T2D and the related metabolic syndrome are the results of interaction between disease risk genes and environmental promoters. This role of gene-environment interaction in the development of metabolic disease is also observed in mice of different strains carrying in their background different disease risk genes. For example, B6 mice are prone to becoming obese as they age on either high- or low-fat diet, whereas 129 mice are relatively resistant to obesity on either diet. Furthermore, on both the low- and high-fat diet, B6 mice have higher insulin levels and inferior glucose and insulin tolerance compared with 129 mice, indicating that B6 mice develop more insulin resistance than 129 mice. This difference in susceptibility to insulin resistance is also apparent when B6 and 129 mice are subjected to a genetic insulin resistance challenge. Thus, B6 mice that are double heterozygous for knockout of the insulin receptor and Irs1 genes exhibit marked hyperinsulinemia and massive islet hyperplasia and develop early hyperglycemia, with more than 85% of mice becoming overtly diabetic by 6 months. By contrast, 129 mice with the same double heterozygous deletions show very mild hyperinsulinemia and minimal islet hyperplasia, and less than 2% of mice develop diabetes by the age of 6 months. The severe diabetes and insulin resistance in B6 double heterozygous mice versus the almost complete lack of disease in the 129 mice carrying the same mutations indicate the strength of the genetic modifiers between these strains. Identifying the specific background genes that modify insulin resistance is therefore an important challenge and could provide novel targets for therapy and prevention of T2D.
Genome-wide scanning of the F2
intercross between B6 and 129 mice identified 8 QTLs on 6 different chromosomes linked to these different responses to HFD and the double heterozygous knockout (6
). The strongest QTL linked with hyperinsulinemia/insulin resistance is on chromosome 14. Consistent with an important genetic determinant at this locus of insulin resistance, mice homozygous for the B6 allele in this region have 6-fold-higher insulin levels than mice homozygous for the 129 allele when placed on HFD, and mice heterozygous for B6 and 129 alleles had intermediate levels. Several candidate genes for insulin resistance are present in this region, the most prominent of which is Prkcd
In the present study, using 3 different in vivo models, we demonstrate that PKCδ is a major regulator of insulin resistance and is at least one of the major genetic modifiers of the diabetic risk between these two strains of mice. PKCδ also acts as a major modulator of the risk of development of hepatic steatosis between these strains of mice. We establish the importance of PKCδ not only by finding differences in expression levels, but also by showing that PKCδ can regulate whole body and hepatic insulin sensitivity and hepatic lipid accumulation through liver-specific overexpression, whole body knockout, and liver-specific reduction in PKCδ. Importantly, PKCδ expression is also increased in livers of obese and obese type 2 diabetic subjects and in obese subjects correlates with hyperglycemia and hypertriglyceridemia, suggesting that this enzyme also plays a role in the development of human metabolic syndrome.
All major insulin-responsive tissues (skeletal muscle, adipose tissue, and liver), like most tissues of the body, express multiple isoforms of each of the 3 PKC classes (25
). Through the use of transgenic and knockout approaches, several members of the PKC family have been implicated in insulin action, the development of insulin resistance, and the regulation of metabolism. In mice with deletion of the Prkca
gene, insulin signaling through IRS-1–dependent activation of PI3K and of its downstream processes including glucose transport are enhanced in muscle and adipose tissue (26
). Likewise, knockout of the Prkcb1
gene mildly enhances overall glucose homeostasis in vivo (27
). Deletion of PKCε improves glucose-stimulated insulin secretion, reduces insulin clearance, and protects against hepatic insulin resistance, whereas muscle-specific inactivation of PKCι/λ has been shown to impair glucose transport (28
). Excessive activation of the atypical PKCζ has been shown to activate SREBP1c and NF-κB and contribute to hyperlipidemia and systemic insulin resistance (29
), and we have also demonstrated the existence of divergent regulation of hepatic glucose and lipid metabolism dependent on PI3K and PKCι/ζ (30
). Studies on PKCθ inactivation have reported conflicting effects on HFD-induced insulin resistance (31
In the current study, we demonstrate that PKCδ has unique characteristics among the PKC family members that define it as a major modifier of the diabetic risk between B6 and 129 mice, as well as in obese and diabetic humans. There are at least two aspects in PKCδ expression that differentiate B6 mice from 129 mice. First, expression analysis performed on liver, muscle, and adipose tissue of B6 and 129 mice on normal chow reveals higher expression of PKCδ in all tissues from B6 compared with 129 mice at 6 months of age (6
). Indeed, PKCδ expression in liver was higher in B6 mice compared with 129 mice at 6 weeks, an age when they are metabolically indistinguishable, as well as at birth, pointing to a genetically programmed difference between these strains as opposed to a change secondary to diet, obesity, or metabolic derangement. Second, upon exposure to HFD, expression of PKCδ in the liver was further increased in B6 mice, but not increased at all in 129 mice, pointing to a genetic impact on the control of the environmental mediated regulation of PKCδ expression in the liver of B6 and 129 mice on top of the basal genetic difference. The exact cause for the difference in PKCδ expression remains unclear, but it is not due to differences in gene copy number or to difference in the sequence of the coding region of the gene. Further genetic analyses are underway to identify possible genetic differences in the promoter region of Prkcd
and/or the methylation state of a large CpG island present in the first intron of the Prkcd
gene. Whether humans also exhibit genetically programmed differences in regulation of PKCδ expression that might contribute to genetic risk of diabetes will also require further study, but obese humans with or without T2D clearly show increased expression of PKCδ in liver. Interestingly, a major locus associated with metabolic syndrome phenotypes has been identified in a region on human chromosome 3p that includes the Prkcd
Recently, several studies have indicated the importance of PKCδ in tissues involved in diabetes pathogenesis, including muscle, pancreas, and vascular tissues. PKCδ has been shown to modify glucose uptake in myotubes (34
), lipogenesis in liver (12
), development of ER stress in hepatoma cells (19
), and development of vascular changes in the retina (35
). A role of PKCδ in the regulation of insulin secretion has also been reported in two studies using PKCδ transgenic mouse models (36
), but the results of the latter were conflicting as to the direction of change. Our findings indicate that PKCδ is a significant modifier of insulin sensitivity. Indeed, in mice on a normal chow diet, whole body PKCδ-null mice exhibited an impressive 7-fold increase in glucose infusion rate during a euglycemic-hyperinsulinemic clamp, associated with increased insulin sensitivity in liver to insulin inhibition of glucose production. Such differences in glucose infusion rates could not be solely due to a reduction in HGP, and suggest an improvement in insulin sensitivity in muscle and/or adipose tissue, where most of the glucose uptake occurs. In ongoing experiments, the contribution of each of these tissues to the observed phenotype will be assessed by the use of mouse models with muscle-specific and adipose tissue–specific deletions of the Prkcd
In the normal liver, insulin suppresses Foxo1 and gluconeogenesis while stimulating SREBP1c and lipogenesis (38
). In T2D, a state of insulin resistance, insulin is unable to suppress gluconeogenesis normally, but SREBP1c and lipogenesis are elevated (39
). How SREBP1c is induced in T2D remains an important but unanswered question (40
). One possibility is that the specific branch of the insulin signaling cascade that activates lipogenesis remains sensitive to insulin, even as the branch regulating gluconeogenesis becomes resistant. Recent evidence has shown that insulin stimulates SREBP1c via mTORC1 (41
), raising the possibility that the combination of hyperglycemia and hypertriglyceridemia in T2D is due to a failure to suppress Foxo1 coupled with increased mTORC1 signaling.
Our data suggest that PKCδ may be a key player in the development of T2D. We find that both genetic and environmental insults can induce PKCδ, and this in turn produces defects in signaling through Foxo1, resulting in increased levels of gluconeogenic genes and hyperglycemia. The improvement in insulin sensitivity in liver of PKCδKO mice is linked to an improvement in the early steps of insulin signaling, in particular an increase in tyrosine phosphorylation of IRS-1 leading to increased signaling down the PI3K/Akt pathway. This improvement is associated with a decrease in the phosphorylation of IRS-1 on Ser307, one of the end points of a signaling cascade attempting to reestablish insulin sensitivity in insulin-resistant liver (42
). The effect of PKCδ appears to be mediated through p70S6K, which shows enhanced phosphorylation/activation in states where PKCδ levels are high and decreased phosphorylation/activation in states where PKCδ levels are low. On the other hand, PKCδ promotes signaling to SREBP1c. This appears to occur via mTORC1 as indicated by activation of the mTORC1 target p70S6K.
PKCδ may also affect insulin sensitivity by regulation of inflammation. For example, in hepatocytes in vitro, PKCδ serves as an intermediate in response to TNF-α activation of NF-κB and the ER stress response (19
). PKCδ has also been shown to play a role in production of ROS in adipocytes from HFD-treated animals, and this could further contribute to insulin resistance and risk of T2D (43
). PKCδ could also participate in the development of inflammation in adipose tissue, as PKCδ activation in mesenteric fat has been shown to lead to secretion of the proinflammatory cytokine IL-6 by adipocytes (45
). In this study, we observed that PKCδ-deficient mice exhibit protection against adipose tissue inflammation associated with insulin resistance (Supplemental Figure 3E). Moreover, using Gene Network Enrichment Analysis (GNEA) to integrate gene expression data for different tissues from B6 versus 129 mice of varying ages and dietary conditions, we have recently shown significant differences in expression of immune system–related genes in the adipose tissue of B6 versus 129 mice, even prior to the onset of obesity (17
). This results in elevated expression of inflammatory markers in B6 as compared with 129 mice and increased infiltration of adipose tissue with macrophages and T cells. As mice age, or when they are subjected to HFD, the disparity in the inflammatory status of the adipose tissue becomes even more pronounced. Thus, the differences in the repertoire of inflammatory cells in the fat tissue are associated with the predisposition to metabolic diseases. This study together with our present findings demonstrate how PKCδ can serve as a major mediator of insulin resistance, as well as the deleterious effects of inflammation on insulin signaling, in liver and other insulin-sensitive tissues.
Some studies have also reported that PKCδ may be a target of or play a role in insulin action. In L6 myotubes, insulin is capable of regulating PKCδ protein levels through both transcriptional (46
) and post-transcriptional (47
) mechanisms. Insulin also increases the amount of PKCδ within the nuclear fraction of L6 myotubes. Finally, some studies suggest that insulin can induce PKCδ activation through Src tyrosine kinase, and that PKCδ plays a positive role in insulin-stimulated glucose transport (34
). Whatever the exact mechanisms involved in these potentially positive roles of PKCδ, our results indicate that the major effect of high PKCδ is inhibition of insulin action in vivo and that deletion of PKCδ will improve insulin action and steatosis in liver and lead to an overall improvement in glucose tolerance and increase in whole body insulin sensitivity, as demonstrated by the euglycemic-hyperinsulinemic clamp. These finding are supported by a recent study showing that inhibition of PKCδ in mice under HFD conditions improves their glucose tolerance (12
In addition to PKCδ, several other PKCs are present and differentially regulated in liver in response to stimuli such as HFD and therefore may also participate in the regulation of hepatic metabolism and insulin sensitivity. For example, PKCα is induced by HFD treatment and can act to inhibit insulin signaling via phosphorylation of IRS-1 (51
). PKCε has also been shown to play a role in hepatic insulin resistance (52
), and our results show that PKCε expression is also enhanced in liver of HFD-treated animals and/or ob/ob
mice. Indeed, the whole subclass of novel PKC isozymes is induced by most environmental stressors associated with the development of hepatosteatosis and hepatic insulin resistance. Thus, one may need to develop agents that target multiple members of this group if one is to use these as targets to treat the hepatic dysregulation associated with the metabolic syndrome.
However, PKCδ appears to play a unique role as a modifier of insulin sensitivity. First, its expression is genetically programmed, allowing PKCδ to serve as an inherited modifier of insulin sensitivity in mice. Second, regulation of PKCδ expression is also genetically controlled, increasing in insulin resistance–prone, but not insulin resistance–resistant, strains of mice subjected to HFD. Third, the PRKCD gene is located in a region of human chromosome 3 previously linked to the metabolic syndrome, and PKCδ expression is upregulated in human obese and obese diabetic subjects to an extent similar to that seen in obesity- and insulin resistance–prone strains of mice. Finally, relatively moderate differences in expression of PKCδ can change the risk for development of whole body and liver insulin resistance and the risk for development of hepatic steatosis. Thus, PKCδ represents a strong candidate for pharmaceutical treatment of insulin resistance, T2D, and the metabolic syndrome, as well as their associated complications such as hepatic steatosis.