Association of nicotinamide-N-methyltransferase mRNA expression in human adipose tissue and the plasma concentration of its product, 1-methylnicotinamide, with insulin resistance

Abstract

Aims/hypothesis: Nicotinamide-N-methyltransferase (NNMT) was recently shown to be upregulated in mouse models of insulin resistance and obesity. So far, it is unknown whether NNMT is regulated in human disease. We have explored the hypothesis that white adipose tissue (WAT) NNMT expression and plasma 1-methylnicotinamide (MNA) concentration are increased in human insulin resistance and type 2 diabetes.

Methods: NNMT expression and plasma MNA were analysed in three groups of individuals: (1) 199 patients undergoing abdominal surgery; (2) 60 individuals on a 12-week exercise programme and (3) 55 patients on a two-step bariatric surgery programme.

Results: Patients with manifest type 2 diabetes have a significantly (approximately twofold) higher NNMT expression both in omental and subcutaneous WAT compared with controls. Notably, plasma MNA correlated significantly with WAT NNMT expression in patients with type 2 diabetes (women, r = 0.59, p < 0.001; men, r = 0.61, p < 0.001) but not in healthy control individuals. In insulin-resistant individuals, there was an inverse correlation between insulin sensitivity and plasma MNA (r = 0.44, p = 0.01) or adipose tissue NNMT mRNA (r = 0.64, p < 0.001). The latter association was confirmed in a second cohort (n = 60, r = 0.78, p < 0.001). Interventions improving insulin sensitivity--exercise and bariatric surgery--were associated with a significant (p < 0.001) reduction in WAT NNMT expression. Bariatric surgery was also associated with a significant decrease in plasma MNA.

Conclusions/interpretation: We demonstrate that WAT NNMT expression is regulated in human insulin resistance and type 2 diabetes and that plasma MNA correlates with increased tissue NNMT expression and the degree of insulin resistance, making it a potential biomarker for loss of insulin sensitivity.

Fig. 1

Differences in WAT NNMT expression with type 2 diabetes and BMI. (a) Expression of NNMT in either omental (light-grey bars) or subcutaneous (dark-grey bars) adipose tissue was significantly higher in individuals with type 2 diabetes compared with non-diabetic individuals. (b) There was no correlation between omental WAT NNMT expression and BMI (r ² = 0.001). AU, arbitrary units; T2D, type 2 diabetes. **p < 0.01, *p < 0.05 for indicated comparisons

Fig. 2

Correlation between plasma MNA concentration and NNMT expression in omental (a–d) and subcutaneous (e–h) WAT. Circles, women; triangles, men; black symbols, non-diabetic individuals; white symbols, individuals with type 2 diabetes. Data were log₁₀-transformed to achieve normal distribution Correlation coefficients were as follows: (a) r ² = 0.04, NS; (b) r ² = 0.09, NS; (c) r ² = 0.35, p < 0.001; (d) r ² = 0.37, p < 0.001; (e) r ² = 0.01, NS; (f) r ² = 0.01, NS; (g) r ² = 0.39, p < 0.001; (h) r ² = 0.35, p < 0.001. AU, arbitrary units; scWAT, subcutaneous WAT

Fig. 3

Association of adipose tissue NNMT expression and insulin resistance for the subgroup of patients for whom there were clamp data available (n = 85). (a) Expression of NNMT in omental (light-grey bars) but not subcutaneous (dark-grey bars) adipose tissue was significantly higher in individuals with GIR <50 μmol kg⁻¹ min⁻¹ (***p < 0.001 vs GIR ≥50 μmol kg⁻¹ min⁻¹). (b) Negative correlation of tissue NNMT expression with insulin sensitivity in insulin-resistant individuals (GIR <50 μmol kg⁻¹ min⁻¹, n = 32, r ² = 0.41, p < 0.001). (c) Positive correlation between NNMT expression in omental adipose tissue and plasma MNA concentrations in insulin-resistant individuals (r ² = 0.44, p < 0.001). (d) Negative correlation of GIR and plasma MNA in insulin-resistant individuals (r ² = 0.19, p < 0.05). AU, arbitrary units

Fig. 4

Interventions that improve insulin sensitivity are associated with reduced adipose tissue NNMT expression and plasma MNA concentrations. (a) NNMT expression in subcutaneous adipose tissue before (light-grey bars) and after exercise intervention (dark-grey bars) for individuals with NGT, IGT or type 2 diabetes (n = 20 each; ***p < 0.001 vs pre-exercise, paired t test). (b) Intra-individual change in NNMT expression (percentage of pre-exercise expression, **p < 0.001 vs NGT, paired t test) upon exercise. (c) Correlation between adipose tissue NNMT expression and GIR during hyperinsulinaemic–euglycaemic clamp before the exercise programme (r ² = 0.61, p < 0.001) and (d) after the exercise programme (r ² = 0.50, p < 0.001). (e) Omental and subcutaneous WAT NNMT expression in the context of a two-step bariatric surgery (n = 45; ***p < 0.001 vs first surgery, paired t test). (f) Change in plasma MNA concentration upon two-step bariatric surgery (n = 38; **p = 0.001 vs first surgery, paired t test). Light-grey bars, samples obtained at first surgery; dark-grey bars, samples obtained at second surgery 12 ± 2 months later. AU, arbitrary units; scWAT, subcutaneous WAT; T2D, type 2 diabetes