Type 2 Diabetes Is Associated with a Different Pattern of Serum Polyamines: A Case–Control Study from the PREDIMED-Plus Trial

Objective: Polyamines are naturally occurring cationic molecules present in all living cells. Dysregulation of circulating polyamines has been reported in several conditions, but little is known about the levels of serum polyamines in chronic metabolic disorders such as type 2 diabetes (T2D). Therefore, the aim of this study was to evaluate the polyamine-related metabolome in a cohort of metabolic syndrome individuals with and without T2D. Design and methods: This was a nested case–control study within the PREDIMED-Plus trial that included 44 patients with T2D and 70 patients without T2D. We measured serum levels of arginine, ornithine, polyamines, and acetyl polyamines with an ultra-high performance liquid chromatography tandem mass spectrometry platform. Results: Our results showed that serum putrescine, directly generated from ornithine by the catalytic action of the biosynthetic enzyme ornithine decarboxylase, was significantly elevated in patients with T2D compared to those without T2D, and that it significantly correlated with the levels of glycosylated hemoglobin (HbA1c). Correlation analysis revealed a significantly positive association between fasting insulin levels and spermine. Multiple logistic regression analysis (adjusted for age, gender and body weight index) revealed that serum putrescine and spermine levels were associated with a higher risk of T2D. Conclusions: Our study suggests that polyamine metabolism is dysregulated in T2D, and that serum levels of putrescine and spermine are associated with glycemic control and circulating insulin levels, respectively.

Emerging evidence suggests that increased levels of polyamines in white adipose tissue, liver or skeletal muscle could stimulate energy expenditure and confer resistance to diet-induced obesity and non-alcoholic fatty liver disease [12,13]. Remarkably, polyamine metabolism has been directly implicated in adipogenesis [20][21][22], suggesting that increased polyamine levels might be involved in adipose tissue expansion in obesity. In diet-induced obesity mouse models, a high-dose daily administration of either spermidine or spermine has been shown to be an effective strategy for weight loss and improvement of glycemic status [23][24][25], indicating a possible therapeutic role for exogenous polyamines in metabolic disorders. In addition, emerging evidence has suggested a beneficial effect of chronic low-dose oral spermidine administration on cardiovascular disease in hypertensive experimental models [26]. In humans, epidemiological studies have shown association of a high consumption of dietary spermidine with a reduction in cardiovascular events and decreased mortality [26,27].
While the above discussion indicates that increased polyamine levels may be clinically beneficial, several studies support a contrary conclusion. For example, polyamine levels in human adipose tissue are elevated in insulin-resistant morbidly obese subjects as compared to their insulin-sensitive counterparts [28]. Furthermore, blood polyamines have been reported to be significantly higher in obese subjects as compared with normal weight controls [29]. However, the role of polyamines in these pathological conditions and in type 2 diabetes (T2D), frequently associated with obesity and insulin resistance, is not fully understood. In the pancreas, polyamines are mainly located within the secretory granules of the islet β cells [30], where they have been implicated in proinsulin biosynthesis and insulin secretion [31]. Islet polyamine levels diminish with both age and obesity [18], suggesting that alterations in the intracellular levels of polyamines might affect β cell function. Remarkably, transgenic mice which overexpress the polyamine catabolic enzyme spermidine/spermine acetyltransferase (SSAT), which contains higher levels of putrescine and reduced levels of spermidine and spermine in the pancreatic islets, display impaired glucose-stimulated insulin secretion [11].
Taking these various studies together, it is clear that further information is needed to confirm the association of polyamines with diseases such as diabetes and obesity. The aim of this study was to determine whether serum levels of polyamines are indeed positively associated with T2D and with the degree of glycemic control in a cohort of subjects with metabolic syndrome.

Study Design and Participants
This is a cross-sectional analysis of 114 subjects (>55 years) included in the PREDIMED-Plus trial (clinical trial registration number 89898870) [32], who were consecutively recruited at the Virgen de la Victoria University Hospital (Malaga, Spain). All study subjects met at least three components of the metabolic syndrome (updated harmonized criteria of the International Diabetes Federation and the American Heart Association and National Heart, Lung and Blood Institute) [33], and all were overweight or obese (body mass index, BMI 27-40 kg/m 2 ).
Out of these 114 subjects, 44 participants had T2D and 70 participants were non-diabetic. T2D was defined by HbA1c levels ≥ 6.5% or fasting plasma glucose ≥ 126 mg/dL according to the American Diabetes Association (ADA) guidelines [34]. All participants provided written informed consent, and the study protocol and procedures were approved according to the ethical standards of the Declaration of Helsinki by the Research Ethics Committees from all the participating institutions.

Laboratory Measurements
Blood samples were obtained from the antecubital vein after an overnight fast and placed in vacutainer tubes after an overnight fast. Excessive agitation of blood specimens was avoided to minimize hemolysis and serum was immediately separated from the blood samples by centrifugation for 10 min at 4000 rpm and frozen at −80 • C until analysis. Enzymatic methods (Randox Laboratories Ltd., Crumlin, UK) were employed to analyze the levels of serum cholesterol, triglycerides, HDL-cholesterol, glucose, and glycosylated hemoglobin (HbA1c) using a Dimension Vista autoanalyzer (Siemens Healthcare Diagnostics, Erlangen, Germany). Serum insulin levels were measured by immunoassay using an ADVIA Centaur autoanalyzer (Siemens Healthcare Diagnostics). Insulin resistance (IR) was calculated from the homeostasis model assessment of IR (HOMA-IR) with the formula: HOMA-IR = [fasting serum insulin (µU/mL) × fasting blood glucose (mmol/L)]/22.5.

Sample Preparation and Ultra-High Performance Liquid Chromatography Tandem Mass Spectrometry (UHPLC-MS/MS) Analysis of Serum Polyamine Levels
Fifty microliters of serum were aliquoted to a 1.5 mL Eppendorf LoBind tube and mixed with 5 µL of internal standard and 167 µL of methanol. Protein in the samples was precipitated by vortexing for 1 min. A volume of 334 µL of chloroform was added and the mixture was vortexed for 1 min and centrifuged for 10 min at 15,000 rpm and 4 • C. The upper layer was collected and transferred to a tube. In order to derivatize the sample, 100 µL of carbonate-bicarbonate buffer (pH 9) and 50 µL of dansyl chloride (10 mg/mL in acetone) were added. The mixture was vortexed and placed in the dark for 1 h at room temperature. The compounds were extracted with 250 µL of ethyl acetate twice; prior to the second extraction, 2.5 µL of trifluoroacetic acid were added. The combined organic phases were evaporated in a SpeedVac at 45 • C and stored at −20 • C until analysis. Samples were reconstituted in 50 µL of ammonium acetate 0.2 M:acetonitrile (30:70).
Chromatography was performed with an Agilent UHPLC 1290 series binary pump (Agilent Technologies, Santa Clara, CA, USA), and separation was carried out on a Kinetex EVO C18 column (2.6 µm particle size, 2.1 mm internal diameter × 150 mm length) (Phenomenex, Torrance, CA, USA) held at 25 • C. The mobile phase for elution was a gradient established between water acidified with 0.1% formic acid (A) and acetonitrile acidified with 0.1% formic acid (B) at a flow rate of 400 µL/min. The injected amount was 2.5 µL.
MS/MS analysis was performed on an Agilent QqQ 6490 Series mass spectrometer operating in AJS + ESI. The ionization source parameters were optimized using MassHunter Optimizer (Agilent Technologies, version 6.0) as follows: nebulizer gas (nitrogen) with a pressure of 15 psi, a gas flow of 15 L/min at 200 • C, a sheath gas flow of 11 L/min at 350 • C, a capillary voltage of 2.5 kV, and a nozzle voltage of 1000 V.

Statistical Analysis
Results are given as the mean ± SD. To test normality of the study variables, a Kolmogorov-Smirnov test was applied. Comparisons between study groups of the parameters with and without a normal distribution were performed using Student's t-test and Mann-Whitney U test, respectively. Spearman correlation analyses were performed to determine associations between serum polyamine levels with several anthropometric and biochemical variables. Multiple linear regression analysis was performed to determine the associations between HbA1c and HOMA-IR with clinical variables and serum polyamines. A multiple logistic regression model was used to estimate odds ratios (OR) and 95% confidence intervals (CI) of gender, age, BMI, and serum putrescine levels, with T2D as dependent variable. All statistical analyses were performed with SPSS version 22.0 for Windows (SPSS Iberica, Madrid, Spain). p < 0.05 was considered as statistically significant.

Results
Clinical characteristics in patients with T2D and without T2D are shown in Table 1. While there were no significant differences in age, blood pressure, HDL-cholesterol or triglyceride levels between study groups, patients with T2D showed significantly higher BMI, waist circumference, HbA1c levels, fasting glucose, insulin, and HOMA-IR. Among the measured polyamine-related metabolites, only arginine, ornithine, and putrescine were significantly different between T2D and non-diabetic subjects ( Table 2). Whereas the serum levels of arginine and ornithine were lower in T2D subjects, putrescine, which is directly produced from ornithine by the enzymatic activity of ornithine decarboxylase (ODC) (Figure 1), was increased. No significant differences in the serum levels of other polyamines such as spermidine and spermine, or their acetylated derivatives acetylputrescine, N 1 -acetylspermidine, N 8 -acetylspermidine, N 1 ,N 8 -diacetylspermidine, N 1 -acetylspermine, and N 1 ,N 12 -diacetylspermine were detected in diabetic patients as compared to those without T2D.  Data are mean ± SD or n (%). p values were calculated for the difference between study groups using either Student's t-test or Mann-Whitney U test depending on the normality of the clinical variables. p < 0.05 was considered significant. Spearman correlation analysis revealed a significant positive correlation between HbA1c and serum putrescine levels ( Figure 2). However, no correlation between putrescine, fasting glucose or other clinical parameters such as BMI or HOMA-IR was observed. In addition, serum levels of arginine were negatively correlated with HOMA-IR, whereas there was a significant positive association between serum spermine levels and HOMA-IR ( Figure 2). Interestingly, spermine significantly correlated with insulin levels (Figure 2) but not with glucose levels (the two variables used to calculate HOMA-IR index). No significant correlation was observed between BMI and the serum levels of polyamines or related metabolites. Multiple linear regression analysis, performed after adjusting for age and gender, showed a significant positive association between HbA1c and putrescine (Table 3), and between HOMA-IR and spermine (Table 4). Further multiple logistic regression analysis (after adjusting for age, gender, and BMI) revealed that both putrescine and spermine levels were associated with a higher risk of T2D (Table 5). Association between putrescine and T2D were still found when fasting insulin was included in the model (Supplementary Table S1). However, no association between serum polyamines and T2D was observed when either HOMA-IR or fasting glucose was added in the logistic model (Supplementary Table S1). Spearman correlation analysis revealed a significant positive correlation between HbA1c and serum putrescine levels ( Figure 2). However, no correlation between putrescine, fasting glucose or other clinical parameters such as BMI or HOMA-IR was observed. In addition, serum levels of arginine were negatively correlated with HOMA-IR, whereas there was a significant positive association between serum spermine levels and HOMA-IR ( Figure 2). Interestingly, spermine significantly correlated with insulin levels (Figure 2) but not with glucose levels (the two variables used to calculate HOMA-IR index). No significant correlation was observed between BMI and the serum levels of polyamines or related metabolites. Multiple linear regression analysis, performed after adjusting for age and gender, showed a significant positive association between HbA1c and putrescine (Table 3), and between HOMA-IR and spermine (Table 4). Further multiple logistic regression analysis (after adjusting for age, gender, and BMI) revealed that both putrescine and spermine levels were associated with a higher risk of T2D (Table 5). Association between putrescine and T2D were still found when fasting insulin was included in the model (Supplementary Table S1). However, no association between serum polyamines and T2D was observed when either HOMA-IR or fasting glucose was added in the logistic model (Supplementary Table S1).   (1)); age (years); and serum putrescine (ng/mL).  (1)); age (years); and serum spermine (ng/mL).     (1)); age (years); and serum spermine (ng/mL).  (1)); age (years); BMI (kg/m2); putrescine (ng/mL); spermidine (ng/mL); and spermine (ng/mL).

Discussion
In this study we have investigated the circulating profile of metabolites related to polyamine metabolism in a cohort of metabolic syndrome patients with and without T2D. Analysis of the serum polyamine profile revealed significantly increased levels of putrescine in T2D patients as compared to non-diabetic subjects. This finding could indicate an upregulation of the polyamine biosynthetic pathway, more specifically, the production of putrescine from ornithine by the biosynthetic enzyme ODC. In fact, putrescine levels have been shown to be elevated in kidney and liver in certain animal models of diabetes [16,35]. Given that the circulating levels of ornithine (the precursor of putrescine) are diminished in diabetic individuals, the higher putrescine levels cannot explained by a higher availability of substrate. In contrast, despite the lack of statistical significance, the serum levels of spermidine tended to be lower in the diabetic group. Since putrescine levels are higher in T2D, this suggests that the conversion of putrescine into spermidine (by spermidine synthase; (Figure 1)) may be impaired in T2D.
Correlation analyses further revealed a significant positive association of serum spermine and with circulating insulin levels and HOMA-IR. Because of the lack of correlation between spermine and glucose levels, these results suggest that spermine levels are more strongly associated with hyperinsulinemia than with insulin resistance in overweight or obese patients with metabolic disease. In pancreatic β cells, extranuclear polyamines mainly accumulate within secretory granules [30]. Whereas putrescine and spermidine appear to affect proinsulin biosynthesis, spermine may have a stimulatory role in long-term insulin release [31]. Thus, the positive correlation between circulating insulin and spermine could be explained by the fact that both molecules are likely co-secreted by pancreatic β cells. On the other hand, recent studies have reported beneficial effects of spermine administration in animal models with impaired glucose homeostasis. For instance, treatment with exogenous spermine has been shown to be effective in decreasing body weight and fasting glucose and improves glucose tolerance in diet-induced obese mice [24]. In addition, in rat adipocytes, spermine has been shown to enhance glucose transport and the conversion of glucose into triacylglycerols [36], and to increase the ability of insulin to bind to its own receptor [37]. Taken together, these studies suggest that spermine might modulate glucose homeostasis by acting as an endogenous insulin sensitizer when it is released into the circulation together with insulin; further studies are required to test this hypothesis.
In conclusion, in this study we demonstrate that serum putrescine levels are significantly increased in T2D and that they are positively correlated with HbA1c levels in a cohort of subjects with metabolic syndrome. In addition, we show that the serum levels of spermine are significantly correlated with insulin levels and HOMA-IR, suggesting the possible relationship of circulating spermine with elevated circulating insulin levels. In order to understand how T2D can influence polyamine metabolism (or vice versa), further studies are needed. was supported by the Miguel Servet Type II program (CPI13/00003) from the ISCIII, and co-financed by the European Regional Development Fund (ERDF), and by the Andalusian Health Public System (C-0030-2018). J.C.F.-G. and F.C. were supported by research contracts from the Andalusian Health Public System (B-0003-2017 and C-0032-2016, respectively). B.R.-M. was the recipient of a Sara Borrell postdoctoral fellowship (CD16/0003) from the ISCIII, and co-financed by the European Regional Development Fund (ERDF).

Conflicts of Interest:
The authors declare no conflict of interest.