The Role of Aeromonas-Goblet Cell Interactions in Melatonin-Mediated Improvements in Sleep Deprivation-Induced Colitis

Background Our previous studies demonstrated that melatonin could effectively ameliorate sleep deprivation- (SD-) caused oxidative stress-mediated gut microbiota disorder and colitis. The research further clarified the mechanism of melatonin in improving colitis from the perspective of the interaction between Aeromonas and goblet cells. Methods A seventy-two hours SD mouse model with or without melatonin intervention and fecal microbiota transplantation (FMT) to explore the vital position of Aeromonas-goblet cell interactions in melatonin improving SD-induced colitis. Moreover, Aeromonas or LPS-supplied mice were assessed, and the influence of melatonin on Aeromonas-goblet cell interactions-mediated oxidative stress caused colitis. Furthermore, in vitro experiment investigated the regulation mechanism of melatonin. Results Our study showed that SD induced colitis, with upregulation of Aeromonas and LPS levels and reductions in goblet cells number and MUC2 protein. Similarly, FMT from SD mice, Aeromonas veronii colonization, and LPS treatment restored the SD-like goblet cells number and MUC2 protein decrease and colitis. Moreover, LPS treatment downregulated the colonic antioxidant capacity. Yet, melatonin intervention reversed all consequence in SD, A.veronii colonization, and LPS-treated mice. In vitro, melatonin reversed A. veronii- or LPS-induced MUC2 depletion in mucus-secreting human HT-29 cells via increasing the expression level of Villin, Tff3, p-GSK-3β, β-catenin, and melatonin receptor 2 (MT2) and decreasing the level of p-IκB, p-P65, ROS, TLR4, and MyD88 proteins, while the improvement effect was blocked with pretreatment with a MT2 antagonist but were mimicked by TLR4 and GSK-3β antagonists and ROS scavengers. Conclusions Our results demonstrated that melatonin-mediated MT2 inhibits Aeromonas-goblet cell interactions to restore the level of MUC2 production via LPS/TLR4/MyD88/GSK-3β/ROS/NF-κB loop, further improving colitis in SD mice.


Introduction
Sleep has a vital effect on the homeostasis of gastrointestinal mucosal barriers and intestinal microbiota [1,2]. Insufficient sleep is closely related to lots of adverse outcomes, including higher risk of cardiovascular disease, diabetes mellitus, coronary heart disease, hypertension [3], and inflammatory bowel disease (IBD) [4]. In fact, some researches have indicated that sleep disorders may disrupt the immune homeostasis in the intestines, further inducing inflammatory response, and resulting in the occurrence of IBD [4]. Conversely, IBD patients also suffer from poor sleep quality and markedly prolonged sleep latency, as well as frequently sleep fragmentation [5], which highlights the closely correlation between IBD and sleep deficiency.
The intestinal homeostasis relies on closely regulated cross-talk between the mucosal immune, intestinal microbiota, and intestinal epithelial cells (IECs) [6]. Mucus layer damage accelerates intestinal epithelium-pathogen interactions and the pathogen invasion [7]. O-glycosylated mucin (MUC)2, produced by goblet cells, constitutes the major component of the intestinal mucus layer of the rodents' intestines and is a mechanical barrier by forming a huge network of mucus polymer barrier [8]. Importantly, the reduction of MUC2 content and goblet cells number means a thinner mucus layer, which is closely related to IBD [9].
Considering that mucin has a positive effect on offering protection resist the multiple inflammation caused by toxins and invading pathogenic bacteria, it is vital to distinguish the elements that regulate MUC2 gene expression exposed to insufficient sleep [10][11][12].
Melatonin (N-acetyl-5-methoxytryptamine, MT), synthesized from tryptophan, often used to regulate sleep. MT can regulate a series of molecular process, such as circadian rhythms, sleep control, immune pathway, oxidative stress, apoptosis, and autophagy [13]. Moreover, MT is a neurotransmitter among intestinal hormones, and it affects physiological functions of the gastrointestinal tract (GI), including bicarbonate secretion, motility, permeability, energy utilization, and tight junction proteins in intestines [14]. Specifically, consider that mucin influences the abundance of intestinal microbiota [15] and that some microbiota treats glycan as a nutrient [16,17], bacteria modulation via melatonin may be due to goblet cells differentiation-mediated mucin regulation. Therefore, MT is expected to treat a variety of GI diseases, such as necrotizing enterocolitis, ischemic injuries, and IBD [13]. However, it remains unknown whether MT mediated the regulation of MUC2 synthesis and secretion improves IBD in response to SD. Thus, we aim to determine the influence of SD on MUC2 depletion and the roles of MT using a continuous 72 h SD mouse model and Aeromonas veronii or LPS-treated mucussecreting human HT-29 cells.

Materials and Methods
All experiments were operated which subject to the Guide for the Care and Use of Laboratory Animals published by the Animal Welfare Committee of the Agricultural Research Organization, China Agricultural University (Approval No. CAU20170911-2).
2.1. Animal Model Establishment. 168 male ICR mice (eight weeks old; Vital River Laboratory Animal Technology Co. Ltd., Beijing, China) were fed in 28 cages (six mice/cage) in general environments (temperature: 21 ± 1°C, relative humidity: 50 ± 10%) with a regular 10 h dark: 14 h light cycle (lights on at 7 : 00 am.). The mice eat and drink freely. One week acclimatization later, the mice were casually distributed to 14 groups and used to carry out four experiments: the sleep deprivation experiment: sleep deprivation (SD), SD + melatonin supplementation (SD + MT), and nonsleepdeprived control (CON) groups, the ZT time of day that SD started and MT administration were previously described by Gao et al. and Zhang et al.; [18,19]; fecal microbiota transplantation (FMT) experiment: F-CON, F-SD, F − SD + MT (F-SM), and F-R group, FMT was performed via oral gavage of a feces into wild mice as described Stebegg et al.; [20] Aeromonas veronii colonization experiment: C-CON, C-Aeromonas (C-A), and C − Aeromonas + MT (C-AM) groups; and LPS experiment: LPS, LPS + melatonin supplementation (LPS + MT), LPS + TAK − 242 supplementation (LPS + TAK − 242), and non-LPS control (CON) groups.
Colitis was evaluated daily based on the overall rating of body weight, stool consistency, and fecal occult blood to count the disease activity index (DAI), which indicated in Table 1. The scoring range for DAI scores is 0-4, as previously recorded by Murthy et al. [21]. Specific study design is supplied in the Supplementary Material.

Fecal Occult Blood Test.
Specific study design is supplied in the Supplementary Material. The judgment criterion is negative: (-) there is no rose red or cherry red after 3 minutes; positive: (+) rose red or cherry red appears within 30-60 s; strong positive: (++) rose red or cherry red appears immediately; and the strongest positive: (+++) a deep rose red or deep cherry red appears immediately.

Intestinal Permeability to Fluorescein
Isothiocyanate-(FITC-) Dextran. Two hours before the end of the experiment (6 : 00 am), all mice were given oral administration of 0.6 mg/g body weight of 4-kDa FITC-dextran at a concentration of 80 mg/mL and fasted for 2 hours. Euthanasia was subsequently carried out. Blood was collected by retroorbital ocular hemorrhage and centrifuged (500 × g, 10 minutes) to collect serum. A fluorescence spectrophotometer with emission at 535 nm and excitation at 485 nm was used to calculate the fluorescence value in the serum. Create a standard curve via diluting FITC-dextran within PBS using a standard curve to calculate the content of FITC-dextran of the serum.

Histological Staining.
Specific study design is supplied in the Supplementary Material. The scoring criteria are (I) 0: no obvious inflammation; (II) 2: low-grade inflammation, scattered infiltrating monocytes (1~2 lesions); (III) 4: multifocal lighter inflammation; (IV) 6: high level of inflammatory response and upregulated blood vessel density, as well as obvious thickening of the wall; and (V) 8: the most severe inflammation, accompanied by transmural leukocyte infiltration.
2.6. Enzyme-Linked Immunosorbent Assay (ELISA). A competitive ELISA assay (Uscn Life Science, Inc., Wuhan, China) was used to assess the inflammatory factors (IL-10,   2.11. Determination of ROS Formation. Consistent with the manufacturer's instructions, we use a ROS test kit, purchased from Sigma-Aldrich to detect the ROS content (n = 9). A flow cytometer with an oxidation-sensitive DCFH-DA fluorescent probe was used to measure the intracellular ROS generation. Specific operation is supplied in the Supplementary Material.

Western
Blotting. The appropriate amount of colon segment was quickly homogenized in liquid nitrogen and stored at -80°C refrigerator for western blotting analysis. Specific operation is supplied in the Supplementary Material.

Statistical
Analysis of Data. We use SPSS 10.0 statistical software (SPSS, Inc., Chicago, IL, USA) to analysed the data and showed as themean ± standard error. Differences between groups were statistically analysed using one-way ANOVA, which was used to indicated the significance of differences among groups (P < 0:05 and P < 0:01).

Gene
Sense Antisense 3 Oxidative Medicine and Cellular Longevity MT-supplied elevated the colitis phenotype, goblet cells number and expression level of MUC2, Villin, Tff3, p-GSK-3β and β-catenin and reduced the DAI score, level of Aeromonas and LPS, and contents of TLR4 and MyD88, and there was no obviously difference between the SD+MT and control groups (P > 0:063).

FMT Promotes Reestablishment of the Intestinal
Microecology. As illustrated in Figure 2, compared with the F-CON group, the body weight, colonic length, the level of IL-10, and IFN-γ, the goblet cells number and expression levels of MUC2, p-GSK-3β, and β-catenin proteins decreased by 2:3 ± 0:669% ( Figure

Melatonin Regulates the MUC2 Level in HT-29 Cells
Treated with A. veronii. To determine the vital effect of MT in MUC2 production, we selected mucin-secreting human HT-29 cells exposed to various proportions (1 : 5~5 : 1) of A. veronii for 12 h, which formed a homogeneous population of polarized goblet cells. Compared to the cells in the CON group, A. veronii treatments from 1 : 3-5 : 1 significantly inhibited the proliferation activity and MUC2 expression level of HT-29 cells, which decreased by 44:2 ± 0:010 − 56:9 ± 0:007% ( Figure S7A) and 38:6 ± 0:011 − 62:1 ± 0:005% ( Figure S7D), respectively. A decrease in proliferation activity and MUC2 level was examined after 8 h of infiltrated with 1 : 1 A. veronii, which reduced by 35:2 ± 0:009% ( Figure S7B) and 34:1 ± 0:040% ( Figure S7E), respectively. To investigate the molecular pathway via which MT regulates colonic MUC2 production, cells were supplied with MT at 10 -10 -10 -8 M concentrations 1 h prior to A. veronii treatment (1 : 1) for 8 h. We observed that the proliferation activity and expression level of MUC2 protein were decreased by 54:8 ± 0:007% ( Figure S7C) and 61:2 ± 0:033% ( Figure S7F) in the A. veronii-treated group relative to the CON group, respectively. Obviously, the inhibitory effect of A. veronii on proliferation activity and MUC2 expression was reversed by pretreatment with MT after 10 -8 M MT supplementation.    Figure S8A), which decreased by 54:3 ± 0:084% and 58:4 ± 0:084%, respectively. A reduction in MUC2 expression level was presented after 24 h of exposed to 100 pg/mL LPS ( Figure S8B), which was reduced by 35:2 ± 0:078%. To verify the molecular pathways by which MT regulates colonic MUC2 production, the cells were treated with 10 -8 M MT thirty min prior to LPS (100 pg/mL) for 24 hours. We observed that the expression level of MUC2 protein was decreased by 12:3 ± 0:065% in the LPS-treated group relative to the CON group. Obviously, the inhibitory effect of LPS on MUC2 expression was reversed by supplementation with MT ( Figure S8C). We also explored the ability of MT to improve cell viability in LPS-treated IECs in an LDH assay. Compared with the control cells, LPS caused a large amount of cell death and released LDH ( Figure S8D). Statistically, the LDH level was obviously upregulated ( Figure S8E) in the LPS-treated group relative to the vehicle group. Conversely, the MTT assay indicated that LPS caused a decrease in proliferation index ( Figures S8F, G). However, after MT supplementation, the LDH index was significantly reduced by 12:5 ± 0:094%, while the proliferation capacity observably increased by 17:1 ± 0:075%, which almost returned to the vehicle level.  (Figure 7(f)) compared with the vehicle group. By contrast, MT-supplied improved the stimulatory effect of LPS, leading to no significant difference between the CON group and the MTsupplied group. Moreover, the LPS-induced effect was blocked by pretreatment with TAK242, which showed an upregulation of MUC2, Tff3, and Villin mRNA and a downregulation of TLR4, MyD88, and ROS contents in the LPS+TAK242-treated IECs relative to the LPS-treated group. Similarly, TW119 supplementation effectively reversed these LPS-induced changes, while it had no effect on the expression levels of TLR4 and MyD88. Furthermore, our results demonstrated that NAC imitated the improvement effect of MT and upregulated the expressions of MUC2 protein, Tff3, and Villin mRNA, as well as downregulated the ROS content in LPS+NAC-treated IECs relative to the LPS-treated group. Yet, it had no influence on the expression levels of TLR4 and MyD88. These studies indicated that the inhibitory influence of MT on oxidative stress closely associates with the suppression of TLR4/MyD88 pathway induced by LPS. In contrast, 4P-PDOT 17 Oxidative Medicine and Cellular Longevity pretreatment counteracted the effects of MT. The expressions of MUC2, Tff3, and Villin were decreased by 33:9 ± 0:081%, 35:9 ± 0:040%, and 32:9 ± 0:076%, while TLR4, MyD88, and ROS contents were increased by 42:9 ± 0:064 %, 35:2 ± 0:070%, and 38:9 ± 0:092% in the LPS+MT+4P-PDOT group versus the LPS+MT group.
3.10. Regulatory Effect of Melatonin on the GSK-3β/βcatenin/NF-κB Loop Mediated by TLR4/MyD88 Pathway Inactivation in HT-29 Cells Treated with LPS. Next, we explored how TLR4/MyD88 activation was related to mucin deficiency. The results showed an upregulation in the expression levels of p-P65 (Figure 8(a)) and p-IκB (Figure 8(b)) and a significant reduction in the expression levels of p-GSK-3β (Figure 8(c)), β-catenin (Figure 8(d)), and MT2 (Figure 8(f)) in the LPS-treated group, relative to that in the vehicle group while the expression level of MT1 protein did not change (Figure 8(e)). While MT supplementation effectively improved these LPS-induced changes. In contrast, after treatment with TAK-242 and TWS119, we observed a downregulation of p-P65 and p-IκB and an upregulation of p-GSK-3β and β-catenin relative to the LPS group while these had no influence on the expression of MT2 protein level. Moreover, our results suggested that NAC, imitated the improvement effect of MT and downreg-ulated the expressions of p-P65 and p-IκB proteins in LPS+NAC-treated IECs relative to the LPS-treated group while it had no influence on the expression levels of p-GSK-3β, β-catenin, and MT2 proteins. However, in treatment with 4P-PDOT, the antagonist of MT2 counteracted the therapeutic influences of MT and failed to reverse the changes induced by LPS, which further confirmed the role of the TLR4/MyD88-mediated GSK-3β/β-catenin/NF-κB/ ROS pathway in LPS-treated IECs.

Discussion
Our previous results showed that SD induced intestinal mucosal barrier damage, including a decrease in the number of goblet cells and MUC2 protein and an upregulation of pathogen (Aeromonas) content [18]. Meanwhile, FMT from SD mice to wild mice recapitulates the SD-like Aeromonas content increase and goblet cells number and MUC2 protein level decreases, while MT improved these changes. These results demonstrated a high correlation between Aeromonas level increase and goblet cells number decrease. In the SD experiment, we found that the relative abundance of Aeromonas in the colon of the SD mice was upregulated most significantly, and after melatonin supplementation, it returned to the control level. Aeromonas species belong to Gram- negative microbiota usually resides in the intestines. Members of the genus Aeromonas are related to lots of interintestinal and extraintestinal infections in rodents [22].
Aeromonas obviously induces intestinal inflammation, while can frequently cause extraintestinal inflammatory responses, e.g., biliary system infection, necrotizing fasciitis, cholangitis,  Oxidative Medicine and Cellular Longevity surgical wounds, abdominal meningitis, and posttraumatic cellulitis [23]. To test the vital role of Aeromonas in SDinduced MUC2 depletion, we established an A. veronii colonization mouse model. The results indicated that A. veroniicolonized mice exhibited a colitis phenotype, including an upregulation of Aeromonas and its cell wall composition LPS levels and a downregulation of goblet cells' number and MUC2 protein level. However, MT supplementation significantly suppressed A. veronii colonization-induced proliferation of Aeromonas and LPS, restored MUC2 deficiency, and improved colitis, which strongly indicated that Aeromonas coupling with goblet cells promoted MUC2 depletion, further inducing colitis. Furthermore, we set up a LPS-treated mouse model with or without MT supplementation to demonstrate that excessive Aeromonas-related LPS has a negative effect on MUC2 depletion in SD mice. The study is consistent with previous researches in mice that were exposed to LPS that caused obvious signs of colitis: shorter colonic length, lighter weight, and more serious intestinal permeability [24]. Meanwhile, there was an elevation in LPS level and a reduction in goblet cells number and MUC2 protein, Tff3, and Villin mRNA expression levels in LPS-treated mice. Importantly, the decrease of goblet cells number, MUC2 content, and a thinner intestinal mucus layer closely associates with IBD [9]. Van der Sluis et al. previously pointed that MUC2 -/mice behave with clinical and histological characteristics of colitis [25]. Furthermore, the present findings also showed that MT supplementation could alleviate LPS-induced colitis and the reduction of goblet cells and MUC2 protein, Villin, and Tff3 mRNA. Similarly, Shah et al. showed that MT could promote the proliferation of goblet cells to secrete mucin to resist pathogenic bacteria [26]. Therefore, we speculated that MT could improve colitis by suppressing Aeromonas-goblet cell interactions, thereby upregulating the goblet cells number and MUC2 protein, further improving SD-induced colitis. Meanwhile, we observed a decrease in the expression levels of p-GSK-3β and β-catenin proteins, as well as an increase in TLR4, MyD88 and p-P65, and p-IκB proteins in LPS-treated mice, while MT supplementation restored these changes. Similarly, TAK-242 (an antagonist of TLR4) supplementation mimicked the improvement effect of MT. Researches demonstrated that inappropriate TLR activation can induce prolonged inflammatory response and even autoimmune and inflammatory diseases [27]. Similarly, in rat biliary epithelia, anti-TLR2 antibodies or anti-TLR4 antibodies supplementation could reduce MUC2 expression treated with LPS [28]. It seems that boosting LPS activates the TLR4-MyD88-GSK-3β-NF-κB pathway and promotes MUC2 depletion, further inducing colitis.
In vitro, we further found that LPS and A. veronii could upregulate the LDH index and ROS abundance and downregulate the expression of MUC2 protein level and cell proliferation activity in mucus-secreting human HT-29 cells. However, MT supplementation effectively suppressed this process. Further, the ameliorate effect of MT could be suppressed by 4P-PDOT. However, TAK242 mimicked the effect of MT in LPS-treated cells, while it had no influence on the expression level of MT2 proteins. LPS binds innate immunity TLRs, which activates signalling cascades related to NF-κB transcription factor and mitogen-activated protein kinases (MAPKs) in adipocytes [29]. Furthermore, we found that TWS119 treatment enhanced the expressions levels of p-GSK-3β and β-catenin, ultimately suppressing NF-κB activation and promoting MUC2 secretion, while it had no influence on the expression levels of TLR4 and MT2 protein and MyD88 mRNA. Emerging date has also indicated that GSK-3β mediated the activation of the NF-κB signalling cascade via enhancing the NF-κB transcriptional activity in the nucleus to promote cancer [30]. Similarly, NAC counteracted the effect of LPS in HT-29 cells, which suppressed the activation of NF-κB pathway. Previous researches indicated that oxidative stress-mediated NF-κB activation ultimately relies on the phosphorylation and proteasomal degradation of its inhibitor IκBα, supporting nuclear NF-κB translocation. Therefore, in our study, a large number of in vivo and in vitro experiments have been demonstrated that melatonin-mediated MT2 inhibits aeromonas-goblet cell interactions to restore the level of MUC2 production via the LPS/TLR4/MyD88/GSK-3β/ROS/NF-κB loop, further improving colitis in SD mice. Certainly, additional loss-of-function experiments using knock-down or knockout mice/cell lines would be useful.

Conclusions
Overall, our study revealed that MT2-mediated MT suppressed Aeromonas coupling with goblet cells and restored MUC2 depletion by inhibiting the TLR4/MyD88/GSK-3β/ β-catenin/ROS/NF-κB loop, ultimately improving SDinduced colitis in mice. Obviously, the study supported original evidence for the useful influence of MT as a physiological controller of Aeromonas-induced MUC2 deficiency and supported the recent enlarging of the definition of psychobiotics to include MT-based strategies.

Data Availability
All data generated or analysed during this study are included in this published article.

Conflicts of Interest
The authors declare that they have no potential conflicts of interest, including any financial, personal, or other relationships, with other people or organizations.