Introduction
Sirtuin 1 (SIRT1), is a (NAD)-dependent deacetylase, a member of the sirtuin family [
1]. SIRT1 can deacetylate many histone and nonhistone proteins. Therefore, it is involved in the regulation of multiple physiological processes, including substrate metabolism [
1,
2].
SIRT1 may influence insulin signaling in multiple insulin sensitive cells [
3]. It increases insulin receptor substrate 2 (IRS2) and protein kinase B (PKB, known as Akt) phosphorylation in response to insulin whereas it decreases the expression of protein-tyrosine phosphatase 1B (PTP1B), a negative regulator of insulin signalling [
4‐
6].
Some studies indicate that SIRT1 enhances insulin signalling at least in part because of its antiinflammatory effects in AT. SIRT1 overexpression prevents macrophage accumulation caused by high-fat feeding [
7]. SIRT1 represses proinflammatory gene expression in adipocytes, possibly through nuclear factor κB (NFκB) deacetylation and inhibition of binding to its target gene promoters [
8]. It also increases adiponectin synthesis/secretion [
9].
SIRT1 may also influence skeletal muscle metabolism through deacetylation of peroxisome proliferator-activated receptor gamma coactivator-1-α (PGC1-α), a mitochondrial fatty acid oxidation activator [
10].
In the context of the above data, it was hypothesized that SIRT1 influences insulin sensitivity [
11] and may be a therapeutic target in the prevention and treatment of disorders related to insulin resistance [
9,
12,
13]. Resveratrol, a SIRT1 activator, had a protective effect against diet-induced insulin resistance in mice [
5,
14] and decreased plasma glucose and triglycerides, homeostatic model assessment (HOMA) index and inflammation markers in humans with obesity [
15]. However, in the studies with SIRT1 overexpression in multiple tissues in mice, both improved and unchanged glucose tolerance was observed [
16,
17].
It remains unclear as to what the possible mechanism of beneficial SIRT1 action is on insulin sensitivity and which tissue is its major target. Furthermore, human data on the potential relationships of SIRT1 expression in different tissues with insulin sensitivity, especially with simultaneous assessment of SIRT1 in adipose tissue and muscle, are very limited.
Therefore, we aimed to assess AT and skeletal muscle SIRT1 expression in young male subjects in relation to body weight, insulin sensitivity; tissue SLC2A4 (encoding GLUT4), AT proinflammatory gene and ADIPOQ and muscle PGC1A expression. We also examined the regulation of tissue SIRT1 expression by hyperinsulinemia and circulating free fatty acids (FFA) elevation.
Discussion
In the present study we found that AT, but not skeletal muscle, SIRT1 expression is decreased in obesity and is positively related to whole-body insulin sensitivity. We found that the relationship between AT SIRT1 and SLC2A4 explains the correlation of the former with whole-body insulin sensitivity. Furthermore, we observed that 6 h hyperinsulinemia decreased AT and increased skeletal muscle SIRT1 expression and that both effects were negated by concurrent Intralipid/heparin infusion.
Our data indicate that AT
SIRT1 may be important for obesity-associated insulin resistance. Decreased AT
SIRT1 expression in obesity had also been reported in other human studies [
20‐
25]. In some, obese subjects had BMI in the range of morbid obesity [
22], whereas in others [
20,
21] mean BMI was similar to the value observed in our obese group. Thus, it is interesting to note that overweight subjects in our study had an AT
SIRT1 expression comparable with those in the normal-weight group, which suggests that there could be a threshold associated with body fat accumulation, which determines a decrease in AT SIRT1. Our results suggest that such threshold may be associated with insulin resistance as overweight subjects had normal insulin sensitivity, which may be due to their young age. The relationship between AT
SIRT1 and indirect indices of insulin sensitivity [
20,
21,
23] and the values from the clamp study in subjects with a family history of type 2 diabetes [
26] was also observed by other researchers. In subjects with morbid obesity, decreased AT
SIRT1 was observed in the insulin resistant vs. insulin sensitive group [
27]. We demonstrated that insulin sensitivity, and not BMI, was an independent predictor of AT
SIRT1, whereas such analysis has not been reported in other studies.
Surprisingly, no differences in muscle
SIRT1 was found among the study groups. Individuals with type 2 diabetes had lower muscle SIRT1 protein compared to control subjects [
4]. However, such difference may be secondary to diabetic conditions, as it was demonstrated in C2C12 myocytes that high glucose significantly reduces the number of SIRT1-positive nuclei and total cellular SIRT1 protein content [
28].
We also did not observe any correlations between muscle
SIRT1 and metabolic parameters. Although muscle
SIRT1 was positively related to muscle
SLC2A4 and
PGC1A expression, these associations did not seem to influence whole-body insulin sensitivity. Both increased and unchanged insulin-stimulated activation of Akt was observed in cultured myotubes with SIRT1 overexpression [
4,
5]. Furthermore, studies with muscle-specific SIRT1 overexpression in rodents showed that it did not enhance insulin-stimulated muscle glucose uptake [
29‐
31]. These data indicate that SIRT1 activation in tissues other than muscle may be important for modulating insulin action. Our data on the relationship between AT, but not skeletal muscle,
SIRT1 expression with insulin sensitivity, fall in line with this hypothesis.
Rutanen et al. [
26] observed a correlation between AT
SIRT1 mRNA expression and insulin sensitivity in subjects with a family history of type 2 diabetes. They suggested that results obtained in AT reflected metabolic changes in skeletal muscle, as they observed positive correlation between AT and muscle SIRT1 expression in a small subgroup of their study subjects (
n = 11) [
26]. Although our results are generally in agreement with those of Rutanen et al. regarding positive correlation between AT
SIRT1 mRNA and insulin sensitivity, they do not support the hypothesis about the role of skeletal muscle
SIRT1 in determining insulin action. We also did not observe any correlation between AT and muscle
SIRT1 expression. This indicates that AT SIRT1 associates with whole-body insulin sensitivity without any relation to muscle SIRT1. In our study AT
SIRT1 was also not related to the local proinflammatory gene and
ADIPOQ expression.
We found correlation between AT
SIRT1 and AT
SLC2A4 expression and this association explained the relationship between AT
SIRT1 and whole-body insulin sensitivity. SIRT1 knock down in adipocytes led to a decrease in GLUT4 translocation and glucose uptake after stimulation with insulin [
8]. AT accounts only for a small fraction of whole-body glucose disposal, however, it was demonstrated that AT-specific GLUT4 depletion resulted in profound metabolic abnormalities, including muscle and liver insulin resistance [
32]; whereas AT GLUT4 overexpression in mice with muscle GLUT4 knockout increased muscle glucose uptake [
33]. The results of other studies that in part were similar to our own, also demonstrated a decreased AT
SLC2A4 expression in animal models of type 2 diabetes [
34] and in insulin-resistant humans [
33], which may serve to support our findings. Thus, SIRT1 may influence insulin sensitivity through its effect on adipose tissue GLUT4 expression. However, the cause-effect relationship cannot be established on the basis of our data and it is also possible that lower AT
SIRT1 expression in obesity may be an effect of hyperinsulinemia, as discussed below.
We next examined the effect of hyperinsulinemia and circulating FFA elevation on tissue
SIRT1 expression. In AT, hyperinsulinemia resulted in a decrease in
SIRT1. AT SIRT1 increased in response to fasting [
22] and weight loss [
35], where a decrease in fasting insulin was also observed, however, no correlation between these changes was reported. On the other hand, an increase in AT
SIRT1 in response to weight loss without significant change in serum insulin was also observed [
20]. It was demonstrated that SIRT1 expression in various tissues, including fat, increased after caloric restriction in rats, and that insulin attenuated this response [
36]. To our knowledge, AT SIRT1 expression after hyperinsulinemia in humans has hitherto not been reported. Insulin stimulates adipogenesis and inhibits lipolysis, whereas SIRT1 exerts opposite effects, so the decrease in SIRT1 expression after insulin infusion might be important for the maintaining of adipocyte differentiation and function. It is interesting to note that this effect was present only in the normal-weight group. It is possible that mild prolonged hyperinsulinemia observed in obesity has already led to a decrease in AT
SIRT1 expression in the obese group and thus our experimental hyperinsulinemia did not promote any additional effect. We did not observe a decrease in AT
SIRT1 in response to insulin in the overweight group, despite baseline
SIRT1 and insulin sensitivity comparable to the normal-weight group. One may hypothesize that lack of response of AT
SIRT1 to insulin represents an early metabolic abnormality in this group, even before the onset of overt insulin resistance, measured as a decreased insulin-stimulated glucose uptake.
FFA elevation, obtained by Intralipid/heparin infusion, negated the insulin effect on AT
SIRT1. This negation of insulin effect may seem paradoxical, when we take into account that a high-fat diet decreases AT SIRT1 [
37], however, it may reflect insulin-desensitizing action of FFA and may indicate that AT SIRT1 response contributes to FFA-induced insulin resistance.
Muscle
SIRT1 expression increased after hyperinsulinemia. This effect was mostly expressed in the group with obesity, however, a similar tendency was also present in other groups. Insulin significantly increased SIRT1 expression in C2C12 myotubes under low glucose conditions, which was associated with an impaired insulin ability to exert myogenic-stimulating action [
28]. However, in our relatively short-term, 6 h experiment, changes in
SIRT1 may influence substrate metabolism. Therefore, we can hypothesize that an increased muscle
SIRT1 expression during hyperinsulinemia might represent an additional mechanism to maintain muscle glucose uptake.
As in AT, Intralipid/heparin infusion negated the insulin effect on muscle SIRT1 expression. Baseline tissue SIRT1 reflect rather long-term processes regulating its expression whereas the effects of insulin and FFA show short-term regulation of SIRT1 expression. Thus, although our data do not indicate muscle SIRT1 as a modulator of long-term insulin sensitivity, it is possible that it is regulated by short-term insulin fluctuations/nutrient availability to exert some metabolic effects.
The strengths of our study include: use of the “gold standard” in measurement of insulin sensitivity, i.e., the hyperinsulinemic-euglycemic clamp; analysis of AT and muscle gene expression in a human study with a large group of participants and the examination of an apparently healthy study population, which allowed us to assess early events in the development of obesity-related insulin resistance, without the confounding effects of hyperglycemia, morbid obesity and other diseases. Also, we for the first time report the relationship between AT SIRT1 and SLC2A4 as well as the tissue-specific effect of hyperinsulinemia on SIRT1 expression.
One limitation of our study arises from the fact that limited tissue availability prevented SIRT1 protein expression being measured in samples from all participants. Nevertheless, we were able to demonstrate an excellent correlation between AT and muscle SIRT1 mRNA and protein expression. We were also able to validate the major findings of our study at the protein level in subgroups of participants, both in AT and in muscle. Another limitation of our study is the fact that one cannot establish cause-effect relationship on the basis of our results.
In conclusion, our data show that AT, but not skeletal muscle, SIRT1 is associated with insulin sensitivity in healthy young humans, possibly because of its correlation with adipose tissue SLC2A4 expression. Insulin differentially regulates AT and skeletal muscle SIRT1 expression in the short-term and circulating FFA elevation negates these effects, which may be associated with lipid-induced insulin resistance.