GH18 family glycoside hydrolase Chitinase A of Salmonella facilitates bacterial invasion and survival by modulating host immune responses

Salmonella is a facultative intracellular pathogen that has co-evolved with its host and has also developed various strategies to evade the host immune responses. Salmonella recruits an array of virulence factors to escape from host defense mechanisms. Previously chitinase A (chiA) was found to be upregulated in intracellular Salmonella. Although studies show that chitinases and chitin binding proteins (CBP) of many human pathogens have a profound role in various aspects of pathogenesis, like adhesion, virulence and immune evasion, the role of chitinase in strict intravacuolar pathogen Salmonella has not yet been elucidated. In this study, we deciphered the role of chitinase of Salmonella in the pathogenesis of the serovars, Typhimurium and Typhi. Our data propose that ChiA mediated modification of the glycosylation on the epithelial cell surface facilitates the invasion of the pathogen into the epithelial cells. Further we found that ChiA aids in reactive nitrogen species (RNS) and reactive oxygen species (ROS) production in phagocytes, leading to MHCII downregulation followed by suppression of antigen presentation and antibacterial responses. In continuation of the study in animal model C. elegans, Salmonella Typhi ChiA was found to facilitate attachment to the intestinal epithelium, gut colonization and persistence by downregulating antimicrobial peptides.

invasion in epithelial cells and SPI2 effectors are required for intracellular survival and 95 proliferation. Surprisingly we found that SPI1 effectors invF and hilA were significantly 96 upregulated during the early phase of infection in the ΔchiA mutant bacteria, whereas no 97 significant difference was observed in the expression of SPI2 effector ssaV after 16 hours of 98 infection ( Fig. 1E-F), suggesting that the reduced bacterial invasion in the epithelial cells by 99 ΔchiA mutant is independent of SPI1 gene expression.  by defect in SCV maintenance. We found that after 16 hours of infection, ΔchiA mutant 135 bacteria did not co-localize with the late-endosomal marker LAMP1 (Fig. 3A, S1G, S1H), 136 suggesting disruption of SCVs in the mutant bacteria infected cells. Upon counting the 137 number of SCV-bound and cytoplasmic bacterial population, we found that 81.6% of STM 138 ΔchiA and 87.2% of STY ΔchiA quit the vacuole, while very less WT bacteria (STM WT 139 12.2%, STY WT 8.2%) did not co-localize with LAMP1 ( Fig. 3B-C). We further quantified 140 the cytosolic bacterial population by chloroquine resistance assay and found significantly  phagocytic cells hints towards that this strain might be highly immunogenic. Phagocytic cells 155 are known to inhibit intracellular bacterial growth by production of reactive nitrogen species 156 and reactive oxygen species [20]. Estimation of nitric oxide produced by the infected BMDCs 157 suggested that ΔchiA mutant infected cells produced significantly less nitric oxide (Fig. 4I). 158 We further checked the survival of the WT and ΔchiA in the absence of NO using NOS2 -/-159 BMDCs and observed that WT bacteria survived similar to the ΔchiA mutants ( Fig. 4J-K), 160 suggesting that chitinase might be involved in the induction of NO in DCs. Furthermore,

161
ΔchiA infected peritoneal macrophages (PM) showed significantly less ROS level as 162 compared to WT infected cells (Fig. 4L), indicating that chitinase might be regulating RNI 163 and ROS level in the infected cells. NO is an important cell signaling molecule that is 164 produced to kill many human pathogens, such as Salmonella, Mycobacterium  be measured by incorporation of 3 H thymidine in the DNA of the proliferating population. 173 We found that with ΔchiA mutant infection, the expansion of CD8 + T cell population was 174 significantly higher in response to the antigen stimulation (Fig. 4N). Since APCs such as 175 macrophages and DCs possess MHC-I and MHC-II molecules on the cell surface in order to 176 induce both CD8 + T cells and CD4 + T cells population, respectively, we detected the surface 177 MHC-II molecules on activated PMs. We found that upon ΔchiA infection, the surface MHC-178 II level was similar to uninfected cells, while WT infection significantly reduced the surface 179 MHC-II level (Fig. 4O, S2E). Further immunofluorescence analysis verified that ΔchiA 180 infection did not change the surface MHC-II molecules on macrophages (Fig. 4M, S2F). 181 Together these data suggests that Salmonella ChiA dampens host antimicrobial responses 182 leading to enhanced pathogen survival. survival as compared to the STM WT infected cohort (Fig. 5A), suggesting a role of chitinase 190 A during infection in vivo. We also found that STM ΔchiA mutant bacteria were shed prior to 191 the STM WT and the ΔchiA mutant was defective in PP colonization after 2 hours of oral 192 gavage ( Fig. 5B-C). Further, we orally infected C57BL/6J mice with sublethal dose of 193 Salmonella strains (10 7 CFU/animal) and bacterial CFU from liver, spleen, mesenteric lymph 194 node (MLN) and PP was enumerated after sacrificing the animals after indicated time 195 intervals. We found that STM ΔchiA mutant infected animals showed less bacterial burden in 196 each of the organs and higher body weight as compared to the STM WT infected animals 197 ( Fig. 5D-H). Since our ex vivo data suggested that the ΔchiA mutant was unable to induce 198 NO production which is known to affect T cell survival, we checked whether chitinase 199 mediated NO upregulation has any effect in vivo. We also found a significant increase in the 200 spleen length after 20 days of infection with STM ΔchiA bacteria, as compared to the STM 201 WT infected mice (Fig. S3A, S3B). We isolated total splenocytes from these infected spleens 202 and checked the activated CD4 + T cell population and T cell proliferation by flow cytometry. 203 We found that STM ΔchiA infection leads to a significant increase in CD4 + T cell population, 204 as well as an increase in the activated CD25 + T cells (Fig. 5K). Analysis of T cell mediated 205 cytokine response revealed that there was a significant increase in the pro-inflammatory 206 cytokines IL2 and IFNγ, in the serum isolated from ΔchiA infected animals ( Fig. 5I-J), 207 whereas there was no difference in the anti-inflammatory cytokine levels (Fig. S3C, S3D). 208 Lesser IL2 in STM WT infected mice serum as compared to that in ΔchiA mutant infected 209 serum further strengthens the previous finding that ChiA aids in dampening of T cell 210 activation, as IL2 is a key marker of CD8 + T cell proliferation [23]. Previous reports 211 suggested that high level of IFNγ can induce B cell proliferation and enhance IgG2a and 212 IgG3 production [24]. Therefore, we looked into the role of chitinase in anti-Salmonella 213 immune response by detecting the anti-Salmonella IgG titer from infected mice serum.

214
Interestingly, we found a significant increase in the anti-Salmonella antibody titer in the 215 serum obtained from STM ΔchiA mutant infected cohort (Fig. 5L). We further used the 216 polyclonal convalescent sera isolated from STM ΔchiA infected mice to probe against STM 217 WT-mCherry whole cell lysate to test the polyclonality of the sera. Multiple dense bands 218 against various Salmonella proteins were obtained after incubating the membrane with sera 219 collected from STM ΔchiA mutant infected cohort (Fig. S3E). Together these data suggest 220 that Salmonella chitinase A is essential for restricting innate and humoral immune responses host to study S. Typhi pathogenesis [26]. Given that C. elegans pharyngeal lumen is rich in 227 chitin, it served as a suitable host to study the role of chitinase in bacterial pathogenesis [27]. 228 We began with checking the bacterial CFU in the infected worms after 24 hours and 48 hours 229 of continuous feeding. We found that the STY ΔchiA and STY ΔchiA:pQE60 strains showed 230 a higher bacterial burden after 24 hours of continuous feeding (Fig. S3F), but the fold change 231 of bacteria were lesser than that of the STY WT and STY ΔchiA:chiA strains (Fig. 6A). We using the transgenic worm FT63 strain that expresses GFP in the epithelial cells. We 239 visualized the bacterial colonization in the worms gut after 24 hours and 48 hours of 240 continuous feeding. We used STM ΔinvC mutant as a control which is known to be invasion 241 defective in nonphagocytic cells [28]. We found that S. Typhi ΔchiA showed less 242 colonization than STY WT after 24 hours continuous feeding, while the colonization was 243 significantly reduced after 48 hours feeding (Fig. 6C), suggesting chitinase is required for 244 successful gut colonization in C. elegans. Percent colonization was measured as the ratio of 245 the diameter of the lumen occupied by the bacteria to the total diameter of the gut (Fig. S3G). 246 Interestingly STM ΔinvC did not show any defect in colonization in the C. elegans gut, 247 suggesting SPI1 effector InvC is not essential for colonization in the worms gut. We next adhere to the gut lumen effectively, it will remain adhered to the gut lumen and proliferate 258 even when the worms are fed with E. coli OP50. We observed that after 24 hours of feeding 259 on STY ΔchiA followed by 24 hours feeding on E. coli OP50, the STY ΔchiA was unable to 260 persist in the gut, whereas STY WT showed significantly higher colonization in the 261 pharyngeal lumen (Fig. S3H). When we further extended the infection for 48 hours, followed 262 by 24 hours of E. coli OP50 feeding, we observed that STY WT showed profound 263 colonization of the gut lumen, while STY ΔchiA colonization was diminished (Fig. 6E), 264 suggesting Salmonella utilizes chitinase to attach to the lumen wall for enhanced persistence 265 in the worms. Interestingly, after 24 hours of continuous feeding STY WT attached to the 266 luminal wall, but not STY ΔchiA strain. (Fig. S4A). After 48 hours of continuous feeding, we Grinder, a part of the terminal bulb, is the complex structure that helps in uptake and grinding 275 of bacteria before it passes to the intestine where the nutrients get absorbed. In a healthy and 276 well-fed state, worms feed at the average rate of 200 pumps/min. Since we observed that 277 Salmonella uses chitinase to colonize chitin-rich organs (Fig. 6D), we next looked into the 278 nutritional state of the worms by counting the number of pharyngeal pumps per min. We 279 found a significant reduction in the number of pharyngeal pumps/min after 72 hours of STY 280 WT and STY ΔchiA:chiA infection (Fig. 7B). Further, in vivo oxidative stress was quantified 281 using CL2166 worms, that possess oxidative stress inducible GFP. STY WT and STY 282 ΔchiA:chiA infected worms showed significantly higher oxidative stress and 'bag of worms' 283 phenotype (Fig. S4C, S4D). We next checked the nutritional fitness of the worms by  conditions. Salmonella enterica serovar Typhi strain CT18 was used as the wild type strain, 365 and was also the parental background for the mutant strain used in this study, i.e. ΔchiA. S. 366 Typhi chiA was trans-complemented in pQE60 plasmid in 5' NotI-chiA-BamHI 3' direction.

367
This plasmid was transferred to STY ΔchiA strain to make complement strain.

368
Complemented strain STY ΔchiA:chiA and empty vector strain STY ΔchiA;pQE60 strains 369 were maintained on LBA supplemented with ampicillin (50 µg/ml). The mCherry expressing 370 strains were cultured in Lennox broth with 50µg/ml Ampicillin at 37˚C in shaking condition.

371
List of strains used in this study. sterile MiliQ water and 10% glycerol, followed by resuspension in 50 µl of 10% glycerol.

399
Kanamycin resistance cassette was amplified from pKD4 plasmid using primers containing   After infection for the specified time, the cells were fixed with 3.5% PFA for 20 min on ice. isolated from the animals after 20 days and the length was measured. analyzed the data, prepared the figures and wrote the manuscript. DC supervised the work.

592
All the authors read and approved the manuscript.  represented as mean + SEM of 3 independent experiments (N=3, n=3). One-way ANOVA