The Role of the Bacterial Muramyl Dipeptide in the Regulation of GLP-1 and Glycemia

The host’s intestinal microbiota contributes to endocrine and metabolic responses, but a dysbiosis in this environment can lead to obesity and insulin resistance. Recent work has demonstrated a role for microbial metabolites in the regulation of gut hormones, including the metabolic hormone, glucagon-like peptide-1 (GLP-1). Muramyl dipeptide (MDP) is a bacterial cell wall component which has been shown to improve insulin sensitivity and glucose tolerance in diet-induced obese mice by acting through the nucleotide oligomerization domain 2 (NOD2) receptor. The purpose of this study was to understand the effects of MDP on GLP-1 secretion and glucose regulation. We hypothesized that MDP enhances glucose tolerance by inducing intestinal GLP-1 secretion through NOD2 activation. First, we observed a significant increase in GLP-1 secretion when murine and human L-cells were treated with a fatty acid MDP derivative (L18-MDP). Importantly, we demonstrated the expression of the NOD2 receptor in mouse intestine and in L-cells. In mice, two intraperitoneal injections of MDP (5 mg/kg body weight) caused a significant increase in fasting total GLP-1 in chow-fed mice, however this did not lead to an improvement in oral glucose tolerance. When mice were exposed to a high-fat diet, they eventually lost this MDP-induced GLP-1 release. Finally, we demonstrated in L-cells that hyperglycemic conditions reduce the mRNA expression of NOD2 and GLP-1. Together these findings suggest MDP may play a role in enhancing GLP-1 during normal glycemic conditions but loses its ability to do so in hyperglycemia.


Introduction
Glucagon-like peptide-1 (GLP-1) is a gastrointestinal hormone secreted from the enteroendocrine L-cell upon food consumption [1,2]. Although GLP-1 has numerous metabolic and protective properties, this hormone is best known for its action on pancreatic β-cells as it enhances glucose-dependent insulin secretion [3,4]. As a result, GLP-1-based therapeutics including GLP-1R agonists, have become an important part of improving glycemia for individuals with chronic metabolic diseases such as obesity and Type 2 Diabetes (T2D) [5].
In humans, L-cells are primarily found in the intestinal crypts of the ileum and colon [6]. Due to the proximity of these cells in relation to the intestinal microbiota, research has sought to define the impacts of the intestinal bacterial community on the development of metabolic diseases through GLP-1 [7,8]. Changes in diversity and richness of the microbial community are associated with the development of obesity and T2D [9,10]. As a consequence, bacterial metabolite production is likely altered and may contribute to host metabolism [11,12]. Therefore, manipulating the intestinal microbial makeup (probiotics and prebiotics) or administering microbial metabolites (postbiotics) in order to regulate host metabolism via gut hormone release is an attractive therapeutic approach [12,13].

L18-MDP Stimulates GLP-1 In Vitro
In order to determine the effects of MDP on GLP-1 secretion, we incubated GLUTag cells with MDP (dosing ranging from 0.001-100 µg/mL) or control for a two-hour period ( Figure 1A). While a slight increase at the 0.1 µg/mL dose was observed, there was no significance in the effect of MDP on GLP-1 secretion. To assist with cell uptake of MDP, we tested the fatty acid modified L18-MDP. GLUTag cells treated with both 5 and 10 µg/mL concentrations of L18-MDP had a significant increases in GLP-1 secretion (p < 0.0001 in post hoc analysis) with the higher concentration leading to a 3.29 ± 0.29-fold increase (from a basal secretion of 125 pg/mL) ( Figure 1B). Conversely, the fatty acid modified nuclear oligomerization domain 1 (NOD1) agonist γ-D-Glu-mDAP (iE-DAP) had no effect on GLP-1 ( Figure 1C). To make certain that the effects L18-MDP were not limited to murine L-cells, we treated the human NCI-H716 cells with this compound. In NCI-H716 cells, L18-MDP caused a similar GLP-1 fold-increase (3.85 ± 0.06 compared to control) at the 10 µg/mL concentration (p < 0.0001 in post hoc analysis, Figure 1D). Cell viability/toxicity of murine and human cells were examined using the neutral red absorbance assay and were not affected by L18-MDP ( Figure 1E,F).

NOD2 is Expressed in L-cells
As MDP is the ligand for the NOD2 cytoplasmic receptor, we examined NOD2 expression in mouse ileum and in the L-cell lines. NOD2 and GLP-1 protein expression and distribution were detected in intestinal L-cells by performing double fluorescent immunohistochemistry in paraffin embedded mouse ileum sections (Figure 2A). The number of cells expressing NOD2, GLP-1, or both were counted using four different tissue samples where 3.6 ± 1.14 per 100 cells were positive for NOD2, 2 ± 0.37 per 100 cells were positive for GLP-1 and 1.6 ± 0.31 per 100 cells were positive for both NOD2 and GLP-1 proteins ( Figure 2B). Moreover, NOD2 mRNA expression was detected in GLUTag cells ( Figure 2C). Finally, we demonstrated the protein expression of NOD2 in NCI-H716 cells by western blot ( Figure 2D). Western blotting for NOD2 in GLUTag cells did not yield a band of predicted size.

NOD2 is Expressed in L-cells
As MDP is the ligand for the NOD2 cytoplasmic receptor, we examined NOD2 expression in mouse ileum and in the L-cell lines. NOD2 and GLP-1 protein expression and distribution were detected in intestinal L-cells by performing double fluorescent immunohistochemistry in paraffin embedded mouse ileum sections (Figure 2A). The number of cells expressing NOD2, GLP-1, or both were counted using four different tissue samples where 3.6 ± 1.14 per 100 cells were positive for NOD2, 2 ± 0.37 per 100 cells were positive for GLP-1 and 1.6 ± 0.31 per 100 cells were positive for both NOD2 and GLP-1 proteins ( Figure 2B). Moreover, NOD2 mRNA expression was detected in GLUTag cells ( Figure 2C). Finally, we demonstrated the protein expression of NOD2 in NCI-H716 cells by western blot ( Figure 2D). Western blotting for NOD2 in GLUTag cells did not yield a band of predicted size.

L18-MDP Causes A Slight Increase in p38 MAPK Phosphorylation
As previous studies have implicated the p38 MAPK signaling pathway (separately) in GLP-1 secretion and in NOD2 signaling, we examined p38 MAPK phosphorylation in MDP treated GLUTag cells. A modest but statistically significant increase in p38 MAPK phosphorylation was observed after the 10 min treatment. No increase in p38 MAPK phosphorylation was observed after the 20 min L18-MDP treatment ( Figure 3).

L18-MDP Causes a Slight Increase in p38 MAPK Phosphorylation
As previous studies have implicated the p38 MAPK signaling pathway (separately) in GLP-1 secretion and in NOD2 signaling, we examined p38 MAPK phosphorylation in MDP treated GLUTag cells. A modest but statistically significant increase in p38 MAPK phosphorylation was observed after the 10 min treatment. No increase in p38 MAPK phosphorylation was observed after the 20 min L18-MDP treatment ( Figure 3).

The In Vivo Effects of MDP on GLP-1 Secretion in Chow-fed Mice
To determine the effects of exogenous MDP on GLP-1 secretion in vivo, fasted chow-fed male and female mice were injected with MDP (5 mg/kg) or saline 30-min prior to blood sampling. Male and female results were combined as no statistically significant differences in the parameters measured were observed between sexes. After a single intraperitoneal (IP) injection of MDP, there was no significant difference in GLP-1 levels between MDP or saline-treated animals ( Figure 4A). We then administered MDP (5 mg/kg) or saline, at the time of fasting (−16 h) as well as at the end of the overnight fast (−30 min). Here, MDP treatment lead to a 1.57 ± 0.13-fold increase in GLP-1 compared to control (from a basal secretion of 52.6 ± 6.44 pM) (p < 0.001, unpaired t-test) ( Figure 4B). However, there were no improvements in glucose tolerance ( Figure 4C) or insulin secretion ( Figure 4D). Since we were interested in the role of MDP and GLP-1 mediating restored glucose tolerance, we repeated the experiment in high-fat diet-fed mice.

The In Vivo Effects of MDP on GLP-1 Secretion in Chow-Fed Mice
To determine the effects of exogenous MDP on GLP-1 secretion in vivo, fasted chow-fed male and female mice were injected with MDP (5 mg/kg) or saline 30-min prior to blood sampling. Male and female results were combined as no statistically significant differences in the parameters measured were observed between sexes. After a single intraperitoneal (IP) injection of MDP, there was no significant difference in GLP-1 levels between MDP or saline-treated animals ( Figure 4A). We then administered MDP (5 mg/kg) or saline, at the time of fasting (−16 h) as well as at the end of the overnight fast (−30 min). Here, MDP treatment lead to a 1.57 ± 0.13-fold increase in GLP-1 compared to control (from a basal secretion of 52.6 ± 6.44 pM) (p < 0.001, unpaired t-test) ( Figure 4B). However, there were no improvements in glucose tolerance ( Figure 4C) or insulin secretion ( Figure 4D). Since we were interested in the role of MDP and GLP-1 mediating restored glucose tolerance, we repeated the experiment in high-fat diet-fed mice.

The In Vivo Effects of MDP on Glucose Tolerance in Obesogenic-Fed Mice
Mice were fed a 45% (calories from fat) high-fat diet (HFD) to drive hyperglycemic phenotypes in the animals. An oral glucose tolerance test (OGTT) was performed after two IP injections of MDP (5 mg/kg) or saline. No significant effect of MDP was detected on glucose tolerance ( Figure 5A-C). Additionally, no effect was observed for GLP-1 and insulin secretion responses. While mice had elevated fasting glucose levels relative to baseline as the HFD continued (8.79 ± 0.50 mmol/L at 100 days, Figure 5D), the levels did not reach the level of marked hyperglycemia previously reported [25]. To induce pronounced levels of hyperglycemia, we treated a new cohort of mice with the high sugar and high fat Western diet (WD). After 70 days mice did exhibit a greater degree of fasting blood glucose levels (9.35 ± 0.42 mmol/L). Nevertheless, we did not observe any improvements in glucose tolerance in these animals when treated with the similar 2-injection MDP protocol ( Figure 5E). Next, we altered the MDP injection schedule to provide three, once-daily IP injection of MDP (100 μg/mouse) or saline to hyperglycemic (10.47 ± 0.85 mmol/L) WD-fed mice. Again, no significant differences in glucose tolerance from MDP injections were observed in these animals ( Figure 5F). Finally, we tested the NOD2 agonist, mifamurtide (50 μg/mouse), under the 2-injection protocol and found no improvements in glucose tolerance.

The In Vivo Effects of MDP on Glucose Tolerance in Obesogenic-Fed Mice
Mice were fed a 45% (calories from fat) high-fat diet (HFD) to drive hyperglycemic phenotypes in the animals. An oral glucose tolerance test (OGTT) was performed after two IP injections of MDP (5 mg/kg) or saline. No significant effect of MDP was detected on glucose tolerance ( Figure 5A-C). Additionally, no effect was observed for GLP-1 and insulin secretion responses. While mice had elevated fasting glucose levels relative to baseline as the HFD continued (8.79 ± 0.50 mmol/L at 100 days, Figure 5D), the levels did not reach the level of marked hyperglycemia previously reported [25]. To induce pronounced levels of hyperglycemia, we treated a new cohort of mice with the high sugar and high fat Western diet (WD). After 70 days mice did exhibit a greater degree of fasting blood glucose levels (9.35 ± 0.42 mmol/L). Nevertheless, we did not observe any improvements in glucose tolerance in these animals when treated with the similar 2-injection MDP protocol ( Figure 5E). Next, we altered the MDP injection schedule to provide three, once-daily IP injection of MDP (100 µg/mouse) or saline to hyperglycemic (10.47 ± 0.85 mmol/L) WD-fed mice. Again, no significant differences in glucose tolerance from MDP injections were observed in these animals ( Figure 5F). Finally, we tested the NOD2 agonist, mifamurtide (50 µg/mouse), under the 2-injection protocol and found no improvements in glucose tolerance. . MDP has no effect on glucose clearance in mice fed obesogenic diets. Oral glucose tolerance test (2 g/kg) (A-C) and fasting blood glucose (D) were measured after 30 days (A), 60 days (B) and 100 days (C) on a 45% high-fat diet (HFD) after mice received two IP injections of MDP (5 mg/kg) or saline. n = 7-10 mice per group. Oral glucose tolerance test (2 g/kg) after 70 days on Western Diet with two IP injections of MDP (5 mg/kg) or saline, n = 8-12 mice per group (E). OGTT in male mice after 110 days on Western Diet with three IP injections of MDP (100 μg) or saline, n = 4-6 mice per group (F). Results were analyzed by one-way ANOVA or unpaired t-test where appropriate. Values shown as mean ± SEM.

The Effects of MDP on Fasting GLP-1 Levels
As no effects of MDP on glucose tolerance were observed in both obesogenic models, we compared the effect of MDP injection on fasting GLP-1 and insulin levels in mice at different time points on the HFD. The initial significant increase in GLP-1 (p < 0.001 in post hoc analysis) in MDP treated mice was no longer observed even as soon as after 30 days on the obesogenic diet ( Figure 6A). MDP had no effects on fasting insulin levels ( Figure 6B). MDP has no effect on glucose clearance in mice fed obesogenic diets. Oral glucose tolerance test (2 g/kg) (A-C) and fasting blood glucose (D) were measured after 30 days (A), 60 days (B) and 100 days (C) on a 45% high-fat diet (HFD) after mice received two IP injections of MDP (5 mg/kg) or saline. n = 7-10 mice per group. Oral glucose tolerance test (2 g/kg) after 70 days on Western Diet with two IP injections of MDP (5 mg/kg) or saline, n = 8-12 mice per group (E). OGTT in male mice after 110 days on Western Diet with three IP injections of MDP (100 µg) or saline, n = 4-6 mice per group (F). Results were analyzed by one-way ANOVA or unpaired t-test where appropriate. Values shown as mean ± SEM.

The Effects of MDP on Fasting GLP-1 Levels
As no effects of MDP on glucose tolerance were observed in both obesogenic models, we compared the effect of MDP injection on fasting GLP-1 and insulin levels in mice at different time points on the HFD. The initial significant increase in GLP-1 (p < 0.001 in post hoc analysis) in MDP treated mice was no longer observed even as soon as after 30 days on the obesogenic diet ( Figure 6A). MDP had no effects on fasting insulin levels ( Figure 6B).

The Effect of A Hyperglycemic Environment on NOD2 and GCG mRNA Expression
Since the in vivo experiments demonstrated a loss of GLP-1 response to MDP during the HFD (which occurs alongside elevated blood glucose), we examined the effects of a high glucose environment on NOD2 and proglucagon (GCG, GLP-1 gene) mRNA expression in GLUTag cells. Cells were incubated in low glucose media (5 mmol/L) or high glucose media (25 mmol/L) for 24 or 48 h. We determined that NOD2 expression was significantly reduced in the high glucose environment after the 48 h incubation (p < 0.05, in post-hoc analysis) ( Figure 7A). In addition, the overall GCG expression was significantly reduced due to the high glucose treatment (p < 0.05, twoway ANOVA) ( Figure 7B).

Discussion
In this study we examined the effects of MDP and derivatives on GLP-1 secretion in L-cells and animals. Initially we found that the acyl-modified MDP stimulates GLP-1 secretion. Next, we demonstrated the presence of the MDP receptor, NOD2, in the cell models as well as in mouse intestinal tissue. Since MDP was found to enhance GLP-1 secretion in vitro we next explored its effect in mice. While we did observe an MDP-induced enhancement in basal GLP-1 secretion, this did not lead to an improvement in glycemia. Furthermore, during an obesogenic diet, the MDP-induced

The Effect of A Hyperglycemic Environment on NOD2 and GCG mRNA Expression
Since the in vivo experiments demonstrated a loss of GLP-1 response to MDP during the HFD (which occurs alongside elevated blood glucose), we examined the effects of a high glucose environment on NOD2 and proglucagon (GCG, GLP-1 gene) mRNA expression in GLUTag cells. Cells were incubated in low glucose media (5 mmol/L) or high glucose media (25 mmol/L) for 24 or 48 h. We determined that NOD2 expression was significantly reduced in the high glucose environment after the 48 h incubation (p < 0.05, in post-hoc analysis) ( Figure 7A). In addition, the overall GCG expression was significantly reduced due to the high glucose treatment (p < 0.05, two-way ANOVA) ( Figure 7B).

The Effect of A Hyperglycemic Environment on NOD2 and GCG mRNA Expression
Since the in vivo experiments demonstrated a loss of GLP-1 response to MDP during the HFD (which occurs alongside elevated blood glucose), we examined the effects of a high glucose environment on NOD2 and proglucagon (GCG, GLP-1 gene) mRNA expression in GLUTag cells. Cells were incubated in low glucose media (5 mmol/L) or high glucose media (25 mmol/L) for 24 or 48 h. We determined that NOD2 expression was significantly reduced in the high glucose environment after the 48 h incubation (p < 0.05, in post-hoc analysis) ( Figure 7A). In addition, the overall GCG expression was significantly reduced due to the high glucose treatment (p < 0.05, twoway ANOVA) ( Figure 7B).

Discussion
In this study we examined the effects of MDP and derivatives on GLP-1 secretion in L-cells and animals. Initially we found that the acyl-modified MDP stimulates GLP-1 secretion. Next, we demonstrated the presence of the MDP receptor, NOD2, in the cell models as well as in mouse intestinal tissue. Since MDP was found to enhance GLP-1 secretion in vitro we next explored its effect in mice. While we did observe an MDP-induced enhancement in basal GLP-1 secretion, this did not lead to an improvement in glycemia. Furthermore, during an obesogenic diet, the MDP-induced

Discussion
In this study we examined the effects of MDP and derivatives on GLP-1 secretion in L-cells and animals. Initially we found that the acyl-modified MDP stimulates GLP-1 secretion. Next, we demonstrated the presence of the MDP receptor, NOD2, in the cell models as well as in mouse intestinal tissue. Since MDP was found to enhance GLP-1 secretion in vitro we next explored its effect in mice. While we did observe an MDP-induced enhancement in basal GLP-1 secretion, this did not lead to an improvement in glycemia. Furthermore, during an obesogenic diet, the MDP-induced enhancement in basal GLP-1 was lost. We demonstrated that this loss in effect may be due, in part, to the downward expression of NOD2 under hyperglycemic conditions.
In order to stimulate GLP-1 secretion in vitro, the synthetic saturated stearoyl fatty acid derivative of MDP (L18-MDP) was required. Although L18-MDP is not naturally found in the gut, several protocols rely on the use of a fatty acid to facilitate the uptake of MDP [26,27]. We also demonstrated that a fatty acid modified NOD1 agonist (C12-iE-DAP) had no effects on GLP-1 secretion. This is important as long chain fatty acids are known ligands for the free fatty acid receptors on GLUTag cells. As no GLP-1 stimulation was observed with the NOD1 agonist, we can confirm that lipid modification, as well as NOD1 activation does not alter GLP-1 release. Our findings are in agreement with previous work demonstrating that injections of iE-DAP did not influence glucose tolerance in diet-induced obese mice [25]. L18-MDP's stimulatory ability was further confirmed by our experiments on NCI-H716 cells which do not express Gpr40 and Gpr120 but exhibited a similar GLP-1 response to GLUTag cells [28][29][30]. The concentration of L18-MDP used in our cell studies is slightly below the amount of colonic luminal MDP recorded from a small healthy-human study (reported as 20-87 µM; Vavricka et al., 2004 [31]). It has also been suggested that under conditions of intestinal mucosal inflammation, MDP concentrations are likely to be much higher due to the rapid degradation of the peptidoglycan by phagocytic cells in the inflamed mucosa [32]. Nevertheless, the amount of luminal MDP that traverses the protective mucous layer and reaches the intestinal brush border as well as the mechanism by which MDP enters the L-cells has yet to be elucidated [33]. Future work with radiolabeled MDP may resolve this.
Next, we established the presence of MDP's receptor, NOD2 in L-cells. Previous studies have localized NOD2 in the intestinal crypts and Paneth cells however, this is the first time NOD2 has been shown to be present in enteroendocrine cells [22,34]. While there were NOD2 positive cells outside of the L-cell population, nearly all of the GLP-1 positive cells were co-localized with NOD2. As the MAPK signaling pathway has been associated with both NOD2 activation [35] and GLP-1 secretion [36], we explored p38 MAPK phosphorylation during L18-MDP treatments. We observed a slight, but statistically significant increase in p38 MAPK phosphorylation after 10-min L18-MDP treatments but this increase was lost at the 20-min timepoint. However, we noted that the increase was modest and may not be of biological relevance. Although this provides some indication of mechanism, additional experiments are required to elucidate the full signaling cascade activated by L18-MDP. Future work should explore other NOD2 signaling cascades including extracellular regulated kinase (ERK) and c-Jun N-terminal kinase (JNK) [37].
In chow-fed animals, we found that MDP stimulated fasting GLP-1 levels only after two IP injections. This could be due to the clearance rate of this molecule as it takes about two hours for MDP to be almost completely excreted [25,38]. Therefore, the first injection may be priming the L-cells while the second injection results in significant GLP-1 release. In our experiments, even with the increase in GLP-1, we did not observe any improvements in insulin secretion or glucose tolerance. The inability of MDP to improve glucose clearance in chow-fed mice is in agreement with results previously described in MDP-treated euglycemic mice [25]. Interestingly, elevated GLP-1 levels were only observed in chow-fed mice and were lost after only 30 days on the HFD suggesting that phenotypic changes associated with the diet interfere with MDP's action on L-cells.
While our goal was to evaluate the mechanism of MDP-improved glucose tolerance, we were unable to reproduce those previous findings [25]. Although we used the same strain of mice on the same 45% HFD and were injecting very similar or the same amounts of MDP, we could neither reach the same level of hyperglycemia even after a longer time period on the diet nor record any changes in glucose tolerance. This led us to repeat the experiment on mice fed a WD (which incorporates high sucrose in addition to high fat) to reach comparable fasting glucose levels. However, no improvements in MDP-induced glucose tolerance were observed. As this was a similar result to the HFD mice which had no alterations in GLP-1 or insulin, we did not further explore GLP-1 and insulin responses in the WD-fed mice. The discrepancy between our findings and previous studies could have occurred due to variations in blood collection techniques, time of fasting, time of year the study was conducted, animal handling, and facility factors. Nevertheless, we did observe a potential explanation for this loss in MDP sensitivity in the animals.
In L-cells, we determined that high glucose contributes to the down regulation of NOD2 and GCG mRNA expression. A reduction in NOD2 in L-cells during hyperglycemia may explain the loss of effect of MDP on GLP-1 in the obese mice. Consistent with our experiments, Grasset et al. (2017) [39] showed that the control of GLP-1 action on insulin secretion and gastric emptying was dramatically impaired in NOD2 knock-out (KO) mice. Furthermore, the deletion of NOD2 exacerbates diet-induced insulin resistance as NOD2 sensing of bacterial peptidoglycan provides protection from metabolic defects [40]. Thus, the peptidoglycan-NOD2 signaling pathway plays an important role in the control of GLP-1-induced insulin secretion and the inhibition of gastric emptying. Future work should analyze intestinal NOD2 expression in HFD-and WD-fed mice to confirm our in vitro results.
In conclusion, MDP was only able to stimulate GLP-1 secretion in vitro when it was in its fatty acid modified form. The exact mechanism for L18-MDP stimulated GLP-1 release remains to be investigated but may act through the p38 MAPK signaling pathway. In vivo, we demonstrated that MDP stimulates fasting GLP-1 however, this effect is lost during the onset of hyperglycemia. Finally, we suggested that this loss of function is due to the down regulation of NOD2 and GCG genes in hyperglycemic environments.

Animals
The animal use protocol (AUP) on this project is 3 September 2019. The approval date was 19 November 2019. Animal ethics were approved by the Laurentian University Animal Care Committee in accordance with the Canadian Council on Animal Care. Male and female C57BL/6 mice aged 6-7 weeks were purchased from Charles River Laboratories (St. Constant, QC, Canada). Mice were all co-housed in standard cages maintained on a 12 h light/dark cycle in the Paul Field Animal Care Facility at Laurentian University. All mice had food and water available ad libitum unless otherwise noted.

45% High Fat Diet and Study Design
Mice were fed chow diet (8640, Envigo Teklad, Indianapolis, ID, USA) until 17 weeks in age when they were randomly divided into treatment and control groups. Mice received two intraperitoneal injections of MDP (5 mg/kg body weight; InvivoGen, San Diego, CA, USA) or phosphate-buffered saline (PBS) at time of fasting (16 h prior to oral glucose tolerance test (OGTT)) and 30 min prior to the OGTT. Dosing was chosen based on previous research [25,38]. After the initial experiment, mice were placed on a 45% High Fat Diet (D12451, Research Diets Inc., New Brunswick, NJ, USA). An OGTT was performed, and blood was collected and at 30, 60, and 100 days on the diet for glucose, insulin, and GLP-1 measurements.

Western Diet and Study Design
Male and female mice, 6-7 weeks old were placed on the Western Diet (D12079B, Research Diets Inc., New Brunswick, NJ, USA) for 10 weeks. Mice received two intraperitoneal injections of MDP (5 mg/kg body weight; InvivoGen, San Diego, CA, USA) or PBS at time of fasting (16 h prior to OGTT) and 30 min prior to OGTT. For 3 day injections, mice received a once-daily intraperitoneal injection of MDP (100 µg/mouse) or PBS for three consecutive days and fasted 16 h prior to OGTT [25]. Blood glucose was measured at 0, 5, 30, 60, and 90 min as described.

Oral Glucose Tolerance Test (OGTT)
Mice were fasted overnight for 16 h in order to measure fasting glucose, GLP-1, and insulin. After fasting, animals received an oral glucose gavage (2 g/kg body weight in PBS). Ninety microliters of blood was collected into ethylenediaminetetraacetic acid (EDTA) coated tubes (Starstedt Inc., Germany) from the saphenous vein of the animal at 0, 15, and 60 min after the oral glucose challenge. An inhibitor cocktail (10% v/v) of Aprotinin (5000 KIU/mL; MilliporeSigma, Oakville, ON, Canada) and Diprotin A (0.1 mM; MilliporeSigma) was added to the blood sample prior to placing it on ice. Blood glucose was measured using a handheld glucometer (OneTouch Verio Flex) to assess glucose tolerance. Blood samples were stored on ice and centrifuged at 5000× g, 4 • C, for 5 min. Plasma was collected and stored at −20 • C (short-term) or −70 • C (long-term) for insulin and GLP−1 enzyme-linked immunosorbent assays (ELISA).

Evaluation of GLP-1 and Insulin Hormone Levels
Total GLP-1 and insulin levels in plasma samples were determined using mouse-specific ELISA kits (Crystal Chem, IL, USA) as per the manufacturer's guidelines. For the GLP-1 ELISA, 10 µL of plasma diluted 1:5 in buffer solution was used for the assay. For the insulin ELISA, 5 µL of plasma was used per sample. Results are presented as absolute values for GLP-1 and change (delta) in insulin compared to 0-min timepoint for insulin.

Gene Expression Analyses
GLUTag cells were grown in six-well plates at a density of 500,000 cells/mL for 24 or 48 h in low glucose (5 mM) or high glucose (25 mM) DMEM media (10% FBS, 1% P/S) at 37 • C, 5% CO 2 . After the incubation period, RNA was extracted from the cells with the BioRad (Ontario, Canada) total RNA kit following the manufacturer's direction. The RNA was quantified with a UV spectrophotometer. cDNA was prepared using SensiFast cDNA synthesis kit (Bioline, ON, Canada) with 1 µg of RNA.
Primers were designed for the NOD2 gene, GCG proglucagon gene, and the RPL13a reference gene and purchased from Integrated DNA Technologies (Coralville, IA, USA) ( Table 1). Table 1. Primer sequences used for gene analysis in murine GLUTag cells.

Fluorescent Immunohistochemistry
Formalin fixed paraffinized mouse ileum tissue sections of C57BL/6 mice were prepared on microscope slides. The tissue sections were deparaffinized as previously described in Gagnon's work [42] excluding the microwave antigen retrieval. Briefly, tissues were blocked with TBS + 5% Normal Donkey Serum (NDS) (Abcam, Toronto, ON, Canada) for 20 min at room temperature (RT). Immediately after blocking, sections were incubated with primary antibodies 1:200 NOD2 (Novus Biologicals, CO, USA) and/or 1:250 GLP-1 (Abcam, Toronto, ON, Canada) and/or only blocking buffer overnight at 4 • C. After 3 × 5 min washes with TBS, the tissue samples were incubated for 45 min at RT with secondary antibodies 1:200 Donkey-anti-mouse 488 nm (Abcam, Toronto, ON, Canada) and 1:150 Goat-anti-rabbit 594 nm (Abcam, Toronto, ON, Canada) in blocking buffer. After 3 × 5 min washes with TBS, the samples were coated with Fluoroshield Mounting Medium with DAPI (Abcam, Toronto, ON, Canada) and a coverslip was fixed using clear nail polish. The tissues were observed in a dark room using an inverted Zeiss Axioplan fluorescent microscope and images were taken using Zeiss AxioVison software (Zeiss, Oberkochen, Germany). Semiquantitative analysis of NOD2 and GLP-1 expression was done using 4 images of different tissue sections at 20× magnification. All cells in the field of view were counted and results were presented as positive cells per 100 cells.

Western Blot
GLUTag cells were seeded into 6-well plates (1,000,000 cells per well) and grown for 48 h, as indicated above. On the day of the treatment, cells were rinsed with secretion media and then treated for 10 or 20 min with secretion media alone or secretion media containing 5 µg/mL L18-MDP at 37 • C, 5% CO 2 . Media was removed and wells were rinsed once with ice-cold PBS. Each well received 100 µL of lysis buffer (Cell Signaling Technologies, Danvers, MA, USA) containing protease and phosphatase inhibitors (Complete Mini/PhosphoSTOP; Roche LifeSciences, Guelph, ON, Canada). Cells were collected, sonicated on ice (10 s, Power 3), and centrifuged (5 min, 13,000× g, 4 • C). Protein concentrations were measured using the Bradford protein method and the expression levels of phospho-p38 MAPK and total p38 MAPK (1:1000, Cell Signaling Technologies, Danvers, MA, USA) were analyzed by Western blotting. BioRad Chemidoc XRS documentation apparatus was used to record the chemiluminescence signal and the signal intensity was measured using the QuantityOne program. Densitometry was used to analyze band intensity to compare phosphorylated p38 MAPK and total p38 MAPK protein expression between treatments.
To detect the NOD2 protein in NCI-H716 cells, cells were grown as previously described and treated after 48 h with RPMI 1640 secretion media for 2 h. Following the protocol described above, the NOD2 primary antibody (Novus Biologicals, Centennial, CO, USA) was diluted 1:1000 and the polyvinylidene fluoride (PVDF) membrane was treated as defined above.

Statistical Analyses
Values are reported as mean ± SEM. Experiments comparing two groups were analyzed using Student's t-test. Experiments using several doses of a treatment were analyzed by one-way ANOVA followed by Dunnett's post hoc test. Studies with two independent variables were analyzed by two-way ANOVA followed by a Sidak post hoc test. Statistical analysis was done using GraphPad Prism software. Values of n for each experiment are reported in the figure legends. p < 0.05 was considered significant.