Chronic calcitriol supplementation improves the inflammatory profiles of circulating monocytes and the associated intestinal/adipose tissue alteration in a diet-induced steatohepatitis rat model

Vitamin D deficiency and up-regulated TNFα-related signals are reported to be involved in abnormalities including intestinal hyper-permeability, bacterial translocation, systemic/portal endotoxemia, intestinal/adipose tissue/hepatic inflammation, and hepatic steatosis in nonalcoholic steatohepatitis (NASH). This study aims to explore the molecular mechanisms and effects of chronic calcitriol [1,25-(OH)2D3, hormonal form of vitamin D] on gut-adipose tissue-liver axis abnormalities using a high-fat diet (HFD)-fed rat model of NASH. In HFD-fed obese rats on a 10-week calcitriol (0.3 μg/kg/TIW) or vehicle treatment (NASH-vit. D and NASH-V rats) reigme, various in vivo and in vitro experiments were undertaken. Through anti-TNFα-TNFR1-NFκB signaling effects, chronic calcitriol treatment significantly restored plasma calcitriol levels and significantly improved vitamin D receptor (VDR) expression in monocytes and the small intestine of NASH-vit. D rats. Significantly, plasma and portal endotoxin/TNFα levels, bacterial translocation to mesenteric lymph nodes, plasma DX-4000-FITC, fecal albumin-assessed intestinal hyper-permeability, over-expression of TNFα-related immune profiles in monocytes, inflammation of intestinal/mesenteric adipose tissue (MAT)/liver and hepatic steatosis were improved by chronic calcitriol treatment of NASH rats. Additionally, in vitro experiments with acute calcitriol co-incubation reversed NASH-V rat monocyte supernatant/TNFα-induced monolayer barrier dysfunction in caco-2 cells, cytokine release from MAT-derived adipocytes, and triglyceride synthesis by lean-V rat hepatocytes. Using in vivo and in vitro experiments, our study reported calcitriol signaling in the gut as well as in adipose tissue. Meanwhile, our study suggests that restoration of systemic and intestinal vitamin D deficiency using by chronic vitamin D treatment effectively reduces TNFα-mediated immunological abnormalities associated with the gut-adipose tissue-liver axis and hepatic steatosis in NASH rats.

In obese animals, intestinal dysbiosis is associated with increased macrophage infiltration and high cytokine release by mesenteric adipose tissue (MAT), which is positioned near the intestine and is drained by the portal vein [8]. Additionally, intestinal hyper-permeability and MAT inflammation are involved in the pathogenesis of portal endotoxemia and hepatic steatosis [9,10]. Chronic intestinal/MAT inflammation and a disrupted intestinal barrier result in bacterial translocation and the progression of NAFLD to NASH [9, 11,12]. Notably, in NASH, intestinal hyper-permeability, systemic/portal endotoxemia, and systemic/intestinal/MAT inflammation are initiated by TNFα-released from activated immune cells [2,4,5,7,9,10]. So, anti-TNFα agents have the potential to simultaneously ameliorate the aforementioned abnormalities in NASH [9, 10,13,14].
In cultured human peripheral blood monocytes, 1,25-(OH) 2 D 3 , the hormonal form of vitamin D, is able to dose-dependently inhibit LPS-stimulated TNFα production [15]. Serum TNFα levels are negatively correlated with serum vitamin D concentrations in healthy women [16]. Among NASH patients, low serum vitamin D concentrations are closely associated with severe hepatic steatosis and inflammation [17]. Vitamin D receptor (VDR) is a nuclear receptor that mediates most of the known functions of vitamin D, including its anti-TNFα effects. Chronic mucosal inflammation and TNFα-induced down-regulation of gut epithelial VDR can be observed in cases of inflammatory bowel disease [18]. Through direct epithelial VDR up-regulation and indirect TNFα inhibition, vitamin D analog displays mucosal protection and anti-inflammatory effects [19].
Through histological evaluation in rat livers, vitamin D treatment from 6-weeks (time point for diet-induced NASH) to 12-weeks (time point for diet-induced fibrosis) after diet feeding has been found to prevent the progression of NASH to hepatic fibrosis [20]. With histological and gene expression measurements, depletion of vitamin D in a westernized diet exacerbates NAFLD through the activation of toll-like receptor (TLR) in NASH rat livers [21]. In addition to prevent the progression of NASH, vitamin D treatment reduced the serum free fatty acid/triglyceride levels, hepatic thiobarbituric acid-reactive substances (TBARS) levels, and hepatocyte apoptosis in rats [22]. Besides reducing hepatic steatosis, supplementation of vitamin D in the diet alleviates high-fat diet (HFD)-induced overweight and hyperinsulinemia by up-regulating hepatic lipolytic genes and adipose-tissue energy expenditure genes expressions in mice [23]. Especially in adipose tissue of obese rats, depletion of vitamin D in diet exacerbated HFD-increased adipose size by up-regulation of lipogenic/inflammatory genes and macrophage infiltration [24]. Further study reported that vitamin D treatment reduces hepatic triglyceride levels, hepatic nonalcoholic fatty liver disease activity score, and hepatic CD68/TGFβ1/αSMA expression, as well as decreases the levels of serum asparate aminotransferase and alanine aminotransferase in NASH rats [25]. Significantly, vitamin D treatment up-regulates nutrition sensing genes expression in adipose tissue of HFD-fed diabetic mice [26]. Chronic administration of vitamin D-enriched mushroom extracts reduces HFDinduced body fat accumulation, hepatic inflammation/steatosis (by histology and EchoMRI) and serum triglyceride/cytokine levels in mice [27].
To summarize the abovementioned studies [20-27], the therapeutic potential-associated with the anti-hepatic steatosis effects of vitamin D in the parallel abnormalities in circulation and the intestinal and adipose tissue of NASH animals have not been fully explored.
Accordingly, the present study aimed to evaluate the molecular mechanisms and effects of chronic vitamin D treatment on the aforementioned gut-adipose tissue-liver axis abnormalities using a rat model of diet-induced NASH.

Animals
Male Sprague-Dawley (SD) rats (4-week old) were purchased from Charles River Japan, Inc. (Yokohama, Japan) and caged at 22˚C on a 12-hour light-dark cycle with free access to water. Normal-chow-diet (NCD, Laboratory Autoclavable Rodent Diet 5010) or high-fat-diet (HFD, D12492) were given for 14 weeks to form the lean or NASH groups. From 4 to 14 weeks after NCD/HFD feeding, lean rats (Lean-V/lean-vit.D, n = 6 in each group) and NASH rats (NASH-V/NASH-vit.D, n = 9 in each group) received 10-weeks of either vehicle or 0.3 μg/kg/TIW of 1,25(OH) 2 D 3 by gastric gavage.
This study was approved by the Animal Experiments Committee of Yang-Ming University and was performed according to the "Guide for the care and use of laboratory animals" prepared by the National Academy of Science, USA and the ARRIVE guidelines [28]. All efforts were made to minimize animal numbers necessary to produce reliable results and suffering was reduced by administering anesthetics (zoletil and xylocaine). At the end of the experiments, the rats were euthanized with 2-3 times the anesthetic dose of zoletil.

Experimental design
Three days before the following experiments, in vivo intestinal permeability was measured in all rats. Subsequently, heparinized portal vein and whole-body blood was collected to separate peripheral blood mononuclear cells (PBMCs). Meanwhile, the liver, the intestine [doudenum, ileum, cecum and colon], mesenteric fats [the fat surrounding the gastrointestinal tract from the gastroesophageal sphincter to the end of the rectum], and the mesenteric lymph nodes [MLN, drained LNs from the terminal ileum, cecum and ascending colon] were collected by aseptic dissected. The macrophage numbers in the intestine and MAT were measured by flow cytometry. Primary monocytes (CD14 + cells) were isolated from PBMCs and adipocytes were isolated from MAT.

Intestinal permeability
This measurement was based on the intestinal permeability to 4,000-Da fluorescent-dextran [DX-4000-fluorescein isothiocyanate (FITC), FD4000; Sigma-Aldrich, St. Louis, MO] by the analysis of time-dependent serum DX-4000-FITC concentration curves and area under curves (AUCs) [4,6,8]. For validation, the intestinal permeability was re-assessed by measurement of albumin content in the rat feces using ELISA kits (MyBioSource, Inc, San Diego, California, USA).
Monocytes were incubated for 20 h and cell free supernatants were harvested to measure TNFα, IL-6 and MCP-1 levels using ELISA kits (R&D, Minneapolis, MN). The supernatants of 48-hour-cultured NASH-V rat monocytes were used as conditioned medium (NASH-V-CM) to evaluate it effect on caco-2 cell monolayer integrity, lean-V rat adipocyte cytokine release and lean-V rat hepatocytes steatosis.

Various intestinal and mesenteric adipose tissues (MAT) markers
Intestinal caspase-3 activity was determined. The protein and mRNA levels in the rat intestine, rat MAT-derived adipocytes, caco-2 cells, and rat hepatocytes were measured using appropriate antibodies and primers (S1 Table). For flow cytometry analysis, red blood cells (RBCs) present in 1 gram of intestinal and MAT homogenate were lysed using Pharm Lyse (BD Biosciences); the remaining cells (3×10 6 cells/gram tissue) were suspended in PBS containing 2mM EDTA and exposed to FcBLOCK (BD Biosciences) for 20 minutes. To indentify macrophage infiltration (×10 3 /gram tissue) in tissues, the cells (3×10 5 cells) were then labeled with FITC-conjugated F4/80 and PE-conjugated CD11b antibodies for quantification of CD11b(+)F4/80(+) cells by flow cytometry using a FACSalibur analyzer (BD Biosciences), whereas dead cells were identified using propidium iodide (BD Biosciences).

Histological examination
NAFLD activity scores (NASs) were evaluated for rat liver samples. Duodenal tissue samples were used for immunohistochemical (IHC) studies of TNFα, TNFRI, and active caspase-7 expression.

Bacterial translocation (BT) and fecal analysis
BT was defined as positivity of a bacterial culture in sterilized liquefied LMN. Three stool pellets (one per day) from each rat were pooled, dried and stored at 80˚C to allow DNA extraction for quantification of the total number of bacterial cells, including intestinal Lactobacillus spp., Bifidobacterium spp. and Bacteroides-Prevotella spp.

NASH-V-CM/TNFα-stimulated cytokines release from rat MAT-derived adipocytes
Lean-V and NASH-V adipocytes were isolated from rat MATs to measure time-dependent cytokine release [29]. Next, cytokines/cytokine receptor levels in the supernatant and mRNAs levels in the cell lysates of lean-V rat adipocytes treated with incremental concentration of 1,25(OH) 2 D 3 -treated were measured after stimulation with either TNFα or NASH-V-CM.

In vitro effects of calcitriol on NASH-V-CM/TNFα-induced caco-2 cell monolayer barrier dysfunction
Differentiated caco-2 monolayer cells (5×10 5 ) were treated with buffer, TNFα, or NASH-V-CM. Subsequently, 1,25(OH) 2 D 3 was applied or was not to the apical and basolateral compartments for 48 hours. Next, barrier integrity was determined by measurement of the apical-to-basolateral flux of a fluorescent marker [fluorescein sulfonic acid (FSA; 200μg/mL; 478Da)] by the formula of FSA clearance (nl/h/cm 2 ) = Fab/([FSA]a)×S) [30]. Fab is the apical-to-basolateral flux of FSA (light units/h), [FSA]a is the concentration at baseline (light units/nl) and S is the surface area (0.3cm 2 ). Higher FSA represents more severe caco-2 monolayer mucosal dysfunction. Additionally, mRNA/protein expression levels in the caco-2 lysates were measured.

In vitro effects of calcitriol on NASH-V-CM/TNFα-induced lipogenesis on lean-V rat hepatocytes
After isolation and standard preparation [29,31], lean-V rat hepatocytes (5×10 5 ) were cultured with buffer, TNFα, or NASH-V-CM in the presence or absence of 1,25(OH) 2 D 3 for 36 hours. Following this, oil red O stain-based measurement of intracellular TG accumulation was carried out. In parallel, expression of various mRNA/protein in corresponding cell lysates was measured.
Statistical analysis. Data were analyzed using Graphpad Prism 4 (GraphPad Software, San Diego, CA) and expressed as means±S.D. Statistical significance for each group was determined using one-way ANOVA with post hoc multiple comparisons being performed using the Newman-Keuls test. When the criteria for parametric testing were validated, Mann-Whitney U-tests were performed. P value less than 0.05 was considered significant.

Results
Notably, 14 weeks of high-fat diet (HFD) feeding induced typical NASH histology as well as high ALT, AST, fasting glucose and triglyceride levels in NASH-V rats ( Fig 1A, Table 1). Significantly, the HE stain-assessed NAS scores, O-red-oil stain-assessed hepatic steatosis, hepatic triglyceride levels, body weight, and liver weight in NASH-V rats were higher than those in the lean-V rats (Fig 1A-1C, Table 1).

Correction of the hypo-vitaminemia D inhibits systemic/portal endotoxemia and systemic inflammatory profiles
Compared to lean-V rats, NASH-V rats were found to have higher levels of plasma TNFα, LBP, and plasma/portal endotoxin, as well as lower levels of plasma/intestinal calcitriol (Table 2). Compared to the NASH-V group, restoration of plasma calcitriol levels in NASHvit.D rats by 10 weeks of calcitriol treatment was accompanied by the suppression of plasma and portal endotoxin levels as well as the reduction of LBP and TNFα levels ( Table 2). Higher TNFα, TNFRI, NFκB, and TLR4 expression in NASH-V rat monocytes than in the lean-V Chronic calcitriol supplementation improves intestinal/adipose tissue alternation in steatohepatitis rats group was observed and this was accompanied by lower vitamin D receptor (VDR) expression (Fig 1D-1F). Significantly, 10 weeks of calcitriol treatment suppressed the TNFα, TNFRI, NFκB and TLR4 expression and normalized VDR expression in the NASH rat monocytes ( Fig  1D-1F). Nonetheless, no significant difference in the aforementioned markers was found when lean-V and lean-vit.D groups were compared.
MCP-1 and IL-6 from MAT-derived adipocytes was higher in the NASH-vit. D group than that in the NASH-V group (Fig 4B). In lean-V rat adipocytes cultures, NASH-V-CM and TNFα induce the release of aforementioned cytokines and up-regulated the TNFR1/NFκB expressions in cell lysates, which were dose-dependently abolished by vitamin D co-incubation (Fig 4C and 4D).  Chronic calcitriol supplementation improves intestinal/adipose tissue alternation in steatohepatitis rats  (Fig 5C  and 5D). Dose-dependently, incubation with vitamin D normalized the aforementioned NASH-V-CM and TNFα-induced changes in the caco-2 cells culture system (Fig 5).

Acute calcitriol incubation prevents NASH-V-CM/TNFα-induced lean-V rat hepatocytes' lipogenesis
Compared to the buffer group, NASH-V-CM and TNFα (25ng/mL) significantly increased the intracellular triglyceride content of lean-V rat hepatocytes (Fig 6A). Furthermore, the HFD-V-CM and TNFα-induced lean-V rat hepatocyte lipogenesis was accompanied by up- Chronic calcitriol supplementation improves intestinal/adipose tissue alternation in steatohepatitis rats regulation of TNFR1/NFκBp65 as well as lipogenic signals in the cell lysates (Fig 6B and  6C). Notably, incubation with vitamin D dose-dependently reversed the aforementioned NASH-V-CM and TNFα-induced changes in the lean-V rat hepatocyte culture system ( Fig  6A-6C).

Discussion
Increased TNFα-TNFR1 gene expression and elevated soluble TNFα levels are well-recognized pathogenic factors for NASH development [32,33]. Vitamin D can suppress TNF-α and IL-6 production by activated-monocytes in type 2 diabetic patients [34]. In cultured peritoneal macrophages of patients undergoing peritoneal dialysis, vitamin D dose-dependently inhibit LPS-stimulated TNFα release [35]. Our study reveals that the cytokines (TNF-α, IL-6, and MCP-1) released from NASH-V rat monocytes is higher than those in the lean-V group. Significantly, in our study, chronic vitamin D treatment inhibited the cytokines levels in monocytes of NASH-vit. D rats compared to those in NASH-V rats. TNFα-mediated inflammatory responses are mainly mediated by TNFR1 [36]. So, it is reasonable to observe that TNFα- Chronic calcitriol supplementation improves intestinal/adipose tissue alternation in steatohepatitis rats related effects were found to be mediated by the TNFR1-NFκB pathway in monocytes and the tissues of NASH-V rats. Vitamin D and vitamin D receptors (VDRs) are important regulators of intestinal inflammation [18,19,31,37]. In particular, vitamin D and VDR deficiency exacerbates experimental inflammatory bowel disease (IBD) [18,19,37]. It is reported that VDR gene-knocknout mice develop severe intestinal inflammation in experimental models of IBD [37][38][39]. In vitro studies have shown that vitamin D protects against dextran sodium sulfate (DSS)-induced disruption of intestinal epithelial tight junctions [40,41]. In our NASH rats, the intestine VDR expression was negatively correlated with the animals' intestinal TNFα levels, intestinal hyper-permeability and tight junction protein expression and these could be corrected by chronic vitamin D treatment.
Binding of TNFα to TNFR, which activates myosin light chain kinase (MLCK) and NFκB in intestinal epithelial cells can result in epithelial mucosal barrier dysfunction [40][41][42]. Intestinal MLCK over-expression has been reported in patients with ulcerative colitis and Crohn's disease [43]. Up-regulated MLCK and NFκB can induce apoptosis and rearrangement of tight Chronic calcitriol supplementation improves intestinal/adipose tissue alternation in steatohepatitis rats junction proteins, including occludin and ZO-1 [40,[41][42][43]. In our NASH rats receiving chronic vitamin D treatment, down-regulation of intestinal TNFα-TNFR1-NFκB signals prevented intestinal apoptotic activity and preserved the integrity of the intestinal mucosal barrier by decreasing MLCK expression and normalizing tight-junction protein expression. In our caco-2 cell system, TNFα and the supernatant of cultured NASH-V rat monocytes could induce monolayer mucosal dysfunction and the corresponding pathogenic signals, which could be inhibited by vitamin D co-incubation.
In experimental colitis models, 1,25(OH) 2 vitamin D 3 has been found to suppress intestinal mucosal injury, decrease intestinal inflammation and maintain the integrity of the intestinal mucosal barrier [38,39]. A negative correlation has been noted between serum vitamin D and bacterial-translocation in HIV/hepattis C virus co-infected patients [44]. High LBP levels and MLN positive culture rates are representative markers for increased bacterial-translocation [44,45]. In our NASH rats receiving vitamin D treatment, normalization of plasma/tissue calcitriol and VDR levels was accompanied by a decrease in the MLN positive culture rate and in plasma LBP levels. In addition to bacterial-translocation, alteration in gut microbiota is known to be involved in the progression of NASH [3, 13,14]. Modulation of gut microbiota ameliorates obesity-associated impaired intestinal mucosal intergrity and bacterial-translocation in rats [13,14]. In our study, the lack of an effect of chronic vitamin D treatment on the gut microbiota indicates that the benefits of chronic vitamin D treatment in our NASH-vit.D rats involved other mechanisms.
In our lean rat mesenteric adipose tissue (MAT)-derived adipocytes, TNFα and the supernatant of cultured NASH-V rat monocytes induced cytokines release, which could be inhibited by vitamin D co-incubation. The blood and lymphatic vessels draining the gut are embedded in MAT [46]. In NASH animals with endotoxemia, there is a positive regulatory loop between intestinal mucosal barrier dysfunction and MAT inflammation [6,8]. In our NASH rats, vitamin D treatment-related correction of intestinal mucosal barrier dysfunction was accompanied by an improvement in MAT inflammation and a reduction in endotoxemia. In fact, a causal link has been reported between the HFD-induced gut inflammation and the activated inflammatory profile in MAT adjacent to the inflamed intestine [10]. In inflamed adipose tissue, vitamin D can suppress the TNFα and NFκB-mediated cytokines release [47,48]. In NAFLD patients, vitamin D deficiency has been reported to increase the risk of NASH via activation of NFκB signals [17,49]. So, it is reasonable to observe in our NASH rats that macrophage infiltration, inflammation, and TNFα-NFκB-mediated cytokines release in gut and adipose tissues were simultaneously inhibited by chronic vitamin D treatment.
Blood from the portal vein drains to the liver and thus, the liver is exposed to relatively high concentrations of TNFα and IL-6 released by the inflammed gut and MAT [50]. Both in vivo and in vitro studies have reported that TNFα and IL-6 can exacerbate hepatic steatosis [51][52][53]. TNFα can activate NFκB and stimulate IL-6 production from hepatocytes [54]. Vitamin D treatment suppresses hepatic lipogenesis by down-regulation of lipogenic signals [23,25,55]. In our study, chronic vitamin D treatment decrease hepatic steatosis by suppressing the levels of TNFα, NFκB and IL-6 in NASH rat livers. Additionally, in an in vitro study, the suppression of TNFα-TNFR1-NFκB signaling and corresponding lipogenesis by vitamin D co-incubation was observed in lean rat hepatocytes.
In conclusion, our study suggests that systemic/portal endotoxemia, intestinal inflammation, intestinal hyper-permeability, together with up-regulation of TNFα-mediated signaling contribute to down-regulation of VDR expression in monocytes as well as within the intestine of NASH rats. Furthermore, intestinal hyper-permeability exacerbates bacterial translocation to mesenteric lymph nodes and mesenteric adipose tissue inflammation; this subsequently leads to TNFα-TNFR1-NFκB-mediated hepatic steatosis in NASH rats. Intriguingly, the intestinal VDR deficiencies as well as the related TNFα-TNFR1-NFκB-mediated gut/adipose/ liver abnormalities can be effectively attenuated by chronic calcitriol treatment in NASH animals. So, in addition to diet control and exercise, chronic use of vitamin D may be a promising strategy for the improvement of NASH.
Supporting information S1 Table. Primer of rat gene used for quantitative realtime PCR analysis.