The Gut Microbiota Metabolite Urolithin B Prevents Colorectal Carcinogenesis by Remodeling Microbiota and PD-L1/HLA-B

Colorectal cancer has risen to the third occurring cancer in the world. Fluorouracil (5-Fu), oxaliplatin, and cisplatin are the most effective chemotherapeutic agents for clinical chemotherapy. Nevertheless, due to chemotherapeutic drug resistance, the survival rate of patients with CRC remains very low. In this study, we used the inflammation-induced or mutation-family-inherited murine CRC models to study the anticancer and immunotherapy effects of urolithin B (UB), the final metabolite of polyphenols in the gastrointestinal tract. The label-free proteomics analysis and the gene ontology (GO) classifications were used to test and analyze the proteins affected by UB. And 16S rDNA sequencing and flow cytometry were utilized to uncover gut microbiome composition and immune defense improved by UB administration. The results indicated that urolithin B prevents colorectal carcinogenesis by remodeling gut microbial and tumor immune microenvironments, such as HLA-B, NK cells, regulatory T cells, and γδ TCR cells, and decreasing the PD-L1. The combination of urolithin B with first-line therapeutic drugs improved the colorectal intestinal hematochezia by shaping gut microbiota, providing a strategy for the treatment of immunotherapy treatment for CRC treatments. UB combined with anti-PD-1 antibody could inhibit the growth of colon cancer. Urolithin B may thus contribute to anticancer treatments and provide a high immune response microenvironment for CRC patients' further immunotherapy.


Introduction
Colorectal cancer (CRC) is the third most prominent cause of cancer-associated mortality worldwide [1]. The World Health Organization reported that there were 1.93 million new cases of cancer in 2020. Colorectal cancer has risen to the third occurring cancer in the world. In 2020, the incidence of colorectal cancer, with 550 thousand, ranked second in China [2]. The prognosis of colorectal cancer is related to its clinical stage. According to the American Joint Committee on Cancer staging system, the five-year survival rate is less than 60% for patients with colorectal cancer and lymph node metastasis (stage III) and less than 10% for patients with distal metastasis (stage IV) [3]. Cytotoxic chemotherapy and radio-chemotherapy are used to treat patients with stage III CRC [4]. Fluorouracil (5-Fu), oxaliplatin, and cisplatin are the most effective chemotherapeutic agents for clinical chemotherapy [5,6]. Nevertheless, due to chemotherapeutic drug resistance, the survival rate of patients with CRC remains very low [7][8][9]. Early discontinuation is common in stage III colon cancer patients receiving chemotherapy [10]. The discovery of anti-colon cancer drugs has become an urgent problem to be solved.
Notably, in the tumor microenvironment of colorectal cancer, the increased proportion of regulatory T (Treg) cells of colorectal patients is the key factor of tumor immune escape [11,12]. Besides, the number of natural killer cells (NK) is positively correlated with the clinical prognosis of colorectal patients [13,14]. What is more, the natural immune cell, γδ T cell, is mainly distributed in the intestinal epithelium and has a protective effect on the intestine [15,16]. In addition, gut microbiota also plays an important role in the colon cancer microenvironment. For example, Alloprevotella is the pathogenic bacteria in CRC development [17,18], and Akkermansia muciniphila [19][20][21][22] is the immunomodulatory bacteria in cancer development. Akkermansia muciniphila was related to the favorable therapeutic efficacy of immunotherapy, and Bacteroides has been considered an unfavorable bacteria [23,24].
Therefore, in this study, we investigated the cancerpreventive properties of urolithin B on colorectal carcinogenesis. The regulations of gut microbiota, tumor immune microenvironment, and PD-L1 were also investigated.

Materials and Methods
2.1. Animal. Male C57BL/6 mice, 8 weeks, weighing 18-20 g, purchased from Shanghai Model Organisms Center (Inc) were used, maintained in Chinese Experimental Animal Resources Research Institute for food and drug control (SPF standards, room temperature at 20-25°C, relative humidity 50%, 12 h light/12 h dark cycle). All groups were fed with a regular diet of standard feedstuff freely. The studies were approved by the Experimental Animal Management Committee at Capital Medical University (AEEI-2021-254).
Apc min/+ mice with a C57BL/6 background were obtained from Shanghai Model Organisms Center (Inc) and maintained in a standard environment in Chinese Experimental Animal Resources Research Institute for food and drug control. The progeny of Apc min/+ intercrosses were genotyped by PCR analysis with DNA isolated from the tail with the Genomic DNA Mini Preparation Kit with Spin Column (D0063, Beyotime) using the following two primers: 5 ′ -GTGCAGCAGCTTTAAGGAA-3 ′ and 5 ′ -AATGGAACT CGGTGGTAGA-3 ′ . The DNA sequencing was performed with the former primer: GACATGATGATAGTAGGTC AGACAATTTCAATACTGGAAACATGACTGTTCTTT CACCATATTTAAATACTACGGTATTGCCCAGCTCT TCTTCCTCAAGGGGAAGTTTAGACAGTTCTCGTTC TGAGAAAGACAGAAGTT (T/A) GGAGAGAGAGCGAG GTATTGGCCTCAGTGCTTACCATCCAACAACAGAA AATGCAGGAACCTCATCAAAACGAGGTCTGCAGAT CACTACCACTGCAGCCCAGATAGCCAAAGTTATGG AAGAAGTATCAGCCATTCATACCTCCCAGGACGACA (wild type-T and mutant-A).

CRC Model of Colorectal Tumorigenesis and Treatment.
Male C57BL/6 mice were used at the age of 8 weeks, administrated with 12.5 mg/kg AOM (A5486, Sigma, i.p.). One week later, 2.5% DSS (MB5535, Meilunbio) was given in drinking water for one week and then regular drinking water for two weeks. The DSS cycle was repeated three times, and the experiment lasted for 70 days.
APC min/+ mice were used at the age of 8 weeks, with 2.5% DSS (MB5535, Meilunbio) given in the drinking water for one week followed by regular drinking water for two weeks. The experiment cycle was repeated twice and lasted for 50 days.
Treatment groups of low, middle, and high or positive control were given a gavage of 10 mg/kg, 20 mg/kg, and 40 mg/kg urolithin B or 30 mg/kg capecitabine dissolved in 0.25% CMC-Na, and the combination group was given both 30 mg/kg capecitabine and 20 mg/kg UB, while the model group was given the same volume of distilled water, with no administration to the control group. For adjuvant immunotherapy, the mice were administrated with 100 μg anti-PD-1 (clone RMP1-14, BioXcell, i.p.) once a week with or without 20 mg/kg UB.
All mice were weighed regularly. In the end, collect the colorectum of the mice after sacrifice and wash with PBS, to calculate tumor load and prepare them for further experiments.
Tumor load was calculated according to the diameter of the tumors. The tumor load of each mouse is equal to the sum of the average diameters of all tumors on its large intestine.
2.3. Label-Free Proteomics Analysis. SW480 cells were treated with 0.15 or 15 μM urolithin B. Next, cells were lysed, and the concentration of extracted protein was tested by the BCA protein assay kit (Beyotime, P0012). Lyophilizate of the proteins was redissolved, centrifuged, and isolated through the C 18 column, tested by LC-MS/MS, followed by STRING network analysis showing proteins regulated by UB.
2.6. Immunohistochemistry. Mouse CRC tissues were embedded in paraffin and made into slicers. Then, the slicers were deparaffinized in xylene and hydrated into water with graded ethanol (5 minutes in 100%, 5 minutes in 100%, and 5 minutes in 95%, then 5 minutes in 75%).
For HE or PCNA staining, procedures were performed according to instructions (G1005, Servicer Bio) provided by the manufacturer. For the immunohistochemical experiment, slicers were pretreated with antigen retrieval with citrate (G1202-250ML) or Tris-EDTA (G1203-250ML) and then incubated with PBST (PBS containing 0.3% Triton X-100) for 20 minutes and incubated with 3% H 2 O 2 in methanol at 37°C for 20 minutes. After being blocked with goat serum (BOSTER, AR1009) in PBST for 1 h, tissue slides were incubated at 4°C overnight with primary antibodies as follows: anti-PD-L1/CD274 antibody (K009918P, Solarbio). The sections were then washed with PBST for 3 times and incubated with goat anti-rabbit IgG antibody (HRP) (ARG65351, Arigo) at 25°C for 30 minutes. Then, the slicers were treated with the DAB Substrate Kit (20x) (Servicebio, G1212-200T) and then counterstained with hematoxylin. All the slides were imaged at the end for 40x amplification.

Western Blot Analysis.
Colon tissues were isolated and dissected from mice, homogenized, and lysed with ice-cold RIPA Lysis Buffer (P0013C, Beyotime) by a high-speed low-temperature tissue grinding machine, then centrifuged at 8000g for 10 min at 4°C. The protein concentrations were measured using the BCA protein assay kit (Beyotime) according to the instructions. Then, the total proteins (50 μg) were separated with 10% SDS-PAGE and transferred onto a PVDF membrane (Millipore). The membrane was then blocked with 5% skimmed milk powder (P1622, Applygen) dissolved in 1x PBST (RBU164-500, Beijing Roby Biotechnology) for 1 h at room temperature and incubated with primary antibodies overnight at 4°C. The prime antibodies including anti-HLA-B antibody (K009916P, Solarbio) was diluted in blocking solution (1 : 1000, P0252-100ml, Beyotime). And β-actin (K006153P, Solarbio) was used as an internal control. After that, HRP-labeled goat antirabbit IgG (L3012, SAB) was used as the secondary antibody to be incubated with the membrane. Finally, the Immobilon Western HRP Substrate (WBKLS0500, Millipore) was used for the chemiluminescence detection of the bounded antibody.

Statistical Analysis.
All experiments were performed at least three times. The data are presented as the mean ± SE.

Oxidative Medicine and Cellular Longevity
Statistical differences between the control group and treated groups were evaluated using Student's t-test or the Student-Newman-Keuls multiple comparison test. Differences between groups are considered statistically significant at p < 0:05.

Results and Discussion
3.1. The Effect of Urolithin B on AOM/DSS and Apc min/+ Colorectal Carcinogenesis. In both the AOM/DSS colorectal carcinogenesis (CRC) model (Figure 1(a)) and the Apc min/+ CRC model (Figure 1(b)), the fur of mice was disordered and their food intake was reduced after DSS treatment. Visible bloody stool began to appear on the fifth day after drinking DSS. Some of the mice in the AOM/DSS model died after drinking DSS water during the second and third cycles (Figure 1(c)). Compared with the CRC model, the administration of 20 mg/kg UB improved the survival rate; however, the administration of capecitabine failed to improve the survival rate (Figure 1(c)). In addition, the administration of UB reduced the tumor size in both AOM/DSS (Figures 1(d) and 1(e)) and Apc min/+ CRC (Figure 1(g)) models. Compared with the AOM/DSS model, administration of UB (p < 0:01) or capecitabine (p < 0:001) could reduce the adenoma burden of AOM/DSS CRC mice (Figure 1(d)). Although the UB combined with capecitabine group shows no changes in the tumor load and survival rate compared to the capecitabine group (Figures 1(c) and 1(d)), the UB combined with capecitabine group has less intestinal hematochezia (Figure 1(f)).
The colons with adenoma from AOM/DSS mice were sliced and stained with hematoxylin and eosin (Figure 1(h)) or stained by immunohistochemistry with anti-PCNA antibody (Figure 1(i)). Compared with the c57 group, the colons of the AOM/DSS model lost the goblet cells, crypts, and epithelial cells and formatted the atypical hyperplasia or tubular adenoma or adenocarcinoma. The administration of UB fixed those disorders and reduced the expression of PCNA on colons of CRC mice compared with AOM/DSS model mice (Figure 1(i), B). Additionally, there Immunity Resistance Apoptosis Autophagy Bacterial invasion of epithelial cells Figure 3: STRING network analysis showing the direct roles of UB on related immunity/drug resistance/autophagic/apoptotic/bacterial invasion-related pathways. Immunity-related proteins (red circles), resistance-related proteins (purple circles), apoptosis-related proteins (green circles), autophagy-related proteins (black circles), and bacterial invasion of epithelial cells-related proteins (blue circles) are shown. 6 Oxidative Medicine and Cellular Longevity

Label-Free Proteomics Analysis of Urolithin B-Regulated
Proteins. Next, the protein expression levels of SW480 cells treated with urolithin B were analyzed by proteomics analysis. Protein abundance ratios > 2:0 and <0.5 combined with a t-test (p < 0:05) between samples from two groups were used to identify differentially expressed proteins. A total of 914 and 960 upregulated proteins in response to 0.15 and 15 μM UB treatments, respectively, were identified by label-free proteomics (Figures 2(a) and 2(b)). The biological process of UB treatment showed that UB upregulated mitochondrial fragmentation that is involved in the apoptotic process, the positive regulation of the intrinsic apoptotic signalling pathway, the intrinsic apoptotic signalling pathway in response to DNA damage, mitochondrial membrane proteins in the apoptotic signalling pathway, the positive regulation of the apoptotic process, the cell cycle, the intrinsic apoptotic signalling pathway by a p53 class mediator, phagosome acidification, G1/S phase transition of the mitotic cell cycle, and apoptotic mitochondrial changes (Figure 2(a)). Autophagy-upregulated proteins with fold changes greater than 2 and apoptosis-upregulated proteins with fold changes greater than 3.5 are shown in Figure 2(c). STRING network analysis showed that UB-regulated immunity-related proteins ( Figure 3, red circles), UB-regulated resistance-related proteins (Figure 3, purple circles), UB-regulated apoptosisrelated proteins (Figure 3, green circles), UB-regulated autophagy-related proteins (Figure 3, black circles) and UB-regulated bacterial invasion of epithelial cells-related proteins (Figure 3, blue circles) interacted with each other. In general, UB was shown to regulate immunity-related proteins, autophagy-related proteins, apoptosis-related proteins, resistance-related proteins, and bacterial invasion of epithelial cells-related proteins in colon cancer cells.  Figure 4 shows the Shannon and Simpson indexes of alpha diversity results concerning the microbia in the AOM/DSS colon compartment with or without urolithin B supplementation. Compared to capecitabine, the first-line therapeutic drug group, the urolithin B group significantly increased the Shannon index of intestinal flora (p < 0:05, Figure 4(d), B). The Chao index of the microbiota of the urolithin B group was higher than that of the AOM/DSS group (p < 0:05, Figure 4(d), A). After the intervention of capecitabine combined with urolithin B, the microbial structure deviated from the CRC group and approached the c57 group (Figure 4(e)). As shown in Figures 4(a)-4(c) and 5(c)-5(e), compared to the AOM/DSS CRC group, urolithin B decreased the Bacteroides at the family (p < 0:05) and genus (p < 0:05) levels; after the combination of capecitabine and urolithin B, the abundance of Verrucomicrobiota (p < 0:05) and Desulfobacterota (p < 0:01) was increased on the phylum level, the abundance of Akkermansiaceae (p < 0:05) was increased at the family level, and the abundance of Alloprevotella was decreased significantly (p < 0:05) at the genus level. LEfSe results (Figures 5(a) and 5(b)) revealed that urolithin B and capecitabine significantly changed the constitution of the gut microbiota in AOM/DSS CRC mice. The dominant flora of the urolithin B group is c_Bacteroidia, p_Bacteroidota, o_Bacteroidales, and f_Muribaculaceae; that of the capecitabine combined with urolithin B group is g_ Dubosiella, f_norank_o_Clostridia UCG-014, o_Clostridia UCG-014, g_norank_f_orank_o_Clostridia UCG-014, and g_Desulfovibrio. Compared with the AOM/DSS group, UB decreased the abundance of Bacteroidaceae and Bacteroides, and so did the UB combined with capecitabine group (Figures 5(c)-5(e)).

Urolithin B Regulated Immune Factors in Colorectal
Notably, UB inhibits colon cancer and regulates the immune response of colon cancer. UB inhibited the expression of PD-L1 (Figure 7(a)). A combination of UB and capecitabine increased the expression of HLA-B (Figure 7(b)). In the experiment of UB combined with anti-PD-1 antibody (Figure 7

Discussion
Generally, UB prevented colorectal carcinogenesis by shaping gut microbial and tumor immune microenvironment, enhancing the vitality of NK cells, inhibiting the activity of regulatory T cells and the expression of PD-L1, and upregulating the expression of HLA and γδ TCR. UB combined with anti-PD-1 antibody could inhibit the growth of colon cancer and be beneficial to the immunotherapy of colon cancer. The combination of UB with first-line therapeutic drugs further improved the anticancer therapy effect and was favorable to the colon cancer immune microenvironment by regulating the composition of immunomodulatory bacteria.
Inverse correlations between the consumption of plantbased foods and mortality as a result of chronic diseases, such as cardiovascular diseases, neurodegenerative diseases, and cancer, have been investigated. Urolithin B, the final metabolite of polyphenols produced by gut microbiota through the loss of one of the two lactones, is a dibenzopyran-6-one derivative [45,46]. Based on in vitro studies, the health benefits attributed to urolithins are numerous and diverse, from antimalarial properties and topoisomerase inhibition to quenching of bacterial quorum sensing. In vivo studies conducted in colon cancer are very relevant since the colon is the portion of the GI tract where urolithins are produced and achieve bioactive concentrations.
As an adjuvant therapy, FOLFOX chemotherapy can improve the postoperative survival rate of patients with stage III/II colon cancer patients [4][5][6]. However, adjuvant chemotherapy leads to the incidence of bone marrow suppression, and some patients fail to complete the established chemotherapy regimen [10]. In recent years, immunotherapy has developed into an effective strategy for treating advanced cancer [47,48]. However, after surgery, the effect of immunotherapy will be affected by the reduction of the immune response [3]. Therefore, improving the tumor immune microenvironment of CRC patients has a very important guiding significance for clinical treatment.
The regulatory T cells (Tregs) are immunosuppressive in the tumor microenvironment of colon cancer. UB is a high content of multiple intestinal metabolites in food, natural drugs, and traditional Chinese medicine. It is produced by intestinal metabolism in the gastrointestinal tract. This study shows that UB can regulate the tumor microenvironment of colon cancer, upregulate NK cells and γδ T cells, inhibit Treg cells, improve immune response, PD-L1, and upregulate the response of immune regulatory flora after combination with capecitabine. Capecitabine is the first-line drug for chemotherapy treatment; after chemotherapy failure, colon cancer patients will further consider improving immune response in immunotherapy. Reducing the immunosuppressive tumor microenvironment and building a high immune response platform will play an important role in the efficient immunotherapy treatment of colon cancer patients. In the present study, the application of UB may provide a high immune response environment for CRC patients' further immunotherapy.
Intestinal flora in the tumor microenvironment of colon cancer is related to immune response. The immunogenicity of colon cancer is weak, and the pathogenic intestinal bacteria increase after colon cancer surgery and chemotherapy [19][20][21][22][23][24]. The immune response of intestinal flora plays a regulatory role in immunotherapy. Studies have shown that Akkermansia muciniphila is immunomodulatory bacteria [19][20][21][22]. Akkermansia muciniphila was related to the favorable therapeutic efficacy of immunotherapy, and Bacteroides has been considered unfavorable bacteria. The high abundance of those immunomodulatory bacteria in the intestine can upregulate the immune response and improve the treatment rate. Akkermansia muciniphila (A. muciniphila), one of the members of immunomodulatory bacteria, is an anaerobic gram-negative strain that colonizes the mucus layer of mucin-rich ileum and colon. A. muciniphila with appropriate abundance provides nutritional support for intestinal epithelial cells, which can protect an intestinal mucosal barrier, enhance intestinal epithelial integrity, reduce inflammatory response, improve glucose and lipid metabolism, and participate in immune response [49][50][51]. The destruction of microbial structure and the loss of specific bacteria interfere with the efficacy of immunotherapy. By comparing the intestinal flora of responders and nonresponders, the researchers found that the relative abundance of A. muciniphila in tumor patients was correlated with the immune response, and the abundance of A. muciniphila increased in patients who responded to immunotherapy [19][20][21][22]. It is worth noting that A. muciniphila can enhance the immune response to chemotherapeutic drugs [21,22]. Routy et al. summarized the positive correlation between A. muciniphila preclinical tumor model and anticancer immune response in cancer patients [19]. In the present study, we showed that UB significantly upregulated A. muciniphila level and downregulated Bacteroidales and Alloprevotella. So, in the primary site of intestinal metabolism, urolithin may upregulate the abundance of immunomodulatory bacteria and further regulate the immune response. Upregulating the abundance of immunomodulatory bacteria and thus regulating the immune response may be one of the mechanisms of urolithin's anti-colon cancer effect.
Although the abundance of A. muciniphila, a member of immunomodulatory bacteria, is related to immune response, some studies have reported that A. muciniphila in the feces exerted specific effects on colorectal cancer [52,53]. However, A. muciniphila regulates immune response in tumor patients [19,[54][55][56][57][58][59]. Akkermansiaceae was significantly lower in the CRC group [60], and the addition of A. muciniphila significantly inhibited the tumorigenic effect of another flora [59]. It is also associated with favorable responses to antiprogrammed death protein 1 (PD-1) therapy [61]. In the present study, we study the changes in tumor volume and the expression of A. muciniphila in mice after oral administration of urolithin B by using two animal models, AOM/DSS chronic inflammation induced and APC min/+ adenoma colon cancer model. The tumor decreased significantly after the administration of urolithin B, followed by 13 Oxidative Medicine and Cellular Longevity the downregulation of PD-L1 and increase of the expression of A. muciniphila. These results support the enhanced immune response of A. muciniphila in colon cancer.
Alloprevotella, which is correlated with higher levels of chromosomal aberrations [62], might contribute to the development of CRC [63]. In addition, the inhibition of PD-1 decreased the abundance of Alloprevotella [64]. Those reported results illustrated that Alloprevotella not only is related to the development of colon cancer but also participates in the immunotherapy of PD-1 of colon cancer. Our present results show that the combination of urolithin B with first-line therapeutic drugs can downregulate the expression of Alloprevotella and inhibit the expression of PD-L1. Notably, UB combined with anti-PD-1 antibody could inhibit the growth of colon cancer, which further provide a strategy for the treatment of immunotherapy for CRC treatments.
Overall, the prevention of colorectal carcinogenesis of urolithin B is attributed to the shape of gut microbial and tumor immune microenvironment. The anticancer concentration of urolithin B is consistent with the GI concentration of polyphenols, suggesting the protective effect of polyphenols against diseases.

Conclusion
In summary, we showed that urolithin B, the final metabolite of polyphenols in the GI tract, prevented colorectal carcinogenesis by remodeling the gut microbial and tumor immune microenvironment ( Figure 8). UB plays an antitumor role in five aspects. Firstly, UB kills tumor cells by enhancing the vitality of NK cells. Secondly, UB inhibits the activity of regulatory T cells to relieve and promote antitumor immunity. Thirdly, UB inhibits the expression of PD-L1, relieves immunosuppression, and promotes antitumor immunity. In addition, UB upregulates the expression of HLA-B and TCR, plays a role similar to the dendritic cell vaccine, enhances antigen presentation, and promotes antitumor immunity. Lastly, UB regulates the immunomodulatory flora in colon cancer and plays an antitumor role. Compared with the single use of capecitabine, the combination of urolithin B with first-line therapeutic drugs showed better body conditions and increased antitumor effects on colorectal carcinogenesis mice, by shaping gut microbia and providing a strategy for the treatment of immunotherapy for CRC treatments. UB combined with anti-PD-1 antibody could inhibit the growth of colon cancer. Urolithin B may thus contribute to anticancer treatments and provide a high immune response environment for CRC patients' further immunotherapy.

Data Availability
Data generated during the processing or analysis that support the findings of this study are available from the author (lixue_wang@ccmu.edu.cn) on reasonable request.

Conflicts of Interest
All authors declare no competing interests.