Recombinant Fusion Protein Vaccine Containing FliC and FliD Protects Mice Against Clostridioides Dicile Infection

Bacterial agella are involved in infection through their roles in host-cell adhesion, cell invasion, auto-agglutination, colonization, and formation of biolms, as well as in the regulation and secretion of non-agellar bacterial proteins involved in the virulence process. In this study, we constructed a fusion protein vaccine (FliCD) containing Clostridiodes dicile agellar proteins FliC and FliD. Immunization of mice with FliCD induce potent IgG antibody responses against FliCD and protected mice against C. dicile infection and decrease C. dicile spores and toxin levels in the feces after infection. Furthermore, we found anti-FliCD serum protected mice against CDI and decreased C. dicile spores and toxin levels in the feces after C. dicile infection. Finally, we found that anti-FliCD serum inhibited the binding of C. dicile vegetative cells to HCT8 cells. These results imply that FliCD fusion protein may represent an effective vaccine candidate for the prevention from C. dicile infection (CDI).


Introduction
Clostridioides di cile (C. di cile) is an anaerobic, spore-forming, Gram-positive anaerobic bacterium, and was identi ed as the leading cause of antibiotic-associated diarrhea and colitis 1,2 . C. di cile produces three protein toxins including toxin A (TcdA), toxin B (TcdB), and binary toxin (CDT), rst two of which are the major virulence factors of C. di cile and are the major drivers of C. di cile infection (CDI) symptoms 3,4 . CDI is spread by bacterial spores found within feces, and infections occur in all areas of the world 5,6 .
In the United States, the burden on the healthcare system was calculated at 8.3 cases per 10,000 patientdays, suggesting that CDI is associated with a large burden on the healthcare system 7 . Currently, very few antibiotics are available for the treatment of CDI, and none of which is fully effective 8 , and antibiotic treatment for CDI is often followed by recurrent infection, leading to nontraditional treatments 9,10 .
Flagella of most pathogens play a role in motility and chemotaxis that increases the occurrence of potential interactions between the pathogen and the epithelial mucosal surface. Moreover, bacterial agella are involved in infection through their roles in host-cell adhesion, cell invasion, auto-agglutination, colonization and formation of bio lms, as well as in the regulation and secretion of non-agellar bacterial proteins involved in the virulence process 11 . C. di cile agellin FliC is the major structural component of the agellar lament, and assembly of a agellum requires other proteins called hook-associated proteins (HAP1, HAP2, and HAP3). The iD gene encodes structural component HAP2 of the agellar cap at the distal end of the lament [12][13][14] . Both the FliC and FliD proteins are implicated in the attachment of C. di cile to the mucus layer of the intestine. Flagella and especially the agellar FliD protein appear to be some of the multiple cell adhesions of this microorganism 15 . Researchers also found that agellated, motile C. di cile attach more e ciently to the cell wall of axenic mice than non-agellated strains of the same serogroup 16 . Interestingly, one study showed that both the iC and iD mutant strains adhered better than the wild-type 630△erm strain to human intestine-derived Caco-2 cells, and were more virulent in hamsters 17 . The con icting reports may implicate a complex role of agella in infection of C. di cile.
Previously, FliC immunization provided partial protection against CDI in mice and hamsters (Ghose et al. 2016). In this study, we constructed a fusion protein vaccine (FliCD) containing FliC and FliD, and evaluated its immunogenicity and protection in mice against CDI.

Animals
All studies followed the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and were approved by the Institutes Animal Care and Use Committee (IACUC) at the University of South Florida and Tufts University. Wild-type C57BL/6 mice were purchased from Charles River Laboratories.
Expression and puri cation of recombinant fusion protein FliCD Gene sequences encoding FliC and FliD from C. di cile R20291 18 werebridged with a linker (ggt ggc tct ggt) sequence, synthesized by Geneart (Germany) and cloned between the BsrGI and EagI sites of the vector pHis1525 19 . FliCD was expressed in B. megaterium and puri ed as described previously 20 .

Preparation of C. difficile spores
Sporulation of the C. difficile R20291 strains were induced in Clospore medium as described previously 21 . Briefly, an overnight 20 ml of C. difficile cultured in Columbia Broth was inoculated into 500 ml of Clospore medium, and incubated for 1-2 weeks at 37 °C in an anaerobic incubator. The spore suspension was centrifuged at 10000g for 20 min, and the pellet was washed 5 times with sterile water and suspended in 10 ml of ddH 2 O. The spore suspension was layered onto the top of 10 ml of 50% (wt/vol) sucrose in water in a 15-ml tube. The gradient was centrifuged at 3200×g for 20 min, after which the spore pellet at the bottom was washed five times to remove the sucrose and was resuspended in water. All spore preparations were >99% pure, free of vegetative cells and debris 22 . The spore concentration was determined by serial dilution on TCCA or BHI plates.
Mouse immunization and mouse model of CDI Ten or twenty micrograms of puri ed FliCD in PBS along with alum as adjuvants were used in immunization experiments in animals. Female C57/BL6 mice were housed under the same conditions at a semi-natural light cycle of 14 hours:10 hours (light: dark) in a speci c pathogen-free (SPF) environment. During immunizations and infection with C. di cile, mice were housed in infection rooms. Mice (n = 10) were immunized 3 times at 12-day intervals via intraperitoneal injection (i.p.) routes. Control mice received the same amount of PBS. Sera were collected, and anti-FliCD IgG titers were determined by ELISA. Seven days after the third immunization, immunized or control mice were given drinking water containing a mixture of six antibiotics including ampicillin (200 mg/kg), kanamycin (40 mg/kg), gentamycin (3.5 mg/kg), colistin (4.2 mg/kg), metronidazole (21.5 mg/kg) and vancomycin (4.5 mg/kg) for 4 days, and then received autoclaved water for 2 days, followed by a single dose of clindamycin (10 mg/kg) intraperitoneal injection before challenge with 10 6 C. di cile R20291 spores/mouse via oral gavage as described previously 23 . After infection, mice were monitored daily for a week for survival, weight changes, diarrhea and other symptoms of the disease. Diarrhea was de ned as wet tails, loosen or watery feces. The death included the numbers of mice died after infection and mice euthanized if weight loss was > 20%.
Evaluation of anti-FliCD sera in protection of mice against CDI Mice (n=10) were immunized 4 times at 12-day intervals via intraperitoneal injection (i.p.) routes with 10 μg of FliCD in PBS along with alum as adjuvants. Fourteen days after the fourth immunization, sera were collected and de ned as hyperimmune serum. The mouse model of C.di cile infection was established as described above. Four hours prior to infection with C. di cile R20291spores, 400 ul hyperimmune sera were administrated to each mouse via i.p. Control mice received PBS or control sera.

ELISA for anti-FliCD IgG
ELISA assays were performed as previously described 20 . Brie y, Costar 96-well ELISA plates were coated with 100 ul/well of FliCD (0.5 ug/ml) at 4 ºC overnight. Following washing of the unbound material, plates were blocked with 300 µl of blocking buffer (PBS + 5% dry milk) at room temperature for 2 hours. After washing, 100 µl of 10-fold diluted sera or fecal samples were added into each well of the plates, and incubated for 1.5 hours at room temperature. Following washing with PBS, 100 µl of mouse IgG-HRP (1:3000) were added to each well and incubated for 30 min to 1 hour. Subsequent to a washing step with PBS, substrate TMB was added to allow color development at room temperature for 5-30 min. The reaction was stopped by addition of H 2 SO 4 to each well, and the OD values at 450 nm were recorded by a spectrophotometer. Anti-toxin and anti-FliCD IgG titers of a given sample (serum or fecal sample from immunized mice was de ned as the dilution factor at which the OD 450nm is greater or equal to 2-fold that of serum or fecal sample from non-immunized mice).
Quanti cation of C. di cile spores in mouse feces Fecal samples were collected on days 0, 1, 3, 5 and 7 postinfection. 50 mg of feces were dissolved with 500 µl sterile water for 16 hours at 4°C, and then treated with 500 µl of puri ed ethanol (Sigma-Aldrich) for 1 hour at room temperature to kill vegetative cells. Samples were vortexed, serially diluted and plated onto selective medium supplemented with taurocholate (0.1% w/v), Cefoxitin (8 µg/mL), D-cycloserine (250 µg/mL). The plates were incubated anaerobically at 37°C for 48 hours, colonies counted, and results expressed as the CFU/gram of feces.

Quantitation of C. di cile toxins in mouse feces
After challenge with C. difficile spores, feces were collected, and dissolved in PBS (0.1g/ml) containing a protease inhibitor cocktail, and the supernatants were collected after centrifugation, and used for determination of TcdA/TcdB concentrations by ELISA. Brie y, 96-well Costar microplates were coated with 100 µl of anti-TcdA antibody (1µg/ml) and anti-TcdB antibody (1µg/ml) overnight in phosphatebuffered saline (PBS) at 4°C. On the next day, each well was blocked with 300 µl of blocking buffer (PBS + 5% dry milk) at RT for 2 hours. Next, standards and samples were added to each well (100 µl) in duplicate, and incubated for 90 min at 25°C. After another set of washings, HRP-chicken anti-C. di cile TcdA/TcdB (1:5,000 dilution in PBS, Gallus Immunotech, Shirley, MA) was added to wells for 30 min at RT. A nal set of 3-washing preceded the addition of the TMB microwell peroxidase substrate for 20 min at RT in the dark. The reaction was stopped with 2 N of H 2 SO 4 , and the absorbance was measured using a plate reader at 450 nm.

Adherence inhibition assays
The adherence of the C. difficile R20291 vegetative cells to human gut epithelial cells was assessed as described previously 24 . Brie y, HCT-8 cells were grown to 95% con uence (1×10 5 /well) in a 24-well plate and then moved into the anaerobic chamber, followed by infecting with 1.5 × 10 6 of log phase of R20291 vegetative cells at a multiplicity of infection (MOI) of 15:1. The plate was cultured at 37 ℃ for 100 min in an anaerobic chamber. R20291 vegetative cells was preincubated with hyperimmune serum (1/50, 1/100, 1/500 and 1/1000) for 30 min before being added to cells. After incubation, the cell-C. di cile mixture was washed three times with 1×PBS via centrifugation at 800×g for 1min to remove any unattached R20291. The supernatants after centrifugation from each wash step were collected to enumerate any R20291 that did not adhere to the cells. The R20291 in the supernatant were enumerated on prereduced BHI agar. The controls included PBS incubated with R20291 and preimmune sera (1/50) incubated with R20291, and the adhesion assays were performed in triplicate. The percentage of R20291 adherence was calculated using the following formula: (initial CFU/ml -eluted CFU/ml)/ initial CFU/ml.

Statistical analysis
Animal survivals were analyzed by Kaplan-Meier survival analysis with a log-rank test of signi cance. When comparing results for two groups, student's unpaired t-test was used for statistical signi cance: when comparing the results of more than two groups, one-way analysis of variance (ANOVA) with posthoc analysis by Bonferroni tests. Results are expressed as means ± standard errors of means.
Differences were considered statistically signi cant if P < 0.05 (*). All statistical analyses were performed using GraphPad Prism software.

Immunization of mice with FliCDinduces signi cant anti-FliCD responses in mice
Recombinant FliCD with a 6xHis-tag (97 kDa) was expressed in Bacillus megaterium, and purified by Niaffinity chromatography to a purity greater than 95% (Fig. 1A).
Immunizations of mice with 10 µg or 20 µg FliCD in combination with alum as an adjuvant via i.p. route induced high levels of IgG antibody responses against FliCD in sera (Fig. 1B). The titers of serum after 2 nd immunization is higher than after 1 st immunization (p=0.0062,**p <0.01), the titers of serum after 3 rd immunization is higher than after 2 nd immunization (p=0.0187,*p <0.05), but no signi cant differences between 3 rd immunization and 4 th immunization (p=0.4433). Also, immunization of mice with 20 µg FliCD got higher titers of serum than with 10 µg FliCD after 3 rd immunization and 4 th immunization, but no signi cant.
Immunizations of mice with FliCD protect mice against C. di cile infection and decrease C. di cile spores and toxin levels in feces from mice challenged with C. di cile spores Protection e cacy of FliCD immunization was further evaluated in a mouse model of CDI. After three immunizations (10 µg or 20 µg per immunization for 3 times at 12-day intervals), mice were challenged with 10 6 spores of C. di cile R20291, a hypervirulent strain of ribotype 027. In vehicle (PBS)-immunized mice, signi cant disease symptoms including weight loss (Fig. 2B) and severe diarrhea (Figs. 2C and 2D) were evident in all mice; approximately 60% of mice succumbed by day 4 (Fig. 4A). In contrast, FliCDimmunized mice developed much less severe disease symptoms including less weight loss (Fig. 2B) and lower diarrhea rates (Figs. 2C and 2D) with a signi cantly higher survival rate (80% for 10 µg FliCDimmunized mice, 90% for 20 µg FliCD-immunized mice; Fig. 2A). FliCD-immunized mice secreted signi cantly lower amounts of toxin A (Fig. 3A), toxin B (Fig. 3B) in the feces, compared to PBS immunization groups (Figs. 3A & 3B). The fecal samples of FliCD-immunized mice contained signi cantly fewer R20291 spores compared to PBS immunization groups (Fig. 3C).
Anti-FliCD hyperimmune sera protect mice against C. di cile infection and decrease C. di cile spores and toxin levels in the feces from mice challenged with C. di cile spores To elucidate how FliCD immunized mice are resistant to CDI, we tested whether anti-FliCD sera are protective in mice against CDI. Anti-FliCD hyperimmune sera were collected from mice immunized with 10 µg FliCD for 4 times, and were administrated to mice (400 ul/mouse) via i.p. 4 h prior to C. di cile R20291 infection (10 6 spores) in a mouse model of CDI. Almost all mice in PBS and control sera group developed diarrhea (90% in PBS group and 80% in control sera group; Fig. 4C and 4D), signi cant weight loss (Fig. 4B) with a 20% survival rate in PBS group and 40% in control sera group (Fig. 4A), while the mice administrated by 400 ul hyperimmune serum developed much less severe disease symptoms including less weight loss (Fig. 4B) and lower diarrhea rate (50%; Fig. 4C) with a signi cantly higher survival rate (80%; Fig. 4A).
Mice administered with anti-FliCD sera secreted signi cantly lower amounts of TcdA ( Fig 5A) and TcdB ( Fig 5B) in the feces from postinfection days 1 to 3, compared to PBS group and control sera group (Figs. 5A & 5B). The fecal samples of hyperimmune serum administrated mice contained signi cantly fewer R20291 spores compared to PBS group and control sera group (Fig. 5C).
Signi cant high levels of anti-FliCD antibodies were also detected in sera and feces from hyperimmune serum administrated mice (Fig. 6). The third and fth mouse in the group whose weight loss was above 20% were among those with lowest anti-FliCD titers in sera and feces, a rming the protective effects of anti-FliCD antibodies in mice against CDI Anti-FliCD serum inhibits the binding of C. di cile to HCT8 cells When the anti-FliCD serum was diluted 1 to 50 or 1 to 100 in the cells medium, adherence rate of the C. difficile R20291 vegetative cells to HCT8 cells was signi cantly decreased (4.95 ± 0.67% or 7.49 ± 0.94% vs 15.86 ± 1.21%). When the serum was diluted 1 to 500, the adherence rate was decreased to 10.87 ± 0.55%, but no signi cant (Fig. 7).

Discussion
Both FliC and FliD may play an important role in cell adherence, colonization, invasiveness and pathogenicity of C. di cile 15 . In this study, we found that both FliCD immunizations and hyperimmune anti-FliCD serum could protect against C. di cile infection in mouse ( Fig. 2 and Fig. 4) and decrease C. di cile spores and toxin levels in the feces from mice challenged with C. di cile spores. Furthermore, we found anti-FliCD serum inhibited the binding of C. di cile R20291 vegetative cells to HCT8 cells (Fig. 7).
Based on our results, we hypothesized that antibodies against FliCD may reduce or block the colonization of C. di cile in gut, and consequently protected host who exposed to C. di cile spores. As shown in Fig. 7, when the anti-FliCD serum was diluted to 1:500, the adherence rate decreased with no signi cance, while the serum was diluted to 1:100, the adherence rate decreased signi cantly. This was also supported by our in vivo experiment. In Fig. 6, we found that anti-FliCD titers in feces and sera from the C. di cile-induced moribund mice were among in the lower titer group.
Interestingly, one study showed that both the iC and iD mutant strains lost agella, but adhered better than the wild-type 630△erm strain to human intestine-derived Caco-2 cells, and were more virulent in hamsters 17 , at least partially caused by producing more toxins. Their data also suggest that neither FliC nor FliD is required for cecal colonization of hamsters. More work is clearly needed to further understand the phenotypic differences between a complete loss of agella by gene silence and directly binding by antibodies of agella proteins.
In summary, we constructed a protein vaccine FliCD (containing FliC and FliD), showed its potent e cacy as a new vaccine candidate in experimental mouse models of CDI. Our data showed that not only FliCD fusion protein represents an effective vaccine candidate, but also anti-FliCD serum may represent an alternative therapy against CDI. Immunizations of mice with FliCD provide mice signi cant protection against infection with C. difficile strain R20291. Mice were challenged with C. difficile R20291 spores (106/mouse) 14 days after the third immunization of groups of mice (n=10) with FliCD at 10 or 20 ug/mouse/immunization or PBS in the presence of alum. Kaplan-Meier survival plots (A), mean relative weight of all surviving mice (up to the day of death) (B) of different groups, and frequency of diarrhea (C, D) are illustrated. Data were presented as mean relative weight ± standard error(* p<0.05).     Anti-FliCD serum inhibits the binding of C. di cile to HCT8 cells. R20291 vegetative cells were preincubated with hyperimmune serum (1/50, 1/100, 1/500 and 1/1000, anti-FliCD serum titers is 107) for 30 min before being added to cells. Experiments were independently repeated thrice. One-way analysis of variance (ANOVA) was used for statistical signi cance. Data are present as "Mean±SD". * p < 0.05.