Background
The human gastrointestinal tract is a diverse microbial community which is composed of hundreds of microbial species [
1]. Studies have focused on the role of gut microbiota on host physiology, which show the microbiota affect gut motility, tissue regeneration (in particular of the villi), gut-associated lymphoid tissue (GALT) maturation, and intestinal barrier function [
2]. Accumulating evidence also suggest that gut microbiome could play a pivotal role in the immune system development, protecting host against invading pathogens, and nutrient reclamation and absorption, thus microbiome might avert carcinogenesis, like in breast cancer [
3]. Alterations in the composition of gut microbiota may lead to dysfunction of motility, which might be one of the pathogenic factors in slow transit constipation (STC) [
4,
5].
The germ-free animals are used in many studies to show the influence of gut microflora on gut motor function. The cecum of germ-free rats was reported to be enlarged up to 10 times its normal size, and both gastric emptying and intestinal transit were prolonged compared to conventionally raised animals [
6]. Barbara et al. [
6] suggested that gut microflora affected intestinal motility mainly by bacterial substances or end products of bacterial fermentation, intestinal neuroendocrine factors, or mediators released by the gut immune system.
The monoamine serotonin (5-hydroxytryptamine [5-HT]) is an important gut neuroendocrine factor which is synthesized mainly by enterochromaffin cells (ECs), and has been demonstrated to regulate gut motility in certain pathways. Furthermore, almost 90% of serotonin is synthesized in colonic ECs [
7]. Tryptophan hydroxylase 1 (TPH1) and TPH2 are rate limiting enzymes in mucosal and neuronal serotonin biosynthesis, respectively [
8]. It is suggested that fecal pellets contributed to the release of 5-HT from ECs through activating local mucosal reflexes and stretch reflexes of colonic migrating motor complex (CMMC) to facilitate propulsion [
9]. Fukumoto et al. [
10] reported that 5-HT from ECs could stimulate 5-HT
3 receptors which were located on the vagal sensory fibers, and the sensory information was transferred to the vagal efferent or it stimulated the release of acetylcholine which affected muscle contraction. 5-HT could also cause the release of Ca
2+ from the sarcoplasmic reticulum after it activated 5-HT
3 receptors and the inositol 1,4,5-trisphosphate pathway, which induced contraction of colonic myocytes [
11]. Yano et al. [
12] suggested that gut microbes regulated the 5-HT level in colon, which impacted intestinal motility and hemostasis. Bile acids (BAs) in intestine were reported to exert opposite functions, which inhibited motility in the small intestine and stimulated motility in the large intestine [
13]. Patients with cholestatic disease had a decreased colonic delivery of free BAs, and they would suffer from constipation [
14]. It is reported that a G protein-coupled receptor TGR5 could be activated by BAs [
15]. Intestinal BAs could promote peristaltic contractions in the colon via TGR5 with the release of 5-HT and calcitonin gene-related peptide (CGRP) [
13]. But few systematical studies focus on the relationship among BAs, 5-HT, microbiota, and gut motility.
Germ-free mice model has been used in several studies which were interested in the relationship between gut microbiota and motility. To the best of our knowledge, the germ-free mice model has several limitations as germ-free mice are born in aseptic conditions, which may not only affect the development of the gut motility system, but also the immunity system and metabolic function [
16]. In some studies, broad-spectrum antibiotics were used to deplete gut commensal microflora for creating pseudo-germ-free mice [
17‐
20]. The antibiotics-treated method has been used to prove the reversibility of microbial effects on host 5-HT and gut motility [
12].
The present study was designed to evaluate whether depletion of gut microbiota by antibiotics could influence gut motility and whether this effect was mediated by regulating host serotonin metabolism.
Methods
Animals and antibiotics treatment protocol
All procedures were conducted according to Model Animal Research Center and were approved by the Experimentation Ethics Review Committee of Nanjing University. Twenty male C57BL/6 mice (6–7 weeks old) were obtained from Model Animal Research Center at Nanjing University and adapted to environmental conditions for one week before experiments. Mice were housed in a temperature and humidity-controlled room with a 12 h light/dark cycle (19.0 ± 1 °C, 55% humidity, and fed on a maintenance diet). For the induction of gut dysbiosis, ten mice were treated with broad-spectrum antibiotics in drinking water for 4 weeks (ampicillin, 1 g/L, Sigma; neomycin sulfate, 1 g/L, Sigma; metronidazole, 1 g/L, Sigma; vancomycin, 500 mg/L, Sigma) [
18].
Analysis of gut microbiota
After antibiotics, stool samples of mice were collected in a sterile micro-tube, which were frozen immediately in liquid nitrogen and stored in a −80 °C freezer until analysis. Total DNA was isolated using a Wizard Genomic DNA Purification Kit, which was following the manufacturer’s instructions (Promega, Madison, USA). The relative abundance of bacteria was measured using 16S rRNA analysis, which was performed at the laboratory of BGI (Huada Gene Institute). Primers were designed to amplify the V3–V4 hypervariable region according to the manufacturer’s recommendations [
21]. Then bioinformatics analysis was performed as described [
22]. The mother v.1.27.0 Standard Operation Procedure (SOP) was used to analyze high quality sequence reads [
23].
Total gastrointestinal transit
After fasted overnight with free access to water, mice were administrated by gavage with a semiliquid solution (0.1 mL) containing Evans blue (5%) and methyl cellulose (1.5%) (M0262, Sigma) [
13]. Then fecal pellets were monitored at 10 min intervals for the presence of the first blue pellet, and the time for expulsion of the first blue pellet was measured [
24].
Measuring colonic transit
After fasted overnight with free access to water, the colonic transit of mice was measured with a bead expulsion test. A 3-mm glass bead was inserted into the colon (2 cm proximal to the anal) using a plastic Pasteur pipette lightly lubricated with lubricating jelly as described [
25]. The time until bead expulsion was measured.
Fecal parameter measurements
Freely feeding mice were observed for 8 h, and the number of pellets was counted every 2 h. Fecal water content was measured by comparing the weight of the pellets at the end of the experiment and after drying (24 h at 60 °C). Finally, fecal dimensions of each mice were measured.
Macroscopic assessment
Animals were weighted every week from the beginning to the end of experiments. Mice were euthanized by cervical dislocation, and the whole intestine (from stomach to anus) were excised. The length of the whole intestine and colon was measured. The weight of cecum with their contents was also measured.
Analysis of smooth muscle contractility
Strips of longitudinal (6 mm in length) muscle from proximal colon were mounted in a 37 °C organ bath with a force transducer (MLT0202; AD Instruments, Spain) connected to a PowerLab (AD Instruments, Australia) recording device. Briefly, Ca
2+-free HEPES-Tyrode (H-T) buffer (140.6 mmol/L NaCl, 2.7 mmol/L KCl, 1.0 mmol/L MgCl
2, 10 mmol/L HEPES, and 5.6 mmol/L glucose, pH 7.4) was used to wash out the lumen contents of proximal colon segments. Then, segments were transferred to the organ bath and equilibrated in H-T buffer (137.0 mmol/L NaCl, 2.7 mmol/L KCl, 1.0 mmol/L MgCl
2, 1.8 mmol/L CaCl
2, 10 mmol/L HEPES, and 5.6 mmol/L glucose, pH 7.4) for 15 min. The resting tension was set to 0.5 g prior to force measurement. The isotonic contractions of longitudinal muscle was recorded for 1 h, with H-T buffer being changed every 15 min. The mean amplitude of the basal tension and the frequency of phasic contractions were measured for 30 min after experiments. The detailed force assays were performed according to our previously described methods [
26,
27].
Serotonin measurements
ELISA was used to assess the serotonin in supernatant of tissue homogenates according to the manufacturer’s instructions (DEE5900, Germany). Readings from tissue samples were normalized to total protein content as detected by BCA assay as described [
12].
RNA preparation and qRT-PCR
The entire length of the mouse colon were washed in H-T buffer, flushed with H-T buffer to remove luminal contents. Then the total RNA was extracted from 1 cm regions of proximal mouse colon using TRIzol reagent (Invitrogen). The qRT-PCR assay was performed by a ABI Prism 7100 Sequence Detection System and 96-well optical plates (Applied Biosystems). The PCR primer sequences were as follows, M-TPH1: forward, 5′TCA AGC CCT TTG ATC CCA AG-3′; reverse, 5′-GAC ATC AAG GTC ATA CCG CAA C-3′.
Immunofluorescence
Whole frozen sections were processed for indirect immunofluorescence and confocal microscopy as described [
12], which were prepared from paraformaldehyde-fixed tissues. Briefly, 10-μm frozen sections of proximal colon samples were collected on coated slides. Then it was washed three times with PBS and blocked with 5% bovine serum albumin (sigma Aldrich) at room temperature for 30 min. After staining using the primary antibodies, rabbit anti-mouse CgA (1:500; Abcam), rat anti-mouse 5-HT (1:50; Abcam), and secondary antibodies conjugated to Alexa fluor 488 or 594 (Molecular Probes). Images were examined using a confocal microscope (Olympus, Germany).
Cecal bile acids determination
Cecal contents from mice were collected, freeze-dried, and pulverized. BAs were analyzed using the LC–MS–MS (Liquid Chromatography–Mass Spectrometry and Liquid Chromatography–Tandem Mass Spectrometry). Each bile acid was identified by its relative retention time compared with that of standard BAs.
Statistical analysis
Statistical analysis was performed using SPSS for windows version 19. Data were measured for normal distribution and plotted in the figures as mean ± SEM. The significance of the differences between groups was determined with Student t test (two comparisons) or one-way ANOVA (multiple comparisons). P < 0.05 was considered significant.
Discussion
The interactions between the composition of the microbiota and gut motility are bidirectional, which indicates that the microbiota play an important role in gut motor function and the gut motor function affects the diversity of microflora in the gastrointestinal tract [
6]. Here, we found that the depletion of microbiota by antibiotics treatments was associated with gut motility, and its relationship might be related to 5-HT biosynthesis in the colon. And we also found that metabolism of bile acids was affected in antibiotic-treated mice, and the secondary bile acids was decreased in our study which has been demonstrated to be associated with 5-HT metabolism and gastrointestinal motility [
13].
Several studies have focus on the interaction between the normal gut microbiota and gut motility. The dysfunction of gut motility has been considered to be one of the main pathophysiology in STC [
31]. Recent years, fecal microbiota transplantation (FMT) has become a new therapy for functional gastrointestinal disorders by engraftment and restoration of normal intestinal microbial community functionality and structure [
32]. One prior study showed that FMT had the high efficiency to improve the syndrome of STC [
33], which suggested that normal gut microbiota could benefit to gastrointestinal motor function. In addition, Xu et al. [
11] also found 5-HT could lead to the contraction of isolated smooth muscle cell, and cell length would be shortened in a 5-HT dose-dependent manner. In our study, after antibiotics treatments, the abundance of gut microbiota decreased significantly in mice, which were also named pseudo-germ-free mice. And antibiotics-treated mice had a decreased intestinal motility, with prolonged gastrointestinal and colonic transit, and poor contraction of smooth muscle in proximal colon. In antibiotics-treated mice, the 5-HT levels in colon was lower, which was consistent with that in a previous study [
12]. Thus our findings revealed decreased intestinal motility of mice treated with antibiotics might result from the decreased levels of colonic 5-HT.
The expression of TPH1 in ECs is essential for 5-HT metabolism. Several studies demonstrated that the expression of TPH1 was decreased, but no difference in expression of other steps in 5-HT metabolism was found, like enzymes for 5-HT release [
12]. After antibiotics treatments, our results suggested that mice had lower expression of TPH1, and gut microbiota might influence expression of TPH1, which resulted in the decreased level of 5-HT in colon. In addition, we found that there was no difference in the colonic CgA
+ ECs in both SPF and antibiotics-treated mice, which suggested that the decreased release of 5-HT in “microbiota depletion” by antibiotics was not due to quantity of ECs. Moreover, metabolism modulated by microbiota could change 5-HT biosynthesis in colon, like α-tocopherol, butyrate, cholate, deoxycholate, p-aminobenzoate, propionate, and tyramine [
12]. Alemi et al. [
13] found that secondary bile acids, like DCA and LCA, caused oral contraction and caudal relaxation, and BAs also led to the release of 5-HT. Interestingly, our study found that secondary bile acids were reduced in antibiotics-treated mice. Thus we proposed that secondary bile acids might influence the expression of TPH1 and then affected the release of 5-HT in colon, which involved in the interaction between microbiota and motility. Further studies are needed to prove whether gut motility would recover after increasing secondary bile acids in mice treated with antibiotics treatments. In addition, gut motility is regulated by multiple signaling pathways with antibiotics, further work is necessary to clarify whether other pathways participate in the regulation of gut motility.
Gradually, studies have started to focus on the relationship between stool consistency and gut microbiota richness [
34,
35]. For example, stool consistency was suggested to be negatively correlated with species richness of gut microbiota in humans, which indicated that the fecal parameter, such as water content, was affected not only by gut motility, but also by gut microbiota richness [
34]. In our study, the pellet frequency was lower in antibiotics-treated mice, which was consistent with the decreased gut motility. However, the water content was not consistent with the decreased gut motility from the traditional perspective. Although decreased gut motility does increase water reabsorbed from feces in the colon, our study suggested that antibiotics resulted in not only decreased gut motility but also gut microbiota richness which might influence intestinal fluid secretion [
36,
37]. Barrett et al. [
36] suggested that commensal microbiota exerted important influences on the gastrointestinal epithelial function, such as transporting fluid and electrolytes in the intestine. Further study is needed to explore whether intestinal fluid secretion is affected in antibiotics-treated mice.
Conclusions
In conclusion, our results from antibiotics-treated mice might differ from some others’, and these might be due to different programs of antibiotics administration, different strains of mice, and environmental conditions. Our findings demonstrated that gut microbiota might play a role in the regulation of secondary BAs and 5-HT metabolism in the colon, which in turn might affect gut motility. These mechanisms may throw new light on the treatment of FMT for functional gastrointestinal disorders.
Authors’ contributions
NL, MZ and JG conceived and designed the experiments; CD, XG, WZ, and HT performed the experiment; LX and XG analyzed the data; XG wrote the paper. JG, JL and NL reviewed the manuscript. All authors read and approved the final manuscript.