Mucin Binding Protein of Lactobacillus casei Inhibits HT-29 Colorectal Cancer Cell Proliferation

Many Lactobacillus casei strains are reported to exhibit anti-proliferative effects on colorectal cancer cells; however, the mechanism remains largely unknown. While there has been considerable interest in bacterial small metabolites such as short chain fatty acids, prior reports suggested that larger-sized molecules mediate the anti-proliferative effect of L. casei. Here, other possible ways of communication between gut bacteria and its host are investigated. LevH1 is a protein displayed on the surface of L. casei, and its mucin binding domain is highly conserved. Based on previous reports that the cell-free supernatant fractions decreased colorectal cell proliferation, we cloned the mucin binding domain of the LevH1 protein, expressed and purified this mucin binding protein (MucBP). It has a molecular weight of 10 kDa, is encoded by a 250 bp gene, and is composed primarily of a β-strand, β-turns, and random coils. The amino acid sequence is conserved while the 36th amino acid residue is arginine in L. casei CAUH35 and serine in L. casei IAM1045, LOCK919, 12A, and Zhang. MucBP36R exhibited dose-dependent anti-proliferative effects against HT-29 cells while a mutation of 36S abolished this activity. Predicted structures suggest that this mutation slightly altered the protein structure, thus possibly affecting subsequent communication with HT-29 cells. Our study identified a novel mode of communication between gut bacteria and their host.


Introduction
Gut microbiota plays an increasingly important role in human physiology, and it is generally believed that short chain fatty acids (SCFAs) produced by gut microbiota bridge their function to the host [1][2][3][4][5]. The production of SCFAs in the gut lumen and their hepatic utilization results in a concentration gradient, decreasing from the lumen to the periphery, where selective uptake of butyrate at the epithelium, propionate at the liver, and acetate in the periphery takes place [6][7][8]. It is certainly unlikely that so many physiological outputs are mediated by the limited number of SCFAs alone. However, it is not yet discovered whether there are other patterns of interactions between gut microbiota and the host [9]. Hence, finding other ways of biological intervention between the gut microbiota and its Primers 5 TAGGACAAAGCTACACTTCTG3 and 5 CAGAAGTGTAGCTTTGTCCTA3 were used for site-directed mutagenesis on a PCR instrument with an annealing temperature of 57 • C [33].The plasmid was sequenced, and the verified plasmid was transformed into E. coli BL21 (DE3) cells for protein expression under the induction of isopropyl-β-dthiogalactopyranoside (IPTG, Shanghai Macklin Biochemical Co. Ltd., Shanghai, China). The over-expressed MucBP protein was purified essentially in the same way as described previously [34].

Protein Structure Characterization
The purified 0.3 mg/mL MucBP protein was dialyzed into 10 mM phosphate buffer saline (PBS) buffer, and its circular dichroism (CD) spectrum was recorded on a Chirascan circular dichroism polarimeter (Applied Photophysics, Leatherhead, UK) by scanning the sample from 200 to 260 nm at 25 • C with a bandwidth of 1 nm and step size of 1 nm, as described previously [36]. Proportions of the secondary structures of the protein were calculated with the Chirascan software (Applied Photophysics, UK). The I-TASSER and RoseTTAFold servers were used to predict the three-dimensional structure of MucBP 36S and 36R variants, as described previously [37,38]. The predicted structures were validated with the ProSA-web Protein Structure Analysis server and the SAVES v6.0 server [39][40][41][42].

Cell Culture
The human colon adenocarcinoma cell line HT-29 used in this study was obtained from the National Infrastructure of Cell Line Resource and cultured at 37 • C in a humidified 5% CO 2 incubator using RPMI-1640 medium (Gibco, New York, NY, USA, 11875093) supplemented with 10% fetal bovine serum (FBS, Gibco, 16140089) and 1% penicillinstreptomycin (Sigma, St. Louis, MO, USA, V900929).

Cell Viability Assay
The effect of different concentrations (from 5 µg/mL to 200 µg/mL) of MucBP 36R and MucBP 36S on HT-29 cells viability was assessed using the Enhanced Cell Counting Kit-8 (CCK8; Beyotime Biotechnology, Shanghai, China, C0042) according to the manufacturers' protocol. HT29 cells were treated with MucBP 36R and MucBP 36S at different concentrations (5,25,50,100, and 200 µg/mL) for 72 h and then seeded onto 96-well plates (approximately 3 × 10 3 cells per well containing 100 µL of culture medium) overnight. The supernatant was removed, and 10 µL CCK8 solution with 100 µL fresh medium was added to each well before the plate was incubated at 37 • C for 2 h. The absorbance of each well was measured at 450 nm on a microplate reader (TECAN, infinite F200). The Dunnett's multiple comparisons test was used to compare experimental groups with the control group.

Cloning, Purification, and Sequence Comparison of MucBP from L. casei
The gene sequence of LevH1 in L. casei CAUH35 was identified (GeneBank, AB185852.1), and the amino acid sequence was analyzed with BLAST, which was highly conserved as in L. casei IAM1045 [43]. There are 1294 amino acid residues in the LevH1 protein, and the MucBP domain ranges from amino acid residue 745 to 808. There is a KxYKxGKxW signal peptide in the N-terminal of the LevH1 protein and an LPXTG cell wall anchoring domain in the C-terminal, suggesting that this protein may be secreted and anchored to the L. casei cell wall. Following the signal peptide is the inulinase domain, which is supposed to hydrolyze the prebiotic inulin, and then the predicted mucin binding domain. Compared with all the L. casei strains with reported genomes, including L. casei IAM1045, L. casei LOCK919, L. casei 12A, L. casei Zhang, L. casei CAUH35, and L. casei ATCC334, the core region of MucBP is highly conserved ( Figure 1A). It is noteworthy that the 36th amino acid residue is serine in most of these bacteria, while it is arginine in L. casei CAUH35 and proline in L. casei ATCC334.
primers was designed to amplify the gene. Agarose gel electrophoresis was employed to purify the gene sequence, and a band of around 250 bp was visualized ( Figure 1B), which was consistent with the theoretical 246 bp. Site-directed mutagenesis was used to study the effect of mutation on the 36th amnio acid residue, and the mutation was examined with DNA sequencing.
The MucBP sequence was inserted into the pET-30b plasmid, transformed into E. coli BL21(DE3) cells, induced with IPTG, and MucBP was purified with Ni-NTA agarose chromatography. Finally, the purified protein was examined with SDS-PAGE gel electrophoresis, and a band below 14 kDa was observed, which was consistent with the theoretical molecular weight of MucBP of 10.3 kDa ( Figure 1C). The calculated isoelectric point of this protein is 6.63, and molar extinction coefficient is 12,950 M −1 cm −1 . There are 10 serine and 13 threonine residues in this protein.

Secondary and Tertiary Structures of MucBP from L. casei
The purified MucBP 36R protein was subjected to circular dichroism analysis for determining its secondary structure. There was a minimum peak at around 210 nm, indicating a mixture of mainly β-strands and disordered regions ( Figure 2A). Quantitative analysis Based on the MucBP domain of the LevH1 protein from L. casei CAUH35, a pair of primers was designed to amplify the gene. Agarose gel electrophoresis was employed to purify the gene sequence, and a band of around 250 bp was visualized ( Figure 1B), which was consistent with the theoretical 246 bp. Site-directed mutagenesis was used to study the effect of mutation on the 36th amnio acid residue, and the mutation was examined with DNA sequencing.
The MucBP sequence was inserted into the pET-30b plasmid, transformed into E. coli BL21(DE3) cells, induced with IPTG, and MucBP was purified with Ni-NTA agarose chromatography. Finally, the purified protein was examined with SDS-PAGE gel electrophoresis, and a band below 14 kDa was observed, which was consistent with the theoretical molecular weight of MucBP of 10.3 kDa ( Figure 1C). The calculated isoelectric point of this protein is 6.63, and molar extinction coefficient is 12,950 M −1 cm −1 . There are 10 serine and 13 threonine residues in this protein.

Secondary and Tertiary Structures of MucBP from L. casei
The purified MucBP 36R protein was subjected to circular dichroism analysis for determining its secondary structure. There was a minimum peak at around 210 nm, indicating a mixture of mainly β-strands and disordered regions (Figure 2A). Quantitative analysis indicated there was 29.4% β-strands, 17.7% β-turns, and 41% random coils within this protein.
surface, roughly consistent with the secondary structure analysis.
The MucBP 36S protein structure was also predicted, and compared with the 36R variant ( Figure 2B). There was a slight structural deviation induced by this mutation, as an RMSD value of 1.036 Å was obtained between these two proteins. The side chain of the 36th serine is pointing differently than the arginine side chain, whose orientation highly likely induced the structural difference between these two protein variants ( Figure 2C).

Antiproliferative Activity of MucBP from L. casei
Different proteins were co-incubated with HT-29 cells to evaluate their possible physiological action on this human colorectal cancer cells. The addition of 0, 5, 25, 50, 100, and 200 µg/mL bovine serum album (BSA) protein and co-incubation for 72 h did not show any effect on HT-29 cell proliferation ( Figure 3A). The same co-incubation was performed with MucBP 36R and MucBP 36S proteins. Initially, MucBP 36R did not show any effect on HT-29 cell growth at 5 µg/mL. Surprisingly, MucBP 36R began to inhibit the proliferation of HT- The I-TASSER and RoseTTA fold servers were employed to predict the tertiary structure of MucBP 36R , and a globular protein containing majorly β-strands and random coils was generated ( Figure 2B). The predicted structures were almost identical at the folded region, and were validated with ProSA-web and SAVES v6.0, respectively. Based on the validation results, structures predicted with I-TASSER were subjected to the following analysis, and MucBP 36R displayed positive, negative, and hydrophobic characters on its surface, roughly consistent with the secondary structure analysis.
The MucBP 36S protein structure was also predicted, and compared with the 36R variant ( Figure 2B). There was a slight structural deviation induced by this mutation, as an RMSD value of 1.036 Å was obtained between these two proteins. The side chain of the 36th serine is pointing differently than the arginine side chain, whose orientation highly likely induced the structural difference between these two protein variants ( Figure 2C).

Antiproliferative Activity of MucBP from L. casei
Different proteins were co-incubated with HT-29 cells to evaluate their possible physiological action on this human colorectal cancer cells. The addition of 0, 5, 25, 50, 100, and 200 µg/mL bovine serum album (BSA) protein and co-incubation for 72 h did not show any effect on HT-29 cell proliferation ( Figure 3A). The same co-incubation was performed with MucBP 36R and MucBP 36S proteins. Initially, MucBP 36R did not show any effect on HT-29 cell growth at 5 µg/mL. Surprisingly, MucBP 36R began to inhibit the proliferation of HT-29 cells at 25 µg/mL, and the inhibition effect was further enhanced at higher protein  Figure 3B). On the contrary, MucBP 36S did not exhibit any effect on HT-29 cell proliferation up to a protein concentration of 200 µg/mL ( Figure 3C). 29 cells at 25 µg/mL, and the inhibition effect was further enhanced at higher protein concentrations ( Figure 3B). On the contrary, MucBP 36S did not exhibit any effect on HT-29 cell proliferation up to a protein concentration of 200 µg/mL ( Figure 3C).

Discussion
We turned to the surface layer protein MucBP in L. casei as a potential candidate exerting anti-proliferative activity against colorectal cancer cells. Although our purified protein was smaller than the predicted >50 kDa functioning fraction, it is only a domain of the LevH1 protein, which was more than 10 folds larger. Thus, it is likely that the MucBP protein is virtually the functioning element in the >100 kDa protein fraction found in the previous study [32]. In bacteria with similar mucin binding capacity, anti-proliferative activity could be found. For example, L. paracasei K5 inhibits Caco-2 cell proliferation and displays efficient adherence capacity to this colon cancer cell, similar to the strains L. casei ATCC 393 and L. rhamnosus GG [44]. Our study is the first one, to the best of our knowledge, proving that the cell surface protein MucBP inhibits human colorectal cancer cell proliferation.
Mechanism wise, it was reported that L. casei ATCC 393 exhibited a dose-and timedependent anti-proliferative effect on HT-29 cells by up-regulating TNF-related apoptosis [45]. L. casei BL23 inhibited colorectal cancer growth via the IL-2 signaling pathway [46,47]. For other types of cancer, cell-free culture supernatant of L. casei SR1, SR2, and SR4 isolated from human milk inhibited cervical cancer by up-regulating the expression of apoptotic genes [48]. L. casei Shirota enhanced the antiproliferative effect of geniposide in human oral squamous carcinoma HSC-3 cells [49]. Here, we show the concentration-dependent antiproliferative effect of MucBP against human colorectal cancer HT-29 cells, which demonstrated the cancer cell inhibition activity of this protein, regardless of whether it is on a cell or in a solution. Compared to the BSA group, protein addition alone is not sufficient to induce the inhibiting effect via perturbation of cell culture conditions. This is in line with observations that surface layer proteins from L. casei Zhang, L. rhamnosus, L. gasseri, and L. acidophilus NCFM protected the HT-29 cells from H2O2-induced oxidative injury [50]. It is obvious that surface layer proteins could interact directly with HT-29 cells. Although we did not investigate the detailed mechanism of how MucBP inhibits HT-29 cell proliferation, the phenomenon that a single mutation of arginine into serine at the 36th amino acid undermined the anti-proliferative capacity of MucBP suggested that there might be a receptor on the surface of HT-29 cells that could bind with MucBP and induce colorectal cancer cell death.
Our study does not intend to limit the probiotic effect of L. casei only to the MucBPinhibited colorectal cancer cell proliferation. There are also other bacteria-related, small or big molecules, exhibiting anti-proliferation effects on these cells. For example, exopolysaccharides produced by L. casei SB27 inhibited the proliferation of HT-29 cells, and

Discussion
We turned to the surface layer protein MucBP in L. casei as a potential candidate exerting anti-proliferative activity against colorectal cancer cells. Although our purified protein was smaller than the predicted >50 kDa functioning fraction, it is only a domain of the LevH1 protein, which was more than 10 folds larger. Thus, it is likely that the MucBP protein is virtually the functioning element in the >100 kDa protein fraction found in the previous study [32]. In bacteria with similar mucin binding capacity, anti-proliferative activity could be found. For example, L. paracasei K5 inhibits Caco-2 cell proliferation and displays efficient adherence capacity to this colon cancer cell, similar to the strains L. casei ATCC 393 and L. rhamnosus GG [44]. Our study is the first one, to the best of our knowledge, proving that the cell surface protein MucBP inhibits human colorectal cancer cell proliferation.
Mechanism wise, it was reported that L. casei ATCC 393 exhibited a dose-and time-dependent anti-proliferative effect on HT-29 cells by up-regulating TNF-related apoptosis [45]. L. casei BL23 inhibited colorectal cancer growth via the IL-2 signaling pathway [46,47]. For other types of cancer, cell-free culture supernatant of L. casei SR1, SR2, and SR4 isolated from human milk inhibited cervical cancer by up-regulating the expression of apoptotic genes [48]. L. casei Shirota enhanced the antiproliferative effect of geniposide in human oral squamous carcinoma HSC-3 cells [49]. Here, we show the concentration-dependent antiproliferative effect of MucBP against human colorectal cancer HT-29 cells, which demonstrated the cancer cell inhibition activity of this protein, regardless of whether it is on a cell or in a solution. Compared to the BSA group, protein addition alone is not sufficient to induce the inhibiting effect via perturbation of cell culture conditions. This is in line with observations that surface layer proteins from L. casei Zhang, L. rhamnosus, L. gasseri, and L. acidophilus NCFM protected the HT-29 cells from H 2 O 2 -induced oxidative injury [50]. It is obvious that surface layer proteins could interact directly with HT-29 cells. Although we did not investigate the detailed mechanism of how MucBP inhibits HT-29 cell proliferation, the phenomenon that a single mutation of arginine into serine at the 36th amino acid undermined the anti-proliferative capacity of MucBP suggested that there might be a receptor on the surface of HT-29 cells that could bind with MucBP and induce colorectal cancer cell death.
Our study does not intend to limit the probiotic effect of L. casei only to the MucBPinhibited colorectal cancer cell proliferation. There are also other bacteria-related, small or big molecules, exhibiting anti-proliferation effects on these cells. For example, exopolysaccharides produced by L. casei SB27 inhibited the proliferation of HT-29 cells, and ferrichrome in the supernatant of L. casei ATCC334 culture induced apoptosis in colon cancer cells [51,52]. This evidence suggests that there are also other bacterial components or metabolites involved in the anti-proliferative effects of L. casei. Moreover, it was reported that L. Casei Lcr35 prevented adjuvant 5-FU chemotherapy-induced intestinal mucositis in colorectal cancer-bearing mice [53]. With all these anti-proliferative effects mediated by many molecules from L. casei, this bacterium is of potential significance for the future treatment of cancer. The MucBP protein is also found in Lactobacillus acidophilus, Lactobacillus fermentum, and Lactobacillus fermentum; the role of this protein in these bacteria-host interaction should also be investigated in the future. In all, our study offers an insight of how probiotics interact with their hosts via other ways than microbial metabolites.