Serotonin-Exacerbated DSS-Induced Colitis Is Associated with Increase in MMP-3 and MMP-9 Expression in the Mouse Colon

Background. 5-HT enhances dextran sulfate sodium- (DSS-) induced colitis and is involved in inflammatory bowel disease (IBD). Matrix metalloproteinases (MMPs) play roles in the process of intestinal inflammation. Aims. To examine whether 5-HT induces MMPs expression in mouse colon to enhance DSS-induced colitis. Materials and Methods. C57BL/6J (B6) mice were treated with either low-dose (1.0 mg/kg) or high-dose (2.0 mg/kg) 5-HT by enema, low-dose (1.0%) or high-dose (2.5%) DSS, or combined low-dose (1.0%) DSS and (1.0 mg/kg) 5-HT. Mouse colitis was analyzed. MMPs and tissue inhibitors of MMPs (TIMPs) mRNA were measured by real-time quantitative RT-PCR in mouse colon and in human Caco-2 cells and neutrophils. MMP-3 and MMP-9 protein levels were quantified from immunohistochemistry (IHC) images of mouse colons. Results. 5-HT exacerbated DSS-induced colitis, low-dose 5-HT induces both MMP-3 and MMP-9, and high-dose 5-HT only increased MMP-3 mRNA expression in mouse colon. Mouse colon MMP-3 and MMP-9 protein levels were also elevated by 5-HT treatment. The MMP-2, TIMP-1, and TIMP-2 mRNA levels were increased in the inflamed colon. 5-HT induced MMP-3 and MMP-9 mRNA expression in Caco-2 and human neutrophils, respectively, in vitro. Conclusion. 5-HT induced MMP-3 and MMP-9 expression in mouse colon; these elevated MMPs may contribute to DSS-induced colitis.


Introduction
The inflammatory bowel disease (IBD), comprised of ulcerative colitis (UC) and Crohn's disease (CD), is the recurrent chronic idiopathic inflammatory disease of the GI tract [1][2][3]. Although the precise etiology of IBD remains unknown, a variety of inflammatory mediators and enzymes, such as cytokines, growth factors, reactive oxygen species (ROS), GI hormones, and matrix metalloproteinases (MMPs), are known to regulate inflammation and are involved in pathogenesis of IBD [4,5]. 5-Hydroxytryptamine (5-HT, serotonin), a monoamine neurotransmitter, is biochemically derived from tryptophan. Tryptophan hydroxylase (TPH) is a rate-limiting enzyme in the synthesis of 5-HT and is located prominently in enterochromaffin (EC) cells [6]. Approximately 90% of 5-HT in humans is synthesized in gastrointestinal (GI) EC cells by tryptophan hydroxylase-1 (TPH1) [7], and 5% of 5-HT is synthesized in myenteric neurons [8] and brain by TPH2 [9]. 5-HT is an important mucosal regulatory molecule, which mediates gut motility and secretary functions, promotes growth and turnover of mucosal epithelium, and maintains intestinal homeostasis [10]. 5-HT is also a neuron transmitter in the central nervous system to regulate mood and cause vasoconstriction [10,11]. In the intestinal mucosa, the serotonin reuptake transporter (SERT) takes up 5-HT into epithelial cells and serotonergic neurons and inactivates it [12]. 5-HT exerts its functions by binding to 5-HT receptors which are widely expressed in intestinal tissues [13,14]. 5-HT is involved in the pathogenesis of intestinal disorders, including IBD [15][16][17] and irritable bowel syndrome [15,18,19]. The immunomodulatory effects of 5-HT are associated with gut inflammation [11]. The changes in EC cell numbers and 5-HT contents in colon mucosa of IBD patients have been reported [15,20]. Mice treated with dextran sulfate sodium (DSS) to induce colitis also have increased 5-HT levels contributed by higher number of EC cells and decreased SERT mRNA expression [21]. 5-HT also exacerbated trinitrobenzene sulfonic acid-(TNBS-) induced mouse colitis and is believed to play a key role in pathogenesis of experimental colitis [22,23].
MMPs are a class of structurally related zinc-and calcium-dependent enzymes, which are responsible for the metabolism of extracellular matrix (ECM) via the remodeling degradation of most components of the ECM [24,25]. To date, 23 distinct MMPs have been identified in humans (24 in mice) [26]. MMPs facilitate cell migration and cell infiltration through degradation of basement membrane and ECM [24,27]. MMPs catalyze proteolytic cleavage of biologically active protein molecules, such as cytokines, growth factors, and chemokines, from a membrane-anchored inactive form into a free active form to maintain tissue homeostasis. Under physiological conditions, MMPs are secreted as a latent form of enzyme at low levels and are conversed to active enzymes by proteolytic cleavage [28]. The MMP proteins are tightly controlled at multiple levels, including transcription, translation, secretion, and activation [25,26]. Additionally, MMP proteins are also regulated by tissue inhibitors of MMPs (TIMPs), which are endogenous inhibitors of MMPs. TIMPs consist of four members, TIMP-1, TIMP-2, TIMP-3, and TIMP-4, all of which regulate MMP activities by forming a 1 : 1 complex with the high zinc binding site of MMPs. TIMPs are important regulators of tissue remodeling and cellular behavior [29,30].

Material and Methods
After euthanasia, the colon length was measured, and the sera were collected. The colon tissue was washed in PBS, and then half was fixed in 10% formalin and embedded in paraffin. Sections were stained by hematoxylin and eosin (H&E) for pathological examination or by immunohistochemistry (IHC). The other half of colon was stored at −80 ∘ C for RNA extraction. RNAlater (Qiagen, Hilden, Germany) was used immediately when the samples were collected to prevent RNA degradation. The severity of colitis was evaluated by microscopic examination of colon tissues. Each mouse was scored for inflammation and pathology in a blinded fashion using a modified system as we described previously [47]. These include lymphocytes and neutrophil infiltration (0-3 points), Paneth cell and goblet cell degranulation (0-2 points), epithelium reactivity such as crypt distortion (0-3 points), and inflammatory foci (0-3 points). The threshold for detection of severe acute inflammation corresponds to scores of 6 in this study.

Quantitative Real-Time PCR (qPCR)
. qPCR was performed as we described previously [31]. Briefly, total RNA was extracted using TRIzol Reagent (Invitrogen, USA) according to the manufacturer's instruction. The cDNA was synthesized using PrimeScript6 RT Master Mix (Takara, Japan). The primer sequences for human and mouse samples (Table 1) were designed with Primer3.0 [48] and the primers were synthesized by Sangon Biotechnology (Zhengzhou, China). The reaction mixture of real-time qPCR was performed using a CFX966 Real-Time PCR system (BIO-RAD, USA), and the cyber green was used to detect PCR products. When the primers annealing temperature was 60 ∘ C or more, a twostep PCR reaction method was used. When the primers annealing temperature was below 60 ∘ C, a three-step method was used. Each sample was assayed in triplicate. The efficiency of PCR amplification was 97% to 105%. A melting curve analysis was performed for the PCR products of each target gene and -actin to evaluate primer specificity. The relative abundance of target gene mRNA level was evaluated using the comparative Ct (2 −ΔΔCt ) method and was normalized toactin mRNA level. The mRNA expression level was presented by the relative fold by comparing the quantity of mRNA in different groups. When the Ct value is higher than 35, the gene is considered as unexpressed.
For 5-HT stimulation experiments, cells were seeded in 12well plates at a density of 1.25 × 10 4 cells/well. The cells were serum-starved for 24 h before adding 5-HT to minimize the effect of 5-HT present in the FBS. 5-HT was added to the serum-free medium at different concentrations for up to 24 h.
Heparinized blood samples were collected from consented healthy volunteers. The blood samples were seeded in 6-well plates at a volume of 1.5 mL/well of DMEM and were stimulated with 10 M 5-HT for 24 h. Leucocytes were isolated from the wells by the addition of erythrocyte lysis buffer, containing 0.16 M NH 4 Cl, 10 mM KHCO 3 , and 0.01 mM K 2 -EDTA (pH 7.4) at 4 ∘ C. The lysate was transferred into a 15 mL centrifuge tube and centrifuged at 1500 rpm, 4 ∘ C for 5 min to pellet leucocytes, which were immediately processed for RNA isolation and determination of MMP-9 mRNA levels.

Statistical Analysis.
Student's -test, one-way analysis of variance (ANOVA), and Mann-Whitney test were used to compare the data from different groups. The significant difference was defined as < 0.05. Data was reported as means ± standard deviations (SD). All statistical analysis was performed by using the SPSS 19.0 statistics package (SPSS Inc., Chicago, IL, USA).

5-HT Exaggerated DSS-Induced
Colitis. Similar to Regmi's study that 5-HT exaggerated TNBS-induced colitis, we found 5-HT also exacerbated DSS-induced colitis [22] ( Table 2, Figures 1 and 2). 5-HT alone, at either low (1.0 mg/kg) or high (2.0 mg/kg) dosage, did not induce colitis determined by DAI and histopathological analysis. Mice treated with 1% DSS had signs of mild colitis, including diarrhea and loss of body weight shown at day 5, but the weight loss was less than 10% at day 6. Mild inflammation was seen in colon tissue sections, including increased infiltration of inflammatory cells in the mucosa and submucosa, mild crypt distortion, and shortened crypt depth. Mice drinking 2.5% DSS suffered severe colitis, diarrhea, bloody stools, and losing 21% of body weight at day six. Mice receiving both 1% DSS and 1 mg/kg 5-HT had a similar extent of colitis as those receiving high-dose DSS. The shortened colon was associated with the increased DAI. The pathological analysis demonstrated severe inflammation, including ulcerative lesions, increased number of infiltrating inflammatory cells, crypt loss, and erosion of the mucosa and submucosa.
While 5-HT alone did not induce colitis, high-dose 5-HT significantly increased the levels of proinflammatory cytokine IL-6 and chemokine IL-8 mRNA in the colons (Table 3). DSS alone strongly induced IL-6 mRNA and TNF-, but not IL-8 mRNA levels. Combined 1% DSS and 1 mg/kg 5-HT treatment produced the same pattern of cytokine/chemokine expression as the DSS treatment, that is, no induction of IL-8 mRNA levels.
The data presents 8 mice of each group and stands for relative fold.

MMP-3 MMP-9
Control Control 2.5% DSS 2.5% DSS MMP-3 mainly located in infiltrating cells and stromal cells, some epithelial cell having weak staining. MMP-9 was increased in low 5-HT group but not in high-dose group and also highly increased in tissue with severe inflammation. MMP-9 was located mainly in inflammatory cells. Six mice from each group were included for IHC staining.
of MMP-3 protein in mice treated with low-dose 5-HT is not significantly higher than the control mice (Table 5). MMP-9 was expressed mainly in the infiltrating inflammatory cells (Figure 4, lower panels). Similar to the mRNA expression pattern, MMP-9 protein was elevated in the colons of mice treated with low-dose, but not high-dose, 5-HT. As expected, the colons of DSS-treated mice had nearly 10-fold higher MMP-3 and MMP-9 proteins than the control colons (Table 5).

5-HT Increases MMP-3 in Caco-2 Cells and MMP-9 mRNA in Leucocytes.
Since MMP-3 is expression in colon epithelial cells and MMP-9 is mainly synthesized in leucocytes [26,35], we tested the direct effect of 5-HT on the expression MMP-3 in Caco-2 cells and MMP-9 mRNA in leucocytes. In Caco-2 cells, 5-HT significantly increased the MMP-3 mRNA levels after 24 h treatment with either 5 or 10 M 5-HT compared to the control cells ( Figure 5(a)). MMP-9 mRNA level was not changed under the same conditions (data not shown). MMP-9 mRNA level in leucocytes was also increased 6-fold after 24 h treatment with 10 M 5-HT ( Figure 5(b)). No MMP-3 mRNA was detected in leucocytes (data not shown). [15][16][17] and is well documented to exacerbate chemical-induced colitis in mice [22,23]. The development of IBD is associated with changes of EC cells [49,50]. The elevated EC cell number and decreased SERT mRNA expression lead to increase in 5-HT content, which is associated with the pathogenesis of DSS-induced colitis [21]. Since MMPs are upregulated in IBD and other inflammatory conditions and DSS treatment induced MMP-2, MMP-3, and MMP-9 mRNA levels in mouse colon [45,46], we analyzed the direct effect of 5-HT on the expression of these MMP genes. We found 5-HT alone induced MMP-3 and MMP-9 but not MMP-2 gene expression in the colon. Furthermore, we showed the direct effect of 5-HT induction of MMP-3 expression in Caco-2 cells and MMP-9 expression in leucocytes. Thus, we provided the first evidence to show that 5-HT has a direct effect on the induction of MMPs in two different cell types.

5-HT is involved in pathogenesis of IBD
5-HT-induced MMP-3 and MMP-9 gene expression may be mediated by induction of proinflammatory cytokines, since IL-6 and IL-8 mRNA levels are significantly elevated in the colons of mice treated with high-dose 5-HT. Induction  of the same proinflammatory cytokines is in rat colons after being treated by 1.0 mg/kg 5-HT [22]. These 5-HT activated cytokines and chemokine are produced by local neutrophils, monocytes/macrophages, dendritic cells, and T cells [51][52][53] and are known to induce MMP-3 expression [37,38]. MMP-9 is predominantly synthesized in inflammatory cells, and the consequence of MMP-9 induction draws more leukocytes from the circulation to colon mucosa to worsen inflammation [54]. Furthermore, MMP-9 can potentiate IL-8 activity tenfold, further enhancing the chemokine activity [55]. Therefore, mice treated with combined 5-HT and lowdose DSS could produce as severe colitis as mice treated with high-dose DSS. We noted that the expression of MMP-9 in gene and protein was no longer increased in the colons of mice treated with high-dose 5-HT in spite of having high levels of IL-6, IL-8, and TNF-expression. It suggested that there is a doseeffect of 5-HT on MMP-9 expression. Apparently there is a negative feedback mechanism between MMP-9 protein and 5-HT biosynthesis, since in Mmp9 −/− mice, the TPH1 gene expression is highly regulated [56]. Therefore, the high level of 5-HT inhibits further induction of MMP-9 gene, although the exact mechanism is unknown.
While MMP-2 has the same function with that of MMP-9 in proteolytical degradation of gelatin, the expression of MMP-2 is ubiquitous and constitutive under physiological conditions [25,26]. The expression of MMP-2 does not respond to inflammatory stimuli because MMP-2 gene lacks the binding site for inflammatory transcription factor [26]. It is possible that the high basal level of MMP-2 expression makes it more difficult to detect a low level of induction by 5-HT. This point is supported by the smaller fold of induction in DSS-treated colons; MMP-2 gene was induced 1.8-to 6.3-fold, whereas MMP-3 was induced 28-to 218-fold, and MMP-9 was 11-to 111-fold. Although we did not test the function of MMPs analyzed in this study, we suspect the elevated expression of MMP-9 aids, while MMP-2 suppresses inflammation as illustrated by several groups [57,58]. The function of MMP-3 is less clear in the colon. 5-HT did not affect TIMP-1 and TIMP-2 gene expression, although DSS-induced strong inflammation induced both TIMP gene expressions. The similar pattern of induction of TIMPs and MMPs has been reported for IBD patients [59]. It was reported that the ratio of MMP-3/TIMP-1 is increased in inflamed colons compared with noninflamed colon tissues of IBD patients [60]. The results from human and animal model indicate that MMPs and TIMPs are involved in the process of inflammation and their expressions are regulated by multifactor and multistep, and cytokines and growth factors are involved in expression regulation of both MMPs and TIMPs [30,61].

Mediators of Inflammation
Besides inhibition of MMPs, TIMPs have many other functions such as regulation of cell proliferation, cell migration, angiogenesis, and apoptosis [29].

Conclusion
Our findings show that 5-HT induces the expression of MMP-3 and MMP-9 in mouse colon. This is a direct effect since 5-HT induces MMP-3 mRNA levels in human colon cancer Caco-2 cells and MMP-9 mRNA in human primary culture of leucocytes. Our results support the notion that 5-HT exacerbates DSS-induced colitis by enhancing the production of MMP-3 and MMP-9. The significance of our findings is that we have provided a link between 5-HTassociated colitis and MMPs.