A novel strain of Bacteroides fragilis enhances phagocytosis and polarises M1 macrophages

Commensal Bacteroides fragilis possesses immune-regulatory characteristics. Consequently, it has been proposed as a potential novel probiotic because of its therapeutic effects on immune imbalance, mental disorders and inflammatory diseases. Macrophages play a central role in the immune response, developing either a classical-M1 or an alternative-M2 phenotype after stimulation with various signals. The interactions between macrophages and B. fragilis, however, remain to be defined. Here, a new isolate of B. fragilis, ZY-312, was shown to possess admirable properties, including tolerance to simulated gastric fluid, intestinal fluid and ox bile, and good safety (MOI = 100, 200) and adherent ability (MOI = 100) to LoVo cells. Isolate ZY-312 cell lysate promoted phagocytosis of fluorescent microspheres and pathogenic bacteria in bone marrow-derived macrophage (BMDM) cells. Gene expression of IL-12, iNOS and IL-1β in BMDM cells was increased after treatment with ZY-312, indicating the induction of M1 macrophages, consistent with enhanced secretion of NO. Cell surface expression of CD80 and CD86 was also increased. This study is the first to demonstrate that B. fragilis enhances the phagocytic functions of macrophages, polarising them to an M1 phenotype. Our findings provide insight into the close relationship between B. fragilis and the innate immune system.

T cells activating adaptive immune responses 14 . Furthermore, cytokines secreted by macrophages mediate a wide range of immunoregulatory effects 14,15 . For example, IL-12 produced by macrophages mainly boosts activation of Th1 cells conferring cellular immunity, while IL-10 activates Th2 cells and suppresses cellular immune responses. To better understand the opposing roles of macrophages, the M1/M2 paradigm was proposed, which correlates to Th1/Th2 responses 16 . The M1/M2 paradigm provides a framework with which to distinguish the states and functions of macrophages, with M1 and M2 macrophages at either end of a continuum and the intermediate states exhibiting overlapping functions 15 . The main distinguishing feature is that M1 macrophages metabolise arginine to NO, whereas M2 macrophages produce ornithine 16 . Furthermore, M1 macrophages are associated with IL-12 and IL-8/CCL production, cell surface expression of CD80/CD86 14 , promoting Th1 responses and possess strong microbiocidal and tumoricidal activity 17 . By contrast, M2 macrophages are linked with IL-10 production, promoting Th2 responses and aiding tissue repair, as well as ameliorating inflammation 14,15 .
In this study, we isolated a novel bacterial strain from the faeces of a healthy baby. After morphological and genetic analysis, this strain was confirmed to belong to B. fragilis and was designated ZY-312. The probiotic properties of this strain were studied, focusing on the effects on macrophages using an in vitro model of bone marrow-derived macrophages (BMDM). Our findings indicated that a cell lysate of ZY-312 significantly enhanced the phagocytic functions of BMDM, both on fluorescence beads and pathogens. Increased gene expression of IL-12, IL-1β and iNOS, as well as the production of NO and cell surface expression of CD86 and CD80, were also detected. We established for the first time that B. fragilis enhances the function of macrophages, polarising them towards the M1 type.

Results
Isolation and molecular identification of ZY-312. The bacterial strain ZY-312 isolated from the faeces of a healthy breast-fed infant was found to be similar to B. fragilis both morphologically and in terms of its growth characteristics, which is shown in Fig. 1a-d. By Gram staining and direct observation by scanning electrochemical microscopy and transmission electron microscopy, this strain was found to be Gram-negative, and rod shaped with rounded ends. The physiological and biochemical features of this strain were also in accordance with B. fragilis as described in Bergey's manual. The results of 16S rRNA gene sequencing demonstrated that the nucleotide sequence of this strain was up to 99% identical to that of type stain B. fragilis ATCC 25285 (GenBank database), and showed high similarity to JCM11019, confirming that the isolate was a strain of B. fragilis.  Fig. 2a. Strain ZY-312 entered exponential phase after 8 h, and progressed to stationary phase at around 16 h. Throughout our study, we selected late logarithmic phase (12-16 h) cells because of the peak in the number and viability of bacteria.
Tolerance to air. As most assays are performed aerobically, we tested the viability of ZY-312 after exposure to air, and the survival rates are presented in Fig. 2b. Isolate ZY-312 showed resistance to air for at least three days, which agreed with a previous report 18 . Viable counts of ZY-312 were increased more than 8-fold at 12 h, then survival rates gradually decreased with few surviving cells at 72 h. This demonstrated that ZY-312 is not a strictly anaerobic bacterium, and sufficiently air resistant for the in vitro assays performed in this study.
Tolerance to stimulated digestive fluid. To investigate the ability of ZY-312 to tolerate digestive juice, an in vitro test was adopted to test the survival rate after exposure to simulated gastric fluid (SGF), simulated intestinal fluid (SIF) and bile under different conditions (pH or concentrations) and time points. ZY-312 resistance levels to SGF an SIF are shown in Table 1a. In this study, ZY-312 was unable to survive on exposure to SGF of pH 2.0 (data not shown), but survived well with SGF at pH 3.0 or pH 4.0, and with SIF at pH 6.8. The survival ability of ZY-312 in ox bile is presented in Table 1b. ZY-312 showed high resistance to bile, which was consistent with a previous study 19 that reported that B. fragilis showed good tolerance to bile salt.   Enhancing the phagocytic functions of BMDM cells. To explore whether ZY-312 affected BMDM phagocytic function, fluorescence microspheres and EHEC were used as foreign entities in a phagocytosis assay.
The results with fluorescent beads (Fig. 3a-d) demonstrated that the ZY-312 cell lysate (initial MOI = 500) not only promoted more BMDM cells to phagocytose beads (p < 0.05), but also encouraged BMDM to phagocytose more beads (p < 0.01), with significant differences observed between the treated and untreated group at 90 min.
In the pathogen phagocytosis assay, we infected BMDM cells with EHEC (MOI = 50). Accordingly, ZY-312 cell lysate had no influence on BMDM cell-associated bacteria (p > 0.05), but improved the phagocytosis of EHEC significantly (p < 0.001) (Fig. 3e,f). These findings indicated that ZY-312 cell lysate modulate and facilitate BMDM cells to phagocytose foreign entities. Polarising BMDM cells to become M1-like macrophages. Gene analysis of M1/M2 related and specific markers expressed by BMDM cells treated by ZY312 cell lysate or ZY312 viable bacteria are shown in Figs 4a-e and 5a-e, respectively. After treatment with ZY-312 cell lysate, IL-1β expression at the transcriptional level in BMDM cells was increased regardless of the initial inoculum of bacteria from which the lysate was obtained, whereas IL-12 and iNOS expression increased in a time-dependent manner. Under the same conditions, IL-10 was expressed at a relatively low level while Arg-1 was unchanged or even down regulated. To further confirm such polarisation, we analysed the total amount of NO in the cell supernatant (Fig. 4f). Consistent with the transcriptional changes, the amount of NO secreted was significantly increased (p < 0.05) in a manner dependent on incubation time and the bacterial inoculum to obtain the lysate. Similar results were obtained in infection experiments with viable bacteria (Fig. 5a-f). All these results demonstrated that ZY-312 promotes IL-12, IL-1β and NO expression in BMDM cells, and polarises them to become M1-like macrophages.
Enhancing cell surface antigen expression in BMDM cells. The cell surface expression of CD86 on BMDM cells was significantly enhanced after treatment both with ZY-312 cell lysate and bacteria (Fig. 6b,d), while CD80 expression was enhanced only with cell lysate (Fig. 6a,c). Taking into consideration the nutrients exhaustion and growth kinetics of ZY-312 in air, 4 h was chosen as the end point for experiments with living bacteria while the effects of cell lysate were assessed over a time course up to 12 h. These experimental differences might explain the lack of CD80 induction by viable bacteria. In conclusion, ZY-312 cell lysate promoted BMDM cells to express CD80 and CD86 after treatment for 12 h, and this effect has been partially reproduced in experiments with viable bacteria (MOI = 500).

Discussion
Bacteroides fragilis, a ubiquitous anaerobic bacterium that colonises the lower gut, was recently proposed as a potential probiotic because of its considerable clinical benefits on inflammatory and mental disorder diseases 6,8 .
In this study, we isolated ZY-312 and confirmed that it is a strain of B. fragilis by morphological analysis and 16 s rRNA sequencing according to Bergey's manual. First, we explored its fundamental properties, and demonstrated its high tolerance to SGF, SIF and bile, with the exception of low pH SGF (pH = 2.0). Given that our experiments were performed in air, we tested the survival rate of this isolate in air over time and found that it survived at least 72 h, a result that was consistent with previous reports 18 . Furthermore, we demonstrated the safety of this isolate in vitro at MOI = 200 and its adhesive ability to colon cells. Taken together, our findings indicate that ZY-312 is a new-isolated strain of B. fragilis with some desirable characteristics. Much research has been performed on the interaction between B. fragilis and the innate immune system, especially for PSA. The reciprocity between DCs and B. fragilis has been explored comprehensively and some insight into the molecular mechanisms has been reported 9,10,20 . However, little is known regarding the relationship between B. fragilis and macrophages. Macrophages are the first line of defence against pathogens when microorganisms invade 11 , and function as the control centre during host defence 12 . Macrophages have been designated as M1 or M2 based on their activation of Th1 or Th2 responses, highlighting the crucial role of macrophages in the host immune system 16 . In this current study, we found that following exposure to the ZY-312 cell lysate for 12 h, BMDM cells phagocytosed more fluorescence beads and pathogens compared with the control group (Fig. 3), implying the enhanced phagocytic function of macrophages treated with B. fragilis. Furthermore, we observed significantly upregulated gene expression of IL-12 and IL-1β in BMDM cells after treatment with cell lysate or viable bacteria. Increased expression of the iNOS gene and release of total NO into the supernatant were also observed, confirming the results of a previous report 21 . It has been reported that purified PSA induced production of IL-12 22 and NO 21,23 by dendritic cells. DCs internalise PSA through the TLR-2 pathway and present it to T cells 23 , and this process is NO dependent since NO is necessary for the degradation of PSA. The expression of CD80 and CD86 were increased on BMDM cells (Fig. 6) after treatment, and similar results have previously been shown for DCs 21 . Exposure to PSA reportedly increases the expression of co-stimulatory molecules CD80 and CD86 1,24 and signalling through CD80 and/or CD86 to activate the generation of CD4 + T cells 25 .
Interestingly, B. fragilis seems to play a complex role in interacting with the immune system, which is dependent on the microenvironment. Once B. fragilis (ATCC 25285) leaks into the abdominal cavity, it leads to inflammation due to the increased production of pro-inflammatory cytokines TNF-α and IL-1β and abscess formation, then PSA is internalised by antigen-presenting cells after antigen presentation loading on the MHC II compartment to generate IL-10 producing T cells for controlling excessive inflammation 6,26,27 . This process is mediated by the complex interaction between B. fragilis and the innate as well as adaptive immune system. B. fragilis appears to initially develop a M1-like response to induce antimicrobial molecules that assemble immune cells and promote microbicidal molecules to kill and confine other gut microorganisms to prevent a more serious infection 6 . Our finding that B. fragilis promotes IL-12 production and phagocytic functions in macrophages provides evidence for this. Then, increased expression of CD80 and CD86 and degraded PSA loading on the surface of antigen-presenting cells, results in the generation of IL-10-producing T cells that aid the transition to an M2-like response to avoid an excessive inflammation response 6 .
The role of B. fragilis has recently been reconsidered and it is now thought that the development of an abscess may not be harmful to the host but may protect it from serious or even life-threatening complications 6 . Moreover, B. fragilis has been proposed as a potential probiotic because of the profound beneficial impacts of this organism on allergic and mental disorder diseases. However, much work is still needed to further elucidate the intricate and intimate relationship between B. fragilis and the host immune system. Our study is the first to systemically study the immune regulatory effects exerted by B. fragilis on macrophages in vitro, and to demonstrate that B. fragilis not only enhances the phagocytic functions of macrophages, but also polarises them to an M1 phenotype, providing new insight into our understanding of potential probiotic characteristics of B. fragilis.

Methods
Isolation of strain ZY-312 and molecular identification. The bacterial strain ZY-312 was isolated from the faeces of a healthy, breast-fed, infant as reported previously 28 . The collection of stool sample was part of a study approved by the Medical Ethics Committee of NanFang Hospital (NFEC-2014-040) and the procedure was performed in accordance to Institutional Review Board guidelines. A consent form was read and signed by parents of that baby. The isolated strain presented morphological and biological characteristics similar to those of B. fragilis, as described in Bergey's manual. Genomic DNA from the strain was extracted using a DNA extraction kit (Qiagen) and 16S rRNA gene amplification was performed by PCR using the universal primers 27 F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492 R (5′-GGTTACCTTGTTACGACTT-3′) (Lanes D-J). Amplified products were sequenced (Biomed) and the BLAST algorithm was used to search for homologous sequences in the NCBI database.
Growth curve and air exposure. Strain ZY-312 was cultured in sterile tubes containing 10 ml of tryptone soy broth (TSB) supplemented with 5% fetal bovine serum (FBS), grown anaerobically at 37 °C in anaerobic glove box (Bugbox, Ruskin) for 24 h. Optical density at 600 nm was detected every 2 h. Given that most of the following tests were performed in air, we first verified the resistance of the strain to the atmosphere. Strain ZY-312 in late log phase was harvested and suspended in TSB (plus 5% FBS) to 10 9 cfu/ml, incubated aerobically at 37 °C for 96 h. The survival was determined by the spread counting.
Simulated gastric fluid, intestinal fluid and bile tolerance assay. Tolerance assays were performed as previously reported 29,30 . For preparation of SGF and SIF, 1 g/L pepsin (Sigma) or 10 g/L trypsin enzyme (Sigma) was supplemented with PBS, respectively. Using 1 M sodium hydroxide (Sinopharm) and 2.75 M hydrochloric acid solution (Sinopharm), SGF was adjusted to pH 2.0, 3.0, 4.0, and SIF to pH 6.8. Bacterial cells harvested after 14 h incubation (late exponential phase) were collected by centrifugation at 2000 g for 5 min at RT, washed twice with PBS and suspended to 10 9 cfu/ml in SGF or SIF. Cell suspensions were then incubated at 37 °C anaerobically for 0, 1, 2 or 3 h. The viability of samples was determined by the spread plate method. Similarly, bacterial cells were collected as described above and adjusted to 10 −4 , 10 −6 or 10 −8 cfu/ml in bile containing 0, 1, 2 or 4% ox gall bladder powder (Sangon) in sterile TSB (plus 5% FBS). The mixture was incubated at 37 °C anaerobically for 0, 1, 2 or 4 h and each samples was serially diluted and spread on agar. The survival rate was determined by the ratio between the counts of viable bacteria and initial bacteria.

Cytotoxicity against colon cells. The in vitro safety of ZY-312 was tested on LoVo cells using a Cytotoxicity
Detection Kit Plus (LDH) (Roche). LoVo cells were seeded at 5 × 10 4 cells per well into 96-well plates and incubated overnight. 100 μl bacteria suspension (5 × 10 7 or 1 × 10 8 cfu/ml, MOI = 100 or 200) was added per well. After co-incubation at 37 °C in 5% (v/v) CO 2 and 95% (v/v) humidified air for 4 h, absorbance values were determined using a microplate reader (SpectraMax, M2) and cytotoxicity was calculated with the following equation provided by the manufacturer's protocol.
BMDM isolation. Bone marrow-derived macrophage (BMDM) cells from mice were stimulated with recombinant mouse M-CSF (R&D) as reported previously 21 . Briefly, the bone marrow cells were stimulated with 20 ng/ml M-CSF. After 7 days culture, more than 90% pure BMDM were obtained. For the following tests, BMDM cells were harvested using a cell scraper (Costar) and cultured at 5 × 10 5 cells/well in a fresh 24-well plates overnight.
Preparation of ZY-312 cell lysate. ZY-312 cell lysates were prepared in advance should be used immediately. ZY-312 cell wall were broke by repeated ultrasonic sonication (Branson) on ice at 50 W for 10-s, at 10-s intervals, until cells were completely ruptured. The efficiency of cell breakage was confirmed by TSA agar plate counting before and after sonication. Centrifugation at 8000 g at 4 °C for 10 min and filtering were performed to ensure bacteria-free.
Fluorescence phagocytosis assay. Fluorescent microspheres (Molecular Probes, 1.0 μm, carboxylatemodified) were pretreated and adjusted with culture medium to about 2 × 10 9 microspheres/ml. Bath sonication for 30 min immediately before the phagocytosis assay was necessary to disperse the beads. After cell lysate (initial MOI = 500) treatment for 12 h, replaced 300 μl of beads to each well (using culture medium as a negative control), incubated at 37 °C for 30, 60 or 90 min. Cells were washed with cool PBS three times and collected, detected and analysed by flow cytometry (BD, Accuri, C6). Nonspecific adherent beads were not considered as microspheres and were due to low nonspecific binding. Phagocytosis rate and phagocytosis index were calculated using the following equations: where, the Sum of beads phagocytosed by BMDM cells = Cells that phagocytosed 1 bead ×1 + Cells that phagocytosed 2 beads ×2 + Cells that phagocytosed 3 beads ×3 + Cells that phagocytosed 4 beads ×4 + Cells that phagocytosed 5 beads ×5 + Cells that phagocytosed more than 5 beads ×6. The numbers of beads phagocytosed was determined by overlying the fluorescence value and peak area with the interval.
Pathogen phagocytosis assay. The pathogen used for this study was Enterohaemorrhagic Escherichia coli (EHEC) (Lab stock, No. CBSLAM00087), a serotype of E. coli accounts for haemorrhagic diarrhoea and kidney failure 32 . A pathogen phagocytic assay was performed according to a previous report 33 . EHEC cells were harvested from overnight culture in LB and suspended in DMEM at 2 × 10 7 cfu/ml. 0.5 ml EHEC suspension were added to BMDM cells treated with cell lysate (initial MOI = 500) for 12 h, incubated at 37 °C for 30 min. To detect cell-associated bacteria (including both adherent and phagocytosed bacteria) after incubation, cells were washed three times with cool PBS, and 0.5 ml per well of 0.1% TritonX-100 solution (Sigma) was added for 3 min to lyse cells (3 min was sufficient for total lysis under these conditions). Additionally, at same time point, the supernatant was replaced with culture medium containing 50 μg/ml gentamycin and incubated for 30 min to kill all extracellular bacteria. Cells were washed, lysed as above, and bacteria were counted after serial dilution and plating on LB agar.
RT-PCR assays. BMDM cells (six replicated wells per group) were treated with ZY-312 cell lysates (initial MOI = 20, 100, 500) for 3, 6, 12 h or ZY-312 viable bacteria (MOI = 20, 100, 500) for 4 h, and total RNA was extracted using the Purelink TM RNA kit (Ambion) following the manufacturer's protocol. The concentration and quality of the extracted RNA was determined by spectrophotometric measurements at OD 260 (ND-1000; NanoDrop). Isolated total RNA (5 μg) was transcribed to cDNA using Superscript II reverse transcriptase (Invitrogen) and oligo (dT) primers (Invitrogen) following the manufacturer's protocol. After purification with the QIAquik PCR purification kit (Qiagen), purified cDNA was used as template in qRT-PCR assays. Quantitation RT-PCR was performed for IL-10, IL-12 IL-1β, iNOs, Arg-1 and the reference gene GAPDH using the Light Cycler system (Roche), the SYBR Green master mix (Roche) and the primers shown in Table 2. Data was analysed using the LCS480 1.50.SP4 software (Roche). Culture medium was included as a background control, and results were represented as log-fold changes. Non-template controls were included in the assay.

Experimental replicates and statistical methods.
All experiments were performed at least in triplicate using independent assays, and values were expressed as the mean ± standard error. Unpaired Student's t-test was performed to determine statistically significant differences in phagocytic assays; one-way ANOVA test and Dunnett's post-hoc test were performed to determine statistically significant differences in other experiments. A p value of <0.05 was considered to indicate statistical significance.