In uence of Saskatoon Berry and Cyanidin-3- Glucoside on Gut Microbiota and Relationship with Metabolism and In ammation in High Fat-High Sucrose Diet-Induced Insulin Resistant Mice


 Background

Type 2 Diabetes (T2D) has become one of most common and harmful chronic diseases worldwide. T2D is characterized as insulin resistant and is often associated with unhealthy dietary habits. The present study assessed the effects of freeze-dried Saskatoon berry powder (SBp) and cyanidin-3-glucoside (C3G, an anthocyanin enriched in SBp) on metabolism, inflammatory markers and gut microbiota in high fat-high sucrose diet (HFHS) diet induced insulin resistant mice.
Results

Male C57 BL/6J mice received control, HFHS, HFHS + SBp (8.0 g/kg body weight/day) or HFHS + C3G (7.2 mg/kg/day, equal amount of C3G in 8.0 g/kg/day SBp) diet for 11 weeks. HFHS diet significantly increased the levels of glucose, cholesterol, triglycerides, insulin resistance and inflammatory mediators in plasma. The results of 16S rRNA gene sequencing demonstrated that HFHS diet increased the ratio of Bacteroidetes/Firmicutes (B/F) phylum bacteria and an elevated abundance of Muriculaceae family bacteria in the feces of mice. SBp or C3G supplementation attenuated HFHS diet-induced disorders in metabolism and inflammatory markers, and increased B/F ratio and Muriculaceae abundance in mouse gut compared to HFHS diet alone. The abundance of Muriculaceae in the gut microbiota negatively correlated with body weight, glucose, lipids, insulin resistance and inflammatory mediators in mice. The results of functional predication analysis suggest that HFHS diet upregulated the genes of gut bacteria involved in inflammation-related cellular processes, and inhibited bacteria involved in metabolism. SBp and C3G partially neutralized the alterations induced by HFHS diet in gut microbiota implicated in metabolism or inflammation.
Conclusion

The findings of the present study suggest that SBp is a potential prebiotic food mitigating Western diet-induced disorders in metabolism, inflammation and gut dysbiosis, and C3G possibly contributes to the beneficial effects of SBp.


Introduction
Diabetes has become one of the most common metabolic disorders worldwide, and the trend of its upsurge is continuing. Nine out of ten diabetic patients in adults are type 2 diabetes (T2D), which is characterized as insulin resistance and often associated with obesity [1]. Genetic factors and multiple environmental factors are implicated in the etiology of T2D. However, the precise mechanism or treatment for T2D remains to unclear [2]. Western pattern diet or foods enriched with fat or sugar plays a crucial role in the epidemic of T2D in the world [3]. T2D is associated with low-grade chronic in ammation [4].
Accumulating lines of evidence suggest that gut microbiota plays an important role in the development of diabetes, obesity and in ammation [5]. Foods are the most common modulator of gut microbiota.
Western diet-induced gut dysbiosis is associated with obesity and T2D [6].
Saskatoon berry (Amelanchier alnifolia Nutt.), a type of fruit-bearing shrubs, belongs to Rosaceae family in Amelanchier genus and is also known as Juneberry and service berry [7]. It natively grows in the southern Yukon, the Prairies in Canada and the west northern regions of the United States of America, and has been more recently planted in Europe [8,9]. Saskatoon berry fruits contain more polyphenols including anthocyanins compared to other common berries, such as strawberry or blueberry [10]. At least four anthocyanins have been detected in Saskatoon berries, among them cyanidin-3-galactoside and cyanidin-3-glucoside (C3G) are most abundant, accounting for about 61-74% and 18-21%, respectively [8,11].
Saskatoon berry fruits possess abundant antioxidant capacity [12,13]. Previous studies in our group demonstrated that Saskatoon berry powder (SBp) was capable of reducing vascular in ammation in leptin receptor-knockout (db/db) diabetic mice. Supplementation of C3G resulted in a 3-times greater in anti-in ammatory activity than C3Ga in cultured endothelial cells [11,14]. Our recent study demonstrated that supplementation of 5% SBp (~ 8.0 g/kg/day) attenuated high fat-high sucrose (HFHS) diet induced hyperglycemia, hyperlipidemia, insulin resistance, in ammation and gut dysbiosis in mice [15]. However, the effect of oral administration of C3G on the composition of gut microbiota, and the comparison of C3G and SBp on gut microbiota, metabolism and in ammatory mediators in mice fed with HFHS diet have not been documented.
The present study compared the effects of supplementation of C3G and SBp on circulating glucose, lipids, insulin, in ammatory markers and gut microbiota in a HFHS diet-sensitive rodent model, C57 BL/6J mice. Differences in the impact of SBp and C3G on gut microbial pro le and the function prediction of the microbial genes in mice were further investigated using 16S rRNA gene sequencing and bioinformatics approaches.

Animals
Male C57 BL/6J mice (n = 32, 6 weeks of age) were obtained from the Jackson Laboratory (Bar Harbor, ME). Mice were housed in standard plastic cages in an air-conditioned room with an alternative 12 h day/night light cycle and received regular mouse chow and tap water for 1 week for stabilization. The protocols of animal experiment have been approved by the Animal Protocol and Ethics Committee in the University of Manitoba.

Dietary intervention
Mice were randomized into 4 groups (n = 8/group, four in a cage) and receive one of following diets for Animal monitoring and sample collection Body weights and food intake of animals was assessed every 2 weeks until 10th weeks after the start of the dietary experiment. About 200 µL of blood was collected from mouse saphenous veins every 2 weeks after an overnight fasting for measuring levels of plasma glucose to monitor the development of diabetes. Mice were euthanized at 11 weeks after the onset of the dietary experiment via the inhalation of 5% iso urane to reduce the pain of animals. Blood was withdrawn subsequently through heart puncture.

Analyses of metabolic variables
The levels of plasma glucose and cholesterol of mice were analyzed using Sekisui Diagnostics SL reagent kits (Charlottetown, PE). Plasma levels of triglycerides were measured using BioAssay Systems reagents (Hayward, CA). Insulin levels in plasma were assessed using enzyme-linked immunosorbent assay (ELISA) kits from EMD Millipore (Billerica, MA) for insulin. The homeostatic model assessment of insulin resistance (HOMA-IR) of the mice was calculated from the simultaneous samples of plasma insulin and glucose as previously described [16].

Fecal sample collection
Mice were housed in a singly housed cage with fresh bedding overnight at the 10th week after the start of the dietary experiment for overnight. Fecal pellets were collected from individual cage and stored in separated tubes at -80 °C before further analysis.
Extraction and sequencing of bacteria DNA DNA extraction was achieved using PowerFecal DNA Isolation Kit (QIAGEN, Germantown, MD). DNA was quanti ed using a NanoDrop 2000 spectrophotometer (Thermo Scienti c). Bacteria DNA in mouse feces was ampli ed using polymerase chain reaction (PCR) with modi ed primers containing 515F (5′-GTGYCAGCMGCCGCGGTAA) and 926R (5′-CCGYCAATTYMTTTRAGTTT) targeting the V4-V5 region of bacterial DNA. A high-throughput Hamilton Nimbus Select robot and Coastal Genomics analytical gels were run to verify the quality of PCR products. Failed amplicons with spurious bands were repeatedly ampli ed after modifying conditions for PCR until quali ed bands were produced. The PCR amplicons were normalized by using a high-throughput Charm Biotech Just-a-Plate 96-well normalization kit, then pooled to construct a library and quanti ed before sequencing on an Illumina MiSeq platform in the Integrated Microbiome Resource in the Dalhousie University [17].

Bioinformatic analysis and statistics
Freshly generated raw data of gut microbiota in the form of fastq le was demultiplexed according to the barcode sequences, followed by trimming using Cutadapt (version 1.17) to remove primers. Trimmed reads were imported as artifact into an open-source bioinformatics pipeline of decentralized microbiome analysis package Quantitative Insights Into Microbial Ecology 2 (QIIME2, version: 2018. 8) [18]. After joining of paired-end reads and ltering out of low-quality reads, amplicon sequence variants (ASVs) were obtained through DADA2 work ow. Taxonomies were assigned to ASVs using a Naive-Bayes approach and SILVA database. Diversity metrics (Core-metrics-phylogenetic) within QIIME2 were used for evaluating α-and β-diversity. Differences between data from multiple groups were examined using the analysis of variance(ANOVA) and post-hoc Tukey test. Signi cant difference was set at p < 0.05. Relative abundances of ASVs assigned with taxonomy in feature table were correlated with physical and clinical parameters. The phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) was used to identify differences in predictive metagenome function. Functions of microbial genes were predicted with the use of Galaxy web application, and Kyoto Encyclopedia of Genes and Genomes [19].

Effects of HFHS diet and supplementation of SBp or C3G on body weights and food intake of mice
Signi cant increases in body weights between mice fed with the control diet and HFHS diet with and without a supplementation of SBp or C3G were detected at ≥ 2 weeks after the start of the intervention (p < 0.05 or 0.01). The trend of body weight increase in mice receiving HFHS diet was continued throughout the experiment. No signi cant difference was detected among mice fed with HFHS diet with or without a supplementation with SBp or C3G (Fig. 1A). Daily food intake of all mice was measured at onset and every 2 weeks during the dietary experiment until 10 weeks after the start of dietary intervention. No signi cant difference in food intake was detected among various groups receiving different experimental diets in the present study (Fig. 1B).

Effects of SBp and C3G on glucose and lipid metabolism in HFHS diet-induced obese mice
The HFHS diet signi cantly increased the levels of fasting plasma glucose (FPG) in mice compared to the control diet (p < 0.01). The supplementation with SBp or C3G in the HFHS diet signi cantly lowered FPG, compared to that in HFHS alone (p < 0.05 or 0.01). However, the levels of FPG in the SBp or C3G group were still signi cantly higher than that in the control group. No signi cant difference in FPG was found between the SBp and C3G group. The HFHS diet also signi cantly elevated the levels of cholesterol and triglycerides in plasma compared to the control diet (p < 0.01). Plasma cholesterol or triglyceride levels in both the SBp and C3G group were signi cantly lower than the HFHS group (p < 0.01). Cholesterol and triglyceride levels in the plasma of mice receiving HFHS diet supplemented with SBp or C3G were signi cantly higher than that in the control group (p < 0.01). No signi cant difference in plasma cholesterol or triglycerides was detected between the SBp and C3G groups (Fig. 2).
Effects of SBp and C3G on fasting plasma insulin and insulin resistance in mice receiving HFHS diet Insulin resistance was assessed using HOMA-IR generated from the levels of FPG and insulin in the plasma samples simultaneously withdrawn from mice. The levels of plasma insulin and HOMA-IR in the HFHS group were signi cantly higher than the control group (p < 0.01). Supplementation with SBp and C3G signi cantly reduced the levels of insulin and HOMA-IR compared to that in mice fed with the HFHS diet alone (p < 0.01). The levels of insulin and HOMA-IR in the SBp or C3G group were still higher than that in the control group (p < 0.01). No signi cant difference in insulin or HOMA-IR was detected between the SBp and C3G group ( Fig. 3A and 3B).

Effects of HFHS diet and the supplementation of SBp or C3G on circulating in ammatory mediators in mice
The HFHS diet elevated the levels of MCP-1 and PAI-1, two in ammatory markers, in the plasma of mice (p < 0.01). The supplementation of SBp or C3G in the HFHS diet signi cantly reduced the circulating levels of MCP-1 and PAI-1 compared to the HFHS diet in mice (p < 0.01). The levels of the in ammatory markers in mice receiving SBp or C3G diet were still signi cantly higher than the control group (p < 0.01). No signi cant difference in plasma MCP-1 or PAI-1 was detected between mice receiving HFHS diet supplemented with SBp or C3G ( Fig. 4A and B).

Impact of supplementation of SBp or C3G on gut microbiota in HFHS diet-fed mice
The results of β-diversity analysis demonstrated that the gut microbial compositions in the stool of mice from the four dietary groups were well separated in principal component analysis (Fig. 5). No signi cant difference in α-diversity variables was detected in mice receiving different diets (Shannon index, chao1). Table 1 demonstrated that Bacteroidetes and Firmicutes represented the vast majority phylum bacteria in mouse feces. Mice fed with the HFHS diet had lower abundances of Bacteroidetes, Actinobacteria, Proteobacteria and Verrucomicrobia phylum bacteria, but a higher relative abundance of Firmicutes phylum bacteria compared to mice receiving the control diet (p < 0.05 or 0.01). Supplementation with SBp or C3G to the HFHS diet augmented the abundance of Bacteroidetes and reduced that of Firmicutes in mice feces (p < 0.01). The abundance of Actinobacteria in the stool of the C3G group was the only type of phylum bacteria signi cantly different from that in SBp group (p < 0.05). Table 1 Effect of different diets on the abundances of gut microbiota on phylum level. Male C57 BL/J6 mice (6 weeks of age) were randomized into 4 groups and received following diets for 11 weeks: control (CTL) group: low-fat diet; HFHS group: HFHS diet; SBp group: SBp (8.0 g/kg/day) supplemented in the HFHS diet; C3G group: C3G (7.2 mg/kg/day) supplemented in the HFHS diet. Values in the tables were expressed in mean ± SD (% of total gut microbiota, n = 8/group). *, **: p < 0.05 or 0.01 versus control group; +, ++: p < 0. The relative abundance of Bacteroidetes phylum bacteria in mice with HFHS diets was signi cantly lower than that in mice fed with the other three types of diets (p < 0.01, Fig. 6A). The relative abundance of Firmicutes phylum bacteria in the HFHS group was signi cantly higher compared to that in the other three groups (p < 0.01, Fig. 6B). Besides, the HFHS group also had a signi cantly lower ratio of Bacteroidetes/Firmicutes (B/F) and higher in Firmicutes/Bacteroidetes ratio compared to that in the control, SBp or C3G group (p < 0.05 or 0.01). The SBp group had a signi cantly higher B/F ratio and lower F/B ratio compared to the HFHS group. C3G supplementation induced a signi cantly lower F/B ratio (p < 0.01), but not in B/F ratio, compared to the HFHS group ( Fig. 6C and 6D).
Statistical differences in the relative abundances of all types of gut family bacteria, except that of Bi dobacteriaceae, were detected among mice receiving various diets (p < 0.01, Fig. 7A). Mice fed with the HFHS diet had evidently higher relative abundances in Erysipelotrichaceae or Lachnospiraceae and lower abundance in Muribaculaceae family bacteria compared to mice in the other groups (heatmap in Fig. 7A). Supplementation of SBp or C3G decreased the abundance of Lachinospiraceae, and increased that of Muribacculaceae compared to HFHS diet (p < 0.05). However, the relative abundances of several family bacteria in the SBp group were different from that in the C3G group (see open triangles in Fig. 7A). Heatmap (Fig. 7B, right) demonstrated correlations between the relative abundance of family bacteria and some diabetes-related biochemical parameters. The abundances of De uviitaleaceae, Eggerthellaceae, Erysipelotrichaceae, Family XIII, Lachnospiraceae, Peptpcoccaceae, Peptostreptococcaeae, Rununococcaceae and Streptococcaceae family bacteria were positively correlated with body weights, glucose/lipid metabolism and in ammatory markers in the mice, while Clostridiaceae 1, Clostridiales vadinBB60 group, Muribaculaceae and Lactobacillaceae were negatively correlated with the physical, metabolic and in ammatory variables. The relative abundance of Muribaculaceae in mice in the HFHS group was the lowest among all groups. Both SBp and C3G supplementations signi cantly increased the abundances of Muribaculaceae family bacteria compared to the HFHS group.
The greatest difference in the mean proporation of family bacteria between the HFHS group and SBp or C3G group was Muribaculaceae family bacteria ( Fig. 7C and 7D). The mean proportion of Muribaculaceae in the HFHS group was signi cantly lower than that in the SBp or C3G group (p < 0.001).
The mean proportion of Lachnospiraceae was signi cantly higher in the HFHS group and lower in the SBp and C3G group (p < 0.05 or 0.001). Similar pattern of changes were detected in the mean proportion of Erysipelotrichaceae, which was higher in the HFHS group but lower in the SBp and C3G group. The mean proportion of Ruminococcaceae was signi cantly higher in the HFHS group compared to the SBp group (p < 0.001, Fig. 7C); while, no signi cant difference in the mean proportion of Ruminococcaceae was detected between the HFHS and C3G groups (Fig. 7D).

Functional predication of the changes in gut microbial genes induced by HFHS diet with and without the supplementation of SBp or C3G
Function prediction using PICRUSt analysis demonstrated that the impact of the HFHS diet and that supplemented with SBp or C3G on the abundances of genes of gut microbes in 15 classes of predicted functions in mice (p < 0.01, ANOVA, Fig. 8). Microbial genes related to membrane transport were the most abundant in the stool of mice in the present study. At least two types of ASV (operational taxonomic unit or OTUs) distribution among the cellular functions in mice receiving different dietary intervention were found: 1) the HFHS diet resulted in the highest microbial abundances compared to that in the control group, and the supplementation with SBp or C3G attenuated the effect of the HFHS diet; 2) mice in the control group had more abundant ASV than in the HFHS group, and the supplementation of SBp or C3G substantially increased ASV compared to the HFHS diet, but still relatively less than that in the control group. Examples for the rst type of cellular functions include membrane transport and cell motility, which may be implicated in in ammation. The second type of cellular functions included the metabolism for energy, carbohydrates, amino acids, lipids, cofactors and vitamins, replication and repair, which may potentially contribute to metabolism. It is noticed that transcription was activated, and the translation pathway was suppressed in mice receiving the HFHS diet. SBp or C3G supplementation attenuated the effect of the HFHS diet in transcription or translation in mice. The ndings suggest that SBp and C3G may neutralize the effect of the HFHS diet in in ammation and promote metabolism in mice via the modulation of gut microbiota.

Discussion
The major novel ndings generated from the present study include: i) the supplementation of C3G signi cantly reduced HFHS diet-induced hyperglycemia, hypercholesterolemia, hypertriglyceridemia, insulin resistance, in ammatory markers and gut dysbiosis in mice in similar extents as those treated with SBp containing comparable amount of C3G; 2) both SBp and C3G supplementation to HFHS diet increased the abundance of Bacteroidetes phylum bacteria and decreased the abundance of Firmicutes phylum bacteria compared to HFHS diet alone; 3) supplementation with SBp or C3G increased the abundance of gut Muribaculaceae family bacteria in comparable intensity, and the abundance of Muribaculaceae bacteria in mouse feces was negatively correlated with body weights, FPG, lipids, and in ammatory markers in the mice; 4) the results of functional predication analysis suggests that the supplementation of SBp or C3G reduced the abundance of gut microbial genes involved in in ammation and enhanced gut microbial genes involved in the metabolic processes in mice receiving HFHS diet.
Previous studies demonstrated that C3G increased the translocation of glucose transporter-4 in skeletal muscle through the activation of insulin and AMP protein kinase pathway in mice [20]. C3G also augmented glucose-induced insulin secretion in INS-1E pancreatic β-cells and glucose uptake in HepG2 hepatocytes [21]. C3G inhibited high glucose-induced cholesterol accumulation in HK-2 kidney epithelial cells through the activation of LXRα pathway [22]. C3G-rich Bayberry extract reduced mitochondrial reactive oxygen species production and necrosis of INS-1 cells, and lowered blood glucose in streptozotocin-induced diabetic mice [23]. Previous studies in our laboratory demonstrated that C3G inhibited glycated low-density lipoproteins induced NADPH oxidase activation, mitochondrial dysfunction and cell viability in cultured vascular endothelial cells [24]. The inhibitory effect of C3G on endoplasmic reticulum stress markers in endothelial cells was at least 3-times stronger than the most abundant anthocyanin, cyandin-3-galacotoside, in SBp [14]. SBp supplementation signi cantly reduced HFHS dietinduced hyperglycemia, hypercholesterolemia, hypertriglyceridemia and insulin resistance in C57 BL/6J mice [15]. The results of the present study demonstrated that C3G induced similar hypoglycemic and hypolipidemic effects as SBp containing comparable amount of C3G in HFHS diet-induced insulin resistant mice. The ndings of the present study are consistent with the previous studies on the effects of Saskatoon berry and C3G in glucose and lipid metabolism reported by our and other groups [11,20,23]. The ndings suggest that C3G potentially plays a critical role in the metabolic bene ts of SBp in glucose and lipid metabolism.
A recent study demonstrated that C3G administration normalized gut dysbiosis induced by a food contaminant, 3-chloro-1,2-propanediol, in rats, which was characterized by increased abundances of Lachnospiraceae and Actinobaceria phylum bacteria compared to 3-chloro-1,2-propanediol alone [25]. C3G is relatively stable in stomach and intestine [26], and it is converted to active metabolites, protocatechuic, vanillic, and p-coumaric acids in gastrointestinal rats with the assistance from gut microbiota [27]. The present study demonstrated that C3G inhibited HFHS diet-induced gut dysbiosis in mice. Both C3G and SBp signi cantly increased the abundance of Bacteroidetes phylum bacteria and Muribaculaceae family bacteria. Muribaculaceae, also known as S24-7, is a family of bacteria under Bacteroidetes phylum, which negatively in uences cellular processes involving in the development of chronic in ammation, diabetes and obesity [28]. The ndings of the present study suggest that Muribaculaceae family bacteria may be the targets of C3G and SBp in gut microbiota for their bene cial effects on metabolism and in ammation.
Although the impact of supplementation of SBp and C3G on the abundances of gut Bacteroidetes phylum bacteria and Muriculaceae family bacteria in HFHS diet fed mice was comparable, but the effects of SBp and C3G on the compositions of at least other 8 families of gut bacteria were signi cantly different (Fig. 7A). For example, the abundance of Clostriaceae 1 was signi cantly increased in the C3G group compared to that in the HFHS group, but its abundance was not signi cantly different between the SBp and HFHS group (Fig. 8). The results suggest that both SBp and C3G supplementation upregulated the abundance of Muriculaceae family bacteria, a type of gut bacteria negatively related diabetes and obesity, in mice gut, but their effects on the regulation of other family bacteria were not consistent, which may result from other components in SBp.
Functional prediction analysis helps to predict the potential of changes in gut microbiome based on metabolic pathways. Previous studies demonstrated that high fat diet reduced lipid metabolism including short chain fatty acids and the substrates for hepatic gluconeogenesis [29]. A recent study found that strawberry supplementation induced marked changes in functional potentials of microbial composition in db/db mice [30]. The present study has demonstrated the HFHS diet upregulated the genes of gut microbes with predicted functions related to in ammation and downregulated microbial genes related to lipids, carbohydrates, amino acid, cofactors and vitamins. SBp or C3G supplementation to HFHS diet tends to normalize the impact of HFHS diet on metabolism and in ammation in gut microbiota in mice. We noticed that HFHS diet increased the abundance of bacterial genes related to transcription, but decreased the genes of bacteria contributing to protein translation in mice, which possibly results from the inhibitory effect of HFHS diet on microbial genes related the metabolism of translation related substrates, such as amino acids and cofactors. SBp and C3G supplementation neutralized the contradictory effects of HFHS diet on microbial genes related to transcription and translation in mice (Fig. 8). The outcome of the functional prediction supports the metabolic and in ammatory changes induced by HFHS diet and the resilient capacity of SBp or C3G in HFHS diet-fed mice in the present study.

Conclusion
The present study demonstrated the in uence of a functional food, Saskatoon berry, and one of its active components, C3G, on the metabolism, in ammation and gut microbiome in HFHS diet-induced insulin resistant mice. C3G supplementation resulted in similar bene cial effects as SBp on HFHS diet induced disorders in metabolism, in ammation and gut microbiota in mice. The results of functional predication analysis indicated that SBp or C3G reversed the changes in microbe genes related to comparable cellular functions induced by HFHS diet in mice. The ndings suggest that SBp is a potential prebiotic food mitigating Western diet-induced disorders in metabolism, in ammation and gut dysbiosis, and C3G possibly contributes to the bene cial effects of SBp. The results of the present study may help to design clinical studies to investigate e cacy of Saskatoon berry and its active components on the prevention and management of diabetes in humans.      Effects of HFHS diet supplemented with or without SBp or C3G on the relative abundance of Bacteroidetes and Firmicutes and their ratios. The experimental regimen was described in the legend of