bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Host mucin is subverted by Pseudomonas aeruginosa during infection to 2 provide free glycans required for successful colonization 3 4 5 Casandra L Hoffman1 and Alejandro Aballay1* 6 7 8 9 1 Molecular Microbiology and Immunology Department, Oregon Health and Sciences 10 University, Portland, OR, United States of America 11

12

13 * Corresponding author

14 E-mail: [email protected]

1 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

15 Abstract 16 The mucosal barrier, found lining epithelial cells, serves multiple functions in a range of animals. 17 The major structural components of mucus are mucins, which are heavily glycosylated proteins 18 that are either membrane bound or secreted by the epithelial cells. Mucins are key components of 19 the innate immune system, as they are involved in the clearance of pathogens from the airways 20 and intestines, and their expression is typically upregulated upon epithelial cell exposure to a 21 variety of pathogens. In this study, we identified the mucin MUL-1 as an innate immune factor 22 that appears to be utilized by P. aeruginosa to colonize hosts. We found that while the 23 expression of several mucins, including MUL-1, increased upon P. aeruginosa infection of the 24 nematode Caenorhabditis elegans, silencing of or deletion of mul-1 resulted in enhanced 25 survival and reduced bacterial accumulation. P. aeruginosa required host sialidase CTSA-1.1 to 26 use mucin-derived glycans to colonize the host, while sialidase-encoding bacteria required host 27 MUL-1 but not CTSA-1.1 to cause a lethal infection. This role of mucins and free glycans in 28 host-pathogen interaction appears to be conserved from C. elegans to humans, as P. aeruginosa 29 binding to human lung epithelial cells was also enhanced in the presence of free glycans, and 30 free glycans reversed the binding defect of P. aeruginosa to human lung cells lacking the mucin 31 MUC1. 32

2 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

33 Author Summary 34 The gastrointestinal, respiratory, reproductive, and urinary tracts, are large surfaces exposed to 35 the exterior environment and thus, these mucosal epithelial tissues serve as primary routes of 36 infection. One of the first lines of defense present at these barriers is mucus, which is a highly 37 viscous material formed by mucin glycoproteins. Mucins serve various functions, but 38 importantly they aid in the clearance of pathogens and debris from epithelial barriers and serve 39 as innate immune effectors. In this study, we describe the ability of Pseudomonas aeruginosa to 40 utilize mucin-derived glycans to colonize the intestine and ultimately cause death in 41 Caenorhabditis elegans. We also show conserved mechanisms of P. aeruginosa virulence traits, 42 by demonstrating that free glycans alter the ability of the bacteria to bind to human lung alveolar 43 epithelial cells. Over the course of host-pathogen evolution, pathogens seem to have evolved to 44 use mucins for their own advantage, and thus one of the biggest questions is which party benefits 45 from pathogen-mucin binding. By gaining a better understanding of pathogen-mucin 46 interactions, we can better protect against pathogen infection. 47

3 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

48 Introduction. 49 The mucus layer found lining mucosal epithelial cells is a crucial first barrier against microbial 50 infections. The main structural components of the protective barrier of mucus are mucins, which 51 are high molecular weight proteins, heavily glycosylated at serine and threonine residues (1-4). 52 Mucins are secreted in large quantities by mucosal epithelial cells and both membrane-tethered 53 and secreted mucins are found on the apical surface of all mucosal epithelium. Mucins aid in the 54 formation of highly viscous, aqueous solutions that protect epithelial cells from physical damage, 55 dehydration, and infection (1,2,5). 56 Epithelial barriers have evolved multiple mechanisms to respond to environmental cues, 57 including those associated with pathogen exposure. At basal levels and in response to microbes, 58 epithelial cells secrete defensive compounds, including mucins, defensins, lysozymes, and other 59 antimicrobial compounds. Together, these compounds can form a physical barrier, have direct 60 antimicrobial activity, and aid in pathogen clearance. Mucins are able to perform all of these 61 activities. In addition to forming an impervious gel, mucus traps microbes, aids in the clearance 62 of microbes, forms a physical barrier, and provides a matrix for a rich array of antimicrobial 63 molecules (6,7). 64 Despite the various defensive properties of mucins, some bacterial pathogens are able to 65 colonize mucosal epithelial barriers (6,7) . Microbes must first penetrate the mucosal barrier to 66 either attach to epithelial cells or release toxins that disrupt the epithelial barrier (7). One of the 67 low-affinity binding mechanisms used by microbes is mediated by hydrophobic interactions with 68 lectins and glycosylated receptors. Bacterial adhesins can bind to oligosaccharides present on 69 mucins to mediate adhesion (8,9). One of the largest questions that remains to be answered is 70 whether the bacterial-mucin binding event favors the bacteria or the host. It would appear that 71 increased mucin secretion during infection would categorize mucins as a component of host 72 defenses, but the coevolution of pathogens and hosts may allow for a range of roles for mucins 73 for both parties. Some components of mucins may facilitate bacterial colonization of the host, 74 including oligosaccharides that act as adhesion sites for bacteria and individual monosaccharides 75 from mucins that can be accessed as energy sources by bacteria with mucolytic activity (10). 76 Finally, mucus components can influence the virulence characteristics of pathogenic bacteria, 77 such as virulence factor expression, adhesion, motility, proliferation, and/or growth (11-15).

4 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

78 Alterations in mucin expression or glycosylation patterns have been linked to various 79 pathologies and diseases, including cystic fibrosis, chronic obstructive pulmonary disease, 80 cancers, and inflammatory bowel disease (1). These diseases, that are linked to changes in mucin 81 expression, glycosylation patterns, and alterations in mucus levels, are associated with higher 82 rates of infection with opportunistic pathogens (2,16,17). Increases in mucus levels often result 83 in a more highly viscous material that is not readily cleared from the mucosal barrier. Decreases 84 in mucus levels remove the protective barrier, which allows pathogens to have direct access to 85 epithelial cells. Many of the diseases linked to alterations in mucus levels have no known cure, 86 and the primary mode of treatment involves controlling mucin expression, which is a key method 87 used to prevent bacterial infections associated with these diseases (16,18). A better 88 understanding of the mechanisms by which pathogens utilize mucins will provide novel 89 approaches to control infectious processes and prevent bacterial colonization. 90 The complexity of mammalian epithelial mucosal surfaces and the additional functions of 91 the innate immune system can mask the roles of individual mucins at epithelial barriers. Using 92 the model organism, Caenorhabditis elegans, we can tease apart the roles of individual mucins at 93 the intestinal epithelial barrier. This model provides the advantage of a simple organism’s small 94 genome size and the absences of adaptive immunity. C. elegans eats bacteria found in 95 decomposing organic matter (19). Several pathogens are present in the environment in which C. 96 elegans feeds, which can impair the growth of the nematode and induce stress responses, 97 ultimately leading to nematode death (20). The ability to distinguish between pathogenic and 98 non-pathogenic microbes is critical for the survival of the nematodes and thus, C. elegans has 99 methods to recognize and counteract pathogens. One of these responses is the upregulation of 100 mucins during infection (21,22). 101 Herein we explore the role of mul-1 during pathogen infection and find that pathogens 102 use the immune factor to gain a competitive advantage in the host. We identified mul-1 as a 103 mucin, that when inhibited, enhances resistance to P. aeruginosa and S. enterica infection. This 104 is despite the fact that mul-1 gene expression is induced upon infection by these pathogens as 105 part of the general immune response. We further explored the role of the mucin during infection 106 and found that MUL-1 provides glycans that are required for attachment to C. elegans intestine 107 and that a similar mechanism is conserved in mammalian epithelial cells. These results indicate

5 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

108 that pathogens have evolved a mechanism to subvert conserved immune effectors to colonize 109 different hosts. 110 111 Results. 112 C. elegans lacking mul-1 exhibit enhanced resistance to P. aeruginosa infection. 113 Little is known about the roles of specific mucins in the C. elegans intestine during infection, but 114 several enzymes and transporters that modify the glycan patterns of mucins have been identified 115 to alter host-pathogen interactions. We set out to better understand the roles of specific intestine- 116 expressed mucins during P. aeruginosa infection. The six identified mucins and mucin-editing 117 enzymes (Table S1) were silenced by RNAi and the survival of the animals was monitored 118 during P. aeruginosa PA14 infection. Upon analysis, RNAi for two genes, mul-1 and gpdh-1, 119 resulted in enhanced resistance to pathogen infection and RNAi for one mucin, let-653, resulted 120 in enhanced susceptibility to pathogen infection (Table S1 and Figure S1). 121 Because mucins are upregulated upon infection and have been characterized as bona fide 122 immune effectors, the enhanced resistance phenotype for two of the genes was unexpected. The 123 most significant phenotype was observed for mul-1 RNAi animals, which showed enhanced 124 resistance to P. aeruginosa compared to RNAi control animals (Figure 1A). mul-1 mRNA 125 transcripts are found at high levels in the intestine (23) and it was previously shown that mul-1 is 126 upregulated upon infection with P. aeruginosa at 4 and 8 hours (21,22). Other upregulated genes 127 from those datasets include innate immune factors, which would suggest that mul-1 is an 128 immune response factor turned on to combat pathogen infection. 129 130 The enhanced resistance of mul-1 RNAi animals to P. aeruginosa is due to lack of bacterial 131 colonization. 132 Because mul-1 RNAi results in delayed death upon P. aeruginosa infection, we aimed to better 133 understand the mechanism by which mul-1 contributes to enhanced resistance to the pathogen. 134 We tested the ability of P. aeruginosa to colonize the nematode intestine. RNAi for mul-1 135 resulted in reduced accumulation of bacteria, determined by obtaining P. aeruginosa Colony 136 Forming Units (CFUs) from infected animals over the course of infection and measuring the 137 fluorescence intensity in the intestines of animals infected with P. aeruginosa expressing GFP. 138 Initially, RNAi mul-1 animals accumulated P. aeruginosa at a similar rate as control animals, but

6 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

139 there were significantly fewer CFUs recovered per nematode after 4 hours (Figure 1B) and little 140 to no fluorescent signal was found in RNAi mul-1 animals compared to control RNAi animals at 141 24 hours post infection (Figure 1C). To determine if the enhanced resistance to P. aeruginosa 142 infection is a general phenotype in the mul-1 RNAi animals, we exposed these animals to 143 Salmonella enterica ser Typhimurium ST1344 and monitored survival. The enhanced resistance 144 phenotype was also observed when the RNAi mul-1 nematodes were exposed to full lawns of S. 145 enterica. RNAi mul-1 nematodes lived longer upon exposure to S. enterica than RNAi control 146 nematodes (Supplementary Figure 1A) and accumulated fewer S. enterica bacteria 147 (Supplementary Figure 1B and 1C). 148 Because of the increase in survival upon pathogen exposure observed in mul-1 RNAi 149 animals, we studied if differences in the lifespan of the mul-1 RNAi animals may be responsible 150 for the enhanced survival in the presence of P. aeruginosa. As shown in Figure 1D, the longevity 151 of mul-1 RNAi animals was indistinguishable from that of control animals, ruling out that the 152 possibility that enhanced longevity was the cause of enhanced survival in the presence of 153 pathogenic bacteria. To confirm the roles of mul-1 during infection, an in-frame, full gene 154 deletion strain was used. The mul-1(ac7) strain also showed enhanced resistance to P. 155 aeruginosa (Figure 1E), significantly fewer CFUs per nematode (Figure 1F), and there was no 156 difference in lifespan in comparison to wildtype N2 nematodes (Figure 1G). 157 Expression of mul-1 has been reported in the intestine, hypodermis, and PVD and OLL 158 neurons (23,24). To address whether intestinal mul-1 is specifically contributing to infection, 159 intestine-specific RNAi was performed using two intestine-specific RNAi strains, VP303 (25) 160 and MGH171 (26). In both cases, we found that mul-1 RNAi in the intestine resulted in enhanced 161 resistance to P. aeruginosa (Figure 2A and 2B). 162 One of the possible explanations for the mul-1 RNAi animals’ enhanced resistance to P. 163 aeruginosa could be due to the increased expression of other intestinal-expressed mucins that 164 compensate for the loss of MUL-1. To rule out the possibility that other mucins are upregulated 165 to compensate for the loss of mul-1, qRT-PCR was performed on RNAi control vector and RNAi 166 mul-1 nematodes. When the RNAi mul-1 nematodes were grown on non-pathogenic E. coli, 167 there were some statistically significant increases in expression levels of the intestine-expressed 168 mucins and mucin-related enzymes compared to RNAi control nematodes. (Figure 3). Many of 169 the significant increases found when the nematodes were grown on E. coli were no longer

7 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

170 observed in mul-1 RNAi nematodes grown on P. aeruginosa compared to control animals also 171 exposed to P. aeruginosa. We don’t think these less than threefold increases could account for 172 the enhanced resistance phenotype of the mul-1 RNAi animals. 173 We have shown that mul-1 RNAi animals are not colonized by P. aeruginosa during 174 infection. Recent work shows that colonization by P. aeruginosa and other pathogens causes 175 bloating of the intestine that results in the expression of innate immune effectors and the 176 elicitation of a pathogen avoidance behavior (27). Because the mul-1 mutant has limited bacterial 177 accumulation in the intestine, we hypothesized that these animals would not bloat and thus 178 would not avoid P. aeruginosa. C. elegans strains were exposed to partial lawns of P. 179 aeruginosa PA14 and lawn occupancy was characterized at 24 hours. Consistent with the 180 previous study, over 80% of the wild type N2 animals were outside of the P. aeruginosa PA14 181 lawn, demonstrating that wild type animals avoid P. aeruginosa (Figure 4A). In comparison, 182 only 16.8% of the mul-1 (ac7) animals were outside of the bacterial lawn (83.2% lawn 183 occupancy), suggesting that the mutant animals do not avoid P. aeruginosa. Because the mul-1 184 (ac7) animals do not avoid, it would be presumed that the animals have enhanced ingestion of P. 185 aeruginosa and die more rapidly than wild type animals. Although the mul-1 (ac7) animals don’t 186 avoid P. aeruginosa, they still demonstrate an enhanced resistance to the pathogen in comparison 187 to wild type when survival was monitored on partial lawns of P. aeruginosa (Figure 4B). 188 189 Mucin-derived glycans are required for successful P. aeruginosa infection. 190 Based on our data that shows mul-1 (ac7) and mul-1 RNAi animals are not colonized by P. 191 aeruginosa, we hypothesized that MUL-1 plays a beneficial role for pathogens in the host. This 192 is contrary to the idea that MUL-1 is a C. elegans innate immune effector used by the nematode 193 to fight and clear infections. To better understand the role of MUL-1 inside the C. elegans 194 intestine, we probed whether P. aeruginosa is able to access glycans from MUL-1 in the C. 195 elegans intestine. 196 In certain cases, pathogens have been reported to utilize glycans such as O-linked 197 oligosaccharides from mucins (28-30), but in order to expose the host glycans, bacteria require a 198 sialidase enzyme to cleave the sialic acid cap from host polysaccharides. P. aeruginosa does not 199 encode a bacterial sialidase, thus we questioned if C. elegans expresses an enzyme that performs 200 a similar function to process polysaccharides for P. aeruginosa use. C. elegans encodes CTSA-

8 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

201 1.1, a homolog to human cathepsin A, which is predicted to have serine-type carboxypeptidase 202 activity. This enzyme is predicted to have neuraminidase activity, which hydrolyzes terminal 203 sialic acid residues on polysaccharide chains, most often exposing a galactose residue. The 204 neuraminidase is expressed in the intestine, but does not change expression upon pathogen 205 infection [21,31,32]. To determine if the neuraminidase contributes to the enhanced resistance to 206 P. aeruginosa infection, RNAi for ctsa-1.1 was performed. RNAi ctsa-1.1 significantly enhanced 207 resistance, but not to the same extent as mul-1 RNAi (Figure 5A). Co-RNAi for both mul-1 and 208 ctsa-1.1 was not additive, which suggests that these two gene products contribute to the same 209 mechanism by which P. aeruginosa establishes an infection and kills C. elegans. There was no 210 difference in PA14 accumulation in the RNAi mul-1, RNAi ctsa-1.1, or co-RNAi, but all three 211 conditions resulted in significantly fewer CFUs per nematode compared to RNAi control (Figure 212 5B). To confirm that the enhanced resistance to P. aeruginosa phenotype was not due to 213 enhanced longevity, we studied the mul-1 RNAi, ctsa-1.1 RNAi, and co-RNAi nematode 214 lifespan. There were no differences in lifespan under all RNAi conditions (Figure 5C). 215 Because the mul-1 RNAi enhanced resistance phenotype was observed for both P. 216 aeruginosa and S. enterica, we asked if both pathogens require the C. elegans neuraminidase, 217 ctsa-1.1. Using the RNAi conditions above, nematodes were monitored for survival on S. 218 enterica. As we have shown earlier, RNAi mul-1 animals display an enhanced resistance 219 phenotype when infected with S. enterica, but the RNAi ctsa-1.1 animals are as susceptible to S. 220 enterica as RNAi control, and accumulate as many S. enterica per nematode as control RNAi 221 animals. (Supplementary Figure 2A-C). S. enterica encodes and expresses functional bacterial 222 sialidases that allow the bacteria to access glycan molecules from mucins (31). Thus, the C. 223 elegans enzyme may not be required by S. enterica during infection. The use of the C. elegans 224 enzyme, CTSA-1.1 appears to be specific to P. aeruginosa, which lacks a sialidase. 225 To further elucidate the mechanism(s) by which mucins benefit pathogens during 226 infection, we tested if individual purified free glycans had any effect on bacterial growth. We 227 also wanted to determine if free glycans had any impact on C. elegans survival on P. aeruginosa. 228 Based on previous studies that aimed to understand the impact of mucin glycosylation patterning 229 on development and infectious disease, we identified several o-linked glycans that may be 230 important for bacterial binding and or growth in vivo in the nematode. These included N- 231 acetylgalactosamine, galactose, N-acetylglucosamine, and N-acetyl neuraminic acid, which is the

9 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

232 predominant sialic acid found in mammalian mucin glycan chains (32-34). Commercially 233 available free glycans were added to Luria Broth (LB) to assess their effect on PA14 growth. 234 There were no significant changes to bacterial growth in LB (Supplemental Figure 3A). We also 235 tested the effects of these free glycans in a minimal bacterial growth media, M9 media, because 236 LB media is rich and provides ample carbon source(s) for the bacteria, and in LB media we may 237 not observe any additional enhancement to growth. None of the free glycans had any effect on 238 bacterial growth in M9 media over the course of 48 hours. In fact, we observed little to no 239 growth of the bacteria in M9 media or M9 media supplemented with glycans (Supplemental 240 Figure 3B), and it was not until 72 hours that we observed statistically enhanced growth of the P. 241 aeruginosa PA14 grown in M9 with 10 and 20 mM N-acetyl-D-glucosamine. 242 Upon determining that various concentrations of these free glycans had no major effect 243 on bacterial growth in vitro, we assessed the effects of free glycans during infection. Free 244 glycans were solubilized in water and added at a concentration of 20 mM to plates to evaluate 245 the role of glycans during infection. The addition of N-acetyl-glucosamine drastically decreased 246 the survival of RNAi control and RNAi mul-1 C. elegans survival on P. aeruginosa PA14 247 (Figure 6A). This was expected, based upon previous studies which have shown that N-acetyl-D- 248 glucosamine can alter pathogen virulence characteristics and increase virulence factor 249 expression, specifically pyocyanin (35,36). N-acetyl neuraminic acid (Figure 6B) and D- 250 galactose (Figure 6C) had non-significant effects on RNAi control or RNAi mul-1 nematode 251 survival. N-acetyl-D-galactosamine supplementation fully restored RNAi mul-1 nematode 252 survival to RNAi control survival and had no effect on RNAi control nematode survival (Figure 253 6D), which suggests that P. aeruginosa specifically utilizes N-acetyl-D-galactosamine from 254 MUL-1 during infection. 255 To determine whether the pathogens must be exposed to the free glycans during 256 infection, or whether the free glycans cause a long-lasting increase in virulence of the pathogens, 257 bacteria were first cultured in the presence of free glycans at concentrations used on the plates 258 used in killing assays. Bacteria were then collected and used to seed full lawn plates, on which 259 survival for the RNAi control and RNAi mul-1 nematodes was tested. Addition of the free 260 glycans, N-acetyl-D-glucosamine (Figure 6E) and N-acetyl-D-galactosamine (Figure 6F), to the 261 media prior to seeding had no effect on nematode survival on PA14 lawns. This suggests that the 262 free glycans must be present and possibly absorbed by the bacteria as entering the nematode, or

10 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

263 absorbed by the nematodes, so that the bacteria have access to the glycans once inside the 264 nematode intestine. 265 266 Free glycans alter binding and internalization of P. aeruginosa to human lung epithelial 267 cells. 268 Mucins play conserved roles in maintaining homeostasis and providing protection from infection 269 at epithelial barriers. Human lung epithelial cells express mucins, and some of these have been 270 implicated in the binding and internalization of P. aeruginosa into various lung cell lines. Human 271 and mammalian mucins contain many of the same O-linked oligosaccharides; thus, we 272 hypothesized that the free glycans identified in the C. elegans system would have similar effects 273 on altering the binding to and cell death of human lung cells. 274 Free glycans, N-acetyl-D-glucosamine and N-acetyl-D-galactosamine, were tested for 275 their effects on bacterial growth in cell culture media. A range of concentrations of both of the 276 glycans were added to media and bacterial growth was monitored over 24 hours. Within the first 277 4 hours, the glycans had no significant effect on bacterial growth (Supplemental Figure 4A and 278 4B). Bacterial binding to lung cells was quantified at time points during which there were no 279 significant changes in bacterial growth. Human lung alveolar epithelial cells, A549 cells, were 280 used to assess bacterial binding in the presence of free glycans because these cells express 281 MUC1 and MUC5a mucins based upon mRNA transcript profiling (37,38). We reasoned that the 282 use of this cell culture model system, with only 2 mucins, would make it simpler to distinguish 283 which specific human mucin is involved in the binding phenotypes observed. The A549 human 284 lung alveolar cells were seeded and grown for approximately 24 hours until 80% confluency. 2 285 mM free glycans were added to cultures at the same time as bacteria were added. Mid log-phase 286 P. aeruginosa PA14 were added to cell cultures at a Multiplicity of Infection (MOI) of 100 and 287 at 4 hours, bacterial binding was assessed. Both N-acetyl-D-glucosamine and N-acetyl-D- 288 galactosamine significantly increased bacterial adherence (Figure 7A). 289 To determine if the free glycans are required to be processed by the cells prior to use by 290 P. aeruginosa, 2 mM free glycans were added to A549 Lung cells at the time of seeding and 291 remained present for the ~24 hours prior to bacteria were added. P. aeruginosa was again added 292 at an MOI of 100 to cells and bacterial binding was measured at 4 hours. When glycans were

11 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

293 added at the time of cell seeding, we saw that bacterial binding (Figure 7C) was increased for 294 cells supplemented with both N-acetyl-D-glucosamine and N-acetyl-D-galactosamine. 295 We have shown that free glycans, specifically N-acetyl-D-galactosamine, can supplement 296 for MUL-1 during P. aeruginosa infection of C. elegans. In order to determine if these free 297 glycans are able to supplement for mucins in human lung cells, siRNA silencing (lentiviral 298 transfection) was used to knock down expression of human muc1 in A549 human lung cell 299 model. Knockdown did not result in any growth defects of the cells, but it reduced the number of 300 bacteria bound per cell (Figure 8). This is similar to observations in C. elegans, suggesting that 301 mucins and glycans play similar roles in different models of bacterial pathogenesis. Thus, as 302 expected, when N-acetyl-D-glucosamine and N-acetyl-D-galactosamine were added to control 303 and muc1 siRNA knockdown they enhanced bacterial binding. Both free glycans increased 304 binding of P. aeruginosa to the control A549 lung cells when the glycans were added to the 305 media at the same time as the bacteria (Figure 8). The free glycans also reversed the binding 306 defect of P. aeruginosa to muc1 siRNA knockdown cells, restoring the number of bacteria per 307 cell to that of control cells in media alone (Figure 8). 308 309 Discussion. 310 This work details the ability of the pathogen P. aeruginosa to use mucin-derived glycans to 311 successfully colonize mucosal epithelial barriers. Our data shows that silencing or deleting a 312 pathogen-induced immune effector gene, mul-1, results in enhanced resistance to P. aeruginosa. 313 These results were unexpected because we and others have shown that mul-1 expression is 314 increased upon pathogen infection as part of a general immune response (21,22). Understanding 315 the mechanisms by which P. aeruginosa binds to and obtains resources from various mucins 316 may provide a better understanding of host-pathogen interactions and allow for the development 317 of targeted therapeutics. 318 P. aeruginosa is able to utilize mucin-derived glycans to colonize the host, which we 319 have demonstrated by adding free glycans to C. elegans growth media during infection. We have 320 shown that one specific glycan, N-acetyl-D-galactosamine, had no effect on control nematodes, 321 but rescued the enhanced resistance phenotype of the mul-1 RNAi nematodes. This phenomenon 322 appears to be conserved from C. elegans to humans, as P. aeruginosa binding to human lung

12 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

323 epithelial cells was also enhanced in the presence of free glycans, and free glycans reversed the 324 binding defect of P. aeruginosa to muc1 siRNA human lung cells. 325 In addition to enhanced resistance to P. aeruginosa infection, resistance to S. enterica is 326 enhanced upon silencing of mul-1, which suggests that the use of mucin-derived glycans may be 327 a conserved mechanism by a variety of pathogens that infect mucosal epithelial barriers. In order 328 to fully understand the specific role that mul-1 plays for pathogens during infection, we tested the 329 ability of P. aeruginosa and S. enterica to access free glycans from MUL-1 during infection. In 330 order for bacteria to access glycans from mucins, pathogens must first enzymatically cleave the 331 sialic acid cap from the end of polysaccharide chains. Both pathogens have been reported to use 332 free glycans as carbon sources, but only S. enterica encodes and expresses a bacterial sialidase 333 (28,30). C. elegans encodes an enzyme, ctsa-1.1, that is predicted to perform a similar enzymatic 334 function as bacterial sialidases, has a predicted secretion signal, and is expressed in the intestine. 335 We have shown that RNAi silencing of ctsa-1.1 results in enhanced resistance to P. aeruginosa 336 but not enhanced resistance to S. enterica. This suggests that S. enterica is able to use its 337 bacterial-encoded sialidase to gain access to glycans, but P. aeruginosa cannot access glycans 338 without the C. elegans enzyme. It is possible that other intestinal mucins serve similar roles in 339 binding and acting as a carbon source for pathogens, which may depend upon the composition of 340 the glycan chains of the mucin and the specific use of the glycans by the pathogens. 341 One of the most interesting implications of our data is that while mul-1 expression is 342 enhanced upon C. elegans infection with P. aeruginosa and S. enterica, the animals would 343 apparently be at an advantage without mul-1 during infection with P. aeruginosa and S. enterica. 344 The bacteria are able to access the glycans present on the glycan chains by using either host- 345 expressed or bacterial expressed sialidase enzymes. These glycans aid in the colonization of the 346 host, suggesting that despite the fact that MUL-1 in an innate immune response factor, pathogens 347 have evolved to use this mucin and the glycans that decorate the peptide. The ability to control a 348 pathogen’s access to mucins and the provided glycans could prove to be a method to prevent 349 and/or disrupt infection. In certain disease states, opportunistic pathogens are able to colonize 350 and establish an infection when there are changes in the levels of mucus and expression levels of 351 mucins. In some cases, only the glycosylation patterns of mucins change and this provides 352 enhanced binding for pathogens. If there were therapeutics available to prevent binding of

13 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

353 pathogens to mucin oligosaccharides or to control mucin expression, this would be extremely 354 beneficial in treating disease and could prevent opportunistic pathogen infection. 355 356 Acknowledgements 357 None.

14 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

358 Methods 359 Bacterial Strains 360 The following bacterial strains were used: Escherichia coli OP50, E. coli HT115 (DE3), 361 Pseudomonas aeruginosa PA14, P. aeruginosa PA14-GFP, Salmonella enterica serovar 362 Typhimurium 1344. All strains were grown on Luria-Bertani (LB) agarose plates or LB broth at 363 37ºC shaking at 250 RPM. 364 365 C. elegans Strains and Growth Conditions 366 C. elegans hermaphrodites were maintained on E. coli OP50 at 20ºC unless otherwise indicated. 367 Bristol N2 was used as the wild-type control strain. mul-1 (ac7) CRISPR/Cas9 ~1650 bp deletion 368 mutant of isoform A and B (565bp and 952bp deleted, with generated termination codon); 369 predicted truncated protein has 46 amino acids. Gut-sensitive RNAi lines: MGH171 alxIs9 [vha- 370 6p::sid-1::SL2::GFP]; kbIs7 [nhx-2p::rde-1 + rol-6(su1006)]. -- were obtained from the 371 Caenorhabditis Genetics Center (University of Minnesota, Minneapolis, MN). 372 373 RNA Interference (RNAi) 374 RNAi was used to generate loss-of-function RNAi phenotypes by feeding nematodes E. coli 375 strain HT115 (DE3) expressing double stranded RNA (dsRNA) homologous to a target gene (39- 376 41). RNAi was carried out as described previously (42). Briefly, E. coli with appropriate vectors 377 were grown in LB broth containing ampicillin (100 µg/mL) and tetracycline (12.5 µg/mL) at 378 37ºC overnight and plated onto NGM plates containing 100 mg/mL ampicillin and 6 mM 379 isopropyl b-D-thiogalactoside (IPTG) (RNAi plates). RNAi-expressing bacteria were allowed to 380 grow overnight at 37ºC. Gravid adults were transferred to RNAi-expressing bacterial lawns and 381 allowed to lay eggs for 8 hours. The gravid adults were removed, and the eggs were allowed to 382 develop at 20ºC to young adults for subsequent assays. unc-22 RNAi was included as a positive 383 control to account for the RNAi efficiency. All RNAi clones except mul-1 were from the 384 Ahringer RNAi library (Open BioSource). The mul-1 clone was obtained from the Vidal RNAi 385 library (Open BioSource). 386 387 Killing Assay, Survival on Pathogens

15 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

388 Bacterial lawns were prepared by inoculating individual bacterial colonies into 2 mL of LB with 389 50 mg/mL kanamycin and growing them for 7-8 hours on a shaker at 37ºC. For the colonization 390 assays, bacterial lawns of P. aeruginosa or S. enterica ser Typhimurium were prepared by 391 spreading 35 µL of the culture over the complete surface of 3.5-cm-diameter modified NGM 392 agar plates (3.5% instead of 2.5% peptone). Young adult animals were transferred to full lawns 393 of P. aeruginosa PA14 or S. enterica ser Typhimurium and nematode survival was monitored 394 daily. Animals were considered dead upon failure to respond to touch. Animals missing from the 395 agar plate were censored on the day of loss. The KaplanMeier method was used to calculate the 396 survival fractions, and statistical significance between survival curves was determined using the 397 log-rank test. 398 399 Avoidance Assay on Pathogens 400 The bacterial lawns were prepared by inoculating individual bacterial colonies into 2 mL of the 401 corresponding broth mentioned above and growing them for 7-8 hours on a shaker at 37ºC. 20 402 µL of the culture was plated onto the center of a 3.5-cm plate and incubated at 37ºC for 12-16 403 hours. For P. aeruginosa PA14, modified NGM (3.5% instead of 2.5% peptone) plates were 404 used. Thirty synchronized young gravid adult hermaphroditic animals grown on E. coli HT115 405 (DE3) containing control vector or an RNAi clone targeting a gene were transferred outside the 406 bacterial lawns. The numbers of animals on and off the lawns were counted at the indicated 407 times for each experiment. Three 3.5-cm plates were used per trial in each experiment. 408 Experiments were performed at 25ºC. The percent occupancy was calculated as (# on lawn/# 409 total). At least three independent experiments were performed. 410 411 P. aeruginosa-GFP Colonization Assay 412 Bacterial lawns were prepared by inoculating individual bacterial colonies into 2 mL of LB with 413 50 mg/mL kanamycin and growing them for 7-8 hours on a shaker at 37ºC. For the colonization 414 assays, bacterial lawns of P. aeruginosa-GFP were prepared by spreading 35 µL of the culture 415 over the complete surface of 3.5-cm-diameter modified NGM agar plates (3.5% instead of 2.5% 416 peptone). The plates were incubated at 37ºC for 12-16 hours and then cooled to room 417 temperature for at least 1 h before seeding with young gravid adult hermaphroditic animals. The 418 assays were performed at 25ºC. At the indicated times for each experiment, the animals were

16 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

419 transferred from P. aeruginosa-GFP plates to fresh E. coli OP50 plates and visualized within 5 420 minutes under a fluorescence microscope. 421 422 Quantification of Intestinal Bacterial Loads 423 P. aeruginosa-GFP lawns were prepared as described above. For quantification of colony 424 forming units (CFU) at various timepoints, bacterial lawns of P. aeruginosa-GFP were prepared 425 by spreading 35 µL of the culture over the complete surface of 3.5-cm-diameter modified NGM 426 agar plates (3.5% instead of 2.5% peptone). The plates were incubated at 37ºC for 12-16 hours 427 and then cooled to room temperature for at least 1 hour before seeding with young adult 428 hermaphroditic animals. The assays were performed at 25ºC. At indicated times for each 429 experiment, the animals were transferred from P. aeruginosa-GFP plates to the center of fresh E. 430 coli plates for 30 min to eliminate P. aeruginosa-GFP stuck to their body. Animals were 431 transferred again to the center of a new E. coli plate for 30 additional minutes to further eliminate 432 external P. aeruginosa-GFP. Afterward, ten animals/condition were transferred into 50 µL of 433 PBS plus 0.01% Triton X-100 and ground using glass beads. Serial dilutions of the lysates (10-1, 434 10-2, 10-3, 10-4, 10-5) were seeded onto LB plates containing 50 mg/mL of kanamycin to select for 435 P. aeruginosa-GFP cells. Plates were incubated overnight at 37ºC. Single colonies were counted 436 the following day and are represented as the number of bacterial cells or CFUs per animal. Three 437 independent experiments were performed for each condition. 438 439 Fluorescence Imaging 440 Fluorescence imaging was carried out as described previously (42) . Briefly, animals were 441 anesthetized using an M9 salt solution containing 30 mM sodium azide and mounted onto 2% 442 agar pads. The animals were then visualized using a Leica 443 M165 FC fluorescence stereomicroscope. 444 445 RNA Isolation and Quantitative Reverse Transcription-PCR (qRT-PCR) 446 Animals were synchronized by egg laying. Approximately 35 N2 gravid adult animals were 447 transferred to 10-cm RNAi plates seeded with E. coli HT115 (DE3) expressing the appropriate 448 vectors and allowed to lay eggs for 8 hours. Gravid adults were then removed, and the eggs were 449 allowed to develop at 20ºC. For gene expression analysis in the P. aeruginosa infection assays,

17 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

450 the animals were first grown on 10-cm RNAi plates seeded with E. coli HT115 expressing either 451 the empty vector RNAi control or mul-1 RNAi until the young adult stage. Subsequently, the 452 animals were collected, washed with M9 buffer, and transferred to 10-cm modified NGM plates 453 (3.5% instead of 2.5% peptone) seeded with 300 µL of P. aeruginosa culture grown overnight. 454 The P. aeruginosa plates were prepared by spreading 300 µL of the P. aeruginosa culture on the 455 surface of the modified NGM plates, followed by an overnight incubation at 37ºC. After transfer 456 of the animals, the P. aeruginosa plates were incubated at 25ºC for 8 hours. Animals on the 457 control E. coli plates were also incubated at 25ºC. After the desired treatment, the animals were 458 collected, washed with M9 buffer, and frozen in TRIzol reagent (Life Technologies, Carlsbad, 459 CA). Total RNA was extracted using the RNeasy Plus Universal Kit (Qiagen, Netherlands). 460 Residual genomic DNA was removed using TURBO DNase (Life Technologies, Carlsbad, CA). 461 A total of 6 µg of total RNA was reverse-transcribed with random primers using the High- 462 Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA). qRT-PCR 463 was conducted using the Applied Biosystems One-Step Real-time PCR protocol using SYBR 464 Green fluorescence (Applied Biosystems) on an Applied Biosystems 7900HT real-time PCR 465 machine in 96-well-plate format. 25 µL reactions were analyzed as outlined by the manufacturer 466 (Applied Biosystems). The relative fold-changes of the transcripts were calculated using the 467 comparative CT(2-∆∆CT) method and normalized to pan-actin (act-1, -3, -4). The cycle thresholds 468 of the amplification were determined using StepOnePlus software (Applied Biosystems). All 469 samples were run in triplicate. The primer sequences are available in Table S1. 470 471 C. elegans Longevity Assays, Cultivation of C. elegans on Heat-Killed E. coli OP50 472 A single colony of E. coli OP50 was inoculated in 100 mL of LB broth in a 500 mL Erlenmeyer 473 flask and incubated at 37ºC at 225 rpm shaking for 24 hours. Bacteria were concentrated 20 474 times and heat-killed at 100ºC for 1 hour. Bacterial death was confirmed by failure to grow on 475 LB plates at 37ºC overnight. The concentrated, heat-killed bacteria were seeded on NGM plates 476 containing 50 mg/mL of kanamycin and 100 mg/mL of streptomycin. Young adult wild-type N2 477 animals grown on E. coli HT115 RNAi control or target gene RNAi plates were washed with M9 478 medium and incubated at room temperature for 1 hour with M9 medium containing 50 mg/mL of 479 kanamycin to remove live bacteria from their intestinal lumen. The animals were then washed 480 with M9 medium and transferred to NGM plates containing heat-killed E. coli OP50 and

18 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

481 incubated at 20ºC for the duration of the assay. Remaining animals were transferred when the 482 heat-killed E. coli OP50 lawn was reduced. 483 484 C. elegans Longevity Assays, Cultivation of C. elegans on Heat-Killed E. coli HT115 (DE3) 485 RNAi 486 Lifespan assays were performed on RNAi plates containing E. coli HT115 (DE3) with the 487 appropriate vector. Animals were synchronized on RNAi plates and incubated at 20ºC. At the L4 488 larval stage, the animals were transferred onto the corresponding RNAi plates. Animals were 489 scored on a daily basis as alive, dead, or gone. Animals that failed to display touch-provoked 490 movement were scored as dead. Experimental groups contained 60 to 100 animals. Young adult 491 animals were considered as day 0 for the lifespan analysis. The assays were performed at 20ºC. 492 Three independent experiments were performed. 493 494 Bacterial Growth Assays 495 Individual bacterial colonies were inoculated into 2 mL of LB and grown overnight, shaking at

496 225 RPM at 37ºC. Overnight cultures were diluted to an OD570 of 0.05 in either LB media, M9 497 Media, or DMEM F12+K + 10% Heat-Inactivated FBS. 100 µL bacterial cultures were placed in 498 individual wells of a 96 well plate with various glycans added at indicated concentrations.

499 Bacterial growth was monitored over time by measuring the OD570 at indicated time points. 500 501 Cell Lines 502 The A549 human type II alveolar epithelial cell line (ATCC # CCL-185, provided by David 503 Lewinsohn’s lab at Oregon Health and Science University); passage numbers 3-10 were used. 504 A549 cells were maintained in Ham's F-12K (Kaighn's) Medium (Gibco 21127022) 505 supplemented with 10% heat-inactivated fetal bovine serum (SAFC Biosciences, Lenexa, KS)

506 without antibiotics. Cells were grown at 37ºC with 5% CO2 and seeded every 4 days when 507 confluency was approximately 85%. 508 509 Lentiviral Transduction with MISSION TRC shRNA Lentivirus Particles 510 The MISSION TRC shRNA Lentiviral Transfection protocol was used. 1 x 106 A549 cells were

511 plated in 60 mm dishes and grown 20 hours at 37ºC 5% CO2 to ~60-70% confluency. Media was

19 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

512 replaced without 8 µg/mL Hexadimethrine bromide. An MOI of ~0.5 (~13 µL of stock 513 lentivirus) was added to plates and incubated for 6 hours. Media was replaced without antibiotics 514 for 24 hours. The following day, media was replaced with medium containing 5 µg/mL 515 puromycin. Media was replaced with fresh puromycin containing media every 3-4 days until 516 resistant colonies grew to confluency. Cells were then used for indicated assays. 517 518 Gentamycin Protection Assay – Bacterial binding to cells 519 The gentamycin protection assay protocol was used, as previously described (43), with some 520 modifications (44). A549 cells were grown in 96-well plates with or without free glycans present

521 for 2 days at 37ºC, 5% CO2 until they reached approximately 85% confluency, corresponding to 522 1 x 105 cells per well. Wells were washed four times with sterile PBS and serum-free DMEM 523 F12 was added for 2 h. P. aeruginosa PA14 was grown overnight in LB at 37ºC shaking, diluted 524 1:10 and allowed to grow under the same conditions for 4 hours (obtaining mid-log phase 525 bacteria). A549 cells were infected with P. aeruginosa PA14 at an MOI of 100 and incubated for

526 2 hours at 37ºC, 5% CO2. From the original bacterial suspension, serial dilutions were plated 527 were prepared to verify the starting concentration of P. aeruginosa. Following incubation, the 528 supernatants were removed, the wells were washed four times with PBS, and the numbers of 529 associated or internalized bacteria were assessed as follows: 30 µL of 0.05% trypsin-EDTA was 530 added to each well, incubated for 2 min at 37ºC. Cells were lysed with 70 µL of 0.1% Triton X- 531 100 for 2 min at 37ºC. Lysates were removed and serial dilutions were plated. Cell viability was 532 measured using Promega Cell-Titre Glo 2.0 reagent (quantitates cellular ATP levels) and CFUs 533 were normalized per cell under each condition. These results represent the number of both 534 Attached and Intracellular bacteria (CFU / # of cells). To quantify internalized bacteria, after the 535 4 hour incubation with P. aeruginosa PA14 and four PBS washes, 200 µg/mL gentamicin sulfate 536 in 50 µL of medium was added to each well and incubated for 90 minutes to eliminate 537 extracellular bacteria. After the incubation, supernatants were removed, and serial dilutions were 538 plated to quantitate the number of extracellular P. aeruginosa PA14 that may have survived 539 gentamycin treatment. A549 cells were then washed four times with PBS, trypsinized, and lysed 540 as described above; serial dilutions were performed and plated onto LB agar plates. Again, CFUs 541 were normalized to viable cells. The results represent the number of intracellular bacteria (CFU / 542 # of cells).

20 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

543 544 Quantification and Statistical Analysis 545 Statistical analysis was performed with Prism 7 (GraphPad). All error bars represent the standard 546 deviation (SD). The two-sample t test was used when needed, and the data were judged to be 547 statistically significant when p < 0.05. In the figures, asterisks (*) denotes statistical significance 548 as follows: *, p < 0.05, **, p < 0.001, ***, p < 0.0001, as compared to the appropriate controls. 549 The KaplanMeier method was used to calculate the survival fractions, and statistical significance 550 between survival curves was determined using the log-rank test. All experiments were performed 551 in triplicate. 552

21 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

553 References

554 (1) Coles SJ, Bhaskar KR, O'Sullivan DD, Neill KH, Reid LM. Airway mucus: composition and 555 regulation of its secretion by neuropeptides in vitro. Ciba Found Symp 1984;109:40-60.

556 (2) Rogers DF. Mucus pathophysiology in COPD: differences to asthma, and pharmacotherapy. 557 Monaldi Arch Chest Dis 2000 Aug;55(4):324-332.

558 (3) Barchi JJ,Jr. Mucin-type glycopeptide structure in solution: past, present, and future. 559 Biopolymers 2013 Oct;99(10):713-723.

560 (4) Foster WM, Bergofsky EH. Airway mucus membrane: effects of beta-adrenergic and 561 anticholinergic stimulation. Am J Med 1986 Nov 14;81(5A):28-35.

562 (5) Sicard JF, Le Bihan G, Vogeleer P, Jacques M, Harel J. Interactions of Intestinal Bacteria 563 with Components of the Intestinal Mucus. Front Cell Infect Microbiol 2017 Sep 5;7:387.

564 (6) Linden SK, Florin TH, McGuckin MA. Mucin dynamics in intestinal bacterial infection. 565 PLoS One 2008;3(12):e3952.

566 (7) Linden SK, Sutton P, Karlsson NG, Korolik V, McGuckin MA. Mucins in the mucosal 567 barrier to infection. Mucosal Immunol 2008 May;1(3):183-197.

568 (8) Boks NP, Norde W, van der Mei, H. C., Busscher HJ. Forces involved in bacterial adhesion 569 to hydrophilic and hydrophobic surfaces. Microbiology 2008 Oct;154(Pt 10):3122-3133.

570 (9) Krasowska A, Sigler K. How microorganisms use hydrophobicity and what does this mean 571 for human needs? Front Cell Infect Microbiol 2014 Aug 19;4:112.

572 (10) Lillehoj EP, Hyun SW, Liu A, Guang W, Verceles AC, Luzina IG, et al. NEU1 Sialidase 573 Regulates Membrane-tethered Mucin (MUC1) Ectodomain Adhesiveness for Pseudomonas 574 aeruginosa and Decoy Receptor Release. J Biol Chem 2015 Jul 24;290(30):18316-18331.

575 (11) Kaplan JB, Velliyagounder K, Ragunath C, Rohde H, Mack D, Knobloch JK, et al. Genes 576 involved in the synthesis and degradation of matrix polysaccharide in Actinobacillus 577 actinomycetemcomitans and Actinobacillus pleuropneumoniae biofilms. J Bacteriol 2004 578 Dec;186(24):8213-8220.

579 (12) Izano EA, Sadovskaya I, Vinogradov E, Mulks MH, Velliyagounder K, Ragunath C, et al. 580 Poly-N-acetylglucosamine mediates biofilm formation and antibiotic resistance in Actinobacillus 581 pleuropneumoniae. Microb Pathog 2007 Jul;43(1):1-9.

582 (13) Lillehoj EP, Kim BT, Kim KC. Identification of Pseudomonas aeruginosa flagellin as an 583 adhesin for Muc1 mucin. Am J Physiol Lung Cell Mol Physiol 2002 Apr;282(4):751.

22 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

584 (14) Lory S, Jin S, Boyd JM, Rakeman JL, Bergman P. Differential gene expression by 585 Pseudomonas aeruginosa during interaction with respiratory mucus. Am J Respir Crit Care Med 586 1996 Oct;154(4 Pt 2):183.

587 (15) Zanin M, Baviskar P, Webster R, Webby R. The Interaction between Respiratory Pathogens 588 and Mucus. Cell Host Microbe 2016 Feb 10;19(2):159-168.

589 (16) Rogers DF, Barnes PJ. Treatment of airway mucus hypersecretion. Ann Med 590 2006;38(2):116-125.

591 (17) Rogers DF. The role of airway secretions in COPD: pathophysiology, epidemiology and 592 pharmacotherapeutic options. COPD 2005 Sep;2(3):341-353.

593 (18) Ha EV, Rogers DF. Novel Therapies to Inhibit Mucus Synthesis and Secretion in Airway 594 Hypersecretory Diseases. Pharmacology 2016;97(1-2):84-100.

595 (19) Schulenburg H, Felix MA. The Natural Biotic Environment of Caenorhabditis elegans. 596 Genetics 2017 May;206(1):55-86.

597 (20) Samuel BS, Rowedder H, Braendle C, Felix MA, Ruvkun G. Caenorhabditis elegans 598 responses to bacteria from its natural habitats. Proc Natl Acad Sci U S A 2016 Jul 599 5;113(27):3941.

600 (21) Troemel ER, Chu SW, Reinke V, Lee SS, Ausubel FM, Kim DH. p38 MAPK regulates 601 expression of immune response genes and contributes to longevity in C. elegans. PLoS Genet 602 2006 Nov 10;2(11):e183.

603 (22) Head BP, Olaitan AO, Aballay A. Role of GATA transcription factor ELT-2 and p38 604 MAPK PMK-1 in recovery from acute P. aeruginosa infection in C. elegans. Virulence 2017 Apr 605 3;8(3):261-274.

606 (23) Spencer WC, Zeller G, Watson JD, Henz SR, Watkins KL, McWhirter RD, et al. A spatial 607 and temporal map of C. elegans gene expression. Genome Res 2011 Feb;21(2):325-341.

608 (24) Smith CJ, Watson JD, Spencer WC, O'Brien T, Cha B, Albeg A, et al. Time-lapse imaging 609 and cell-specific expression profiling reveal dynamic branching and molecular determinants of a 610 multi-dendritic nociceptor in C. elegans. Dev Biol 2010 Sep 1;345(1):18-33.

611 (25) Espelt MV, Estevez AY, Yin X, Strange K. Oscillatory Ca2+ signaling in the isolated 612 Caenorhabditis elegans intestine: role of the inositol-1,4,5-trisphosphate receptor and 613 phospholipases C beta and gamma. J Gen Physiol 2005 Oct;126(4):379-392.

614 (26) Melo JA, Ruvkun G. Inactivation of conserved C. elegans genes engages pathogen- and 615 xenobiotic-associated defenses. Cell 2012 Apr 13;149(2):452-466.

23 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

616 (27) Singh J, Aballay A. Microbial Colonization Activates an Immune Fight-and-Flight 617 Response via Neuroendocrine Signaling. Dev Cell 2019 Apr 8;49(1):99.e4.

618 (28) Zuniga M, Monedero V, Yebra MJ. Utilization of Host-Derived Glycans by Intestinal 619 Lactobacillus and Bifidobacterium Species. Front Microbiol 2018 Aug 17;9:1917.

620 (29) Koropatkin NM, Cameron EA, Martens EC. How glycan metabolism shapes the human gut 621 microbiota. Nat Rev Microbiol 2012 Apr 11;10(5):323-335.

622 (30) Flynn JM, Niccum D, Dunitz JM, Hunter RC. Evidence and Role for Bacterial Mucin 623 Degradation in Cystic Fibrosis Airway Disease. PLoS Pathog 2016 Aug 22;12(8):e1005846.

624 (31) Arabyan N, Weis AM, Huang BC, Weimer BC. Implication of Sialidases in Salmonella 625 Infection: Genome Release of Sialidase Knockout Strains from Salmonella enterica Serovar 626 Typhimurium LT2. Genome Announc 2017 May 11;5(19):17.

627 (32) Darby C, Chakraborti A, Politz SM, Daniels CC, Tan L, Drace K. Caenorhabditis elegans 628 mutants resistant to attachment of Yersinia biofilms. Genetics 2007 May;176(1):221-230.

629 (33) Parsons LM, Mizanur RM, Jankowska E, Hodgkin J, O Rourke D, Stroud D, et al. 630 Caenorhabditis elegans bacterial pathogen resistant bus-4 mutants produce altered mucins. PLoS 631 One 2014 Oct 8;9(10):e107250.

632 (34) Bergstrom KS, Xia L. Mucin-type O-glycans and their roles in intestinal homeostasis. 633 Glycobiology 2013 Sep;23(9):1026-1037.

634 (35) Korgaonkar AK, Whiteley M. Pseudomonas aeruginosa enhances production of an 635 antimicrobial in response to N-acetylglucosamine and peptidoglycan. J Bacteriol 2011 636 Feb;193(4):909-917.

637 (36) Hall S, McDermott C, Anoopkumar-Dukie S, McFarland AJ, Forbes A, Perkins AV, et al. 638 Cellular Effects of Pyocyanin, a Secreted Virulence Factor of Pseudomonas aeruginosa. Toxins 639 (Basel) 2016 Aug 9;8(8):10.3390/toxins8080236.

640 (37) Kato K, Lillehoj EP, Kai H, Kim KC. MUC1 expression by human airway epithelial cells 641 mediates Pseudomonas aeruginosa adhesion. Front Biosci (Elite Ed) 2010 Jan 1;2:68-77.

642 (38) Berger JT, Voynow JA, Peters KW, Rose MC. Respiratory carcinoma cell lines. MUC 643 genes and glycoconjugates. Am J Respir Cell Mol Biol 1999 Mar;20(3):500-510.

644 (39) Lehner B, Tischler J, Fraser AG. RNAi screens in Caenorhabditis elegans in a 96-well 645 liquid format and their application to the systematic identification of genetic interactions. Nat 646 Protoc 2006;1(3):1617-1620.

647 (40) Timmons L, Tabara H, Mello CC, Fire AZ. Inducible systemic RNA silencing in 648 Caenorhabditis elegans. Mol Biol Cell 2003 Jul;14(7):2972-2983.

24 bioRxiv preprint doi: https://doi.org/10.1101/675538; this version posted June 20, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

649 (41) Wang J, Barr MM. RNA interference in Caenorhabditis elegans. Methods Enzymol 650 2005;392:36-55.

651 (42) Singh J, Aballay A. Endoplasmic Reticulum Stress Caused by Lipoprotein Accumulation 652 Suppresses Immunity against Bacterial Pathogens and Contributes to Immunosenescence. MBio 653 2017 May 30;8(3):17.

654 (43) Zaas DW, Duncan MJ, Li G, Wright JR, Abraham SN. Pseudomonas invasion of type I 655 pneumocytes is dependent on the expression and phosphorylation of caveolin-2. J Biol Chem 656 2005 Feb 11;280(6):4864-4872.

657 (44) Hawdon NA, Aval PS, Barnes RJ, Gravelle SK, Rosengren J, Khan S, et al. Cellular 658 responses of A549 alveolar epithelial cells to serially collected Pseudomonas aeruginosa from 659 cystic fibrosis patients at different stages of pulmonary infection. FEMS Immunol Med 660 Microbiol 2010 Jul 1;59(2):207-220. 661

25 RNAi control A. 100 E. 100 N2 RNAi mul-1 mul-1 (ac7)

p < 0.0001 50 50 p < 0.001 Percent survival Percent survival 0 0 0 24 48 72 96 120 144 168 192 0 24 48 72 96 120 144 168 192 Hours Hours B. 105 F. 105 RNAi control N2 104 RNAi mul-1 104 mul-1 (ac7)

103 103

102 102 CFUs per Worm CFUs per Worm 101 101 0 2 4 6 8 0 2 4 6 8 C. Hours Hours control control mul-1 RNAi mul-1 RNAi RNAi RNAi

PA14 PA14 GFP PA14 GFP GFP PA14 GFP control control mul-1 RNAi mul-1 RNAi RNAi RNAi

PA14 GFP PA14 GFP PA14 GFP PA14 GFP D. G.

100 RNAi control 100 N2 RNAi mul-1 mul-1 (ac7)

50 50 Percent survival Percent survival p = 0.5540 p = 0.5953 0 0 0 4 8 12 16 20 24 28 32 36 40 0 4 8 12 16 20 24 28 32 36 40 662 Days Days 663 Figure 1. RNAi silencing and deletion of mul-1 results in an enhanced resistance phenotype 664 to P. aeruginosa PA14. (A) N2 wild-type animals were exposed to two generations of RNAi. 665 Young adult animals were transferred to full lawns of P. aeruginosa PA14 and nematode survival

26 666 was monitored daily. Animals were considered dead upon failure to respond to touch. Animals 667 missing from the agar plate were censored on the day of loss. The KaplanMeier method was used 668 to calculate the survival fractions, and statistical significance between survival curves was 669 determined using the log-rank test. 3 biological replicates, 180 total animals per condition. P values 670 compared to RNAi control: mul-1 p < 0.0001, let-653 p < 0.001, gly-8 p = 0.5594, cwp-4 p = 671 0.1336, gpdh-1 p = 0.0079. (B) After RNAi, young adult nematodes were transferred to full lawns 672 of P. aeruginosa PA14-GFP (kanr). At indicated time points, nematodes were transferred to fresh 673 E. coli OP50 lawns to remove P. aeruginosa. Worms were ground and serial dilutions were plated 674 on LB-kanamycin plates to calculate Colony Forming Units (CFUs) per worm (nematode). 3 675 biological replicates, 90 total animals per condition. (C) After RNAi, young adult nematodes were 676 transferred to full lawns of P. aeruginosa PA14-GFP (kanr). At 24 hours, nematodes were 677 transferred to fresh E. coli OP50 lawns to remove P. aeruginosa. Animals were anesthetized using 678 an M9 salt solution containing 30 mM sodium azide and mounted onto 2% agar pads. The animals 679 were visualized using a Leica M165 FC fluorescence stereomicroscope. Representative images of 680 ~50 animals per condition from 2 biological replicates. (D) N2 wild-type animals were exposed to 681 two generations of RNAi and L4 larval stage animals were transferred to fresh plates with the 682 corresponding RNAi clone, heat-killed E. coli HT115(DE3). Survival was monitored daily. 683 Worms were transferred to new plates as food was depleted. 3 biological replicates, 180 total 684 animals per condition. (E) Young adult N2 wild-type and mul-1 (ac7) deletion animals were 685 transferred to full lawns of P. aeruginosa PA14 and nematode survival was monitored daily. 686 Animals were considered dead upon failure to respond to touch. Animals missing from the agar 687 plate were censored on the day of loss. The KaplanMeier method was used to calculate the survival 688 fractions, and statistical significance between survival curves was determined using the log-rank 689 test. 3 biological replicates, 180 total animals per condition. (F) Young adult N2 wild-type and 690 mul-1(ac7) deletion animals were transferred to full lawns of P. aeruginosa PA14-GFP (kanr). At 691 indicated time points, nematodes were transferred to fresh E. coli OP50 lawns to remove P. 692 aeruginosa. Worms were ground and serial dilutions were plated on LB-kanamycin plates to 693 calculate Colony Forming Units (CFUs) per worm (nematode). 3 biological replicates, 90 total 694 animals per condition. (G) L4 N2 wild-type and mul-1 (ac7) deletion animals transferred to fresh 695 plates with heat-killed E. coli OP50. Survival was monitored daily. Worms were transferred to 696 new plates as food was depleted. 3 biological replicates, 180 total animals per condition. 697

27 A. B. VP303 RNAi Control MGH171 RNAi Control 100 100 VP303 RNAi mul-1 MGH171 RNAi mul-1

50 p < 0.0001 50 p < 0.0001 Percent survival Percent survival 0 0 0 24 48 72 96 120 144 168 192 0 24 48 72 96 120 144 168 192 698 Hours Hours 699 Figure 2. The enhanced resistance phenotype observed in mul-1 silenced animals is intestine- 700 specific. Intestine-specific RNAi animals, (A) VP303 and (B) MGH171, were exposed to two 701 generations of RNAi. Young adult animals were transferred to full lawns of P. aeruginosa PA14 702 and nematode survival was monitored daily. Animals were considered dead upon failure to 703 respond to touch. Animals missing from the agar plate were censored on the day of loss. The 704 KaplanMeier method was used to calculate the survival fractions, and statistical significance 705 between survival curves was determined using the log-rank test. 3 biological replicates, 180 706 animals per condition total. 707

28 10 **** ** RNAi control - E. coli 8 RNAi mul-1 -E. coli n.s. RNAi control - P. aeruginosa 6 n.s. ** RNAi mul-1 - P. aeruginosa 4 *** ** n.s.

Fold Change n.s. 2 ** 0

mul-1 gly-8 cwp-4 708 let-653 gpdh-1 709 Figure 3. Silencing of mul-1 by RNAi has little effect on the expression of other intestine- 710 expressed mucins during infection. qRT-PCR analysis of the expression of mul-1, let-653, gly- 711 8, gpdh-1, and cwp-4 during growth on both E. coli OP50 and P. aeruginosa PA14 for 8 hours. 712 N2 wild-type animals were exposed to two generations of RNAi. L4 larval stage animals were 713 transferred to full lawns of E. coli OP50 or P. aeruginosa PA14 for 8 hours. RNA was isolated as 714 described in methods section. Expression for indicated gene, under each RNAi condition, is 715 compared to N2 wild-type RNAi control animals grown on E. coli OP50. 3 biological replicates, 716 each with 3 technical replicates. ***p < 0.001 and **p < 0.01 via the t test. n.s., non-significant. 717

29 A. B. N2 100 100 mul-1 (ac7)

p < 0.001 50 50 N2

mul-1 (ac7) Percent survival % Lawn Occupancy 0 0 0 4 8 12 16 20 24 28 0 24 48 72 96 120 144 168 192 Hours 718 Hours 719 Figure 4. mul-1 animals fail to avoid P. aeruginosa PA14 and failure to avoid has no effect 720 on the enhanced resistance phenotype. (A) Young adult N2 wild-type and mul-1 (ac7) deletion 721 animals were transferred to partial lawns of P. aeruginosa PA14. Percent (%) Lawn Occupancy 722 was calculated at indicated time points as 100 x (the total number of animals inside the P. 723 aeruginosa PA14 lawn / total number of animals outside the P. aeruginosa PA14 lawn). (B) Young 724 adult N2 wild-type and mul-1 (ac7) deletion animals were transferred to partial lawns of P. 725 aeruginosa PA14 and nematode survival was monitored daily. Animals were considered dead 726 upon failure to respond to touch. Animals missing from the agar plate were censored on the day 727 of loss. The KaplanMeier method was used to calculate the survival fractions, and statistical 728 significance between survival curves was determined using the log-rank test. 3 biological 729 replicates, 180 total animals per condition. 730

30 A. B. 14000 **** RNAi control ****

*** 12000

100 **** RNAi mul-1 10000 RNAi ctsa-1.1 8000 Co-RNAi 50 3000 **** **** 2000 CFUs per Worm Percent survival 1000 0 0 24 48 72 96 120 144 168 192 0 Hours mul-1 C. ctsa-1.1 co-RNAi

RNAi RNAi control 100 RNAi control RNAi RNAi mul-1 RNAi csta-1.1 co-RNAi 50 Percent survival 0 0 4 8 12 16 20 24 28 32 36 40 Days 731 732 Figure 5. The ability of P. aeruginosa to access O-linked glycans provided by mul-1 in the 733 intestine alters the resistance phenotype of C. elegans nematodes. (A) N2 wild-type animals 734 were exposed to two generations of RNAi targeting mul-1, F41C3.5, and an equal mixture of mul- 735 1 and F41C3.5. Young adult animals were transferred to full lawns of P. aeruginosa PA14 and 736 nematode survival was monitored daily. Animals were considered dead upon failure to respond to 737 touch. Animals missing from the agar plate were censored on the day of loss. The KaplanMeier 738 method was used to calculate the survival fractions, and statistical significance between survival 739 curves was determined using the log-rank test. 3 biological replicates, 180 total animals per 740 condition. (B) After RNAi, young adult nematodes were transferred to full lawns of P. aeruginosa 741 PA14-GFP (kanr). At 8 hours post P. aeruginosa PA14 exposure, nematodes were transferred to 742 fresh E. coli OP50 lawns to remove P. aeruginosa. Worms were ground and serial dilutions were 743 plated on LB-kanamycin plates to calculate Colony Forming Units (CFUs) per worm (nematode). 744 3 biological replicates, 90 total animals per condition. (C) N2 wild-type animals were exposed to 745 two generations of RNAi and L4 larval stage animals were transferred to fresh plates with the 746 corresponding RNAi clone, heat-killed E. coli HT115(DE3). Survival was monitored daily. 747 Worms were transferred to new plates as food was depleted. 3 biological replicates, 180 total 748 animals per condition. 749

31 A. N-acetyl-D-glucosamine C. N-acetyl-D-galactosamine 100 100

50 50

Percent survival 0 Percent survival 0 0 48 96 144 192 240 0 48 96 144 192 240 Hours Hours

B. N-acetyl-neuraminic acid D. D-galactose 100 100

50 50 Percent survival Percent survival 0 0 0 48 96 144 192 240 0 48 96 144 192 240 Hours Hours E. F. N-acetyl-D-glucosamine N-acetyl-D-galactosamine 100 100

50 50 Percent survival 0 Percent survival 0 0 48 96 144 192 240 0 48 96 144 192 240 hours Hours RNAi control RNAi control RNAi control + 20 mM glycan RNAi control + 20 mM glycan RNAi mul-1 RNAi mul-1 RNAi mul-1 + 20 mM glycan 750 RNAi mul-1 + 20 mM glycan 751 Figure 6. Free glycans in nematode growth medium alters the resistance phenotype of C. 752 elegans to P. aeruginosa PA14. N2 wild-type animals were exposed to two generations of RNAi. 753 Young adult animals were transferred to full lawns of P. aeruginosa PA14 on nematode growth 754 media supplemented with (A) 20 mM N-acetyl-D-glucosamine, (B) 20 mM N-acetyl-neuraminic 755 acid, (C) 20 mM N-acetyl-D-galactosamine, or (D) 20 mM D-galactose and nematode survival 756 was monitored daily. Animals were considered dead upon failure to respond to touch. Animals

32 757 missing from the agar plate were censored on the day of loss. Red lines with triangles represent 758 the survival curves of RNAi control animals, without glycans present (solid line) and with glycans 759 present (dotted line). Blue lines with circles represent the survival curves of RNAi mul-1 animals, 760 without glycans present (solid line) and with glycans present (dotted line).The KaplanMeier 761 method was used to calculate the survival fractions, and statistical significance between survival 762 curves was determined using the log-rank test. 3 biological replicates, 180 total animals per 763 condition. To determine if the glycans needed to be present in the nematode growth medium for 764 C. elegans access, P. aeruginosa PA14 was grown for 8 hours at 37ºC in shaking culture with or 765 without (E) 20 mM N-acetyl-D-glucosamine or (F) 20 mM N-acetyl-D-galactosamine prior to 766 seeding full lawns of the bacteria. N2 wild-type animals were exposed to two generations of RNAi. 767 Young adult animals were transferred to full lawns of P. aeruginosa PA14 on non-supplemented 768 nematode growth medium. Red lines with triangles represent the survival curves of RNAi control 769 animals, without glycans present (solid line) and with glycans present (dotted line). Blue lines with 770 circles represent the survival curves of RNAi mul-1 animals, without glycans present (solid line) 771 and with glycans present (dotted line). The KaplanMeier method was used to calculate the survival 772 fractions, and statistical significance between survival curves was determined using the log-rank 773 test. 3 biological replicates, 180 total animals per condition. 774

33 A. 500 p = 0.0017 400 p = 0.0019

300

200

CFUs/1000 Cells 100

0 Media N-acetyl- N-acetyl- Control D-glucosamine D-galactosamine B. 600

p = 0.0102 p = 0.0258 400

200 CFUs/1000 Cells

0 Media N-acetyl- N-acetyl- 775 Control D-glucosamine D-galactosamine 776 Figure 7. Free glycans alter binding and internalization of P. aeruginosa PA14 to A549 777 human lung cells. (A)Approximately 1 x 105 A549 lung cells (~85% confluency) were grown for 778 2 days with or without 2 mM N-acetyl-D-glucosamine or 2 mM N-acetyl-D-galactosamine 779 supplemented media and infected with an MOI of 100 for 4 hours with P. aeruginosa PA14. 780 Intracellular and extracellular bacteria were quantified similarly, as described in methods, but cells 781 were not treated with gentamycin prior to isolating bacteria. Cell viability was measured using 782 Promega Cell-Titre Glo 2.0. Serial dilutions were plated to calculate CFUs and normalized to the 783 total number of live cells. 3 biological replicates, each with 3 technical replicates. P values denoted 784 in figure, calculate via the t test. (B) Approximately 1 x 105 A549 lung cells (~85% confluency) 785 were grown for 2 days without glycans preset in the media. On the day of the experiment, 2 mM 786 N-acetyl-D-glucosamine or 2 mM N-acetyl-D-galactosamine was added to the media at the same 787 time as the cells were infected with an MOI of 100 for 4 hours with P. aeruginosa PA14. 788 Intracellular and extracellular bacteria were quantified similarly, as described in methods, but cells 789 were not treated with gentamycin prior to isolating bacteria. Cell viability was measured using 790 Promega Cell-Titre Glo 2.0. Serial dilutions were plated to calculate CFUs and normalized to the 791 total number of live cells. 3 biological replicates, each with 3 technical replicates. P values denoted 792 in figure, calculate via the t test. 793

34 p = 0.0101 p = 0.0013 600 n.s. p = 0.0035 Control siRNA 400 muc1 siRNA

200 CFUs/1000 Cells

p = 0.008770 p = 0.000549 p = 0.001058 0 Media N-acetyl- N-acetyl- 794 Control D-glucosamine D-galactosamine 795 Figure 8. Free glycans reverse the P. aeruginosa binding defects to muc1 siRNA knockdown 796 human lung cells (A549). Sigma MISSION TRC control and muc1 shRNA Lentiviral Particles 797 were used to create control and muc1 knockdown cells. Approximately 1 x 105 A549 lung cells 798 (~85% confluency) were grown for 2 days without glycans preset in the media. On the day of the 799 experiment, 2 mM N-acetyl-D-glucosamine or 2 mM N-acetyl-D-galactosamine was added to the 800 media at the same time as the cells were infected with an MOI of 100 for 4 hours with P. 801 aeruginosa PA14. Intracellular and extracellular bacteria were quantified, as described in methods. 802 Cells were not treated with gentamycin prior to isolating bacteria. Cell viability was measured 803 using Promega Cell-Titre Glo 2.0. Serial dilutions were plated to calculate CFUs and normalized 804 to the total number of live cells. 3 biological replicates, each with 3 technical replicates. P values 805 denoted in figure, calculate via the t test. n.s. is non-significant. 806

35 807 Figure S1. RNAi silencing of several intestine-expressed mucins alters the resistance 808 phenotype to P. aeruginosa PA14. (A) N2 wild-type animals were exposed to two generations of 809 RNAi. Young adult animals were transferred to full lawns of P. aeruginosa PA14 and nematode 810 survival was monitored daily. Animals were considered dead upon failure to respond to touch. 811 Animals missing from the agar plate were censored on the day of loss. The KaplanMeier method 812 was used to calculate the survival fractions, and statistical significance between survival curves 813 was determined using the log-rank test. 3 biological replicates, 180 total animals per condition. 814 (B) Because RNAi for let-653 is larval lethal, L4 larval stage N2 wild-type animals were 815 transferred to full lawns of both control and let-653 E. coli HT115(DE3) RNAi bacteria. Young 816 adults were transferred to full lawns of P. aeruginosa PA14 and nematode survival was monitored 817 daily. Animals were considered dead upon failure to respond to touch. Animals missing from the 818 agar plate were censored on the day of loss. The KaplanMeier method was used to calculate the 819 survival fractions, and statistical significance between survival curves was determined using the 820 log-rank test. 3 biological replicates, 120 total animals per condition. (C) Time (hours) to 50% 821 death upon P. aeruginosa PA14 exposure was calculated for each of the survival curves in (A) and 822 (B) using Graphpad Prism 8 software and is reported as TD50. 823 824 825 Figure S2. The ability of S. enterica ser Typhimurium to access O-linked glycans provided 826 by mul-1 in the intestine alters the resistance phenotype of C. elegans nematodes. (A) N2 wild- 827 type animals were exposed to two generations of RNAi targeting mul-1, and F41C3.5. Young adult 828 animals were transferred to full lawns of S. enterica ST1334 and nematode survival was monitored 829 daily. Animals were considered dead upon failure to respond to touch. Animals missing from the 830 agar plate were censored on the day of loss. The KaplanMeier method was used to calculate the 831 survival fractions, and statistical significance between survival curves was determined using the 832 log-rank test. 3 biological replicates, 180 total animals per condition. (B) After RNAi, young adult 833 nematodes were transferred to full lawns of S. enterica ST1334-GFP (kanr). At 24 hours post S. 834 enterica ST1334 exposure, nematodes were transferred to fresh E. coli OP50 lawns to remove S. 835 enterica. Worms were ground and serial dilutions were plated on LB-kanamycin plates to calculate 836 Colony Forming Units (CFUs) per worm (nematode). 3 biological replicates, 90 total animals per 837 condition. (C) After RNAi, young adult nematodes were transferred to full lawns of P. aeruginosa 838 PA14-GFP (kanr). At 24 hours post S. enterica ST1334 exposure, nematodes were transferred to 839 fresh E. coli OP50 lawns to remove S. enterica. Worms were then transferred to fresh E. coli OP50 840 plates and at indicated time points, nematodes were ground and serial dilutions were plated on LB- 841 kanamycin plates to calculate persistant Colony Forming Units (CFUs) per worm (nematode). 3 842 biological replicates, 90 total animals per condition 843 844 Figure S3. Free glycans have limited effects on P. aeruginosa growth in bacterial growth 845 media. Individual bacterial colonies were inoculated into 2 mL of LB and grown overnight, 846 shaking at 225 RPM at 37ºC. Overnight cultures were diluted to an OD600 of 0.05 in either (A) LB 847 media or (B) M9 Media supplemented with varying concentrations of free glycans. 100 µL 848 bacterial cultures were placed in individual wells of a 96 well plate with various glycans added at 849 indicated concentrations. Bacterial growth was monitored over time by measuring the OD600 at 850 indicated time points. 851

36 852 Figure S4. Free glycans have limited effects on P. aeruginosa growth in cell growth media. 853 Individual bacterial colonies were inoculated into 2 mL of LB and grown overnight, shaking at 854 225 RPM at 37ºC. Overnight cultures were diluted to an OD600 of 0.05 in DMEM F12+K + 10% 855 Heat-Inactivated FBS with varying concentrations of (A) N-acetyl-D-glucoasmine or (B) N- 856 acetyl-D-galactosamine. 100 µL bacterial cultures were placed in individual wells of a 96 well 857 plate with various glycans added at indicated concentrations. Bacterial growth was monitored over 858 time by measuring the OD600 at indicated time points.

37