Hepatic insulin resistance manifested as impaired suppression of glucose production is a key metabolic derangement that contributes to fasting hyperglycemia and increased HbA1c levels in type 2 diabetes (1). Genetic manipulation to overexpress the key enzymes in gluconeogenesis (particularly PEPCK) in preclinical models has shown that a primary defect in this system can result in increased hepatic glucose production (2,3) but, importantly, can lead to defects in insulin resistance in other tissues (muscle/fat) and impairments in insulin secretion (4,5). Despite this, medications that specifically target hepatic glucose overproduction in diabetes are lacking. Currently, the only medication (other than insulin) that targets the liver is the biguanide metformin, and despite being used for more than half a century, the mechanism by which it does this is still not completely understood (6–9). Thus, understanding the mechanism causing hepatic insulin resistance will lead to more effective and targeted therapeutic options.
Cyclin D1 is involved in cell cycle regulation and DNA synthesis and has been shown to be upregulated in a variety of tumors (including breast cancer). In addition, a number of articles has shown that cyclin D1 can affect metabolic processes, particularly glycolysis, lipogenesis, and mitochondrial function (10–12). Specifically, the loss of cyclin D1 caused an upregulation of hexokinase II, pyruvate kinase, fatty acid synthase, and acetyl-CoA carboxylase, leading to increased lipogenesis. Furthermore, a reduction in cyclin D1 was associated with increased mitochondrial function as well as mitochondrial size. All these changes can be reversed by (over)expression of cyclin D1, lending support to the thesis that this gene that is conventionally linked to proliferation and cancer may also play a role in metabolism.
In this issue of Diabetes, Bhalla et al. (13) investigated the role of cyclin D1 in the regulation of gluconeogenesis and glucose metabolism. In a series of elegant in vivo and in vitro genetic and pharmacological interventions, this study showed that the absence of cyclin D1 increased the expression of genes associated with oxidative phosphorylation in addition to the gluconeogenic genes PEPCK and G6Pase in the liver. Conversely, hepatic overexpression of cyclin D1 caused a downregulation of these gluconeogenic genes, in addition to reducing plasma glucose levels, particularly in the fasted state. Furthermore, the authors explored the mechanism by which cyclin D1 caused the changes in gluconeogenic and oxidative phosphorylation genes and showed that this is predominantly through its binding partner cyclin-dependent kinase 4 (Cdk4), regulating the expression of peroxisome proliferator–activated receptor γ coactivator (PGC)-1α. Thus relatively convincingly, Bhalla et al. show that cyclin D1 can regulate hepatic gluconeogenic gene expression via PGC1α. Indeed, a recent study by Lee et al. (14) corroborated the role of the cyclin D1/Cdk4 complex in controlling gluconeogenic gene expression and hepatic glucose production via phosphorylation of the acetyltransferase GCN5 regulating the acetylation state of PGC1α (Fig. 1). Furthermore, they showed that both dietary amino acids and insulin increased cyclin D1 levels, which consequently reduced gluconeogenic gene expression (14).
To provide relevance to diabetes, Bhalla et al. interrogated a publicly available microarray data set (NCBI GEO accession GSE15653) performed in liver samples from lean control, obese nondiabetic, and obese diabetic subjects (both well-controlled and poorly controlled individuals). It is important to note that we are not provided with any details about these patients (whether they are BMI and sex matched, HbA1c levels, duration of diabetes, medications, etc.). The data presented in Fig. 1H and I in Bhalla et al. suggest that cyclin D1 expression levels were decreased while PEPCK and G6Pase levels were increased in livers of patients with type 2 diabetes. There is no confirmation of the data using quantitative PCR (or some other independent method) to provide confidence in these results. In addition, the authors showed that liver transgenic cyclin D1 mice had reduced gluconeogenic enzyme and fasting glucose levels (see Fig. 3 in Bhalla et al.). It would have been useful to know whether these transgenic mice were protected from the expected increase induced by high-fat feeding in gluconeogenesis. Interestingly, Lee et al. (14) showed that cyclin D1 levels were elevated in C57BL/Ks db/db and high-fat–fed mice and that further increasing in cyclin D1 levels reduced gluconeogenic gene expression and improved hepatic insulin sensitivity. The reason for the discrepancy with the current study showing decreased cyclin D1 levels in livers of individuals with type 2 diabetes is not clear and deserves additional examination.
Curiously, earlier studies suggested that the absence of cyclin D1 was associated with enhanced glycolysis (as assessed by increased hexokinase II and pyruvate kinase expression levels) and lipogenesis in the liver, while the study by Bhalla et al. showed increased gluconeogenic enzymes. These data seem to be contradictory as glycolysis/lipogenesis and gluconeogenesis would be expected to be regulated in opposite directions. Whether the source of the cyclin D1–deficient mice is different or the background strain (which we know can significantly influence the metabolic phenotype of transgenic/knockout mouse [15,16]) has changed over time is not clear.
What then is the importance of cyclin D1 in the etiology, or perhaps treatment of, hepatic insulin resistance in type 2 diabetes? It is difficult to assess this from the data presented by Bhalla et al. (13) for the reasons discussed here. It is of interest that the lack of cyclin D1 can induce PEPCK and G6Pase, while overexpression elicits the opposite response. The study by Lee et al. (14) suggests that increasing cyclin D1 levels can indeed improve glycemic control in preclinical animal models of insulin resistance and diabetes. However, caution needs to be exercised because of the established proliferative/tumorigenic effect of cyclin D1, as we know obesity and diabetes are associated with increased risk of certain cancers (including breast cancer). If cyclin D1 levels are reduced in type 2 diabetes, it would be logical that the therapeutic strategy would be to induce cyclin D1. However, would this increase the risk of cancer further? Given the recent concern of incretin-based therapies in diabetes and cancer (17), further investigation and caution are warranted.
See accompanying article, p. 3266.
Funding. S.A. is supported by a Senior Research Fellowship from the National Health and Medical Research Council of Australia.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.