The relationship between the structural characteristics of lactobacilli-EPS and its ability to induce apoptosis in colon cancer cells in vitro

Colon cancer is one of the most common cancer around the world. Exopolysaccharides (EPSs) produced by lactobacilli as potential prebiotics have been found to have an anti-tumor effect. In this study, lyophilized EPSs of four Lactobacillus spp. for their impact on apoptosis in colon cancer cells (HT-29) was evaluated using flow cytometry. The relationship between capability of a lactobacilli-EPS to induce apoptosis and their monosaccharide composition, molecular weight (MW), and linkage type was investigated by HPLC, SEC, and NMR, respectively. Changes in apoptotic-markers were examined by qPCR and Western Blotting. EPSs were capable of inhibiting proliferation in a time-dependent manner and induced apoptosis via increasing the expression of Bax, Caspase 3 and 9 while decreasing Bcl-2 and Survivin. All EPSs contained mannose, glucose, and N-acetylglucosamine with different relative proportions. Some contained arabinose or fructose. MW ranged from 102–104Da with two or three fractions. EPS of L. delbrueckii ssp. bulgaricus B3 having the highest amount of mannose and the lowest amount of glucose, showed the highest apoptosis induction. In conclusion, lactobacilli-EPSs inhibit cell proliferation in HT-29 via apoptosis. Results suggest that a relationship exists between the ability of EPS to induce apoptosis and its mannose and glucose composition.

however, Lactobacillus are the most common genera used as probiotics 14 . Studies have demonstrated that bacterial secondary products can also alter the gut environment and affect cancer development 15,16 . In a study evaluating the impact of cell fractions and exopolysaccharides (EPSs) from Lactobacillus on colon cancer cells, Liu et al. reported that EPS has the highest cytotoxic effect among the tested fractions and reduced proliferation of colon cancer cells 17 . Another study investigating the effect of lactobacilli EPS on cervical tumor cells demonstrated that lactobacilli EPS induced apoptosis in tumor cells and showed an anti-proliferative effect 15 . Microbial EPSs are primary or secondary metabolites produced by microorganisms and they have been used as prebiotics, which defined as the specific fermented ingredient resulting in changes in the gastrointestinal microbiota and provide health benefit 18,19 . There is a great diversity among EPS produced by lactic acid bacteria (LAB) 20 . LAB EPSs can be divided into homopolysaccharides (HoPSs) consisting one type of monosaccharide, and heteropolysaccharides (HePSs) consisting a backbone of repeating units that are composed of two or more types of monosaccharides 21,22 . LAB mostly produces HePSs which consist of different sugars such as pentose (D-arabinose, D-ribose, D-xylose), hexose (D-glucose, D-galactose, D-mannose), or uronic acids (D-glucuronic acid, D-galacturonic acid). They mostly consist different types of linkages and branches such as α- (1,2) or α- (1,6) linkages which are rigid, β- (1,4) or β- (1,3) which are less rigid 23,24 . Studies reported that the composition and the structure of EPSs tend to be strain dependent 25,26 . Moreover, it has been shown that structural and compositional diversity among EPSs might be responsible for the variation in their health benefits 20,27 . Li et al. purified three fractions of EPS isolated from L. helveticus MB2-1 and evaluated their structure and antioxidant activities in vitro 28 . They reported that although the molecular weights of EPSs were similar, their sugar compositions and antioxidant effect were different as well as their anti-cancer impact on colon cancer cells 28,29 . Additionally, anti-cancer activity of polysaccharides can be affected by other physicochemical properties, such as presence of β-type glycosidic linkages, uronic acid, sulfate groups, and glucose increasing anti-cancer activity [30][31][32][33] . Therefore, the determination of chemical composition and structure of EPS must be taken into consideration when predicting potential applications of EPSs 30 .
Although lactobacilli EPSs have been reported to have cytotoxic effect on various cancer cell lines, the mechanism of action and the impact of its structure on cytotoxic effect have not been understood yet. In this study, we investigated the EPS produced by various Lactobacillus spp. and their impact on proliferation and apoptosis in colon cancer cells. We performed a chemical and structural characterization of lactobacilli EPSs including molecular weight, monosaccharide composition, and linkage type. Additionally, we evaluated their structural, characteristic effects on apoptosis.
Isolation and lyophilization of exopolysaccharide. The method of Frengova et al. was followed to isolate EPS 36 . The growth culture with an optical density of 0.6 at 600 nm (~8.5 log CFU/ml) was heated at 100 °C for 15 min. After cooling, the cell suspension was treated with 17% (v/v) of 85% trichloracetic acid solution and centrifuged at 15, 493 × g for 20 min to remove cells and proteins. The exopolysaccharide was precipitated using two volume of cold absolute ethanol followed by centrifugation at 15, 493 × g for 15 min. The resulting pellet containing EPS was suspended in deionized water. Total carbohydrate was measured at 490 nm by phenol-sulfuric acid method 37 using glucose as standard. The EPSs isolated were stored at −80 °C until being lyophilized in Christ Alpha 2-4 freeze dryer (Marin Christ Co. FL, USA). The freeze-dried EPS powder was stored at 4 °C 38 .
physico-chemical characterization of the eps produced by Lactobacillus strains. Monosaccharide composition. The method of Ledezma et al. was followed to hydrolyze EPSs isolated from Lactobacillus spp. 39 . Briefly, EPSs (10 mg/ml) were incubated with 1 M H 2 SO 4 for 3 hours at 90 °C and then neutralized with 1 M NaOH to pH 7. After the complete hydrolysis, monosaccharide composition of EPS isolated from Lactobacillus spp. was quantified by high pressure liquid chromatography (HPLC) using an AGILENT 1260 system equipped with a refractive index detector at the Middle East Technical University, Central Laboratory. The separation (25 µl volume of injection) was carried out in the Metacarb 67 C columns (300 mm × 6.5 mm) maintained at 90 °C. For N-acetylglucosamine composition, the separation was carried out in the Metacarb 87 H (300 mm × 7.8 mm) column.
Anti-proliferation activity. Impact of EPSs from lactobacilli on HT-29 cell proliferation was evaluated using a WST-1 cell proliferation assay kit (Cayman Chemical Company, Ann Arbor, Michigan, USA). The lyophilized EPSs were dissolved in distilled water and filtered using a 0.2 µm syringe filter prior to analyses. HT-29 cells were seeded into a 96-well plate at a density of 1 × 10 4 cells/well and treated with EPSs at a final concentration of 400 µg/ml followed by 24 h or 48 h incubation at 37 °C with 95% air and 5% CO 2 . After incubation, 10 μl of the WST-1 mixture was added to each well and the plates were incubated for 2 h at 37 °C with 95% air and 5% CO 2 . Formation of formazan was measured at 450 nm by a microplate reader (Epoch, Biotek, Winooski, VT, USA) and the absorbance was correlated with the cell number. The anti-proliferative effect was evaluated by comparing to viability of the treated samples with the untreated control (ultrapure water and DMEM mix without test sample for EPS). The percentage of viability was calculated as follows: Cell distribution by flow cytometry. Briefly, treated and untreated cells (control cells) were washed twice with PBS and harvested by scraping from 6 well plate using a cell scraper in PBS and collected by centrifugation (367 × g, 4 min). Flow cytometry analysis were performed according to kit manufacturer's directions 42,43 . The cell pellets were resuspended into 1 ml of DMEM. Following that, 4 ml of Annexin V binding buffer was added (Cat. No: BB10X-50ml, Immunostep, Spain) and centrifuged (500 × g, 5 min). After centrifugation, the supernatant was aspirated and the cells were resuspended in 200 µl of 1X BB (BB10X diluted 1X with distilled ultrapure water). 100 µl of the cell suspension was incubated for 30 minutes at room temperature and in the dark by adding 5 µl of Annexin V-FITC (FITC Annexin V, Immunostep, Spain) and propidium iodide (PI) (as in the final concentration of 40 µg/ml, Immunostep, Spain). 1X BB (100 µl) was added into each cell tube and the cells were analyzed using ACEA NovoCyte 3000 Flow cytometer. Data analysis was performed using ACEA NovoExpress software.  Table S1. PCR conditions were as follows; 5 min at 95 °C for initial activation, 10 s at 95 °C for 40 cycles of denaturation and 30 s at 60 °C for combined annealing-extension. All reactions were performed in triplicate and repeated at least 2 times. Cyclophilin A (PPIA) gene was used as an internal control to normalize the target transcripts by the 2 −ΔΔCT method 44 .

Western blot analysis. Total protein extracts from untreated cells or cells treated with EPS at different
time intervals (24 h or 48 h) were subjected to Western blot analysis as described by Huang et al. 45 . HT-29 cells at a density of 1 × 10 6 cells were treated with EPSs at a final concentration of 400 µg/ml and incubated for 24 and 48 h. After treatment, the medium was aspirated and the cell culture plate placed on ice washed twice with ice-cold phosphate buffer saline (PBS). After PBS was drained, the cells were lysed by 250 µl of lysis buffer (NP-40 buffer-150 mM NaCl, 50 mM Tris, pH 8.0, and 1% NP-40) containing protein inhibitor cocktail. The cells scraped from the plate gently transferred into the pre-cooled centrifuge tubes. The tubes were kept on ice for 30 min with constant agitation. The lysates were centrifuged at 13,201 × g for 20 minutes at 4 °C and the supernatant was stored at 4 °C until use 46,47 . The total protein was determined using the Bradford assay (Sigma-Aldrich). 40 μg of protein lysates denatured in loading buffer at 95°C for 10 min was separated by 4-12% Acrylamide-Bisacrylamide gel and then transferred onto PVDF membranes using iBlot ® (ThermoFisher). The membranes were then blocked in blocking solution (Western Breeze, ThermoFisher) at room temperature before incubating with antibodies for one hour. The expression patterns of Bax, Bcl-2, Caspase 3, Caspase 9, Survivin were detected using specific antibodies and β-actin was used as loading control 48 . After washes in Antibody Wash, the membrane was incubated in secondary antibody (anti-rabbit IgG) for 30 min. After second washes, the membrane was incubated in Chromogenic Substrate until the bands develop on the membrane. The molecular weight of the protein bands was determined using BIORAD ImageLab 5.21 program compared to the protein marker (Thermo Scientific PageRuler Prestained Protein Ladder 26616).
www.nature.com/scientificreports www.nature.com/scientificreports/ statistical analysis. All experiments were carried out with three replicates and values were reported as means ± standard deviation (SD), unless otherwise indicated. Statistical analysis was performed using SPSS 16.0 (SPSS Inc., Chicago, IL, USA). Statistical difference was assessed with one-way analysis of variance (ANOVA) followed by Tukey test. For Western blot analysis, t-Test (Excel 2007) was performed. Additionally, post hoc Dunnett's test for pair-wise comparison was ran to analyze flow cytometer data. Statistical difference was determined at a P value of 0.05 or less. Monosaccharide composition of EPSs produced by Lactobacillus spp. were used to generate a dendrogram by the Ward method of hierarchical clustering (JMP version 12, SAS Institute Inc., Cary, NC).

Results and Discussion
production of eps by Lactobacillus spp. in culture medium and their sugar composition analysis. The bidirectional interaction between host and probiotic bacteria results in health benefits to the host 49 .
Several factors in lactobacilli have been shown to impact human health in vitro and in vivo, including cell surface components and metabolites 49 . EPSs produced by LAB are one of the important component that have a key role in probiotic activity including anti-proliferative effect, immunomodulation, and adhesion 13,17,50 . Here we investigated the production of EPSs by four Lactobacillus strains (L. plantarum GD2, L. rhamnosus E9, L. brevis LB63, and L. delbrueckii ssp. bulgaricus B3) and their compositional analysis. EPSs were freeze-dried and lyophilized. EPS production by various lactobacilli strains differed significantly from each other (p < 0.05). L. plantarum GD2, L. rhamnosus E9, L. brevis LB63, and L. delbrueckii ssp. bulgaricus B3 yielded 397 ± 4, 298 ± 5, 347 ± 4, and 449 ± 4 mg/l, respectively. L. delbrueckii ssp. bulgaricus B3 produced the highest amount of EPS among other lactobacilli. L. delbrueckii ssp. bulgaricus B3 was isolated from yogurt while the others were isolated from feces samples. Mozzi et al. reported that most of the HePS are produced by Lactobacillus from food origin as seen in our results 51 . Studies on EPS production by Lactobacillus to date reported varying amount of EPS. Sungur et al. stated that two of the L. gasseri strains produced 242 ± 3 mg/l and 255 ± 4 mg/l of lyophilized EPS 15 . In another study, EPS production by L. crispatus in different carbon sources ranged from 200 to 400 mg/l 52 . Van Geel-Schutten et al. examined the production of EPSs by a total of 182 Lactobacillus strains in MRS medium with relatively high sugar concentration 53 . Only 60 of them produced EPS and from those strains only 10% produced EPS more than 100 mg/l which is referred to as the large amount of production. The lactobacilli strains studied here produced in a decent amount of EPS relative to the lactobacilli EPSs studied in the literature 15,52,53 .
EPS produced by LAB exhibit a large variation in their sugar composition 54 . Here, we examined monosaccharide composition of EPSs produced by various Lactobacillus spp., using HPLC. Different types of sugar unite were found in the lactobacilli EPSs studied. The EPSs were mainly composed of mannose and glucose ranging between 71-88% and 10-28%, respectively. Additionally, fructose, arabinose, sucrose + maltose, and N-acetylglucosamine were detected at different ratios ( Table 1). Detection of maltose and sucrose, disaccharide molecules, might be due to incomplete hydrolysation of EPS. A review on functional properties of EPSs derived from yeast reported that EPSs consisting mannose with more than 50% are classified as biologically active 55 . Anti-cancer activity of EPSs has thought to be related with high amount of mannose in sugar composition [56][57][58] . Shao et al. showed that polysaccharides consisting of glucose and mannose can interact with Toll-like receptors and activate host immunity 56 . Vidhyalakshmi and Vallinachiyar reported that macrophages carry mannose and glucose specific receptors which are important in triggering anti-cancer activities and suppressing cell proliferation in tumor 57 . Similarly, Jin et al. studied anti-tumor activity of EPS and concluded that anti-tumor activity of polysaccharides tends to correlate with mannose presence as major monomer in the sugar composition 59 .
The presence of different monomers among the EPSs suggest that EPSs of lactobacilli strains studied here are heteropolysacharides as identified mostly in other LAB 54 . Additionally, cluster analysis of sugar composition in EPSs of Lactobacilli spp. revealed that the L. delbrueckii ssp. bulgaricus B3 clustered by itself separated from the other strains (Table 1). Suggesting that this strain tends to have different biological functions compared to the other strains. Based on the variations in the ratio of monomers among various Lactobacillus, there is a compositional diversity of EPSs and this diversity likely contributes strain to strain variation in their biological functions including proliferation inhibition and apoptosis induction. structural analysis of epss produced by Lactobacillus spp. It has been shown that an impressive differences exist in molecular weights and structure of EPSs among LAB 50,60 . To understand better the diversity among Lactobacillus spp. and compare the structure of EPSs produced we also performed a molecular weight www.nature.com/scientificreports www.nature.com/scientificreports/ analysis by SEC and a structural analysis by NMR. SEC results showed that molecular weight of the EPSs ranged from 10 2 to 10 4 Da consisting of two fractions except LB63_EPS (Table 2). Interestingly, LB63_EPS has three fractions. Similarly, Tallon et al. isolated EPS from L. plantarum EP56 that has two fractions with molecular mass of 8.5 × 10 5 and 4 × 10 4 Da 61 . However, a study on structure analysis of EPS isolated from another strain of L. plantarum (YW32) reported that L. plantarum EPS has only one fraction with molecular weight of 1.03 × 10 5 Da 62 . Hamet et al. examined the EPSs of 28 different Lactobacillus and showed that the fraction number ranged from one to three with molecular weight distribution being strain dependent 63 . Biofunctinality of EPS has been shown to be affected by molecular weight. Xu et al. reported that EPS from Bifidobacterium animalis RH has a stronger atioxidant activity due to its low molecular weight 64 . Based on literature, EPS ≤10 4 Da is considered as low molecular weight [64][65][66] . It has been reported that low molecular weight polysaccharides can easily pass through the host cell membrane barriers and exhibit biological activity better 30 . On the contrary, in other studies investigating chemical composition and anti-tumor activity of polysaccharides, high molecular weight polysaccharides tend to have more anti-tumor impact than those of low molecular weight 67,68 .
Furthermore, NMR chemical shifts were determined as described in literature 41,69 by performing 1 H-NMR, COSY and NOESY NMR analysis. Generally, the 1 H NMR spectrum of a polysaccharide can be divided into three main regions: the anomeric region (δ 4.5-5.5), the ring proton region (δ 3.1-4.5) and the alkyl region (δ 1.2-2.3) 70 . In the present study, the chemical shift of the anomeric H1 protons of the EPS was observed at δ 4.8, δ 4.9 and δ 5.2 ppm. 1 H chemical shifts of the EPS from Lactobacillus spp. is shown in Table 3.
Similar H-H interactions was observed in all NOESY spectra. The released peaks were observed as a result of interaction between H1 proton of α-D-mannose (5.2 ppm) and H2 proton of α-D-glucose (4.0 ppm protons). Furthermore, the H3 proton of α-D-mannose and H1 proton of β-D-mannose were found to be in interaction. These results helped us to understand the sequencing and binding stereochemistry of monosaccharide units of EPSs of Lactobacillus spp. The results showed that all of the lactobacilli studied here produced EPS with the same type of sugar linkage which are β-H1-H3 (β-D-Mannose-α-D-Mannose) /α-H1-H2 (α-D-Mannose-α-D-Glucose), designated as β-1,3 (β-D-Mannose-α-D-Mannose) and α-1,2 (α-D-Mannose-α-D-Glucose), respectively. The primary structure of EPSs is shown in Fig. 1.
Two types of linkage were found among Lactobacillus spp. Main sugar linkage types present in the EPS fractions isolated from Lactobacillus spp. are β-(1,3) and α-(1,2) ( Table 2). Compared to β-linkages, α-linkages result in more flexible polymers 22 . It has been shown that molecular properties including type of linkages of the polysaccharides strongly impact the interactions with proteins 71 . Biological activities such as anti-tumor and apoptosis inducing activity by polysaccharides are strongly associated with their structures. It has been reported that anti-tumor polysaccharides mainly show β-1,3-linkages and polysaccharides containing mainly β-1,6linkages have less anti-tumor activities 71,72 . Another study mainly focused on the structure of β-glucan demonstrated that (1,3)-β-glucan with the (1,6)-β-glucan branches increased immuno competent cell activity and have an important role in anti-tumor activity of the polysaccharides 67 . However, anti-tumor activity of EPS with different linkage have also been reported such as EPS from L.plantarum with β-D-(1-4), β-D-(1-6)-linked glucose residues 73 . As a result, similar linkage structure was determined in all EPSs used in this study. Variations in the ratio of the monomers and molecular weight among Lactobacillus spp. strains suggest that the compositional diversity of EPSs isolated from Lactobacillus spp. likely contributes strain to strain variation in their ability to inhibit proliferation and induce apoptosis. To better understand the contribution of this variation, we further analyzed the impact of EPSs of Lactobacillus spp. on colon cancer cells, HT-29.   www.nature.com/scientificreports www.nature.com/scientificreports/

Anti-proliferative impact of epss produced from Lactobacillus spp. strains on HT 29 cells.
Studies have suggested that LAB products including EPS have anti-tumor activity 21,41,74 . Researches on the impact of EPS on therapeutic functions including anti-tumor activity and immunomodulation brings new expectations to biomedical fields 18,73,75 . Here, we examined the cytotoxic effect of EPSs from Lactobacillus spp. on HT-29 cells at two time points, 24 h and 48 h by WST-1 assay. EPSs showed an anti-proliferative effect on HT-29 cells in a time dependent manner and differed significantly (p < 0.05) from the control (Fig. 2). The cell death in the cells exposed to GD2_EPS, E9_EPS, LB63_EPS, and B3_EPS was higher at 48 h time point than at 24 h time point, with 80.7 ± 1.8%, 71 ± 1.6%, 78.7 ± 1.9%, and 75.3 ± 1.7% cell viability, respectively. Wang et al. tested the inhibitory effect of EPS from L. plantarum strain against HT-29 cells for two time intervals, 24 h and 72 h. While they barely saw an impact at 24 h time point, the strongest anti-proliferative impact was seen after 72 h 62 . However, it is not known that if the cell death happened due to necrosis or apoptosis. As mentioned previously, type of the linkage in lactobacilli EPS may correlate with the anti-tumor activity. β-1,3-linkage has been shown to have a better anti-tumor activity 71,72 . All of the strains we studied here contain β-(1,3), α-(1,2) as the main linkage. The capability of the Lactobacillus to inhibit proliferation in HT-29 cells are likely to be related with their structure. In addition to the chain linkage, anti-proliferative effect of polysaccharides has been demonstrated to be related to their chemical characteristics including molecular weight and molecular composition 28 . Here in this study we showed that Lactobacillus varied in molecular weight and sugar composition of their EPSs. This variation might have impact on their different anti-proliferative effect. Overall, these results suggest that all of the EPSs from the lactobacilli strains studied here are capable of inhibiting proliferation of HT-29 cells in a time dependent manner.

Impact of EPS produced from Lactobacillus spp. strains on apoptosis in HT-29 cells.
Apoptosis is a programmed cell death and is crucial in development and tissue homeostasis 76    www.nature.com/scientificreports www.nature.com/scientificreports/ spp. induced apoptosis in HT-29 cells at both time points with an apoptosis percentage ranging from 28 to 43 at 24 h and 33 to 41 at 48 h. (Figs 3B, 4B). EPSs from L. delbrueckii ssp. bulgaricus B3 showed the highest apoptosis percentage (42.9 ± 2.4% and 40.6 ± 3.3, respectively) at both time points. Distribution of viable, early apoptotic, late apoptotic and necrotic cells showed that EPSs of Lactobacillus spp. induced both early and late apoptosis in HT-29 cells with the early apoptosis being higher (Figs 3C, 4C). Our results suggest that anti-proliferative impact of Lactobacillus studied here could be due to their capability to induce apoptosis (Figs 2-4). This is important in developing a new anti-cancer drug with good efficacy leading the cancer cells to apoptosis.

Genes and proteins involved in apoptotic pathway induced by epss of Lactobacillus spp.
To further understand the underlying mechanism of EPS-induced apoptosis, we next investigated changes in gene expression of apoptosis markers in HT-29 cells treated with EPSs of Lactobacillus spp. We targeted five genes associated with the apoptotic pathway, which are Bax, Bcl-2, Caspase 3, Caspase 9, and Survivin and measured both relative gene expression and protein expression.
Bcl-2 is a large family of proteins that regulate cell death and cell survival. Bcl-2 family proteins take an important role in mitochondria-mediated apoptotic pathway and control the integrity of the mitochondrial outer membrane 79 . In this study, two of Bcl-2 family proteins were examined, Bcl-2 which is an anti-apoptotic protein and Bax which is a pro-apoptotic protein. Effect of EPSs from Lactobacillus spp. on both anti-apoptotic and pro-apoptotic Bcl-2 family proteins were analyzed. The gene expression results showed that Lactobacillus spp. significantly increased the Bax gene expression at two time points, relative to the control, non-treated HT-29 cells (Fig. 5). However; a significant decrease in expression of Bcl-2 gene was observed in HT-29 cells treated with EPSs of Lactobacillus spp. for 24 h or 48 h, relative to the control (Fig. 5). The induction of Bax gene was higher at 24 h time point compared to 48 h time point. The highest increase of Bax gene expression at 24 h time point was observed in the HT-29 cells treated with EPS of L. rhamnosus E9, which was 6.79 ± 0.12 -fold change relative to the control (Fig. 5). The highest increase of Bax gene expression at 48 h time point was observed in the HT-29 cells treated with EPS of L. delbrueckii ssp. bulgaricus B3, which was 4.3 ± 0.01-fold change relative to the control (Fig. 5). Once apoptosis is induced by any agent, Bax proteins reach the mitochondrial outer membrane from cytoplasm. At the same time, anti-apoptotic Bcl-2 family proteins such as Bcl-2 binds Bax to prevent their transfer to the mitochondrial membrane. Therefore, in a cell underwent apoptosis, an increase in Bax gene and a decrease in Bcl-2 gene is expected. Bax/Bcl-2 ratio, hence, could be used to determine the fate of the cells in the apoptotic system 80 . Protein expression data confirmed these results (Figs 6 and S1).
Most of the molecules taking role in cell death are controlled by caspases. Apoptotic caspases can be grouped in two categories; caspases that initiate apoptosis are Caspase 2, 8, 9, and 10; and effector caspases with Caspase 3, 6, and 7 81  Protein expression results also confirmed the gene expression data (Figs 6 and S1).
Studies have shown that polysaccharides induce apoptosis in cancer cells in a time dependent manner 32,45 . In this study, administration of lactobacilli EPSs to HT-29 cells resulted in a time dependent induction in expression of caspases with a higher impact at 48 h compared to the impact at 24 h. The results indicate that EPSs induce apoptosis by caspase activation. Mitochondrial depolarization was dependent on caspase activation, suggesting a positive amplification loop for mitochondrial dysfunction. Loss of mitochondrial membrane potential would lead to release of cytochrome C and activation of Bax and Caspase 3/9 82 .
Survivin is a member of apoptosis inhibitor family 76 and inhibits Caspase 3 and 9. The expression of Survivin gene was suppressed by the EPS of L. rhamnosus E9 (only 0.14 ± 0.01 fold) at 24 h (Fig. 5). However, all of the EPSs resulted in a significant suppression (p < 0.05) in the gene expression of Survivin at 48 h (Fig. 5). This suggest that suppression of Survivin just started at 24 h and reached a high suppression level at 48 h. However, the protein suppression of Survivin has started earlier than the suppression of the gene expression did (Figs 6 and S1). Similarly, Stolfi et al. showed that reduction of Survivin protein was seen as early as 8 h following the application of an apoptosis inducing agent whereas the inhibition of Survivin gene expression occurred at later time points (i.e., 32 hours) 83 suggesting that a posttranscriptional control of Survivin could be involved in this process. EPS from Aphanothece halaphytica has been shown to induce apoptosis in cancer cells by modulating p53-survivin pathway and target unfolded Protein Response Regulator Grp78 84 . The EPS of A. halaphytica induce the expression of CHOP and suppress the expression of Survivin, which leads p53-survivin pathway and cause apoptosis by activating Caspase 3.
The gene and protein expression results demonstrated that the ability to induce apoptosis by EPSs of Lactobacillus spp. associated with an upregulation of Bax, Caspase 3, and 9 and a downregulation of Bcl-2 and Survivin. In vitro studies have suggested that colon cancer could be inhibited by activation of Caspase 3 and 9 and inhibition of Bcl-2 [85][86][87] . In another study, EPS isolated from A. halophytica induce apoptosis similar to our study. They reported that endoplasmic reticulum (ER) signaling pathway leads apoptosis. Based on their explanation for the molecular mechanism, ER targets Grp78 regulating cell respond UPR and suppresses Survivin and Bcl-2 expression while induces Caspase 3 expression, as a results, leads the cells to apoptosis 84 . Our results suggest that EPSs released by Lactobacillus spp. studied here inhibit proliferation via apoptosis in HT-29 cells. It has been reported that biofunctional activities of polysaccharides including anti-tumor and apoptosis inducing activity are strongly associated with their molecular weight, monomer composition, structure, and linkage type 67,72,73,88 .
www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ www.nature.com/scientificreports www.nature.com/scientificreports/ The capability of EPSs from Lactobacillus spp. to induce apoptosis at a high level, relative to the EPSs inducing apoptosis in the literature, might be associated with the main monomer in their structure, mannose, and their linkage type which are β-(1,3), α-(1,2) 75,88-90 . Glucose and mannose are known to have highly specific receptors on macrophages, which is important in tumor immunology. The ability of all EPSs studied here to induce apoptosis in colon cancer cells could be related to their sugar composition having mannose and glucose as major component. B3_EPS, which has the highest amount of mannose in sugar composition, showed the highest apoptosis induction on HT-29 cells. Suggesting that there might be a relationship exists between the ability of an EPS to induce apoptosis and its mannose composition. Additionally, B3_EPS has the lowest amount of glucose relative to the other strains. Having low amount of glucose with a high mannose content might have a role in the ability of B3_EPS to induce apoptosis strongly. www.nature.com/scientificreports www.nature.com/scientificreports/

Conclusion
EPSs produced by LAB are one of the important component that have a key role in probiotic activity including anti-tumor effect, immunomodulation, and adhesion 13,17,50 . In this study we evaluated EPSs of four previously described Lactobacillus spp. isolated from healthy infant feces (GD2_EPS, E9_EPS, and LB63_EPS) and yogurt (B3_EPS) for their health effect on colon cancer cells (HT-29) and for their physicochemical properties 34,35 . We demonstrated that a compositional and structural diversity exists within Lactobacillus spp., and this diversity www.nature.com/scientificreports www.nature.com/scientificreports/ likely contributes to variation in the ability to inhibit proliferation and induce apoptosis. Relative proportions of the individual sugars among Lactobacillus spp. are different and mannose of which portion size impact biological activities 55 , is the major sugar component in EPSs. All of the EPSs contain β-1,3-linkage which has been found in anti-tumor polysaccharides 71,72 . The results showed that EPSs of Lactobacillus spp. inhibit proliferation in colon cancer cells via apoptosis. The level of their capability to induce apoptosis was time dependent. While B3_EPS and GD2_EPS have better impact on inducing Caspases, E9_EPS and B3_EPS showed better impact on Bax and Survivin modulation. B3_EPS showed the highest apoptosis induction on HT-29 cells and has the highest amount of mannose in sugar composition with a quite low amount of glucose. There might be a relationship exists between the ability of an EPS to induce apoptosis and its high mannose and low glucose composition. Collectively these findings are important for further evaluating Lactobacillus spp. EPSs for cancer therapy depending on EPS structure and monomer composition. Particularly mannose ratio in EPS composition should be taken into consideration when designing anti-cancer agents. Further research is required to elucidate the mechanism of action of mannose and glucose on anti-cancer functionality.

Data Availability
All data generated or analysed during this study are included in this published article (and its Supplementary  Information files).