Effect of Shenling Baizhu San on Lipid Metabolism and Intestinal Barrier of UC with Spleen Deciency

Background: Ulcerative colitis (UC) is a chronic inammatory bowel disease and spleen-deciency (SD) is considered as the common Chinese medicine syndrome of UC. However, the mechanism of Shenling Baizhu San (SLS) on UC with spleen-deciency has not been claried. The aim of this study was to explore the inuence of SD on lipid metabolism and intestinal barrier in UC mice and reveal the underlying mechanism of SLS on UC mice with spleen-deciency. Methods: The male C57BL/6 mice were randomly divided into control group, Dextran Sulfate Sodium (DSS) group, DSS+SD group and SLS group. The data of disease activity index, body weight and length of colon were recorded to evaluate the colitis, the spleen and thymus weight were recorded to evaluate the spleen-deciency, the serum metabolites were detected by untargeted metabolomics, the tight junction and intestinal mucus barrier were evaluated by Alcian Blue-Periodic acid Schiff staining and immunochemical, the activity of ATPase were detected by spectrophotometer, the proinammatory factors and the copy number of mitochondrial DNA were detected by RT-PCR, the relative expression of Sirtuin 1 (SIRT1), Peroxisome proliferator-activated receptor γcoactivators-1α (PGC-1(cid:0)) and Mitochondrial transcription factor A (TFMA) were detected by RT-PCR and western blot. Results: The mice treated by compound method showed decreased spleen and thymus weight as well as more serious intestinal inammation than DSS group. Compared with DSS+SD group, SLS improved intestinal inammation and spleen deciency, ameliorated lipid metabolism and intestinal barrier as well as promoted expression of SIRT1, PGC-1(cid:0) and TFMA in UC mice with spleen-deciency. Conclusion: SLS could be considered as an effective treatment for spleen-deciency in UC and its underlying mechanism might be related to improve lipid metabolism and intestinal barrier through mediating SIRT1/PGC-1(cid:0)


Background
Ulcerative colitis (UC), a chronic in ammatory bowel disease, is featured as abdominal pain, diarrhea and bloody stools on clinical manifestations and continuous and diffuse in ammation of the colorectal mucosa on pathological manifestations [1]. The pathogenesis of UC has not been clari ed and it is thought that genetic susceptibility individuals were in uenced by environmental, microbial and immunemediated factors and then lead to intestinal in ammation [2].
Lipid has an important role in constituting biological membranes and maintaining the integrity of the intestinal barrier [3]. The dysfunction of lipid metabolism was founded in UC [4] and our previous clinical research showed there was abnormal alteration of lipid metabolism in UC patients with spleen-de ciency (SD). SD is a common syndrome of UC [5], however, the in uence of SD on lipid metabolism and intestinal barrier in UC has not been clari ed.
There has been amounts of studies reported that improving mitochondrial function is helpful to maintain lipid metabolism balance [6,7]. Mitochondrial biogenesis refers to adaptive increase of the number and quality of mitochondria to produce ATP when human body was in energy-de ciency [8] and Peng et al found sulforaphane can improve lipid utilization and regulate lipid metabolism by promoting mitochondrial biogenesis [9]. SIRT1/PGC-1αpathway has the effect of promoting mitochondrial biogenesis. Peroxisome proliferator-activated receptor γcoactivators-1α(PGC-1α) is involved in regulating the expression of mitochondrial antioxidant genes and preventing oxidative damage and mitochondrial dysfunction [10]. Sirtuin 1 (SIRT1) can active PGC-1α by deacetylation and enhance mitochondrial biogenesis [11,12]. Mitochondrial transcription factor A (TFAM) has in uence in regulating the transcription of mitochondrial nuclear genes and improving the e ciency of mitochondrial biogenesis [13].
Shenling Baizhu San (SLS) is a well-known prescription used to treat diarrhea with SD and was reported as effective treatment on UC in previous researches [14,15]. However, the mechanism of SLS on UC with SD had not been researched. This study aimed to explore the in uence of SD on lipid metabolism and intestinal barrier in UC mice and reveal the underlying mechanism of SLS intervene on UC mice with spleen-de ciency. All the crude medicinal materials were purchased from Jiangsu Province Hospital of Chinese Medicine. The composition ratios of SLS were shown as follow: Ren Shen 20g, Bai Zhu 20g, Fu Ling 20g, Shan Yao 20g, Lian Zi 10g, Bai Bian Dou 15g, Yi Yi Ren 10g, Sha Ren 10g, Jie Geng 10g, Gan Cao 20g according to "Chinese Pharmacopoeia" [16]. All herbs were soaked in 10-fold volumes pure water for 1 hour and decocted at 100℃ for 40min. The liquid medicine was collected and the process was repeated. Then the twice liquid was put together and centrifuged 4000g for 10min. After that, the centrifuged liquid medicine was rotary evaporate (60°C 70rpm) in a rotary evaporator to 64.5ml with a concentration of 2.4g/ml. Previous study found the dosage of 24g/kg/day achieved the best therapeutic effect, so this study applied the dosage.

Animals and grouping
A total of 24 male C57BL/6 mice were purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. All mice were housed in a condition with temperature (22-26℃) and relative humidity (50-60%) on a reverse 12 h light/dark circle. After acclimating for 1 week, the mice were randomly divided into four groups: control group (Ctrl group) were treated by free drinking and eating on day1-day21; UC mice group (DSS group) were treated by free drinking and eating on day1-day14 and treated by 3% Dextran Sulfate Sodium (DSS) on day15-day21; UC mice with spleen de ciency group (DSS+SD group) were induced to spleen de ciency on day1-day 14 and were treated by 3%DSS on day15-day21; Shenling Baizhu San group (SLS group) were induced as spleen de ciency on day1-day 14 and treated by 3% DSS and SLS on day1-day21.
Modeling of UC in mice 3g DSS powder was dissolved into 100ml autoclaved tap water to make 3% DSS solution. The 3% DSS solution were free given to mice 5ml per mouse per day to induce UC [17]. The bottles were emptied and lled with fresh DSS solution every other day.
Modeling of spleen-de ciency in mice Spleen-de ciency was induced by limited feeding and overstrain according to published studies [18]. Food was provided on odd-numbered days and fast on even-numbered days to limit feeding and mice were forced to swim until they lose the ability to rise to the water surface within 10 seconds every day to induce overstrain.

Macroscopic and histological analysis
Body weight, length of colon and disease activity index (DAI) of mice were determined and hematoxylineosin staining (H&E) were used to evaluate the histological alteration. The spleen and thymus were washed using pre-cooled 0.9% saline, dried on lter paper and weighed respectively. The spleen index and thymus index were calculated according to the following formula: spleen index=spleen weight/mouse weight, thymus index=thymus weight/mouse weight.

Untargeted metabolomics
The LC/MS analysis was performed as previous described. The serum was extracted with 550μl methanol and the liquid was vibrated for 20 min. After centrifugation at 15000rpm for 10min, 500μl supernatant was transferred to 1.5ml Eppendorf tube and evaporated under vacuum to completely dryness. The extract was dissolved with 200 μl methanol and vibrated for 20 min. After centrifugation at 18000rpm for 10min for twice, 60μl supernatant was transferred to injection tube to perform metabolomics analysis using UPLC/MS SYNAPT G2-Si equipped with Waters ACQUITY UPLCH-Class (Waters Technologies, USA).
The T3 chromatographic column (100 mm × 2.1 mm ×1.8μm, Waters, USA) was used for compound separation at 55 °C. The mobile phases were 0.1% formic acid in 60% acetonitrile (solvent A) and 0.1% formic acid in 90% isopropanol (solvent B). The ow rate of the mobile phase was set at 0.4 mL/min. The gradient conditions were used as follows: 0-2min gradient 60% A, 2-2.1min gradient 57% A, 2.1-12min gradient 50%A, 12-12.1min gradient 60%A, 12.1-18min gradient 60%A, 18-18.1min gradient 60%A, 18.1-20min gradient 60%A. Data were acquired in negative mode. The data was imported into Progenesis QI and calibrated with internal standards, then a multivariate data matrix containing precursor ions (m/z), retention time, and peak area was imported into EZinfo data. MetaboAnalyst was used to perform the orthogonal projections to latent structure discriminant analysis (OPLS-DA), heatmap analysis and VIP score analysis.
The metabolites were identi ed by comparing the m/z and MS/MS fragmentation patterns from the METLIN (http://metlin.scripps.edu) and Kyoto Encyclopedia of Genes and Genomes (https://www.kegg.jp/).

Real-time PCR
Total RNA was extracted from colon tissue using Trizol reagent (Invitrogen) and the RNA concentration was determined by Nanodrop 2000. cDNA was synthesized using HiScript Ш RT SuperMix (Vazyme). qPCR assays were performed using ChamQ SYBP qPCR Master Mix (Vazyme) with LightCycler96 (Roche). The primers were provided by Invitrogen. The relative mRNA expression was calculated by normalizing the expression of each target gene to that of β-actin using the 2 -△△Ct method. Table 2 The primers used for RT-PCR

Western blotting analysis
Total protein of colon tissue was extracted using lysis buffer and the protein concentration was measured by BCA kit (Beyotime, Shanghai). The proteins were separated by SDS-PAGE and transferred to PVDF membrane (Millipore). The membranes were blocked in the blocking buffer for 1 hour at 37℃, and incubated with the primary antibodies at 4℃ overnight. After that, the membranes were incubated with the second antibodies for 90 min at room temperature. Finally, the proteins were imaged using ECL reagent kit (Millipore) with Chemiluminescence Imaging System (Bio-Rad) and the images were measured by Image Lab Software.

Statistics analysis
The statistics analysis was conducted by SPSS 16.0 software package. The data was analyzed using one-way ANOVA or non-parametric test according to data distribution. The pairwise comparison was conducted using LSD method. P <0.05 was be considered that the difference was statistically signi cant. The differential metabolites with P <0.05 and normalised abundance> 1000 were screened by using Kruskal-Wallis test.

SLS reduced loss of weight and DAI of model mice
The time-related change of body weight of mice was shown in Fig1A. There was gradual downregulation of body weight in DSS group compared with Ctrl group, which was consisted with previous studies. DSS+SD group exhibited greater loss of body weight than DSS group. However, SLS reduced the loss of weight UC mice with spleen-de ciency. The time-related change of DAI was showed in Fig 1B. In comparison with Ctrl group, DAI in DSS and DSS+SD group both increased, while DAI in SLS group decreased signi cantly in comparison with the DSS+SD group. It was suggested that spleen-de ciency might exacerbate the macroscopic manifestations in mice intervened by DSS, and SLS could relieve the manifestations in mice of UC with spleen-de ciency.

Length of colon in different groups of mice
The comparation of colon length was shown in Fig1C and D. Compared with Ctrl group, the colon length in DSS and DSS+SD group both reduced signi cantly and it was even shorter in DSS+SD group, while the mice in SLS group exhibited longer colon than that in DSS+SD group. It was suggested that DSS shorten the colons of mice successfully and spleen-de ciency exacerbated this alteration, however, SLS treatment ameliorated the shortening of colon in mice of UC with spleen-de ciency.
Spleen and thymus weight in different groups of mice As shown in Fig 2, the spleen weight and spleen index increased in the DSS group and decreased in the DSS+SD group compared with the Ctrl group. After intervention with SLS, both the spleen weight and spleen index increased signi cantly compared with DSS+SD group. Compared with the Ctrl group, both thymus weight and thymus index decreased signi cantly in DSS and DSS+SD group, however those in SLS group exhibited signi cant elevation compared with DSS+SD group. It was suggested the intestinal in ammatory induced by DSS upregulated the spleen weight and downregulate the thymus weight in mice, while spleen-de ciency both downregulated the spleen and thymus weight.
Histopathological observation of colon tissue in different groups of mice The histopathological alteration in different groups was observed by H&E staining. As Fig 3 showed, there was complete colonic epithelial structure and normal crypt morphology in colon tissue of Ctrl group. However, colon tissue of DSS and DSS+SD group were damaged. The histopathological alteration of DSS group were charactered as destroyed colonic epithelial structure, incomplete crypt morphology and in ammatory in ltration in mucosal layer, and that of DSS+SD group were charactered as completely destroyed colonic epithelial structure, disappeared crypt morphology and a large amount of in ammatory in ltration in the mucosal layer and submucosa layer. After intervention with SLS, the colon tissue was shown as relatively complete colonic epithelial structure and crypt morphology and in ammatory in ltration in the mucosal layer. It was indicated that spleen-de ciency aggravated histopathological damage of colon tissue induced by DSS and SLS ameliorated the damage at a microscopic level.

The level of in ammatory cytokines in different groups
In comparison with the Ctrl group, the relative mRNA expression of IL-1β, TNF-and IL-6 in DSS and DSS+SD group increased signi cantly and those in DSS+SD group was higher. After SLS treatment, lower mRNA expression of in ammatory cytokines was shown than DSS+SD group (Fig 4), which signi ed spleen-de ciency promoted the intestinal in ammation in UC mice and SLS reduced the intestinal in ammation.

The relative intensity of serum metabolites in different groups of mice
The result of sPLS-DA analysis showed there were differences in metabolites among the four groups ( Fig  5A), and the heatmap veri ed the difference (Fig 5B). After intervention with SLS, the mice showed consistent tendency with Ctrl group (Fig 5D-L) and it was indicated that UC mice with spleende ciency suffered from imbalance of metabolites, while SLS had an effect on regulating the metabolites balance.

The intestinal barrier in different groups of mice
As shown in the AB-PAS staining results, colonic goblet cells and mucus decreased in DSS and DSS+SD group, while SLS treatment relieved the reduction of goblet cells and restored the mucus. As shown in the IHC results, the expression of ZO-1 and Muc-2 decreased compared with Ctrl group, while SLS treatment upregulated the expression of ZO-1 and Muc-2 compared with DSS+SD group (Fig 6). It was indicated that spleen-de ciency exacerbated the damage of tight junction and mucosal barrier in colon and SLS ameliorated the damage effectively.
The expression of SIRT1 PGC1 TFAM mRNA in different groups of mice The relative mRNA expression of SIRT1, PGC1 and TFAM were detected by RT-PCR as shown in Fig 7. Compared with Ctrl group, both the DSS and DSS+SD group exhibited signi cant reduction in the relative mRNA expression and DSS+SD group exhibited further reduction. Compared with DSS+SD group, the relative mRNA expression of SIRT1, PGC1 and TFAM increased signi cantly in SLS group. The results signi ed UC mice existed mitochondrial biogenesis dysfunction and spleen de ciency might aggravate this dysfunction. Besides, SLS could promote mitochondrial biogenesis in UC mice with spleen-de ciency through SIRT1/PGC-1 pathway.
The protein expression of SIRT1 PGC1 TFAM in different groups of mice The relative protein expression of SIRT1 PGC1 TFAM were detected by western blot as shown in Fig 7. The relative expression of above proteins performed the same trend as mRNA expression. Compared with Ctrl group, DSS and DSS+SD group exhibited signi cant reduction in the relative protein expression and DSS+SD group exhibited further reduction. After SLS intervention, the relative protein expression of SIRT1 PGC1 TFAM in mice upregulated signi cantly in comparison with DSS+SD group.
As shown in above results, spleen-de ciency was considered as an unfavorable factor to mitochondrial biogenesis in UC mice and SLS could be acted as an effective intervention to promote mitochondrial biogenesis through SIRT1/PGC-1 pathway.
The activity of Na + -K + -ATPase and Ca + -Mg + -ATPase in different groups of mice Compared with Ctrl group, Na + -K + -ATPase and Ca + -Mg + -ATPase activity in DSS and DSS+SD group reduced signi cantly (P < 0.05, P < 0.01) and those in DSS+SD group showed further reduction, while the ATPase activity in SLS group performed upregulation compared with DSS+SD group. (Fig 8A and B) The results suggested there was cell transport dysfunction in UC mice and spleen-de ciency aggravated the dysfunction and SLS was suggested could improve cell transport and mitochondrial function in UC mice with spleen-de ciency.
The copy number of mitochondrial DNA in different groups of mice Compared with the Ctrl group, the copy number of mitochondrial DNA (mtDNA) in DSS and DSS+SD group were signi cantly reduced (P < 0.05, P < 0.01) and the DSS+SD group showed further reduction. However, the copy number of mitochondrial DNA in SLS group exhibited signi cant elevation compared with DSS+SD groups after SLS treatment. (Fig 8C) The results suggested there was mitochondrial transcription dysfunction in UC mice and spleen-de ciency exacerbated the dysfunction and SLS was considered as an effective measure to improve mitochondrial DNA transcription in UC mice with spleende ciency.

Discussion
The animal model of Chinese medicine syndrome is a kind of biological characterization model that uses some biological characteristics of animals to simulate the characteristics of human syndrome [19]. There had been no uni ed method to preparing animal models of UC with spleen-de ciency syndrome so far and it was reported that the common methods included bitter cold purging, limited feeding, overstrain, reserpine and so on. These methods can be used alone or in combination. The evaluation indicators of spleen-de ciency syndrome included macroscopic characterization and objective indicators. The macroscopic characterization included decreased food intake, loss of weight, loose stools, dry hair, arched back, unresponsiveness and so on [20]. The objective indicators included decreased D-xylose absorption rate, decreased spleen weight, decreased thymus weight and so on [21]. The compound method consists of DSS, limited feeding and overstrain was applied to prepare model of UC mice with spleen-de ciency in this study was referred to a published study [18] and it was adjusted according to the feature of UC. The results of this study showed UC mice with spleen-de ciency exhibited macroscopic manifestation such as decreased food intake, decreased mobility, dry and yellow hair and abnormal objective indicators such as decreased spleen and thymus weight. Furthermore, the model mice showed bloody stools and increased expression of pro-in ammatory factors, which suggested the compound method is an effective and reliable method to prepare animal model of UC with spleen de ciency.
In the theory of Chinese medicine, the main physiological function of the spleen (Pi) includes govern transportation and transformation and control blood circulating in the vessels. Spleen-de ciency lead to dysfunction of transporting and transforming nutrient substances as well as blood circulating, which nally resulted in diarrhea and bloody stools. For the reason that spleen-de ciency is thought as an unfavorable factor to patients with UC. The spleen weight of mice in DSS group signi cantly increased compared with control group, suggesting that UC mice exhibited splenomegaly and hyperfunction of immunity. However, the spleen weight in DSS + SD group signi cantly decreased compared with DSS group, suggesting that the UC mice with spleen de ciency exhibited hypofunction of immunity. The reduction of thymus weight in DSS group suggested UC mice exhibited thymus atrophy and the thymus weight of mice in DSS + SD group further decreased, which indicated spleen de ciency aggravate the thymus atrophy of UC mice. The thymus index in DSS + SD group were higher than that in DSS group, it might be related to the body weight in DSS + SD group were lower than DSS group. SLS was recorded in "Taiping Huimin Heji Jufang" in Song Dynasty. Lin et al reported SLS relieved colitis associated colorectal cancer through inhibiting epithelial mesenchymal transition induced by transforming growth factor β1 [22]. Chao et al reported SLS improved DSS-induced UC in mice via MAPK/NF-kB and pyroptosis signaling pathway [23]. In this study, mice treated by SLS exhibited increased spleen and thymus weight and decreased expression of pro-in ammatory cytokines, which con rmed the effect of SLS on UC with spleen-de ciency and disproved the reliability of compound method to prepare animal model.
The lipid metabolism imbalance was observed in UC with spleen-de ciency. There were decreased relative intensity of PC(O-7:0/O-7:0), Phosphocholine, PC(O-18:1(1E)/0:0), PC (14:0/0:0), LysoSM(d18:0), LysoPC(20:0), Docosatrienoic Acid and increased peak area of LysoPC(22:6) and Glycerophospho-N-Palmitoyl Ethanolamine in DSS + SD group. As a component of lipid, PC was reported accounts for 35%-72% of phospholipid in animal gastrointestinal mucus and plays a key role in enhancing the resistance of mucus barrier to cavity attack [3]. PC depletion may lead to decrease of surface hydrophobicity and invasion of harmful substances, especially bacteria, in the cavity. It was also reported that oral PC can improve the intestinal barrier function in rats [24]. Additionally, the anti-in ammatory effect by inhibiting TNF-α and NFκB has been detected in PC [25,26], and exogenous PC makes a positive impact on ulcerative colitis and intestinal barrier defense against clostridium di cile toxin [27]. The results indicated that the upregulation of PC may be the reason of SLS improve intestinal barrier of DSS + SD mice.
Mitochondria is the "energy factories" and where β-oxidation occurs. When mitochondrial biogenesis was inhibited, the increased NAD+/NADH ratio activated SIRT1 and PGC-1α. Activated PGC-1α is translocated to the nucleus and combined with the key downstream factor nuclear respiratory factor-1/2, and then activated TFAM [10]. TFAM could protect mtDNA from oxidative damage and further improve the e ciency of mitochondrial biogenesis [28]. Na+-K+-ATPase and Ca+-Mg+-ATPase located at the mitochondrial membrane are the core enzymes for energy conversion in organisms and copy number of mtDNA represents mitochondrial transcription function. Previous study showed resveratrol can upregulate the expression of SIRT1 and PGC-1α in rats with type 2 diabetes to regulate mitochondrial biogenesis[6]. Our study showed reductive expression of SIRT1, PGC-1α, TFAM and declining mitochondrial function in DSS + SD group, suggesting there was mitochondrial biogenesis dysfunction in the colon tissue of UC mice with spleen-de ciency. However, SLS upregulated SIRT1, PGC-1 , TFAM expression and promote mitochondrial function, implying that SLS regulated lipid metabolism and intestinal barrier function by mediating SIRT1/PGC-1α pathway.

Conclusions
In conclusion, spleen-de ciency was an unfavorable factor for the intestinal in ammation, lipid metabolism and intestinal barrier of UC mice with spleen de ciency and SLS could be considered as an effective intervention to regulate the abnormal alteration. The underlying mechanism of SLS treatment might be related to activation of SIRT1/PGC-1 pathway.    Effect of SLS on in ammatory cytokines. a Comparison of TNF-in different groups. b Comparison of IL-1β in different groups. c Comparison of IL-6 in different groups. The data were shown as the mean ± S.E.M. #P < 0.05, ##P < 0.01 vs. the Ctrl group; *P < 0.05, **P < 0.01 vs. the DSS+SD group.

Figure 5
The relative intensity of serum metabolites in different groups of mice. a sPLS-DA analysis. b Heatmap. c VIP score. d-l Comparisons of serum metabolites. The data were shown as the mean ± S.E.M. #P < 0.05, ##P < 0.01 vs. the Ctrl group; *P < 0.05, **P < 0.01 vs. the DSS+SD group.   Comparison of Ca+-Mg+-ATPase activity in different groups. c Comparison of mitochondrial DNA in different groups. The data were shown as the mean ± S.E.M. #P < 0.05, ##P < 0.01 vs. the Ctrl group; *P < 0.05, **P < 0.01 vs. the DSS+SD group.