Identification of Lactobacillus Fermentum Strains with Potential against Colorectal Cancer by Characterizing Short Chain Fatty Acids Production , AntiProliferative Activity and Survival in an Intestinal Fluid : In Vitro Analysis

The diagnosis and primary prevention strategies employed for colorectal cancer (CRC) have shown this disease to be a common public health problem especially in developing countries [1,2]. CRC accounts for 8.0 9.7% of all cancer cases and cancer-related deaths [3] and is considered not only a common type of cancer but also a complex and multifactorial disease [4,5] Despite the appreciable understanding of the disease’s pathogenesis, as the environment is considered to play a vital role in its progression, the identification of reliable markers for primary preventive measures for CRC is still deficient [6]. Nevertheless, reports have shown that CRC incidence was reduced to a large extent (up to 80%) by a healthy lifestyle and environmental factors, with diet being a major controlling factor [7]. Dietary interventions have recently attracted increased attention from researchers and clinicians for the prevention and management of CRC [8]. Within this domain of dietary supplements, probiotics have emerged as attractive biotherapeutic agents with nutritional and health benefits. Probiotics, comprised of live microbial food supplements capable of beneficially affecting the gut microbiome, have long been known to augment a variety of immunological and metabolic parameters through diverse mechanisms [8]. A prominent class of probiotics, found to confer health-promoting attributes to the host are lactic acid-producing microorganisms. The Lactobacillus spp. is commonly found in fermented foods as well as in the gastrointestinal (GI) ecosystem. Several probiotic formulations containing L. fermentum, typically those surviving in both GI [9,10] and genital environments [11], were found to reduce infection [12] and overgrowth of harmful bacteria [13]. Also, they retained their beneficial metabolic activities when exposed to intestinal conditions, suggesting their potential for targeted colon delivery and increased colon bioproduction of anti-carcinogenic compounds [14]. L. fermentum have also shown to attribute potential beneficial GI health including anti-inflammatory [15,16] and anti-tumorigenic [17,18] activities. Some L. fermentum strains have shown greater or comparable effects than other probiotic bacteria, such as L. reuteri [19], Bifidobacterium longum [20] and L. plantrum [21].


Introduction
The diagnosis and primary prevention strategies employed for colorectal cancer (CRC) have shown this disease to be a common public health problem especially in developing countries [1,2]. CRC accounts for 8.0 -9.7% of all cancer cases and cancer-related deaths [3] and is considered not only a common type of cancer but also a complex and multifactorial disease [4,5] Despite the appreciable understanding of the disease's pathogenesis, as the environment is considered to play a vital role in its progression, the identification of reliable markers for primary preventive measures for CRC is still deficient [6]. Nevertheless, reports have shown that CRC incidence was reduced to a large extent (up to 80%) by a healthy lifestyle and environmental factors, with diet being a major controlling factor [7]. Dietary interventions have recently attracted increased attention from researchers and clinicians for the prevention and management of CRC [8]. Within this domain of dietary supplements, probiotics have emerged as attractive biotherapeutic agents with nutritional and health benefits. Probiotics, comprised of live microbial food supplements capable of beneficially affecting the gut microbiome, have long been known to augment a variety of immunological and metabolic parameters through diverse mechanisms [8]. A prominent class of probiotics, found to confer health-promoting attributes to the host are lactic acid-producing microorganisms. The Lactobacillus spp. is commonly found in fermented foods as well as in the gastrointestinal (GI) ecosystem. Several probiotic formulations containing L. fermentum, typically those surviving in both GI [9,10] and genital environments [11], were found to reduce infection [12] and overgrowth of harmful bacteria [13]. Also, they retained their beneficial metabolic activities when exposed to intestinal conditions, suggesting their potential for targeted colon delivery and increased colon bioproduction of anti-carcinogenic compounds [14]. L. fermentum have also shown to attribute potential beneficial GI health including anti-inflammatory [15,16] and anti-tumorigenic [17,18] activities. Some L. fermentum strains have shown greater or comparable effects than other probiotic bacteria, such as L. reuteri [19], Bifidobacterium longum [20] and L. plantrum [21].
Several bacterial products were found responsible for the mechanisms associated with these appreciable effects. Among them, short chain fatty acids (SCFAs) produced by the gut microflora are known for their ability to induce cancer cell death and provide a source of energy for colonocytes [22]. The SCFAs resulting from the microbial metabolism of non-digestible carbohydrates in the gut, play a central role in the intestinal homeostasis [23]. They also have shown certain effects, such as; anti-cancer cell-apoptotic effect, promotion of cancer cell cycle arrest, inhibition of cancer cell invasion, and inflammation in the colon [24]. A recent in vitro study showed the adherence property of L. fermentum to cancer cells and the associated anti-proliferative effect through the bioproduction of SCFAs [25]. However, comparative studies investigating the anti-proliferative effect of these bacteria in vitro against CRC cells and their activity in intestinal conditions are infrequent or inconclusive [14,26,27]. Thus, the current study screened a number of L. fermentum bacterial strains (NCIMB -5221, -2797, and -8829) in order to evaluate their biotherapeutic potential against CRC. These strains were previously investigated for the production of certain anti-inflammatory acids [28], cholesterol assimilation [14] in relation to targeted colon delivery [29], and for use in metabolic syndrome (MS) [30]. The aim of this study is to provide insight into SCFA production and anti-proliferative effects against colon cancer cells as well as the bacterial stability in intestinal conditions for L. fermentum bacteria NCIMB -5221, -2797, and -8829.

Materials
Cell culture media including Dulbecco's modified Eagle's medium (DMEM), Eagle's Minimum Essential Medium (EMEM), fetal bovine serum (FBS), and phosphate-buffered saline (PBS) were purchased from Invitrogen. Bacterial culture broth De Man Rogosa Sharpe (MRS) and agar used for plating and growth were obtained from Fisher Scientific (Ottawa, ON, Canada). Water was purified with two systems from Barnstead (Dubuque, IA, USA): an EasyPure reverse osmosis system then a NanoPure Diamond Life Science (UV/UF) ultrapure water system. Reagents and acids such as propionate, acetate, and butyrate, and sodium L-Lactate, were obtained from Sigma (St. Louis, MO, USA).

Bacterial cultures
L. fermentum NCIMB -5221, -8829, and -2797 were obtained from the National Collection of Industrial and Marine Bacteria (NCIMB, Aberdeen, Scotland, UK). L. acidophilus ATCC 314 was purchased from Cederlane Laboratories (Burlington, ON, Canada). To maintain the bacterial cultures, they were inoculated daily in new MRS broth at 1% (v/v). Growth and viability of bacterial cells were determined at OD 620nm (Perkin Elmer 1420 Multilabel Counter, USA) and colony counting using agar plates.

Mammalian cultures
Caco-2 human epithelial CRC adenocarcinoma cell line was purchased from American Type Culture Collection (ATCC, Manassas, VA, USA). Cells were maintained in EMEM + 20% FBS and incubated in a CO 2 incubator (37°C, 5% CO 2 ) for up to two weeks for full differentiation. Caco-2 colon cancer cells were left to attach for up to 24 h to reach a confluence of 50-60% in 96-well plates in DMEM + 10% FBS (37°C, 5% CO 2 ), before experiments. During assays, cell culture medium was substituted by probiotic conditioned medium (CM) mixed with serum and antibiotic-free media (DMEM + 10% FBS).

Preparation of probiotic treatments
For the probiotic treatment used on colon cancer (Caco-2) cells, a conditioned cell culture medium (CM) was prepared according to Grabbing et al. [24] and Kim et al. [25] with slight modifications. Bacterial cultures of L. fermentum and L. acidophilus were passaged for 72 h (37°C, 5% CO 2 ) to reach a late exponential phase (~16 h). The bacterial cells were collected from the culture broth by centrifugation (1000 × g, 15 min, 4°C) and washed with PBS. This bacterial pellet (10 7 -10 9 cru/mL) was incubated in DMEM for 2 hours (37°C, 5% CO 2 ). The medium was also centrifuged (1000 × g, 15 min, 4°C) to remove the bacteria, then sterile-filtered (0.2 µm-pore-size filter, Millipore). The pH was adjusted to 7 using 2 M NaOH and 2 M HCl. Before use, the CM of each bacterium was diluted twice with DMEM.

Bioactivity of L. fermentum bacteria
It was necessary to determine if L. fermentum bacteria were metabolically active in CM or SIF. Since all bacteria are lactic acid bacteria, the concentrations of lactic acid, potentially produced by bacterial cells, were separated and measured by HPLC method, adapted from Dubey and Mistry (1996) [32,33] (described below in detail).

Analysis of lactic acid and SCFAs
Lactic acid and SCFAs were separated using a slightly modified HPLC method [32,33]. The HPLC system used (Model 1050 UV, Hewlett-Packard HP1050 series, Agilent Technologies, USA) was equipped with a UV-vis detector and diode array detector (DAD, 210 ± 5 nm). The column used was a prepacked Rezex ROA -organic acid H+ (8%) column (150 mm x 7.80 mm, Phenomenex, Torrance, CA, USA) attached to an ion-exclusion microguard refill cartridge and heated to 35°C. Data were obtained using ChemStation equipped with LC3D software (Rev A.03.02, Agilent Technologies, CO, USA). The mobile phases (0.05 M H 2 SO 4 and 2% of acetonitrile) were pumped at an isocratic gradient with a 0.7 -0.8 mL/min flow rate. A 100 μl of sample was injected through an autosampler. Lactic, acetic, propionic, and butyric acids were used to prepare standard solutions at concentrations of 1, 10, 100, 500, and 1000 ppm. The concentrations of samples were calculated using the linear regression equations (R 2 ≥ 0.99) from each standard curve.

Cancer cell proliferation assay
The growth of colon cancer cells was determined using an ATP bioluminescence-based assay (CellTiter-Glo® Luminescent Cell Viability Assay, Promega). Caco-2 cells were seeded at 5 × 10 3 cells/ well onto 96-well culture plates and left to attach for 24 -48 h for the formation of an epithelium-like monolayer (37°C, 5% CO 2 ). Caco-2 cells were incubated with the probiotic treatments for 24, 48 and 72 h, (37°C, 5% CO 2 , pH 7). Cell growth inhibition and viability were determined according to the manufacturer's protocol [34]. After incubation, the plate was equilibrated at RT (30 min) and the media was replaced with 100 µL of luminescent reagent and 100 µL of DMEM. The plate was agitated on an orbital shaker (200 rpm, 3 min), followed by incubation at RT for 10 min. Signals were recorded using a multilabel microplate reader (Perkin Elmer, Victor 3, MA, USA).

Determination of bacterial stability in SIF
Each bacterial culture in MRS broth passaged for 72 h was used to inoculate 15 ml of SIF at 3% (v/v), sealed and incubated microanaerobically. At 0, 4, 8, 12, 16, and 24 h, samples were taken to determine the density (OD 620 nm ) and viable bacterial cell count in SIF. The bacterial supernatant was collected by centrifugation (1000 × g, 30 min, 4°C), using 5 ml of bacterial culture, filtered (0.22 μm sterile filters), then stored at -80°C until use.

Relevance of SCFAs produced by L. fermentum strains
To determine whether the concentrations of SCFAs present within the bacterial cell-free extract were the active factors behind suppressing CRC cell growth, the anti-proliferative effect of SCFAs alone was determined. First, lactic, acetic, propionic, and butyric acids produced by each L. fermentum strain were quantified in CM. Mixtures containing the same composition were formulated in DMEM, then added to the colon cancer cells (37°C, 5% CO 2 , pH 7, 72 h). Cell viability was determined using an ATP bioluminescence assay, as described above.

Statistical analysis
Results were presented as means ± standard error of the mean (SEM). Statistical significance was calculated using one-way analysis of variances (ANOVA) with the Tukey's comparison test and Student's t-test. Pearson's correlation method was followed to determine correlation between variables. SPSS statistics software package (version 20.0, IBM Corporation, NY, USA) was used. P-values of p < 0.05 were considered significant.

L. fermentum bacteria produce lactate in the conditioned medium (CM)
Before using the CM of L. fermentum bacteria as a probiotic treatment in vitro, the activity of the bacterial cells incubated in the CM was established by quantifying the level of lactic acid produced. All bacterial strains were active in CM and produced variable amounts of lactic acid ( Figure 1). Data showed that L. fermentum NCIMB 5221 (455.3 ± 9.3 mg/L, p < 0.001) produced the highest amounts of lactic acid when compared with L. fermentum NCIMB -2979 and -8829. All L. fermentum strains produced significantly less lactic acid than L. acidophilus ATCC 314 (1947.7 ± 23.3, p < 0.0001).

The inhibition of colon cancer cells correlates with SCFAs production
To relate the action of L. fermentum bacteria in suppressing CRC cell growth with respect to the production of SCFAs, a correlation analysis was conducted ( Figure 4). Regression analysis showed that the suppression of colon cancer cell proliferation by L. fermentum-CM significantly correlated with the levels of total SCFAs produced by the bacteria in the CM (r = 0.87, p < 0.001, Figure 4d). Cancer cell inhibition correlated with the production of butyric (r = 0.89, p < 0.001) and acetic

The action of probiotic SCFAs is strain-dependent
Establishing a correlation between L. fermentum bacteria SCFA production and their anti-proliferative effect against CRC cells is not sufficient to demonstrate that the inhibition of CRC cell growth is due to SCFAs. Therefore, an additional approach was taken using synthetic SCFAs.
After addition of lactic acid to each formulation, the inhibitory effect of "SSF+LA" was up to 50%, lower than either L. fermentum-CM or SSFs (p < 0.001, Figure 5b), indicating a loss of SCFA efficacy against cancer cells.

L. fermentum strains produced SCFAs in SIF
Despite the decrease in the viability of L. fermentum bacteria in SIF, the bacteria were still able to produce an anti-colon-cancerproliferative effect in a simulated intestinal fluid environment. To

Efficacy of the levels of SCFAs produced in SIF
To verify that L. fermentum bacteria could produce an antiproliferative activity against colon cancer in an intestinal environment, the same concentrations of bacterial SCFAs as produced in the SIF were tested on CRC cells. SCFA synthetic formulations corresponding to the levels of SCFAs produced by the L. fermentum (NCIMB -2797 and -5221) in SIF (SSF-SIF-f) were reconstituted. Additionally, separate concentrations of propionic and acetic acids at the same levels as produced in SIF were tested.

Discussion
CRC is a leading cause of death and an economic burden with a therapeutic market worth billions of dollars worldwide [35]. However, thanks to the preventive potential of this disease [36] it was found that a lifestyle and dietary measures, supplemented with digestive enzymes and probiotics, can substantially decrease CRC incidence [37]. It is  proposed, that increasing the rate of SCFA production through higher gut bacterial carbohydrate fermentation is essential for the maintenance of a healthy colon, with reduction of intestinal injuries, and abnormal cell growth in the lining of the intestines. However, a limited number of probiotic bacteria have been investigated as novel candidates against CRC [38]. This study investigated three L. fermentum strains that have demonstrated antioxidant and anti-inflammatory potential by the production of ferulic acid [39,40]. L. fermentum NCIMB -2797, -8829 and -5221 were investigated for anti-cancer-associated features, such as the production of SCFAs and anti-colon-cancer-cell-proliferative effects in vitro. For this, the cell culture conditioned medium (CM) of each bacterium was used as a probiotic extract treatment for the in vitro study. The metabolic activity of these LAB, when incubated in the CM was verified by the concentrations of lactic acid produced. It was observed that L. fermentum NCIMB 5221 produced significantly high levels of lactic acid as represented in Figure 1. Lactic acid is used by lactate-utilizing butyrate-producing bacteria in the gut [41] and is considered an anti-inflammatory component [42], which has the ability to increase anti-tumor immunoreactivity [43]. SCFAs secreted by gut bacteria induce apoptosis in CRC cells and may, therefore, be relevant for the prevention and therapy of CRC. For example, microbial-derived butyrate was found to promote the stabilization of transcription factors related to epithelial barrier protection [44]. Butyrate and propionate inhibited the activity of histone deacetylases (HDACs) in colonocytes and immune cells and induced anti-inflammatory effects via the differentiation of regulatory T-cells [45]. Thus, SCFAs secreted by L.
fermentum, were quantified and produced at significantly different concentrations ( Figure 2). L. fermentum NCIMB -2797, -8829, and -5221 produced significantly higher amounts of total SCFAs in their CM, compared with L. acidophilus ATCC 314 (p < 005, Figure 2d), but significantly lower amounts of lactate in their respective CM (p < 0.001, Figure 1). This result suggests that L. fermentum may act as an anti-colon cancer agent due to the production of higher quantities of SCFAs distinctively from L. acidophilus ATCC 314. Consequently, L. fermentum may produce anti-tumorigenic and anti-inflammatory activities as shown in a CRC Apc Min/+ mice model [46]. The higher levels of lactate produced may provide more substrate for antioncogenic bacteria in the gut. Therefore, L. fermentum bacteria may play a vital role in CRC prevention through SCFAs production rather than by modulating the gut microbiota. This effect may also provide growth support for other beneficial microbiota, or inhibition of CRCassociated bacteria due to the production of lactic acid [47]. This study also showed that the concentrations of acetic acid and propionic acid measured are about half of the optimal doses suggested in the literature to induce inhibitory effects on Caco-2 cells [48], which predicts a more efficient cancer-suppressive effect of the probiotic treatment by the L. fermentum bacteria.
The role of microbial SCFAs in colon carcinogenesis is debatable and poorly understood. Several reports have provided evidence on the effect of probiotic bacterial supernatants or separately tested pure SCFAs in the mechanism of cancer cell inhibition. Many of these studies associated the potential anti-cancer activity of probiotic bacteria with the production of SCFAs; however, few have validated this theory [49]. In this study, L. fermentum-CM significantly inhibited CRC cell proliferation, in a time-dependent manner, compared with untreated cells and cells treated with L. acidophilus ATCC 314 (p < 0.05, Figure 3). between them (Figures 4e and 4f). To identify potential factors other than SCFAs involved in this activity, concentrations of synthetic SCFAs prepared as a mixture were tested on CRC cells. Figure 4a demonstrates that artificially prepared doses of pure SCFAs have significantly less effect when compared with the probiotic bacterial extracts CM (p < 0.01). This fact supports the ability of a particular naturally produced SCFA to induce inhibitory effects ( Figure 4). Overall, the synthetically prepared mixtures of SCFAs showed a closer effect to L. fermentum-CM (Figure 5b). More specifically, L. fermentum NCIMB 5221 had the same effect as its corresponding SCFA formulation. The L. fermentum NCIMB -2797 and -8829 significantly inhibited colon cancer cell growth less than the corresponding SCFAs synthetic formulations (p < 0.05), indicating that the bacteria have potentially secreted additional anti-cancer products. Nonetheless, L. acidophilus ATCC 314 was significantly less effective than its SCFA synthetic cocktail. This indicates the presence of other bacterial factors, produced in the CM, which hindered the effect of the naturally produced probiotic SCFAs. The data produced indicates that the anti-proliferative effect of the CM is possibly due, in a minor part, to the concentration of bacterial SCFAs; however the effect is not solely related to the presence of SCFAs. As described in Table 1, lactic acid was added to each SCFA synthetic formulation. These lactic acid-containing SCFA mixtures had significantly less effect than either SCFA synthetic formulation or probiotic CM (p < 0.001). This implies that the presence of lactic acid may have reduced the efficacy of SCFAs on the metabolism of cancer cells. This is supported by a study where L-lactate significantly inhibited uptake of butyrate in cancer cells [41], suppressing the anti-cancer effect of the latter. Hence, the lactate, added later to the SSFs, could have suppressed the ability of cancer cell to uptake SCFAs resulting in the decreased action of SSF containing lactate. Some of the bacterial products released by L. fermentum bacteria were indicated as surface [50] and adhesive [51] proteins that bind to the intestinal and gastric mucus as DNA fragments, or lipopolysaccharides [52]. As explained, the anti-proliferative effect of L. fermentum may not only be based on the activity of SCFAs but also on the release of other bacterial products that may have preserved or enhanced the effect of SCFAs.
Another feature related to probiotic strain selection was the loss of viability of L. fermentum bacteria in simulated human intestinal conditions as well as the ability to produce SCFAs. Interestingly, L. fermentum NCIMB -5221 and -8829, which exhibited higher anti-colon cancer potential, showed similar densities /absorbances ( Figure 6a) and resistance to the bile exposure for 4 h, which was significantly higher than for L. acidophilus ATCC 3 (p < 0.05, Figure  6b). Some studies have shown that L. fermentum have resistance to gut conditions; however, this feature varied according to the glucose and other nutrient availability in the gut. L. fermentum tolerance to intestinal conditions was observed, mainly, for a maximum of 4 h, compared with other probiotic bacteria [53]. Between 12 h and 16 h,  Figure 7e). This data implied that L. fermentum bacterial cells are more active and have the potential to produce efficiently higher concentrations of anti-cancer bioactive compounds than L. acidophilus ATCC 314. Testing those concentrations separately on CRC cells (Figure 7) [54] confirms this finding. The levels of SCFAs produced by L. fermentum bacteria in SIF were shown to significantly reduce CRC cell proliferation, compared with L. acidophilus ATCC 314, in adherence with the superior inhibitory effect of the L. fermentum cellfree extract described in Figure 3. Notably, the only SCFA L. acidophilus ATCC 314 that did not produce detectable levels was propionate ( Figure 2b). Nevertheless, the propionic acid concentration produced in the SIF seemed significantly more effective in decreasing the Caco-2 viability than acetic acid SIF concentrations (p < 0.001, Figure 8a), suggesting that propionate production is a major mechanism for colon cancer inhibition by L. fermentum in the intestinal environment.  (Table 2). Data are presented as mean ± SEM (n = 5). *p < 0.05 and ***p < 0.005, compared with control or L. acidophilus ATCC 314. SSF-SIF-f: formulation of SCFAs produced in SIF corresponding to both L. fermentum bacteria (NCIMB -5221 and -2797); SSF-SIF-a: SCFA formulation of SCFAs produced in SIF by L. acidophilus ATCC 314.

Conclusion
This present study is the first to explore and compare the potential suitability of L. fermentum NCIMB -5221, -2797, and -8829 as CRC biotherapeutics in vitro (Figure 9). These strains were characterized for their production of active molecules relevant to CRC and their tolerance to intestinal stress. They also exhibit the production of SCFAs in different environments (supernatant CM or intestinal fluid SIF) and the suppression of CRC cell growth. We were able to compare the antiproliferative effect of L. fermentum probiotic bacterial strains in vitro while evaluating the efficacy of SCFAs bioproduction as a mechanism. Our findings identified a significant effect of L. fermentum strains in inhibiting colon cancer cells which correlate with the ability of these bacteria to produce SCFAs. These strains also showed significant efficiency in producing SCFAs in intestinal conditions, suggesting an ability to generate an appreciable anti-carcinogenic effect in the colon.