Butyrate Impedes the Recruitment of MDSCs to Alleviate CAC Development by Inhibition of the TLR2/MyD88/NF-κB Signaling Pathway

Background: The colitis-associated colorectal cancer (CAC) with inammatory bowel disease (IBD) serving as its prelude often has a poor prognosis due to the hysteretic diagnosis. As a representative of short chain fatty acids (SCFAs), butyrate has been proved to have obvious antitumor effect. Here, we aimed to examine its effect on CAC and possible mechanism in tumor microenvironment (TME). Method: The establishment of CAC mouse model was mainly based on the combination of AOM intraperitoneal injection and DSS three cycle. HE staining was used to analyze the degree of colonic inammation and tumor dysplasia. The proportion of MDSCs population was mainly evaluated by ow cytometry assay. RT-PCR, immunohistochemical staining and western blot analysis was carried out to detect protein molecular expression. Results: In our current study, the AOM-DSS induced CAC mouse model was utilized to evaluate the effect of butyrate on CAC. The administration of butyrate signicantly improved the weight loss, falling survival rate, higher DAI index and anal prolapse caused by the AOM-DSS during the CAC modeling process. Anatomical results including the size and number of tumors and histological results including the abnormal hyperplasia shown by HE staining also conrmed the inhibitory effect of butyrate on CAC. In addition, the proportion of myeloid-derived suppressor cells (MDSCs) assisting tumor immune escape in tumor microenvironment (TME) decreased under the intervention of butyrate. And inammatory mediators including CCL2, IL-6 and TNF-α in TME that induce the recruitment of MDSCs showed the same trend as MDSCs. Toll-like receptor 2 (TLR2) as a receptor molecule related to inammation and immune function was also up-regulated in CAC, accompanied by the synchronous up-regulation of downstream Myd88 and NF-κB molecules, while the use of butyrate signicantly inhibited the upregulation of these molecules. Conclusions: Butyrate might reduce the release of CCL2, IL-6


Background
Colorectal cancer (CRC) due to concealment and high mortality has become the third large malignancy with the changes of contemporary eating habits and lifestyle worldwide [1]. Colitis-associated colorectal cancer (CAC) as a subtype of CRC is the result of interleaving by multifactor including infections, immune response, pro-in ammatory mediators and gut microbiota dysbiosis, etc. And accumulated evidence strongly indicated chronic in ammation is a fundamental factor that contributes to the initiation and progression of CAC. [2]. Due to repetitive cycles of damage and repair in the intestinal epithelium, clinically in ammatory bowel disease (IBD) patients covering ulcerative colitis (UC) and Crohn's disease (CD) are accompanied by a higher risk of developing CAC compared to the general population [3].
Therefore, anti-in ammatory treatment is the most frequently chosen treatment means to contain the development of CAC. For instance, regular usage of aspirin as a kind of non-steroid anti-in ammatory drugs (NSAID) can effectively reduce CRC incidence and international guidelines has recommended aspirin as a primary preventive drug for CRC in speci c populations, which re ects the association between chronic in ammation and CRC from the side [4,5].
Although the exact mechanisms of chronic in ammation eventually developing into CAC remain to be elucidated, the continuous evolution of chronic in ammation into malignant tumor is bound to be accompanied by the changes of tumor microenvironment (TME) [6]. The important components of TME including some recruited various immune cells and in ammatory mediators, play an indispensable regulatory role in tumor formation, proliferation, apoptosis, and migration [7] [7]. As a heterogeneous group of immature myeloid cells, myeloid-derived suppressor cells (MDSCs) can usually be further divided into monocytic MDSCs (Mo-MDSCs) and polymorphonuclear MDSCs (PMN-MDSCs) in both mice and humans, which support tumor development by suppressing antitumor immunity and help tumor cells to evade immune surveillance in the TME [8,9].Moreover, this immunosuppressive effect of MDSCs can be mediated by means of the activation of two enzymes including inducible No Synthase (INOS) and Arginase-1 (Arg1) [10,11]. On the other hand, growing evidence suggest various in ammatory mediators play a key role in promoting IBD as well as inducing CAC through similar mechanism, largely because the in ammatory environment has a similar cytokine composition to TME [12]. For instance, interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) are obviously elevated in CRC in the serum of CRC patients and make a contribution to CRC development and prognosis [13,14]. And CC chemokine ligand 2(CCL2) is considered to promote colorectal carcinogenesis by recruiting PMN-MDSCs [15].
Toll like receptors (TLRs) as a kind of pattern recognition receptors (PRRs), can recognize conserved antigen molecules in the evolution of pathogenic microorganisms, that is, pathogen-associated molecular patterns (PAMPs), and then exert the corresponding functions in innate immune responses [16]. Signals triggered by TLRs are transduced by downstream Myd88/NF-κB signaling pathway to recruit proin ammatory cytokines, and nally promote in ammatory responses and tumor formation [17]. TLR2 as an important part of TLRs family, is a type I transmembrane protein receptor composed of 784 amino acids. Studies have shown that the occurrence of CRC is closely related to the increased TLR2 expression level [18]. Existing studies have proved that TLR2 expression in CRC patients was obviously higher in cancerous tissue than in noncancerous tissue by TaqMan RT-PCR and immunohistochemical analysis [19,20]. C. Xin, et al. found that TLR2 was detected as a direct target of miR-154 as a tumor suppressor in CRC cells. Moreover, overexpression of TLR2 could reverse the tumor suppressive effects of miR-154 on CRC cells [21]. Taken together, TLR2 has been considered as a crucial molecule in the mechanism of mediating CRC formation.
Short chain fatty acids (SCFAs) represented by butyrate are the metabolic end products synthesized from the fermentation of undigested dietary ber, and the metabolites are evaluated as a nutritional support to maintain the balance between gut microbiota and host [22]. SCFAs has attracted extensive attention of many researchers in view of its positive roles including anti-in ammatory and anti-cancer effects, providing energy source for colonocytes, reducing oxidative stress and maintaining the integrity of intestinal epithelium [23]. The increased risk of CRC is closely related to the alterations in the composition of the intestinal ora and the reduction in the SCFAs content [24]. Butyrate as the most studied shortchain fatty acid, can suppress the proliferation and induce apoptosis of colon cancer cell lines [25]. In addition, as a histone deacetylase (HDAC) inhibitor, the main anti-cancer mechanism of butyrate is to promote histone acetylation and ultimately reduce the risk of enteritis and CRC by largely activating the immunomodulatory activity of various tumor suppressors [26]. However, the effect of single butyrate on CAC is lack of direct experimental evidence in vivo, and the exact mechanisms remain to be clari ed. We constructed a mouse model mimicking human CAC by the exposure of azoxymethane (AOM) and dextran sulfate sodium (DSS) to evaluate the alleviating effect of butyrate on intestinal in ammation and transition to CAC. Our main research focus is to determine whether butyrate can affect the components of TME to retard the progress of CAC by inhibition of the TLR2/MyD88/NF-κB signaling pathway.

Animals and induction of CAC murine model
Six to eight-week-old male BALB/c mice were purchased from SPF (Beijing) Biotechnology Co., Ltd. All animal experiments were approved by the Ethics Committee of Jiangsu Vocational College of Medicine (Yancheng, China), and strictly comply with experimental animal operation speci cations. All mice were housed in plastic cages under standard laboratory conditions (23 ± 2°C, humidity of 50 ± 5%, and a 12hour light/dark cycle). Mice were pre-adapted to the environment for one week and provided free access to standard diet and sterile water at regular intervals. The induction of CAC murine model mainly refers to previous studies [27]. All mice (n=10 for each group) were randomly and medially assigned into three groups including control group, model group and butyrate group. Mice of model group and butyrate group were injected intraperitoneally with a single dose of AOM (10mg/kg, Sigma-Aldrich, USA) in PBS on day 0, and followed by three DSS (2.5%, MP Biomedicals, Canada, M.W.=36000-50000 kDa) cycles. In addition, mice in butyrate group were treated butyrate (2%, Sigma-Aldrich, USA) in the drinking water throughout the process. In view of the stability of butyrate, the butyrate solution was renewed weekly. Correspondingly, the control group mice were only given normal drinking water without any treatment (Fig. 1A).
The clinical symptoms of all mice were assessed and recorded weekly during the CAC modeling process, including body weight, diarrhea, bloody stools, death rate, and tumor formation. In addition, the speci c calculation of the disease activity index (DAI) was mainly based on previous experiments according to the three parameters of stool consistency, gross bleeding and weight loss [28]. Mice were sacri ced at day 70 by cervical dislocation. The target colorectal segment (from the ileocecal junction to the anal verge) were removed and cut open along the longitudinal axis. The contents were washed off with PBS (pH 7.4). The three groups of colons were photographed. The number and size of tumors in mouse colon were measured and statistically analyzed by Image J software. After cutting the colon into parts, one part was xed in 10% formalin for histology, and the remaining intestinal tissue was used for total protein and RNA extraction respectively. A similar procedure was used for the spleen and the spleen weight was additionally recorded.

Histopathological And Immunohistochemical Analysis
The obtained intestinal and spleen tissue (4 µm thick) was xed with 10% formalin and embedded in para n to make para n sections, followed by hematoxylin and eosin (H&E) staining according to standard procedure. The degree of tumor atypia and in ammation of CAC mice was graded. The pathological scoring criteria of HE staining in mouse colon was based on the previous standards [29]. For immunohistochemistry analysis, para n sections were dewaxed in water, and then antigen retrieval was carried out by antigen repair solution in a pressure cooker. After the sections were cooled to room temperature, they were incubated with 3% hydrogen peroxide and then blocked with 10% goat serum, so as to exclude non-speci c background staining. and incubate at room temperature for 30 minutes. Finally, the sections were visualized by means of the chromogenic reaction of diaminobenzidine (DAB) substrate, followed by hematoxylin counterstaining for microscopy. Images were captured by uorescence inverted microscope (Nikon, Japan).

Flow Cytometric Analysis
After sacri ce, the removed intestinal tissue was cut along the longitudinal axis, and then the luminal contents were washed with PBS. The intestinal tissue was cut into small 1 cm pieces, followed by with collagenase IV digestion (Worthington, USA). The digested cells were then dispersed into ocs with a pipette and ltered through 100-µm strainers. Finally, the complete single cell suspension was obtained by 40% percoll solution separation solution. As for spleen, splenocytes are mainly obtained by grinding, ltering, and then lysing erythrocytes with ACK lysis buffer (Elabscince, China).
The single cell suspension was counted, and the appropriate numbers of cells were selected for blocking with rat serum for 10min and uorescent antibody staining at 4°C for 30 min. (5ul / test). These antibodies mainly include anti-mouse CD11b-FITC (E-AB-F1081C, Elabscince, China) and anti-mouse Gr1-APC antibodies (E-AB-F 1120E, Elabscince, China). After washing twice with PBS, the cells were analyzed on a BD FACS Verse ow cytometer (BD Biosciences). Data were analyzed by FlowJo (Tree Star Inc.).

Reverse Transcriptionquantitative Polymerase Chain Reaction (Rtqpcr)
Total RNA was extracted mainly by Trizol reagent. The extracted RNA was reverse transcribed into cDNA by reverse transcription kit (RK20402, ABclonal, China), which could be stably stored in the refrigerator at -80 ℃. The speci c procedures were as follows: 25°C for 5 min, 42°C for 15min, 85°C for 5 s, and nally hold at 12°C. The mRNA expressions of target gene were detected by 2X Universal SYBR Fast qPCR Mix (RK21203, ABclonal, China) with β-actin as an internal control. The ampli cation procedure was as follows: pre-denaturation at 95 ℃ for 3 min, 1 cycle; then 45 cycles of 95℃ for 5 s, 60℃ for 30 s; nally end at 25℃. The primer sequences were listed in table 1.
Western blot analysis.

Statistical analysis
All data was presented as mean ± SEM (standard error of the mean). The statistical signi cance of differences in three groups was evaluated by one-way ANOVA using GraphPad Prism 5. P value<0.05 was considered to indicate a statistically signi cant.

General observations
The CAC mouse model was performed in strict accordance with Figure 1A. We rst made a macro level evaluation of the three groups of mice. The symptoms of three mice presented in Figure 1B are representative experimental results. After AOM-DSS treatment, mice showed obvious rectal prolapse and perianal tumor invasion. Interestingly, through butyrate prevention, AOM-DSS treatment did not cause this symptom compared with NC group. In addition, the mice in the experimental group died successively at the 7th and 8th weeks, while the other two groups did not ( Figure 1C). We then evaluated the effect of butyrate on body weight and DAI of AOM-DSS mice. As for weight monitoring, the weight of mice in NC group increased steadily with time, while the weight of mice in experimental group and butyrate group decreased precipitously after each DSS cycle. And the treatment intervention of butyrate effectively slowed down the weight loss (

Butyrate inhibits the expression of in ammatory mediators in both colon and spleen tissues of CAC mice
The in ammatory environments caused by the three cycles of DSS must have a great promoting effect on the induction of CAC, and various pro-in ammatory mediators have played a key role. Therefore, we next detected the pro-in ammatory mediators including TNF-α, IL-6 and CCL2 in colon and spleen tissues using RT-PCR.

Butyrate inhibits AOM-DSS induced CAC by down-regulating the expression of TLR2 receptor
Colitis-associated cancer (CAC) as an in ammation-driven carcinogenesis is inseparable from the activation of toll-like receptors (TLRs)signals in in ammatory environment [30]. As one of TLRs, the relationship between TLR2 and CAC is still controversial. To explore whether butyrate can alleviate in ammation and inhibit the formation of CAC by regulating TLR2 of intestinal epithelium. Through immunohistochemical analysis, we found that intestinal epithelial dysplasia was obvious in AOM-DSS group, and TLR2 was signi cantly up-regulated. The intestinal hyperplasia of CAC mice treated with butyrate was alleviated, and the positive rate of TLR2 was signi cantly reduced ( Figure 6A-B). The results of RT-PCR also showed a consistent conclusion ( Figure 6C). The activation of TLR2 signal will trigger the transmission of downstream signals. MyD88 dependent signal pathway is the main signal pathway for TLR2 to play an immunomodulatory role. It can also induce the production of proin ammatory cytokines by activating NF-κB, and nally drive the transition from in ammation to cancer [31]. Therefore, we used western blot to detect the effect of butyrate on TLR2/MyD88/NF-κB signaling pathway. Western blot results showed that the use of AOM-DSS signi cantly activated this signal pathway, while butyrate inhibited the activation of the pathway in CAC mice [ Figure 6D-G]. In conclusion, butyrate alleviated AOM-DSS induced CAC by inhibiting the expression of TLR2 in intestinal epithelium and blocking the activation of downstream MyD88 dependent signaling pathway.

Discussion
It is indisputable that chronic colitis is an important risk factor for colitis-associated colorectal cancer (CAC), and IBD is usually the precancerous basis of colorectal cancer. Clinically, IBD patients are usually accompanied by the gut microbiota dysbiosis and the decrease of SCFAs content [32]. In addition, epidemiological studies have con rmed that high-ber diets can effectively reduce the incidence rate of colorectal cancer in contrast to high-fat diets, in which the key link for high-ber diets to exert anti-cancer effects is the SCFAs produced by gut microbiota digestion of dietary ber [33]. Butyrate, as the most representative SCFAs, can provide energy source for colon cells to exert mucosal anti-in ammatory and anticancer effects, and the anticancer effect of butyrate has been proved by many studies in both vitro and vivo [34,35] greatly alleviated these apparent symptoms. Further anatomical and histological analysis also con rmed the remission effect of butyrate on CAC referring to the size and number of tumors and the abnormal hyperplasia shown by HE staining. In addition, the increased expression in Ki67, BCL2 and CD31 was found in AOM-DSS induced CAC mice and butyrate signi cantly decreased the positive rate of Ki67, BCL2 and CD31, suggesting that this remission effect of butyrate was achieved by inhibiting the proliferation and promoting the apoptosis of colorectal cancer cells. These ndings are in consistent with our previous study, in which butyrate administration was protective in DSS-induced colitis [37]. In comprehensive view, butyrate plays an important role in inhibiting the progression from chronic colitis to carcinogenesis.
The formation and growth of tumor are often inseparable from the nourishment of tumor microenvironment (TME). They affect each other and co-evolve in a complementary relationship. TME is generally a chronic in ammatory environment with massive in ltration of in ammatory related cells and in ammatory mediators. The diversity of in ammatory cells and in ammatory related factors, tumor cells and stromal cells constitute a complex regulatory network [38]. The cells and molecules within the TME are in a dynamic change process, re ecting the essence of the evolution of the TME. As the main immunosuppressive cell population present in the TME, MDSCs are generated and recruited to the TME to promote the establishment of an immunosuppressive TME that facilitates tumor escape. In addition, MDSCs also assist tumor invasion, angiogenesis and metastasis [39]. Due to the anti-tumor immune effect, MDSCs have become the major obstacle to many cancer immunotherapies and the immunotherapy targeting MDSCs to overcome immune evasion has attracted a lot of attention [40][41][42].
The existing view holds that the immunosuppressive function of MDSCs in CRC is mainly achieved by their ability to inhibit T-cell proliferation and to stimulate Treg development [43]. Similarly, accumulating evidence also shows that MDSCs activity plays a vital role in the promotion of CAC through multiple mechanisms and even obstructs CAC immunotherapy [44]. Consistent with these conclusions, our current study found that MDSCs were up-regulated in the colon and spleen of AOM-DSS induced CAC mice, suggesting the role of MDSCs in promoting the development of CAC. Interestingly, the increase of MDSCs in CAC mice was suppressed after butyrate intervention. Therefore, it can be inferred that the inhibitory effect of butyrate on the progression of CAC is achieved by hindering the recruitment of MDSCs in the intestine.
In addition to the recruitment of immunosuppressive cells dominated by MDSCs, tumor in ammatory microenvironment created by the release of various in ammatory mediators also contributes to the development of CAC. And these tumour-derived in ammatory mediators including TNF-α, IL-6 and CCL2 that accumulate in the TME have also been shown to induce the recruitment and expansion of MDSCs [45]. In particular, chemokine CCL2 can promote tumor growth, progression and metastasis by inducing the accumulation of MDSCs and enhancing their immunosuppressive function during colorectal carcinogenesis. And the deletion of CCL2 in mouse model also leads to the reduction of the MDSC level [15,46]. Our RT-PCR results showed that AOM-DSS induced CAC mice had higher mRNA expression of TNF-α, IL-6 and CCL2 than the NC mice, while butyrate treatment decreased the mRNA expression level of these in ammatory factors of the spleen and colon tissue. It can be considered that butyrate inhibits the accumulation of MDSCs by reducing the release of various proin ammatory and chemokines, thus ultimately inhibiting the development of CAC. In summary, the nal outcome is conceivable that a large number of immunosuppressive cells including MDSCs, Treg cells as well as tumor-associated macrophages (TAM), and in ammatory related factors such as CCL2, TNF-α and IL-6 gather in the TME and jointly lead to a muted immune response that guide tumor immune escape and progression in CAC.
Toll-like receptors (TLRs) mediated MyD88 dependent signals usually represent in ammatory signals, which eventually activate NF-κB and induce the release of various in ammatory mediators and promote the establishment of tumor in ammatory microenvironment. The contribution of TLRs to CAC has also been con rmed that TLRs can promote colon cancer cell proliferation, invasion and metastasis, and conversely inhibition of the expression of TLRs or MyD88 will restrain the growth of colon cancer cells [47]. Paarnio et al found that TLR2 of carcinoma cells was highly expressed while TLR4 expression was lower than that of normal epithelial cells by immunohistochemistry detection in 118 CRC patients [48]. It is also demonstrated that TLR2 stimulation promotes colorectal cancer cell growth via NF-κB signaling pathway and knockout and knockdown of TLR2 can inhibit the proliferation of CAC through animal experiments and cellular experiments, which further proves that TLR2 plays an indispensable role in promoting the progress of CAC [49,50]. The above conclusions are consistent with our current results. We detected TLR2 in the colon of three groups of mice by immunohistochemistry, Western blot and RT-PCR. It was found that the expression level of TLR2 in AOM-DSS group was signi cantly up-regulated compared with the NC group, while the administration of butyrate signi cantly inhibited the expression of TLR2 in colon cancer cells. Similarly, the downstream protein detection of Myd88 and p65 was consistent with the expression of TLR2. These results are highly consistent with the aforementioned results of in ammatory mediators and MDSCs. Therefore, it was concluded that TLR2/MyD88/NF-κB signaling pathway not only played an indispensable role in the progression of CAC, but also was an important mechanism for butyrate to inhibit the progress of CAC.

Conclusion
Our data also revealed demonstrated that butyrate had a signi cant inhibitory effect on the progress of the AOM-DSS induced CAC, and the inhibitory effect of butyrate involved the regulation of MDSCs and in ammatory mediators in TME. Butyrate suppressed the transduction of the TLR2/MyD88/NF-κB signals of colon epithelial cells to reduce the release of in ammatory media such as TNF-α, IL-6 and CCL2, which thereby inhibited the recruitment and aggregation of MDSCs in TME. Once the recruitment of MDSCs in TME was limited, the growth of CAC tumor cells dependent on tumor immune escape was also restrained. Nevertheless, the mechanism by which butyrate inhibit the conversion from IBD to CAC largely remains to be elucidated. Our current research focuses on how butyrate restrains TLR2 signals on colonic epithelial cells of CAC to reduce the aggregation of MDSCs in TME and nally relieve the development of CAC, which emerges a new insight into the mitigation mechanism of butyrate on CAC (Figure 7).  Data are represented as the mean ± standard. *P < 0.05, **P < 0.01 and ***P < 0.001. NS, not signi cant;

Figure 7
Schematic diagram of inhibition of butyrate on AOM-DSS induced mouse CAC.