Administration of Melatonin and Metformin Prevents Deleterious Effects of Circadian Disruption and Obesity in Male Rats

Abstract

Circadian disruption and obesity synergize to predispose to development of type 2 diabetes mellitus (T2DM), signifying that therapeutic targeting of both circadian and metabolic dysfunctions should be considered as a potential treatment approach. To address this hypothesis, we studied rats concomitantly exposed to circadian disruption and diet-induced obesity (CDO), a rat model recently shown to recapitulate phenotypical aspects of obese T2DM (eg, circadian disruption, obesity, insulin resistance, and islet failure). CDO rats were subsequently treated daily (for 12 wk) by timed oral gavage with vehicle, melatonin (a known chronobiotic), metformin, or combination treatment of both therapeutics. Melatonin treatment alone improved circadian activity rhythms, attenuated induction of β-cell failure, and enhanced glucose tolerance. Metformin alone did not modify circadian activity but enhanced insulin sensitivity and glucose tolerance. Importantly, the combination of melatonin and metformin had synergistic actions to modify progression of metabolic dysfunction in CDO rats through improved adiposity, circadian activity, insulin sensitivity, and islet cell failure. This study suggests that management of both circadian and metabolic dysfunctions should be considered as a potential preventative and therapeutic option for treatment of obesity and T2DM.

Environmental conditions associated with disruption of circadian rhythms are becoming increasingly prevalent in today's society. This is partly attributed to increased exposure to artificial light at night due to changes in working environments, shift work, and increased nightly use of electronic devices (1). Importantly, disruption of circadian rhythms is increasingly linked to diverse human pathologies, and particularly, metabolic syndrome, and type 2 diabetes mellitus (T2DM) (2–4). These observations suggest the need for development of novel therapeutic and preventative strategies to counteract deleterious effects of circadian disruption on metabolic health.

Epidemiological studies report that individuals engaged in shift work exhibit increased prevalence of obesity, metabolic syndrome, glucose intolerance, and T2DM (5, 6). Experimental work shows that acute exposure (days to weeks) to circadian misalignment in humans causes glucose intolerance and hyperglycemia associated with impaired β-cell function, insulin resistance, and increased inflammation (7, 8). Consistent with these observations, exposure of animals to shift work-like conditions disrupts behavioral, physiological and molecular circadian rhythms, and recapitulates many metabolic abnormalities seen in T2DM (3, 9, 10). Importantly, deleterious metabolic effects of circadian disruption are significantly amplified when animals are kept on energy-rich diets (eg, high-fat diet), which recapitulate the composition of obesogenic human nutrition (11–13). Therefore, successful therapeutic strategies designed to combat metabolic syndrome associated with circadian disruption will require targeting of both circadian and metabolic dysfunctions.

Melatonin is a hormone synthesized and secreted from the endocrine cells in the pineal gland, which serves as an endocrine signal contributing to entrainment/regulation of organisms' circadian rhythms (14). Melatonin production and secretion is primarily influenced by the duration and intensity of light exposure (15), and thus nightly exposure to artificial lighting suppresses melatonin release (1). Indeed, damaging effects of circadian disruption on human health have been attributed in part to inhibition of nocturnal melatonin production and release (1, 16). This marks melatonin replacement strategy as an attractive option to attenuate circadian disruption. On the other hand, metformin is one of the most commonly used diabetic pharmaceutical agents which improves metabolic function primarily by repressing endogenous glucose production and enhancing insulin sensitivity (17). Furthermore, both metformin and melatonin impart protective cellular effects on the β-cell through attenuation of oxidative stress-induced apoptosis (18, 19).

In the current study, our primary objective was to establish whether melatonin or metformin alone, or as a combination treatment, favorably modify progression of metabolic dysfunction during concomitant exposure to circadian disruption and diet-induced obesity (CDO) in rats and, if so, through which mechanisms? It was recently demonstrated that CDO rats present an optimal model to address chronotherapeutic potential of melatonin and metformin, as this novel animal model recapitulates phenotypical characteristics associated with circadian disruption, obesity and aspects of T2DM (13). Specifically, dysregulation of glucose metabolism in CDO rats is characterized by 1) melatonin-deficient state, 2) dampening of circadian rhythms, 3) increased adiposity, 4) diurnal hyperglycemia and glucose intolerance, 5) insulin resistance, and 6) islet failure (13).

Research Design and Methods

Animals

A total of 41 male Sprague-Dawley rats were purchased (Charles River Laboratories) at 2 months of age and housed under standard environmental conditions and fed chow diet. One month before initiation of experimental protocols, all rats were synchronized to standard 12-hour light, 12-hour dark (LD) cycle and habituated to routine handling, restraint, and oral gavage to reduce stress and risk of injuries associated with daily oral gavage. By convention, time-dependent experiments for all groups will be referenced based on circadian time (CT) from 0 to 24 hours. For example, CT 0 will coincide with 6 am Pacific standard time (PST) and was chosen to reflect the onset of “light” circadian cycle during the preceding 30-day acclimatization period. Subsequently, CT 12 will coincide with 6 pm PST and reflects the onset of “dark” circadian cycle during the preceding acclimatization period.

Experimental study design

At 3 months of age, rats were randomly assigned into 5 protocols for a duration of 12 weeks: 1) control: 12 weeks of standard LD cycle (lights on CT 0, lights off CT 12) on standard chow diet (14% fat, 32% protein, and 54% carbohydrates; Harlan Laboratories); 2) CDO: 12 weeks of 24 hours of constant light exposure (25-W fluorescent tubes 12” above the cage at >100 lux light intensity) fed a purified high-fat Western diet (60% fat, 20% protein, and 20% carbohydrates; Research Diets, Inc) designed to disrupt endogenous circadian rhythms and promote obesity and glucose intolerance; 3) CDO-melatonin (MEL): 12 weeks of 24 hours of constant light exposure on purified high-fat Western diet with concomitant daily 20 mg/kg of body weight (b.w.) melatonin oral gavage; 4) CDO-metformin (MET): 12 weeks of 24 hours of constant light exposure on purified high-fat Western diet with concomitant daily 150 mg/kg of b.w. metformin oral gavage; and 5) CDO-MEL-MET: 12 weeks of 24 hours of constant light exposure on purified high-fat Western diet with concomitant daily 20 mg/kg of b.w. melatonin and 150 mg/kg of b.w. metformin oral gavage. Melatonin (Q-1300; Bachem Americas) and/or metformin HCl (M258815; Toronto Research Chemicals) were suspended into ORA Plus Oral Suspending Vehicle (Perrigo) at concentrations of 4 and 30 mg/mL, respectively, and administered daily by oral gavage 30 minutes before the onset of previously maintained dark/active phase (ie, CT 12 or 6 pm PST). Control and CDO animals were gavaged daily with vehicle Ora Plus solution. Pharmacological doses for each drug were chosen based on previously reported efficacy in modulating circadian (melatonin) or metabolic (metformin) function without deleterious effects on motor performance in rats (20–22). Efficacy of drug absorption and distribution after an oral gavage was also confirmed (Supplemental Figure 1). The timing of drug administration was purposefully chosen to coincide with increased sensitivity to melatonin entrainment in rats (14). The University of California Los Angeles Institutional Animal Care and Use Committee approved all experimental procedures described in the study.

Circadian behavioral activity analyses

A subset of rats was housed in cages equipped with infrared detectors, and 24-hour circadian activity was recorded in 3-minute intervals. Circadian activity data was analyzed and plotted by using percentile normalized format in ClockLab software (Actimetrics). χ² periodogram algorithm analysis (ClockLab, Actimetrics) was used to assess circadian rhythmicity in locomotor activity as previously described (13).

Assessment of diurnal metabolic profiles

Body weights and food intake were measured weekly. During week 10 of treatment, diurnal plasma samples were collected at CT 4 (10 am PST), CT 10 (4 pm PST), CT 16 (10 pm PST), and CT 22 (4 am PST) with at least a 12-hour gap between individual blood draws on the same rat to avoid confounding effects of stress and blood loss. Also, the order of blood sampling from individual rats was changed during blood collections to avoid confounding effects of sampling bias. Blood was collected via lateral saphenous venipuncture (on previously shaved leg using sterile 21-gauge, 1-inch syringe needle) into chilled, EDTA-treated microcentrifuge tubes and immediately centrifuged at 4°C for subsequent analysis.

Assessment of diurnal glucose tolerance

During the final week of treatment, all rats underwent an oral glucose tolerance test (OGTT) performed twice in each rat at CT 4 (10 am PST) and CT 16 (10 pm PST) with at least 6 days between tests on the same rat. OGTT was administered after an acute food deprivation (2 h), and blood was collected via lateral saphenous venipuncture into chilled, EDTA-treated microcentrifuge tubes before (0 min) and 15, 30, 60, 90, and 120 minutes after oral gavage administration of sterile 50% glucose in water solution (1.5-g/kg body weight). Blood was immediately centrifuged at 4°C for subsequent collection of plasma and frozen at −80°C for future analyses.

Assessment of islet morphology and turnover

At the end of metabolic tests, rats were euthanized by isoflurane inhalation overdose and exsanguination via cardiac puncture after exposing the diaphragm. The pancreas was quickly excised and weighed, then fixed in 4% paraformaldehyde overnight at 4°C and subsequently embedded in paraffin. Deparaffinized pancreatic sections (3 sections per animal spaced through the width of the pancreas) were stained for insulin (guinea pig antiinsulin, 1:100, ab7842; Abcam) in conjunction with secondary biotin-conjugated donkey antiguinea pig antibody (1:100; Jackson ImmunoResearch) and detected using a diaminobenzidine peroxidase (HRP) substrate kit (SK-4100; Vector Laboratories). In addition, adjacent deparaffinized sections were coimmunostained for terminal deoxynucleotidyl transferase TMR-dUTP nick-end labeling (TUNEL) (In Situ Cell Death Detection kit, TMR Red, 12156792910; Roche Diagnostics) and insulin (as above) for quantification of β-cell apoptosis, and insulin and Ki-67 (mouse anti-Ki-67; BD Pharmingen) for determination of β-cell replication. All slides were cover slipped with Vectashield-DAPI mounting medium (Vector Laboratories), stored in dark at 4°C, and analyzed within 1–3 days after staining. Slides were viewed, imaged and analyzed using a Zeiss Axio Observer Z1 microscope (Carl Zeiss Microscopy, LLC) and ZenPro software (Carl Zeiss Microscopy, LLC.).

Analytical methods

Plasma glucose was measured by the glucose oxidase method using the YSI 2300 STAT Plus Glucose and Lactate Analyzer (YSI Life Sciences). Plasma insulin and C-peptide were measured in duplicate for each sample according to the manufacturer's instructions using the ALPCO diagnostics rat insulin and C-peptide ELISA.

Statistical analysis and calculations

Activity recordings were analyzed using ClockLab software (Actimetrics). Diurnal insulin resistance was estimated by using homeostasis model assessment index of insulin resistance (HOMA-IR) ([glucose] × [insulin]/405; insulin in mU/L and glucose in mg/dL) (23). Insulin sensitivity during the OGTT was estimated using Matsuda index (10 000/square root of [fasting glucose × fasting insulin] × [mean glucose × mean insulin] during OGTT), which is highly correlative with whole-body insulin sensitivity (24). β-Cell function during OGTT was estimated by previously established index as area under the curve (AUC) for C-peptide over plasma glucose during the first 30 minutes of an OGTT (25, 26). Statistical analysis was performed using ANOVA with post hoc tests wherever appropriate (GraphPad Prism v.6.0). Data are presented as mean ± SEM and assumed statistically significant at P < .05.

Results

CDO rats demonstrated characteristic phenotypic circadian, metabolic, and islet alterations consistent with previous studies (13). CDO exhibited increase in body mass gain and enhanced tissue adiposity compared with control (P < .05 vs control for body mass, and adiposity) (Figure 1). Body mass gain was, however, attenuated by approximately 20% in either melatonin or metformin treated CDO rats (P < .05) (Figure 1, A and B), whereas the combination treatment had a synergistic effect of decreasing body mass by approximately 50% compared with untreated CDO rats (P < .05 for CDO vs CDO-MEL-MET) (Figure 1, A and B). Similarly, both melatonin and metformin alone had a comparable effect on decreasing epididymal adiposity (18%–29%) and liver mass (13%), whereas combined treatment decreased epididymal adiposity by 50% and liver mass by 30% compared with CDO (P < .05) (Figure 1, C–E). Daily caloric intake was comparable in all CDO groups except in CDO-MEL-MET group, where it was modestly attenuated (P < .05 vs all groups) (Figure 1F).

Figure 1.

Melatonin and metformin synergistically attenuate body mass gain, caloric intake, and fat accumulation in CDO rats. A, Body mass gain trajectory during the 12-week study period. Mean % body mass gain (B), mean mass of epididymal (C) and retroperitoneal fat depots (D), and mean hepatic mass (E) and mean daily caloric food intake (F) in control (open bar, n = 4), CDO (black bar, n = 9), CDO-MEL (gray bar, n = 10), CDO-MET (yellow bar, n = 9), and CDO-MEL-MET (blue bar, n = 9) rats. Bar graphs represent mean ± SEM; #, P < .05 vs control; *, P < .05 vs CDO; †, P < .05 vs CDO-MEL and ‡, P < .05 vs CDO-MET.

When housed under baseline LD conditions, all rats displayed robust 24-hour circadian locomotor activity rhythms (Figure 2A) with no differences in circadian period or average daily locomotor activity (P > .05 between all groups) (Figure 2). In contrast, concomitant exposure to constant light and diet-induced obesity in CDO rats led to dampening of circadian activity rhythms (P < .05 vs control) (Figure 2). Importantly, daily melatonin administration alone, or in combination with metformin, improved circadian activity in CDO-MEL and CDO-MEL-MET groups (Figure 2, A–C and Supplemental Figure 2). On the other hand, metformin administration alone failed to impact circadian activity rhythms in CDO rats (Figure 2, A–C). Notably, no significant difference among all experimental groups were observed with respect to average total daily activity during the treatment period (P > .05) (Figure 2D).

Figure 2.

Melatonin administration improves behavioral circadian rhythms in CDO rats. A, Representative 24-hour locomotor activity (double plotted) actograms in rats monitored for 8 days at baseline under standard LD cycle followed by 6 weeks of recordings under either control, CDO, CDO-MEL, CDO-MET, and CDO-MEL-MET conditions. Shaded areas represent periods of dark. Temporal profiles are presented corresponding to CT as outlined in the methods section. B, Representative χ² periodograms of activity recordings in control, CDO, CDO-MEL, CDO-MET, and CDO-MEL-MET condition rats. Broken horizontal lines represent statistically significant threshold in determination of dominant circadian period. Note dampening of circadian rhythms in CDO and CDO-MET groups. Mean χ² periodogram-derived circadian amplitude (C) and average total daily (24 h) activity (D) during corresponding drug treatments (Rx) in control (open bar, n = 4), CDO (black bar, n = 8), CDO-MEL (gray bar, n = 8), CDO-MET (yellow bar, n = 6), and CDO-MEL-MET (blue bar, n = 6) rats. Each bar represents mean ± SEM; #, P < .05 vs control; *, P < .05 vs CDO.

Next, we assessed effects of melatonin and metformin on diurnal regulation of glucose homeostasis by examining diurnal profiles of plasma glucose, insulin, as well as calculated index of insulin resistance (HOMA-IR) (Figure 3 and Supplemental Table 1). As expected, CDO exhibited diurnal hyperglycemia (132 ± 2 vs 124 ± 1 mg/dL), overt hyperinsulinemia (613 ± 55 vs 319 ± 30 pmol/L), and insulin resistance (28 ± 2 vs 14 ± 1 HOMA-IR) evident across the 24-hour circadian day (P < .05, CDO vs control for all parameters) (Figure 3). Notably, although both melatonin and metformin alone demonstrated comparable, but modest effects on restraining hyperglycemia, hyperinsulinemia, and insulin resistance, the combination treatment had a synergistic effect on normalization of diurnal hyperglycemia (132 ± 2 vs 127 ± 1 mg/dL), hyperinsulinemia (613 ± 55 vs 323 ± 38 pmol/L), and complete reversal of insulin resistance (28 ± 2 vs 14 ± 2 HOMA-IR) (P < .05 for CDO vs CDO-MEL-MET for all parameters) (Figure 3 and Supplemental Table 1).

Figure 3.

Effects of melatonin and metformin alone, or in combination, on diurnal glucose homeostasis in CDO rats. Diurnal profiles in plasma glucose (A), plasma insulin (B), and calculated index of insulin resistance HOMA-IR (C) in control (open bar, n = 4), CDO (black bar, n = 9), CDO-MEL (gray bar, n = 10), CDO-MET (yellow bar, n = 9), and CDO-MEL-MET (blue bar, n = 9) rats. Statistical analysis (graphs on the left) was performed by two-way repeated-measures ANOVA for treatment, time, and interaction. Bar graphs (right) display mean ± SEM of AUC for measures of plasma glucose (A), insulin (B), and HOMA-IR (C) across the 24-hour circadian day in respected groups. Plasma samples were obtained at 6-hour intervals. #, P < .05 vs control; *, P < .05 vs CDO; †, P < .05 vs CDO-MEL and ‡, P < .05 vs CDO-MET.

To further delineate effects of melatonin and metformin on glucose homeostasis, we next assessed glucose tolerance, β-cell function, and insulin sensitivity after an OGTT performed on each rat at CT 4 and CT 16 time points (Figures 4 and 5). CDO rats were glucose intolerant (P < .05 vs control at CT 4 and CT 16) (Figures 4A and 5A), displayed diminished β-cell functionality (P < .05 vs control at CT 16) (Figure 5B), and exhibited significant reduction in insulin sensitivity (P < .05 vs control at CT 4 and CT 16) (Figure 5C). Notably, metabolic benefits of melatonin or metformin during OGTT displayed diurnal patterns resulting in synergistic benefits when 2 drugs were administered in combination (Figures 4 and 5). At CT 4 time point, CDO-MEL group showed improved oral glucose tolerance (∼15%, P < .05 vs CDO) (Figure 5A) associated with enhanced β-cell function (∼40%, P < .05 vs CDO) (Figure 5B) with no change in insulin sensitivity. In contrast, at CT 16, CDO-MET group exhibited improved oral glucose tolerance (∼15%, P < .05 vs CDO) (Figure 5A) associated with enhanced insulin sensitivity (∼40%, P < .05 vs CDO) (Figure 5C) and improved β-cell function (∼50%, P < .05 vs CDO) (Figure 5B). Rats treated with the combination therapy (CDO-MEL-MET) demonstrated improved glucose tolerance (∼20%, P < .05 vs CDO) and complete normalization of insulin sensitivity (∼100%, P < .05 vs CDO) irrespective of CT of OGTT (Figure 5, A–C), thus highlighting synergistic therapeutic potential of melatonin and metformin treatment under CDO conditions.

Figure 4.

Effects of melatonin and metformin alone, or in combination, on glucose, insulin, and C-peptide fluxes during an OGTT in CDO rats. Plasma glucose (A), insulin (B), and C-peptide concentrations (C) at baseline (time 0) and after an OGTT performed twice in each rat at CT 4 (left) and CT 16 (right) in control (open circle, n = 4), CDO (black circle, n = 9), CDO-MEL (gray circle, n = 10), CDO-MET (yellow circle, n = 9), and CDO-MEL-MET (blue circle, n = 9) rats. Data are presented as mean ± SEM.

Figure 5.

Effects of melatonin and metformin alone, or in combination, on diurnal glucose tolerance and estimates of β-cell function and insulin sensitivity in CDO rats. A, Mean plasma glucose concentrations during an OGTT performed twice in each rat at CT 4 (left) and CT 16 (right) in control (open circle, n = 4), CDO (black circle, n = 9), CDO-MEL (gray circle, n = 10), CDO-MET (yellow circle, n = 9), and CDO-MEL-MET (blue circle, n = 9) rats. B, Estimate of pancreatic β-cell function expressed as AUC for C-peptide over plasma glucose during the first 30 minutes of an OGTT at CT 4 (left) and CT 16 (right) in control (open circle, n = 4), CDO (black circle, n = 9), CDO-MEL (gray circle, n = 10), CDO-MET (yellow circle, n = 9) and CDO-MEL-MET (blue circle, n = 9) rats. C, Estimate of insulin sensitivity derived from Matsuda index during an OGTT performed at CT 4 (left) and CT 16 (right) in control (open circle, n = 4), CDO (black circle, n = 9), CDO-MEL (gray circle, n = 10), CDO-MET (yellow circle, n = 9), and CDO-MEL-MET (blue circle, n = 9) rats. Data are presented as mean ± SEM; #, P < .05 vs control; *, P < .05 vs CDO; †, P < .05 vs CDO-MEL and ‡, P < .05 vs CDO-MET.

We next set out to examine tissue-specific benefits of melatonin and metformin treatment in CDO rats, with emphasis on the pancreatic islet given its significance in the pathogenesis of T2DM. As expected, increased metabolic and insulin demand associated with CDO provoked significant islet size expansion (with no change in islet density), disruption of normal islet architecture and was associated with increased β-cell fractional area (0.7 ± 0.1 vs 1 ± 0.1%, P < .05 for control vs CDO) (Figures 6 and 7, A–C). Islet morphology in CDO rats was also characterized by a substantial increase in β-cell turnover highlighted by an approximately 3-fold increase in β-cell proliferation and by an approximately 3.5-fold rise in DNA damage-induced β-cell apoptosis (P < .05 vs control) (Figures 6 and 7). Melatonin and metformin treatment alone, and particularly in combination, resulted in preservation of islet architecture and significant attenuation in the frequency of β-cell apoptosis (0.033 ± 0.05 vs 0.015 ± 0.01 β-cells/islet for CDO vs CDO-MEL-MET; P < .05) (Figures 6 and 7) and β-cell proliferation (0.23 ± 0.03 vs 0.05 ± 0.01 β-cells/islet for CDO vs CDO-MEL-MET; P < .05) (Figures 6 and 7). Subsequently, β-cell fractional area in CDO-MEL-MET group was also reduced compared with vehicle-treated CDO (1 ± 0.1 vs 0.7 ± 0.1%, P < .05 for CDO vs CDO-MEL-MET) (Figures 6 and 7). Interestingly, increase in body mass during the 12-week experimental period was significantly correlated with the rate of β-cell demise in vehicle-treated CDO rats (r = 0.6, P < .05) (Figure 7F). However, this relationship was abolished in CDO rats treated with MEL, MET, or MEL-MET (Figure 7F), thus emphasizing the efficacy of these treatments for prevention of CDO-induced β-cell failure.

Figure 6.

Effects of melatonin and metformin administration alone, or in combination, on pancreatic islet morphology in CDO rats. A, Representative examples of pancreatic sections imaged at ×5 magnification (top; scale bars, 2000 μm) and individual islets imaged at ×20 magnification (bottom; scale bars, 50 μm) stained for insulin (brown) and hematoxylin (blue) in control, CDO, CDO-MEL, CDO-MET, and CDO-MEL-MET rats. B, Representative examples of islets stained by immunofluorescence for insulin (green) and counterstained (red) with either glucagon, replication marker Ki-67, or DNA damage-induced apoptosis marker TUNEL counterstained with nuclear marker DAPI (blue) imaged at ×20 magnification (scale bars, 50 μm) in control, CDO, CDO-MEL, CDO-MET, and CDO-MEL-MET rats. White arrowheads highlight examples of Ki-67 and TUNEL-positive β-cells.

Figure 7.

Effects of melatonin and metformin administration alone, or in combination, on islet size distribution, islet density and β-cell turnover in CDO rats. A, Scatter plot of islet size distribution obtained from the analysis of whole pancreatic sections in control (open circles, n = 4), CDO (black circles, n = 5), CDO-MEL (gray circles, n = 5), CDO-MET (yellow circles, n = 5), and CDO-MEL-MET (blue circles, n = 5) rats. Mean islet density (B), β-cell fractional area (C), frequency of β-cell proliferation (D), and frequency of DNA damage-induced β-cell apoptosis (E) in control (open bar, n = 4), CDO (black bar, n = 9), CDO-MEL (gray bar, n = 10), CDO-MET (yellow bar, n = 9), and CDO-MEL-MET (blue bar, n = 9) rats. Bar graphs represent mean ± SEM. F, Linear regression analysis between body mass and the frequency of β-cell apoptosis in CDO (black line, r = 0.6: P < .05), CDO-MEL (gray line, r = 0.1: P = .8), CDO-MET (yellow line, r = 0.1: P = .6), and CDO-MEL-MET (blue line, r = 0.2: P = .4) rats. #, P < .05 vs control; *, P < .05 vs CDO; †, P < .05 vs CDO-MEL and ‡, P < .05 vs CDO-MET.

Discussion

Among recognized environmental triggers epidemiologically associated with T2DM, circadian disruption and obesity have demonstrated parallel rise in recent decades (4). Furthermore, a strong bidirectional relationship between circadian disruption and obesity has also been established in animal models (27, 28) and, more recently, humans (29). Thus, novel therapeutic approaches to combat T2DM associated with circadian disruption should require the ability to target both circadian and metabolic dysfunctions. In our study, use of combination therapy of melatonin and metformin was prompted by previously documented beneficial effects of these compounds in regulation of circadian (14, 30) and metabolic function in T2DM (17, 31). In addition, because nocturnal melatonin secretion is blunted in obese T2DM patients (32), accumulating evidence suggest that restoration of physiological melatonin signaling may attenuate metabolic abnormalities present in T2DM (18, 33). Subsequently, our current results provide the first evidence that daily combination treatment with melatonin and metformin synergizes to prevent onset of metabolic and islet dysfunction in CDO highlighted by 1) reduction in tissue adiposity, 2) decreased diurnal glycemia, 3) enhanced insulin sensitivity, 4) improved oral glucose tolerance, and 5) prevention of β-cell failure.

It was previously demonstrated that concomitant exposure to CDO amplifies deleterious metabolic effects and recapitulates features of islet cell failure in T2DM (11–13). Specifically, CDO rats exhibit diminished circadian control of insulin release, and impaired glucose-stimulated insulin secretion (GSIS) (13). CDO islets display disrupted architecture characterized by increased frequency of β-cell apoptosis as well as increased β-cell turnover, observations consistent with models of obesity-induced β-cell failure (34) and humans with T2DM (35). Alterations in islet biology observed in CDO rats are likely driven by a combination of factors, which include 1) increased insulin demand due to induction of obesity/insulin resistance, 2) increased inflammation, and 3) impaired circadian function. Although disruptive effects of insulin resistance and inflammation on the β-cell have been previously documented (36, 37), more recent studies highlight the importance of circadian clock gene expression in regulation of β-cell function, survival, and response to oxidative stress (38–40). Importantly in our study, the combination treatment of melatonin and metformin abrogated islet failure in CDO rats. These beneficial effects were likely mediated in part through reduction in obesity/insulin resistance leading to the attenuation of insulin demand on the β-cell. However, melatonin or metformin treatment alone also displayed beneficial effects on β-cell turnover and apoptosis, consistent with previously described effects of these compounds in human islets (18, 19).

Endogenous melatonin production and secretion is driven by the suprachiasmatic nucleus (SCN) of the hypothalamus (41), which is a master regulator of the mammalian circadian system (42). Thus, nightly melatonin secretion serves as a hormonal output signal of the circadian clock (14). However, when administered exogenously, melatonin appears to exert beneficial chronobiotic effects via direct effects on the SCN and peripheral circadian clocks mediated through high-affinity melatonin receptors (43). Subsequently, timed exogenous melatonin administration has been shown to be effective in enhancing circadian rhythms in both rodents and humans (44, 45). Consistently in our study, daily melatonin replacement alone improved circadian rhythms in activity in CDO rats, thus suggesting a potential effect of melatonin treatment on the circadian clock. However, our study did not specifically assess circadian variations in clock gene expression in the SCN as well as in metabolic tissues (eg, islet and liver); thus our results cannot definitively assign observed beneficial effects of melatonin to improvements in the circadian clock function. Our study also did not specifically assess melatonin-induced changes in feeding activity patterns. Given accumulating evidence for pleotropic benefits of restricted feeding, some melatonin-mediated metabolic effects might be attributed to improvements in circadian control of feeding (46).

Activation of melatonin signaling may also result in beneficial cellular effects independent of the circadian system. Specifically, exposure of β-cells in culture to melatonin results in sensitization and enhanced activation of the cAMP-protein kinase A-cAMP-responsive element-binding protein signaling (18, 47), a cellular pathway known to be critical for regulation of β-cell function and survival (48). Importantly, enhancing melatonin signaling in β-cell lines and human islets in vitro attenuates induction of oxidative stress and improves GSIS (18). Our current in vivo observations are consistent with previously observed in vitro effects of melatonin in β-cells and demonstrate for the first time improved β-cell survival, attenuation of β-cell turnover and enhanced β-cell function in melatonin-treated CDO rats. Of note, rare melatonin receptor variants impairing melatonin receptor function (particularly mutations effecting activation of cAMP-protein kinase A-cAMP-responsive element-binding protein signaling), are associated with increased propensity for development of T2DM (49). Furthermore, decreased nocturnal melatonin secretion in humans is also associated with increased incidence of T2DM (33), thus lending further support to potential importance of the melatonin pathway in regulation of β-cell function, survival, and glucose homeostasis.

Despite purported benefits of melatonin on metabolic system, recent studies in humans show that acute administration of melatonin reduces glucose tolerance mediated in part through reduction in insulin secretory response and disposition index (50). This effect is amplified in individuals with common T2DM risk variant in melatonin receptor 2 (MTNR1B) (51). Indeed, increased expression of MTNR1B in human β-cells attenuates GSIS consistent with acute effects of melatonin on reducing cAMP levels and potential physiological role in restraining nighttime insulin release (52, 53). Notably, nightly repression of insulin secretion (via somatostatin infusion) in T2DM patients has been previously shown to restore morning insulin secretory response and provide beneficial metabolic effects (54). Taken together, it is clear that additional clinical studies of prospective melatonin therapy in T2DM should carefully evaluate potential benefits/complications of timed melatonin administration (ie, meal, day/night) as well as additional complications arising with risk alleles for MTNR1B.

Metformin is a first line therapeutic choice for the treatment of T2DM and has a long record as an effective antihyperglycemic therapeutic agent (17, 55). Its antidiabetic effects are largely attributed to the suppression of endogenous glucose production and enhancement of peripheral insulin sensitivity (17). Molecular mechanisms underlying metformin's action are complex but are largely attributed to alterations of cellular mitochondrial function and suppression of tissue lipogenesis (31). Interestingly, metformin has recently been shown to modulate the circadian clock function through 5' AMP-activated protein kinase-dependent degradation of a key clock protein, period 2 (56). Although we cannot exclude that some observed beneficial metabolic effects of metformin may be attributed to its effects on the circadian clock, in our study, metformin had no effect on behavioral circadian rhythms, thus excluding any potential effects on the SCN. However, despite the failure of metformin to improve circadian activity rhythms, metformin-treated CDO rats displayed significant improvements in tissue adiposity, insulin sensitivity, β-cell function, and glucose tolerance. It is also important to note that relatively short fasting period used in our study before metabolic testing, likely reduced characteristic benefits of metformin treatment on fasting hyperglycemia in T2DM (17).

Metabolic benefits of melatonin or metformin treatment revealed diurnal efficacy patterns resulting in synergistic benefits when 2 drugs were administered in combination. For example, melatonin-treated rats selectively displayed improved glucose tolerance and β-cell function at the CT 4 time point, whereas metformin treatment enhanced glucose tolerance, β-cell function and insulin sensitivity selectively at CT 16. Subsequently, combination treatment resulted in enhanced glucose tolerance at both time points thus contributing to synergy of the 2 compounds to enhance metabolic function in CDO. Indeed there is an increased appreciation for the concept of chronopharmacology signifying that therapeutic outcomes are dependent on the CT of drug administration, circadian regulation of drug transport and metabolism, as well as circadian control of target gene/protein expression (57). Recently, it was reported that a majority of biological targets of the top 100 selling therapeutics in the United States target circadian gene products (58). Interestingly, this list includes the drug Junamet, which consists of a combination of metformin and a dipeptidyl peptidase-4 inhibitor, Sitagliptin (58). Similarly, melatonin treatment has been known to exert circadian-dependent antihypertensive benefits in humans (59). Additional data are clearly needed to further characterize the chronotherapeutic potential of metformin alone or in combination with melatonin in human subjects.

It is important to acknowledge some limitations of our study. Firstly, melatonin dose used (20 mg/kg·d) exceeds common recommended doses for humans with circadian/sleep disturbances (eg, ∼5 mg/d). However, studies have shown that in humans, melatonin appears to be well tolerated in doses up to 100 mg with some studies establishing tolerability at doses as high as 300 mg (60, 61). Additional clinical studies are warranted to determine therapeutic dosage and potential efficacy of melatonin for circadian and metabolic disorders in humans. Secondly, melatonin exhibits nocturnal secretion irrespective of specie's circadian activity patterns (ie, nocturnal vs diurnal). Thus, melatonin secretion peaks in humans at night coinciding with rest phase and fasting metabolic state, whereas rodents secrete melatonin coincident with onset of activity and feeding and postprandial metabolic state. This apparent mismatch between fast/feeding cycles and circadian phases of melatonin secretion in humans vs rodents complicates interpolation of our results for human therapeutic application. Indeed, acute effects of melatonin on stimulation of insulin secretion, although incompletely understood, appear to vary between nocturnal and diurnal species (53). Nevertheless, beneficial effects of chronic melatonin treatment on metabolic, cardiovascular, and circadian function have been previously observed in both rodent and human studies (62–64).

In summary, the current study shows for the first time that combination therapy of melatonin and metformin synergistically attenuates deleterious circadian, metabolic, and islet effects of concomitant exposure to CDO. Given suggested presence of sleep and circadian abnormalities as well as loss of melatonin secretion in obese T2DM populations (32, 65–67), treatment strategies that include enhancement of both circadian and metabolic functions should be considered as preventative and therapeutic options for future treatment of obesity and T2DM.

Appendix

See Table 1.

Table 1.

Antibody Table

Peptide/Protein Target	Antigen Sequence (if Known)	Name of Antibody	Manufacturer, Catalog Number, and/or Name of Individual Providing the Antibody	Species Raised in; Monoclonal or Polyclonal	Dilution Used
Insulin		Antiinsulin	Abcam, ab7842	Guinea pig; polyclonal	1:100
Glucagon		Antiglucagon	Sigma, g2654	Mouse; monoclonal	1:1000
Ki-67		Anti-Ki-67	BD Pharmingen, 550609	Mouse; monoclonal	1:50