Effects of vitamin D deficiency on the improvement of metabolic disorders in obese mice after vertical sleeve gastrectomy

Vertical sleeve gastrectomy (VSG) is one of the most commonly performed clinical bariatric surgeries for the remission of obesity and diabetes. Its effects include weight loss, improved insulin resistance, and the improvement of hepatic steatosis. Epidemiologic studies demonstrated that vitamin D deficiency (VDD) is associated with many diseases, including obesity. To explore the role of vitamin D in metabolic disorders for patients with obesity after VSG. We established a murine model of diet-induced obesity + VDD, and we performed VSGs to investigate VDD's effects on the improvement of metabolic disorders present in post-VSG obese mice. We observed that in HFD mice, the concentration of VitD3 is four fold of HFD + VDD one. In the post-VSG obese mice, VDD attenuated the improvements of hepatic steatosis, insulin resistance, intestinal inflammation and permeability, the maintenance of weight loss, the reduction of fat loss, and the restoration of intestinal flora that were weakened. Our results suggest that in post-VSG obese mice, maintaining a normal level of vitamin D plays an important role in maintaining the improvement of metabolic disorders.

. It provides evidence for the new pathway of CYP11A1 to initiate D3 metabolism in vivo, and its products show organ/cell type specificity and are modified by CYP27B1 activity. These findings define intermediates as natural products/endogenous biological regulators, which is different from the current dogma that vitamin D is only activated by the sequence D3 → 25(OH)D3 → 1,25(OH)2D3 18 . Circulating 25(OH)D 3 and 1,25(OH) 2 D 3 both bind to vitamin D in the blood, and are inactivated via 24-hydroxylation by CYP24A1 and ultimately catabolized 19 . Circulating 25(OH)D 3 is the predominant form of vitamin D. The Institute of Medicine (IOM) defines vitamin D deficiency (VDD) as a serum 25(OH)D 3 level ≤ 20 ng/mL, whereas the Endocrine Society defines VDD as ≤ 30 ng/mL 20,21 . Currently, due to air pollution, insufficient sunlight exposure, and altered dietary composition, vitamin D deficiency become more common, and it is associated with many diseases including asthma, heart disease, depression, colorectal cancer, and obesity [22][23][24][25][26][27][28][29] .
The mechanisms underlying the effects of bariatric surgery are not yet clear. It has been proposed that potential mechanisms of bariatric surgery may be involved in the body's carbohydrate metabolism, including gut hormones, hypocaloric restriction, intestinal flora metabolism [30][31][32] , bile acid physiology [33][34][35] , and a reduction in the level of lipopolysaccharides (LPS) after surgery 36,37 . Our previous research showed that a high-fat + vitamin D-deficient diet hampers the enterohepatic circulation of bile acids, leading to non-alcoholic steatohepatitis (NASH) 38 , and we reported that this diet aggravated the impairment of intestinal function and affected nutrient absorption in mice with metabolic syndrome 8 . In the present study, we established a murine model of dietinduced obesity with VDD, and we performed VSG surgery to investigate whether VDD affects the improvement of metabolic disorders after VSG.

Research design and methods
Animals and treatment. All of the animal experiments complied with the U.S. National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and complied with the ARRIVE guidelines, the procedures were approved by The Institutional Animal Care and Use Committee of Sichuan University. We purchased 4-week-old male C57BL/6J mice from Beijing HFK Bioscience. The mice were housed under SPF conditions at the animal center at the College of Life Sciences, Sichuan University, and they were given an acclimatization period of 1 week upon their arrival at the animal center. The mice were maintained in a controlled environment (12:12 h light-dark cycle) with free access to food and water.
The diet-induced obesity (DIO) model was established as described 39 . Each mouse was fed one of four types of diet for 14 weeks: (1) control chow with VD 3 at 1000 IU/kg (standard AIN93 formula), comprising the control (Ctrl) group (n = 6); (2) vitamin D-depleted control chow, as the VDD group (n = 6); (3) a high-fat diet (60% calories from fat) with a VD 3 supplement at 1000 IU/kg, as the high-fat diet (HFD) group (n = 12); or (4) high-fat chow without the vitamin D supplement, comprising the HFD + VDD group (n = 12). The compositions of the four types of diet and the vitamin kit are listed in Tables 1 and 2, respectively. After the 14-week diet period, we randomly subdivided the HFD mice and the HFD + VDD mice into two body weight-matched groups (VSG and sham) and perform the VSG surgery or a sham surgery for each mouse as described below. After the recovery during the immediate postoperative period, each mouse was then fed its presurgery diet for 10 weeks. The food intake and body weight of each mouse were recorded weekly.
VSG and sham surgeries. The VSG surgery was performed with the mouse under general anesthesia.
Inhalational anesthesia was induced and maintained using isoflurane and oxygen. The lateral 80% of the stomach was resected, leaving a tubular gastric remnant in continuity with the esophagus superiorly and the pylorus and duodenum inferiorly as described 33 . For the sham procedure, the mouse underwent the same surgical protocol except that the abdominal cavity was simply held open for approx. 30 min (the estimated duration of the gastrectomy) so that surgical stress that was similar to that of the VSG group was experienced by the mouse.  Histological analysis of adipose tissues. Adipose tissues were fixed in 4% paraformaldehyde and then embedded in paraffin. Oil red O staining and hematoxylin and eosin (H&E) staining were performed as described.
RT-qPCR analysis. We extracted RNA according to the previous practices 8 . Firstly, total RNA was isolated using Trizol (Transgen, Beijing, China), and transcribed into cDNA using the PrimeScriptTM RT reagent Kit (Cat. RR047A, TaKaRa, Shiga, Japan). The RT-qPCR was performed using a qPCR instrument (#cfx96, Bio-Rad, Hercules, CA). The primer sequences are listed in Table 3. All the expression levels of genes were normalized by GAPDH.
Quantification of short-chain fatty acids. Short-chain fatty acids (SCFAs) were extracted from the cecal contents of the mice as described 40 . One and half gram of cecum contents was added to 5 mL of water, and the mixture homogenized for 3 min. The pH of the cecal suspension was adjusted to 2.0-3.0 by adding 5 M HCl solution. This mixture was then transferred to polypropylene tubes, and centrifuged for 30 min at 3000 rpm. Finally, the supernatant was filtered with a 0.45-μm microfilter so that we can inject it into a gas chromatograph for the quantification of SCFAs.

Gut microbiota analysis.
For the analysis of the gut microbiota of the mice, the sequencing library preparation, sequencing, and data analysis were performed by Novogene (Beijing, China). Detailed information on the microbiome analysis can be found in the Supplementary Information.  www.nature.com/scientificreports/ Statistical analyses. The data recording, processing, and calculation were completed using Microsoft Excel 2010 and GraphPad Prism 5.0. The data are presented as mean ± standard error of the mean (SEM). Comparisons across groups were assessed with Student's t-test or a one-way analysis of variance (ANOVA) followed by Tukey multiple comparison testing. Statistical significance was accepted at p < 0.05.
Ethics approval and consent to participate. This study was approved by the institutional review board of Sichuan University and Nanjing Drum Tower Hospital Clinical College of Nanjing Medical University. All methods were carried out in accordance with relevant guidelines and regulations in manuscript.

Results
VDD attenuated the improvement of hepatic steatosis, the maintenance of weight loss, and the reduction of fat loss following VSG in obese mice. To determine whether VDD affects the changes in related physiological effects following bariatric surgery, we performed the VSG surgery on mice with diet-induced obesity (DIO). After the 14 weeks of the high-fat diet, the body weights of the DIO mice (i.e., both the HFD mice and the HFD + VDD mice) were significantly increased compared to those of the mice fed a normal diet as the high calories and fat ( Supplementary Fig. S1A,B), this result shows HFD could successfully induced diet-induced obesity (DIO) in mice compared to normal diet; and the plasma VD 3 concentrations of the HFD + VDD mice were significantly lower than those of the HFD mice (p < 0.01) ( Supplementary Fig. S1C,D). The blood glucose levels of the DIO mice were also increased ( Supplementary Fig. S1E). We next performed VSG or sham surgery in both HFD mice and HFD + VDD mice. In the first 6 weeks after the surgery, the HFD mice and HFD + VDD mice displayed similar losses in body weight compared to their respective sham-operated mice (Fig. 1A). Over the course of the first 6 weeks, decreased food intake was observed in the mice that had undergone VSG surgery compared to the sham-operated mice (Fig. 1B). By week 7, the food intakes among all mice were similar. In addition, the body weights of both the HFD-VSG mice and the HFD + VDD-VSG mice remained lower than those observed in the sham-operated mice, despite the lack of a significant difference in all of the groups' food intake after week 7. These data are consistent with those of previous studies 33,41 . The HFD-VSG mice and the HFD + VDD-VSG mice maintained their weight loss throughout the post-VSG period (Fig. 1A), but the body weights of the HFD-VSG mice remained lower than the weights of the HFD + VDD-VSG mice after week 8 (Fig. 1A).
Our measurements of the visceral fat weights of the HFD-VSG mice and HFD + VDD-VSG mice showed that compared with the corresponding groups of sham mice, the visceral fat weights of the VSG mice fell dramatically post-VSG (p < 0.01, each), and there was a significant difference in the visceral fat weights between HFD-VSG mice and HFD + VDD-VSG mice (p < 0.05) (Fig. 1C).
By conducting the histological analysis of adipose tissue, we observed that the VSG surgery significantly reduced the adipocyte hypertrophy steatosis in the mice fed the HFD compared to the sham mice, but this change was not clear in the HFD + VDD mice (Fig. 1D). By conducting the hematological testing, we observed decreased serum triglyceride levels (Fig. 1E) and decreased serum ALT and AST levels after VSG. In addition, compared to the values of the HFD + VDD-VSG mice, the ALT and AST levels among the HFD-VSG mice were www.nature.com/scientificreports/ lower (Fig. 1F,G). These results demonstrate that VDD weakened the role of the VSG surgery in reducing hepatic steatosis in mice, maintaining their weight loss, and reducing their fat mass, which indicated that maintaining a normal level of vitamin D is important in obese mice after VSG surgery.
VDD affected the post-VSG glucose control. Previous study showed that baseline glucose was significantly higher in mice which were fed with high-fat diet for 1 week, and a stable hyperglycemia and insulin resistance appeared over time 42 . However, VSG surgery could provides an substantial improvement of glucose 43 . Therefore, in our study, to investigate whether VDD affects the glucose homeostasis after VSG surgery, we analyzed the results of the glucose tolerance tests and the metabolite values in all of the mice that had undergone the VSG. Prior to the VSG surgery, no significant difference in fasting blood glucose levels or in glucose tolerance was observed between the obese HFD and HFD + VDD mice ( Supplementary Fig. S2A,B), but the fasting blood glucose and insulin levels measured postoperatively were lower in the HFD-VSG mice compared to those in the HFD-sham controls; moreover, the fasting blood glucose and insulin levels of the HFD + VDD-VSG mice were also lower than those of the HFD + VDD-sham controls ( Fig. 2A,B). The fasting blood glucose levels of the HFD-VSG mice were significantly lower than those of the HFD + VDD-VSG mice (p = 0.0403). The insulin levels of the HFD-VSG mice were also lower than those of the HFD + VDD-VSG mice, but the difference was not significant (p = 0.0849) ( Fig. 2A,B).
In line with these findings, the results of the HOMA-IR analysis demonstrated that the insulin resistance of the HFD-VSG mice and HFD + VDD-VSG mice were diminished compared to that of the HFD-sham and HFD + VDD-sham mice respectively. The insulin resistance values also differed between the HFD-VSG and HFD + VDD-VSG mice (Fig. 2C). In the intraperitoneal GTT, the glucose tolerance was improved in the HFD-VSG mice challenged after surgery compared to the HFD-sham mice (Fig. 2D), and the values of the HFD + VDD-VSG and HFD + VDD-sham mice also differed (Fig. 2D). Notably, the HFD-VSG mice exhibited an improved ability to clear glucose, an observation reflected as a 43% reduction in the area under the curve (AUC) relative to the values of the HFD-sham mice. However, compared to the HFD + VDD-sham mice, there was only a 38% reduction in the AUC of the HFD + VDD-VSG mice (Fig. 2E). These results shown that VDD changed the VSG's effect of improving the glucose tolerance of the mice, which demonstrate a normal level of vitamin D is crucial for the regulation of glucose in post-VSG obese mice.

VDD weakened the effects of VSG on intestinal inflammation and permeability improvement in obese mice.
To investigate the effect of VDD on small intestinal inflammation and permeability after VSG, we measured the plasma LPS of the mice after VSG. The plasma LPS levels were significantly decreased after the VSG surgery; in addition, compared to the values of the HFD + VDD-VSG mice, the LPS levels of the HFD-VSG mice were lower (Fig. 3A). www.nature.com/scientificreports/ Since tight junctions (TJs) maintain the enteral epithelial barrier and defects in TJs may result in gut impairment, we analyzed the changes of key components of the TJs expressed in the ileum. Compared to the expressions in the HFD-sham and HFD + VDD-sham mice, the mRNA relative expressions of occludin, ZO-1, and claudin-2 in the ileum regions of the HFD-VSG and HFD + VDD-VSG mice were increased. The mRNA relative expressions of occludin and ZO-1 in the ileum regions of the HFD + VDD-VSG mice were significantly decreased compared to those of the HFD-VSG mice (p < 0.05), but there was no difference in claudin-2 expression between the HFD-VSG and HFD + VDD-VSG mice (p = 0.0652) (Fig. 3B).
Taking these results into account along with the changes of plasma LPS, we speculate that the VSG postoperative intestinal permeability change is associated with the change of LPS levels, and the increased expression of TJs is associated with a decreased LPS level. We also suspect that after a VSG, VDD suppresses the increased expression of TJs and leads to the elevation of LPS, ultimately affecting the integrity of the ileal mucosal lining. We next measured the changes of SCFAs of the cecal contents of mice after they had undergone a VSG. The changes of SCFA in cecal contents are illustrated in Fig. 3C. Compared to the contents in the sham mice, the average SCFA content in the cecal contents of the VSG mice was higher, among which acetic acid, propionic acid, and butyric acid were more readily observed; the other SCFA including isobutyric acid, lactic acid, and caproic acid also showed some between-group differences, but these were not significant. Regarding acetic acid, the HFD-VSG mice and HFD + VDD-VSG mice showed significantly higher mean values compare to the respective HFD-sham and HFD + VDD-sham groups. In addition, the acetic acid values of the HFD-VSG mice differed from those of the HFD + VDD-VSG mice. The same type of findings were observed for propionic acid and butyric acid. The results illustrated that a normal level of vitamin D had effects on intestinal inflammation and permeability in post-VSG obese mice.
VDD affected the alleviation of post-VSG gut dysbiosis. The elevation of the plasma LPS levels by the high-fat feeding may have been the result of gut dysbiosis and impairment of the intestinal interface. We observed that at the phylum level, Firmicutes was dominant over Bacteroidetes in the HFD-sham mice (Fig. 4A). This result was in agreement with clinical findings of increased Firmicutes and decreased Bacteroidetes in obese patients 44,45 . In contrast, in the HFD-VSG mice the abundance of Firmicutes was reduced and that of Bacteroidetes was greatly increased. These observations indicated that the VSG surgery can restore eubiosis, reducing Firmicutes and restoring Bacteroidetes, a major group of Gram-negative microbes that generate endotoxin at their death. www.nature.com/scientificreports/ In the group of HFD + VDD-sham mice, at the phylum level Firmicutes was increased and Bacteroidetes was decreased compared to the HFD-sham mice (Fig. 4A). Among the HFD + VDD-VSG mice, Firmicutes was increased and Bacteroidetes was decreased compared to the HFD-VSG mice.
The symbiotic bacterium Akkermansia muciniphila, which accounted for 3% of fecal bacteria in the HFDsham mice, was increased to 7% in the HFD-VSG mice. In the HFD + VDD-sham group, the ratio of Akkermansia muciniphila decreased to 2%, whereas it was increased to 5% in the HFD + VDD-VSG group. We also applied 16S rDNA high-throughput sequencing to examine the gut microbiota of the mice. Through hierarchical clustering (Fig. 4B,C), we observed increased phylum Firmicutes and decreased Bacteroidetes phylum of the VSG mice compared to the sham-operated mice. Again, the increased phylum Firmicutes and decreased Bacteroidetes phylum of the HFD + VDD-VSG mice were suppressed compared to the HFD-VSG mice. The loss of Bacteroidetes and Akkermansia by the obese mice was restored after VSG (Fig. 4D,E). The results shown that a normal level of vitamin D could alleviate gut dysbiosis in obese mice after VSG sugery. www.nature.com/scientificreports/

Discussion
Obesity is becoming increasingly prevalent worldwide, and obesity and its associated comorbidities have imposed great health challenges and an enormous economic burden on societies. Epidemiologic studies have shown that vitamin D deficiency is associated with obesity [27][28][29] . Obese people are more likely to have vitamin D deficiency, and an inverse relationship between body mass index (BMI) and serum 25(OH)D was found in obese humans. But the precise mechanisms is not clear. A report has demonstrated that adipose tissue plays an vital active role in the regulation of vitamin D metabolism observed during obesity 46 . The other study indicated that abnormal regulation of serum 1,25(OH)2D is associated with major vitamin D metabolizing enzymes 1a hydroxylase and 24-hydroxylase in obesity, which are modified under high-fat diet, thus contributing to the obesity-related reduction of free 25(OH)D 42 . Obesity is also characterized as a chronic low-grade inflammatory disease 47,48 . Studies have shown that LPS is the key factors, and play a crucial in obesity, such as LPS translocation could be enhanced by high-fat meals probably through chylomicron [49][50][51] . In addition, many evidence indicated that rodent models showed that a high-fat diet induces microbiome modifications in order to increase the intestinal permeability, which lead to enhance LPS translocation [52][53][54][55][56] .
Bariatric surgery has been reported to result in important changes in the composition of the gut microbiota, including increases in Proteobacteria and decreases in Clostridium, Bacteroides, and Prevotella 57 . A clinical study 57 described that after a sleeve gastrectomy in subjects with obesity, an increase in the abundance of Akkermansia muciniphila which has a significant ability to degrade mucin and prevent inflammation and adipose tissue alterations 58 . Bariatric surgery is also able to modify the gut microbiota and increase the bacterial diversity, which might influence the circulating levels of LPS 59 .
Previous study showed that baseline glucose was significantly higher in mice which were fed with high-fat diet for 1 week, and a stable hyperglycemia and insulin resistance appeared over time 42 . However, VSG surgery could provides an substantial improvement of glucose 43 . Our present findings showed that the metabolic benefits of a VSG including considerable weight loss, reduced visceral adipose tissue, and improved glucose tolerance and insulin sensitivity are consistent with the results of clinical trials and other rodent studies [31][32][33]60,61 . In addition, the body weights of the HFD mice and the HFD + VDD mice were significantly different at 9 weeks after VSG, and compared to the HFD-VSG mice, the HFD + VDD-VSG mice had more visceral adipose tissue and higher transaminase and plasma LPS levels. Our analysis of the TJ proteins related to intestinal permeability revealed that the mRNA expressions of occludin, ZO-1, and claudin-2 of the ileum of the HFD + VDD-VSG mice were lower than those of the HFD-VSG mice.
Airu Zhu et al. indicated that vitamin D deficiency aggravated the impact of high-fat diet, and lead to robust metabolic disorders 62 . And Su et al. demonstrated that vitamin D could maintains gut eubiosis to antagonize Metabolic syndrome through induction of Paneth cell specific α-defensins 8 . Futhermore, some studies showed the mechanism that vitamin D could improve metabolic syndrome in high-fat diet is up regulation of tight junction components and antimicrobial peptides, partly 63 . Therefore, in our study, we detected the function of normal level of vitamin D in metabolic disorders of VDD in post-VSG obese mice.
SCFAs are produced from non-digestible polysaccharides by gut microbes 64 . SCFAs can also alter the composition of the microbiota 65,66 . The most well-known SCFAs are acetate acid, butyrate acid, propionate acid, and lactate acid. It is well-known that the increased SCFA levels are important for weight regulation because SCFAs can cause satiety so as to decrease the food intake 51 . Further, SCFAs may play important roles in body weight regulation via increasing the total energy expenditure 67 . Also, SCFAs facilitate the absorption of minerals via adjusting the intestinal pH. Meanwhile, SCFAs serve as a substrate of colonocytes (particularly butyrate) to induce a increased portion of beneficial bacteria [68][69][70] .
We analyzed the pH and the changes in the SCFAs of the cecal contents of mice that had undergone a VSG, and the results showed increased contents of acetic acid and butyric acid in the HFD-VSG mice and HFD + VDD-VSG mice, whereas the pH values were decreased in both groups, which is consistent with another report 71 . In our comparison of the HFD-VSG mice and HFD + VDD-VSG mice, we observed that the contents of acetic acid and butyric acid in the feces of the HFD + VDD-VSG mice were decreased, and the pH was increased. Our analysis of the changes in fecal microbes revealed that the abundance of Firmicutes was reduced whereas the abundance of Bacteroidetes was greatly increased in the HFD-VSG mice and HFD + VDD-VSG mice; the abundance of Akkermansia muciniphila was also increased in these mice. However, we also analyzed the changes in fecal microbes between the HFD-VSG and HFD + VDD-VSG mice. The increase in the abundance of Bacteroidetes and Akkermansia muciniphila in the HFD + VDD-VSG mice was less than that in the HFD-VSG mice, and the reduction in the abundance of Firmicutes in the HFD + VDD-VSG mice was smaller than that in the HFD-VSG mice.
Taken together, our findings demonstrated that a deficiency in vitamin D can weaken the effects of VSG including maintaining weight loss, insulin resistance, intestinal inflammation and permeability improvement, and a restoration of the gut dysbiosis. We thus propose that a normal circulating concentration of vitamin D is important for the improvement of metabolic status after VSG.

Data availability
All experimental data used to support this findings are included within the article. www.nature.com/scientificreports/ Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.