms of insulin receptor substrate-1 in non-insulin-dependent diabetes mellitus.
Lancet
 
342
:
828
–832,
1993
33.
Laakso M, Malkki M, Kekalainen P, Kuusisto J, Deeb SS: Insulin receptor substrate-1 variants in non-insulin-dependent diabetes.
J Clin Invest
 
94
:
1141
–1146,
1994
34.
Stumvoll M, Fritsche A, Stefan N, Hardt E, Häring H: Evidence against a rate-limiting role of proinsulin processing for maximal insulin secretion in subjects with impaired glucose tolerance and β-cell dysfunction.
J Clin Endocrinol Metab
 
86
:
1235
–1239,
2001
35.
Fritsche A, Stefan N, Hardt E, Häring H, Stumvoll M: Characterisation of beta-cell dysfunction of impaired glucose tolerance: evidence for impairment of incretin-induced insulin secretion.
Diabetologia
 
43
:
852
–858,
2000

Address correspondence and reprint requests to [email protected].

Accepted for publication 21 June 2001.

ANOVA, analysis of variance; AUC, area under the curve; GDM, gestational diabetes mellitus; GLP, glucagon-like peptide; IGT, impaired glucose tolerance; IPF, insulin promoting factor; IRS, insulin receptor substrate; OGTT, oral glucose tolerance test.

The symposium and the publication of this article have been made possible by an unrestricted educational grant from Servier, Paris.

2002