Introduction

Viruses are ubiquitous and abundant^1,2. Infection can have profound consequences for the health of an individual host. The ability of viruses to transmit from one individual to another can scale these consequences causing morbidity and mortality throughout whole populations. Many factors influence virus transmission^3–5. Abiotic factors such as temperature⁶ and humidity⁷ and biotic factors such as viral load^8,9 and host immune status^10,11 all interact to determine transmission rates. Despite these findings, the determinants of virus transmissibility remain incompletely understood.

The microbiota has emerged as a host-associated factor that modulates multiple aspects of virus infection and thereby alters transmission rates among host organisms^12–14. In general, the microbiota is critical for the proper development of the immune system and for the effective activation of antimicrobial immune responses even at sites distal to microbiota colonization¹⁵. More specifically, bacteria and bacterial surface structures such as lipopolysaccharide and peptidoglycan have been shown to stabilize poliovirus and reovirus in vitro^13,16. These observations likely explain why microbiota depletion by antibiotic treatment was sufficient to provide protection against the same viruses in mice¹³. The bacterial symbiont Wolbachia protects numerous insect species from multiple viruses either by upregulating antiviral defenses or competing for intracellular nutrients^17–21. Individual bacteria have also been found to enhance viral infection; Serratia marcescens promoted infection of the mosquito Aedes aegypti by Dengue, Zika, and Sindbis viruses by secreting a protein, enhancin, that degrades the mucus layer covering epithelial cells¹⁴.

Caenorhabditis elegans is a nematode often found in microbially rich environments such as rotting vegetation²². The first naturally occurring virus capable of infecting the C. elegans was isolated from Orsay, France²³. Orsay virus is a part of a group of nematode infecting viruses closely related to the Nodaviridae family of viruses which infect arthropods and fish^23,24. Like other Nodaviruses, Orsay virus is a positive-sense, single-stranded RNA virus with a bipartite genome²³. Transmission of Orsay virus occurs horizontally through the fecal-oral route. Fluorescence in situ hybridization and immunofluorescence imaging has revealed that Orsay virus solely infects C. elegans intestinal cells^23,25. Viral infection activates host defense mechanisms including the RNA-interference response and a transcriptional program known as the intracellular pathogen response^23,26–29.

C. elegans is a bacterivore and is propagated in the laboratory on lawns of Escherichia coli OP50. Pseudomonas aeruginosa, an opportunistic pathogen of humans, is found in the soil and water and can also infect C. elegans^30–32. Infection of C. elegans with P. aeruginosa activates innate immunity and stress responses, as well as behavioral avoidance responses^33–35. Recently, the bacteria resident in the natural environment of C. elegans in the wild have been of increasing interest^22,36–38. The intestinal lumen of free-dwelling C. elegans is occupied by taxonomically and functionally diverse bacteria that can affect C. elegans fitness and physiology^36,37,39.

In this study, we sought to understand how bacteria that are constituents of the C. elegans microbiota quantitatively affect the transmission of Orsay virus in C. elegans. We observed that monoaxenic cultures of different bacterial species had widely divergent effects on the transmission of Orsay virus and conducted genetic analysis of the bacterial determinants involved in modulating virus transmission. Our data point to a key species-specific role for bacteria as critical determinant of virus transmission.

Results

Wide variation in the effects of bacteria on the transmission of Orsay virus

Orsay virus spreads horizontally through the fecal-oral route and transmission can spread from a single animal to a population of animals on a plate⁴⁰. We sought to assess the impact of bacteria on Orsay virus transmission rates and utilized a collection of bacteria isolated from the environment with wild C. elegans to assemble a panel of Gram-negative bacteria for comparison with the standard laboratory bacterial food, Escherichia coli OP50^{36,37,39,41,42}. All animals were raised to young adulthood on lawns of E. coli OP50 to eliminate differences in developmental rates caused by different bacteria³⁹. We set up a transmission assay by placing infected animals (“spreaders”) together with uninfected animals on a monoaxenic lawn of each test bacterium. We quantified the incidence proportion, defined as the number of new infections produced after 24 h as determined by detection of induction of the pals-5p::GFP reporter, which is induced by infection with Orsay virus (Fig. 1A-C)²⁷. We observed a wide range in the measured incidence proportion (Fig. 1B and 1C). Exposure to many of the naturally associated bacterial strains resulted in transmission that was comparable to the incidence proportion observed with E. coli OP50 with some prominent exceptions (Fig. 1B and 1C). Two Ochrobactrum species promoted virus infection in nearly all individuals in the transmission assay and increased the incidence proportion 2.7-fold and 2.9-fold compared to E. coli OP50 (Fig. 1B and 1C). We confirmed that transmission from the initial spreader animals in the assay, and not multiple rounds of infection, was responsible for the increased incidence proportion observed in the presence of Ochrobactrum vermis MYb71 (Supp. Fig. 1A-D). On the other hand, the presence of P. lurida MYb11 reduced the incidence proportion 4.1-fold compared to E. coli OP50 and 11-fold compared to O. vermis MYb71 (Fig. 1B and C). These results present a striking divergence in the effects of individual bacterial constituents of the C. elegans microbiota on the level of transmission of Orsay virus from infected animals.

Figure 1:

We reasoned that increased transmission of Orsay virus could be due to bacterial modulation of host susceptibility to viral infection. To quantify the degree of host susceptibility to Orsay virus we added Orsay virus in doses that ranged across four orders of magnitude and then measured the fraction of individuals that became pals-5p::GFP positive after 24 h (Fig. 1A and 1D, Supp. Fig. 2 and Supp. Fig. 3). We also specifically calculated the ID₅₀, or the dose of Orsay virus required to infect 50% of the population after 24 h of exposure (Fig. 1D and 1E, Supp. Fig 2). On average, 3.6 μL of Orsay virus stock prepared as indicated in the Methods was required to infect 50% of a population of ZD2611 animals in the presence of E. coli OP50 after 24 hours (Supp. Fig. 2). We observed variability in the measured ID₅₀ even when using the same batch of Orsay virus (e.g. Fig. 1D vs Supp. Fig. 3A vs. Supp. Fig. 3B). We therefore chose to control all our experiments internally rather than attempt to normalize between Orsay virus batches and we report doses used in arbitrary units (a.u.), however 1 a.u. corresponds to 1 μL of Orsay virus filtrate (Methods).

To determine quantitatively how different bacterial species modulate host susceptibility to viral infection independently of the potential differential effects of bacteria on host shedding of viruses, we added fixed doses of Orsay virus to plates of each bacterial species and monitored infection using the pals-5p::GFP reporter. At lower doses of Orsay virus, the presence of O. vermis MYb71 resulted in a higher fraction of infected animals than those observed in the presence of E. coli OP50 or P. lurida MYb11 (Fig. 1D, Supp. Fig. 3). At higher doses of Orsay virus, the presence of P. lurida MYb11 resulted in a reduced fraction of infected animals compared to O. vermis MYb71 and E. coli OP50 (Fig. 1D, Supp. Fig 3). We further quantified the ID₅₀ in the presence of each bacterium and on average we observed that the ID₅₀ in the presence of E. coli OP50 was 14-fold higher than the ID₅₀ in the presence of O. vermis MYb71 (Fig. 1D and 1E, Supp. Fig. 3). The ID₅₀ observed in the presence of P. lurida was 120-fold and 1800-fold higher than the ID₅₀ in the presence of E. coli OP50 or O. vermis MYb71 respectively (Fig. 1D and 1E, Supp. Fig. 3).

We corroborated our observations from the pals-5p::GFP reporter by scoring the same samples using fluorescence in situ hybridization to detect the RNA1 segment of the Orsay virus genome in the intestinal cells of infected animals (Fig. 1F). As expected, we observed that at lower doses the presence of O. vermis MYb71 resulted in a greater fraction of infected animals compared to E. coli OP50 or P. lurida MYb11, while at higher doses the presence of P. lurida MYb11 resulted in a reduced fraction of infected animals (Fig. 1F). Together these data establish that individual members of the C. elegans microbiota can modulate host susceptibility to Orsay virus over three orders of magnitude with dramatic consequences for the transmissibility of Orsay virus.

We also assessed whether bacteria altered host feeding behavior as this may influence the transmission and infection rate of Orsay virus. We quantified the number of pharyngeal pumps per minute (ppm), a metric of C. elegans feeding, in the presence of each bacterium. After 6 h of exposure to E. coli OP50, P. lurida MYb11, or O. vermis MYb71 there were no differences in pharyngeal pumping rates (Supp. Fig. 4A). At 24 h pumping rates were decreased in the presence of O. vermis MYb71 to an average of 242 ppm compared to 261 ppm in the presence of P. lurida MYb11 suggesting that changes to feeding behavior are unlikely to be responsible for observed effects on host susceptibility to Orsay virus (Supp. Fig. 4B).

P. aeruginosa attenuates Orsay virus transmission

In view of the effect of P. lurida MYb11 on attenuating Orsay virus infection of C. elegans, we examined the effect of the distantly related bacterium, Pseudomonas aeruginosa, an opportunistic pathogen of humans that has been characterized extensively^30,43. We observed that in the presence of either P. aeruginosa strains PA01 or PA14, transmission from spreader animals to uninfected individuals was nearly completely blocked and the incidence proportion was 11-fold and 32-fold lower than that observed in the presence of E. coli OP50, respectively (Fig. 2A). The attenuating effect of P. aeruginosa PA01 and PA14 on virus transmission was further reduced 3.1-fold and 9.5-fold, respectively, compared to the incidence proportion observed in the presence of P. lurida MYb11 (Fig. 2A).

Figure 2:

We confirmed that host susceptibility to Orsay virus was reduced in the presence of P. aeruginosa by performing susceptibility assays at two doses of exogenous Orsay virus (Fig 2B, Supp. Fig. 5). At each dose, fewer animals were infected following exposure to equivalent doses of Orsay virus in the presence of P. aeruginosa PA01, PA14, or P. lurida MYb11 as compared to the fraction of animals infected in the presence of E. coli OP50 (Fig. 2B, Supp. Fig. 5). At the highest dose of virus used, we still observed robust attenuation of infection in the presence of P. aeruginosa PA14 as 0.5% of the animals were infected as compared to 54%, 70%, and 97% in the presence of P. aeruginosa PA01, P. lurida MYb11, or E. coli OP50, respectively (Fig. 2B). We confirmed that no individuals were detectably infected with Orsay virus while in the presence of P. aeruginosa PA14 using FISH staining (Fig. 2C). FISH additionally confirmed that the presence of P. aeruginosa PA01 and P. lurida MYb11 attenuated average infection to 24% and 60% of the population respectively compared to E. coli OP50 which supported infection of 86% of the population. Further, P. aeruginosa PA01, P. lurida MYb11, and E. coli OP50 all supported higher levels of infection than P. aeruginosa PA14. (Fig. 2C). Together these results demonstrate a striking capacity of P. lurida MYb11 and P. aeruginosa strains to sharply reduce and even effectively block, host susceptibility to infection with Orsay virus.

We again assessed whether bacterial modulation of host feeding behavior may be responsible for the observed effects on host susceptibility to Orsay virus. There were no differences observed in pharyngeal pumping rate after 6 h of exposure to each bacterium (Supp. Fig. 4A). After 24 h pharyngeal pumping rate increased in the presence of P. aeruginosa PA14 to an average of 264 ppm compared to 248 ppm in the presence of E. coli OP50 which suggests that changes to feeding behavior are unlikely to account for the attenuation of infection observed in P. aeruginosa (Supp. Fig. 4B).

P. aeruginosa is a pathogen of C. elegans and numerous other organisms^30,44,45. Moreover, while both PA01 and PA14 are pathogenic, PA14 is more virulent towards C. elegans³⁰. P. lurida MYb11 promotes developmental rate and fitness, although does so at a minor cost to overall lifespan^39,46. The effect of O. vermis MYb71 on lifespan has not been assessed. We assessed host lifespan under our assay conditions to examine links between bacterial pathogenicity and host susceptibility to Orsay virus. Under our assay conditions, including a full lawn of bacteria, which prevent C. elegans from avoiding undesirable bacteria, and a temperature of 20°C, which is below the optimal virulence temperature of P. aeruginosa, substantial mortality is not observed until 72 hours after exposure (Supp. Fig. 6). All bacteria enhanced mortality in comparison to E. coli OP50 (Supp. Fig. 6). Interestingly, O. vermis MYb71 exposure led to similarly enhanced mortality when compared to P. aeruginosa PA14, and both bacteria enhanced mortality in comparison to P. lurida MYb11 or P. aeruginosa PA01 (Supp. Fig. 6). P. lurida MYb11 also enhanced mortality compared to P. aeruginosa PA01 under these conditions (Supp. Fig. 6). The similar lifespans of animals exposed to P. aeruginosa PA14 and O. vermis MYb71 stand in contrast to their divergent effects on host susceptibility to Orsay virus. This suggests that general pathogenicity is unlikely to be responsible for effects on host susceptibility, although pathogenic consequences specific to exposure to each bacterium may still contribute to the observed effects on host susceptibility to Orsay virus.

Attenuation of Orsay virus transmission by P. lurida and P. aeruginosa is not due to effects on Orsay virus replication, stability, or shedding

We considered the possibility that the attenuation of Orsay virus transmission in the presence of P. lurida and P. aeruginosa strains might be due to inhibitory effects of the bacteria on the replication of Orsay virus once transmitted to a susceptible animal host. To evaluate this possibility, we made use of a plasmid-based system in which viral RNA1 is expressed through a transgene introduced into C. elegans, so that replication of RNA1 can be assessed independent of the entry of exogenous virus into the host⁴⁷. In this system, the Orsay virus RNA1 segment, which encodes the RNA-dependent RNA polymerase (RdRP), is expressed following heat-shock. The expressed RNA1 may then be translated to produce the RdRP which can then replicate RNA1 through a negative-strand intermediate. The expression of RNA1 via heat-shock bypasses any differences in viral entry or pre-replication steps, allowing for a direct test of RNA1 replication efficiency under different conditions⁴⁷. A plasmid expressing a mutated RdRP (RNA1[D601A]) incapable of supporting further RNA1 replication after the initial heat-shock serves as a control for heat shock efficiency⁴⁷. Using this system, we observed that Orsay virus RNA1 replication efficiency was unaffected by the presence of Pseudomonas species relative to RNA1 replication observed in the presence of E. coli OP50 (Fig. 3A). Additionally, there were no differences observed in heat-shock efficiency between the different bacteria (Fig. 3A). These data suggest that bacteria-induced differences in RNA1 replication in the host do not explain the substantial attenuation of Orsay virus transmission and infection rates caused by P. aeruginosa PA01 or PA14 and P. lurida MYb11.

Figure 3:

We sought to confirm the successful replication of the plasmid expressed RNA1 of Orsay virus by assessing whether we could detect the replication of Orsay virus RNA1 in each Pseudomonas species. We exposed young adult wild-type (N2) or RNAi defective animals (rde-1(ne219)) to exogenous Orsay virus and quantified the amount of virus present at 2 h and 24 h post-exposure, with the difference in RNA levels at these points reflecting viral genome replication. Animals of the RNAi defective rde-1(ne219) background have a defective antiviral response leading to higher viral loads than wild-type animals²³. The rde-1(ne219) mutant therefore provides a more sensitive genetic background in which to detect viral replication following rare infection events within a population as is expected in the presence of P. aeruginosa PA14. In rde-1(ne219) animals viral RNA1 levels increased at 24 h relative to 2 h post exposure regardless of the bacteria present (Fig. 3B and 3C). On the other hand, in the wild-type background replication was not observed in animals exposed to P. aeruginosa PA14 and was attenuated in animals exposed to P. aeruginosa PA01 compared to animals exposed to E. coli OP50 or P. lurida MYb11 (Fig. 3B and 3C). In a wild-type background individuals exposed to P. aeruginosa PA01 and P. aeruginosa PA14 are less likely to become infected from exogenous Orsay virus (Fig. 2B and Supp. Fig. 5). It is therefore likely that the viral load we observed in the wild-type populations is due to a smaller proportion of the population being infected, rather than hampered replication²³.

We also assessed whether P. lurida MYb11, P. aeruginosa PA01, or P. aeruginosa PA14 was capable of rapidly degrading Orsay virus or preventing viral shedding, either of which might account for attenuation of Orsay virus transmission. We did not observe a substantial decrease in viral stability in the presence of any bacterium suggesting Orsay virus is not rapidly degraded (Fig. 3D and 3E). We did observe an approximately 2-fold decrease in viral levels between E. coli OP50 and P. lurida MYb11, however, this difference already existed within 30 min of exposure to each bacterium and did not grow further over the course of 24 h suggesting we are measuring differences in recovery of Orsay virus from each bacterial lawn, rather than enhanced degradation in the presence of P. lurida MYb11 (Fig. 3D and 3E). We measured rates of Orsay virus shedding during a transmission assay. We did not observe a substantial decrease in viral shedding in the presence of any bacterium (Fig. 3F). We observed approximately 4-fold lower levels of Orsay virus shed in the presence of P. lurida MYb11 vs E. coli OP50, but as suggested by our virus stability data, we believe these differences at least partly reflect differences in Orsay virus recovery from each bacterial lawn rather than actual differences in viral shedding (Fig. 3F).

Attenuation of Orsay virus transmission by Pseudomonas species is dependent on regulators of bacterial quorum sensing

Extensive studies on Pseudomonas species have demonstrated the importance of quorum sensing for regulating many community-level behaviors in response to growing population density⁴⁸. P. aeruginosa relies upon three quorum sensing systems: las, rhl, and pqs. These systems are arranged hierarchically, however crosstalk between them is extensive (Fig. 4A)⁴⁸. In P. aeruginosa, an additional layer of regulation stems from two-component regulatory systems such as gacA/gacS which regulates numerous genes that together influence quorum sensing, virulence, and biofilm development (Fig. 4A)^49–51. P. aeruginosa possesses three exopolysaccharide biosynthetic clusters that each contribute to biofilm formation: pel, psl, and alg. However, P. aeruginosa PA01 preferentially produces Psl while P. aeruginosa PA14 is unable to synthesis Psl and produces Pel^52–56. We hypothesized that quorum sensing might mediate the effect of Pseudomonas strains to attenuate virus transmission, while differences in exopolysaccharide production might mediate the enhanced attenuation of virus transmission observed in the presence of P. aeruginosa PA14 compared with what was observed in the presence of P. aeruginosa PA01. Therefore, we tested a panel of P. aeruginosa PA01 and P. aeruginosa PA14 quorum sensing and biofilm mutants to determine whether quorum sensing or biofilm formation was involved in the attenuation of Orsay virus infection mediated by P. aeruginosa.

Figure 4:

Mutations of any the regulators of the las, rhl, or pqs quorum sensing systems suppressed the attenuation of Orsay virus infection caused by the presence of wild-type P. aeruginosa PA01 (Fig. 4B, Supp. Fig. 7). Knockout of gacA in P. aeruginosa PA01 also suppressed the attenuation of Orsay virus infection (Fig. 4B, Supp. Fig. 7). On the other hand, mutation of any of the three exopolysaccharide production pathways had no potent effect on the attenuation of infection (Fig. 4B, Supp. Fig. 7). These data support a role for quorum sensing regulated processes in reducing Orsay virus infection rates but do not implicate a role for individual exopolysaccharides.

For P. aeruginosa PA14, mutation of gacA or rhlR suppressed the attenuation of Orsay virus infection observed in the presence of wild-type P. aeruginosa PA14 (Fig. 4C, Supp. Fig. 8). Loss of the rhlR regulators rhlI or pqsE alone had no effect on the attenuation of Orsay virus infection, but simultaneous mutation of both rhlI and pqsE did suppress the attenuation of Orsay virus infection observed in the presence of wild-type P. aeruginosa PA14, similar to that observed for the rhlR mutant, suggesting rhlI and pqsE function redundantly to regulate rhlR in this context (Fig. 4C, Supp. Fig. 8)⁵⁷. Mutation of lasI or lasR suppressed the attenuation of Orsay virus infection to a lesser extent (Fig. 4C, Supp. Fig. 8). Independent mutation of two genes responsible for pel or alg exopolysaccharide production had no effect on infection rates (Fig. 4C, Supp. Fig. 8). We further confirmed that mutation of rhlR and gacA in P. aeruginosa PA01 or P. aeruginosa PA14 suppressed the attenuation of Orsay virus infection via FISH (Fig. 4D and 4E). These data suggest a role for quorum sensing in mediating P. aeruginosa PA14 suppression of Orsay virus infection, as we observed for P. aeruginosa PA01. However, our results obtained in the presence of P. aeruginosa PA14 suggest that there may be some differential regulation of the bacterial effectors responsible in comparison to P. aeruginosa PA01, or additional non-quorum sensing related factors that also mediate suppression.

As P. aeruginosa mutants could suppress the attenuation of infection observed by wild-type P. aeruginosa in the presence of exogenous virus, we next confirmed that these P. aeruginosa mutants similarly affected transmission from infected spreader animals. All P. aeruginosa PA01 quorum sensing mutants suppressed the attenuation of transmission by increasing the incidence proportion >6-fold compared to the wild-type P. aeruginosa PA01 (Fig. 4F). P. aeruginosa PA14 rhlR and gacA mutants suppressed the attenuation of transmission by increasing the incidence proportion 19-fold and 17-fold respectively compared to wild-type P. aeruginosa PA14, but a lasI mutant had minimal effect consistent with the pattern we observed in the susceptibility assay using exogenous Orsay virus (Fig. 4F). Likewise, the magnitude of suppression of the attenuation of Orsay virus transmission observed in the P. aeruginosa rhlR or gacA mutants was greater in the P. aeruginosa PA01 background compared to the P. aeruginosa PA14 background, potentially suggesting the existence of additional strain specific factors that act in P. aeruginosa PA14 to attenuate Orsay virus transmission.

Our findings suggest the possibility that quorum sensing molecules could act directly to attenuate infection. To explore this hypothesis, we prepared liquid E. coli OP50, O. vermis MYb71, P. lurida MYb11, P. aeruginosa PA01, and P. aeruginosa PA14 cultures and added culture supernatant to plates containing E. coli OP50 and Orsay virus. We did not observe any potent effect on host susceptibility to infection by Orsay virus from any supernatant (Supp. Fig. 9), although it is difficult to rule out the possibility that the compounds may act at concentrations that are locally higher than the concentrations at which the compounds are present in our experiments.

P. aeruginosa genes linked to virulence reduce Orsay virus transmission and infection

We designed a candidate-based screen using a non-redundant transposon insertion library to gain further insight into the genetic regulation of P. aeruginosa mediated reduction of Orsay virus transmission and infection (Fig. 5A)⁵⁸. We identified candidate genes to include in our screen from three sources: gacA-regulated genes, rhlR-regulated genes (but rhlI- or pqsE-independent), and the set of genes required for full virulence in C. elegans previously identified by Feinbaum et al. (Fig 5A)^51,59,60. Of the 211 genes tested, 18 putative hits, including lasI, rhlR, and gacA, were identified with the corresponding P. aeruginosa PA14 mutants exhibiting suppression of the attenuation of Orsay virus infection by wild-type P. aeruginosa PA14 (Fig. 6A, Supp. Table 3). Using the 15 novel candidate genes, we performed an additional set of susceptibility assays which confirmed six of the hits (Fig. 5B, Supp. Fig. 10, Supp. Table 3). The identified hits grouped into two clusters based on the strength of their suppression. Transposon insertion in the genes ptsP, prpC, and kinB led to marked suppression of the attenuation of Orsay virus leading to infection of 97%, 89%, and 82% of the population respectively compared to infection in only 4.2% of the population in the presence of wild-type P. aeruginosa PA14 (Fig. 5B, Supp. Fig. 10). Transposon insertion in three additional genes, clpA, glnK, and fabF1 resulted in weaker, but robust suppression of the attenuation of Orsay virus leading to infection of 34%, 28%, and 17% of the population respectively (Fig 5B, Supp. Fig. 10). We additionally tested whether transposon insertion into these genes suppressed attenuation of Orsay virus transmission. While we observed a trend towards mutations affecting susceptibility also affecting virus transmission, we observed a high degree of variation in the transmission assay, such that only mutation of ptsP led to statistically significant suppression of Orsay virus attenuation (Fig. 5C).

Figure 5:

Figure 6:

We noted that all six of the hits originated from the set of genes required for full P. aeruginosa PA14 virulence in C. elegans⁶⁰. These hits represented only six out of the 41 tested genes that are known to influence virulence, indicative of some degree of specificity in the consequences of these mutations beyond their general effect on P. aeruginosa PA14 virulence⁶⁰.

P. lurida MYb11 gacA is required for attenuation of Orsay virus transmission

We next sought to determine whether our findings from the interaction of C. elegans and Orsay virus in the presence of P. aeruginosa could inform us further regarding the mechanisms underlying the attenuation of Orsay virus transmission in the presence of P. lurida MYb11. In particular, we identified gacA, ptsP, prpC, and kinB orthologs using Orthovenn2 and generated a putative knockout allele of each gene by removing the coding potential for all but 10 amino acids from the N and C termini (Fig. 6A)⁶¹. Following exposure to exogenous Orsay virus, knockout of gacA in P. lurida MYb11 led to infection of 77% of the population compared to infection of 43% of the population in the presence of wild-type P. lurida MYb11 from exogenous Orsay virus (Fig. 6B, Supp. Fig. 11A and 11B). These observations were confirmed via FISH, as the gacA mutant supported infection of 62% of the population versus 37% infection in the presence of wild-type P. lurida MYb11 (Fig. 6C). The gacA mutation also suppressed the attenuation of Orsay virus transmission, increasing the incidence proportion 2.9-fold compared to wild-type P. lurida MYb11 (Fig. 6D). These data suggest that gacA has a conserved role between P. aeruginosa and P. lurida MYb11 in the attenuation of Orsay virus transmission and infection of C. elegans. On the other hand, knockout of the P. lurida MYb11 orthologs of ptsP, prpC, or kinB failed to suppress the attenuation of Orsay virus infection or transmission by P. lurida MYb11 (Fig. 6E, Supp. Fig. 11C and 11D)^60,62. One explanation of these results is that these genes play different roles within P. lurida MYb11 and P. aeruginosa and therefore their mutation did not have the same consequences for Orsay virus attenuation. Alternatively, these results might suggest that ptsP, prpC, and kinB act to attenuate Orsay virus infection via some specific effect on P. aeruginosa PA14 virulence but do not influence the interaction of P. lurida MYb11 with C. elegans.

O. vermis MYb71 closely interacts with the C. elegans intestinal brush border

P. lurida MYb11 and O. vermis MYb71 can both be isolated from C. elegans natural environment^36,39. We were therefore curious how exposure to both bacteria simultaneously would impact Orsay virus infection rates. When cultures of O. vermis MYb71 and P. lurida MYb11 were concentrated to 25mg/mL and then mixed, we observed that a 50% mixture of each bacterium (v:v) attenuated infection to the same degree as pure P. lurida MYb11 (Fig. 7A, Supp. Fig. 11A-B). Moreover, we also noted that mixing O. vermis MYb71 with E. coli OP50 also lead to infection rates similar to pure E. coli OP50 (Fig. 7A, Supp. Fig. 11A-B). Previous studies have identified that Ochrobactrum is found in large numbers in the C. elegans intestine following exposure^36,39. We investigated how the levels of Ochrobactrum in the intestine changed following exposure to pure or mixed lawns of bacteria. When alone, GFP expressing Ochrobactrum BH3 was readily observed within the intestine at 4 or 24 h of exposure. Interestingly, exposure to a mixed lawn of Ochrobactrum BH3 and P. lurida MYb11 reduced intestinal accumulation of Ochrobactrum BH3 compared to the pure Ochrobactrum BH3 treatment (Fig 7B and 7C, Supp Fig. 11C-F). On the other hand, exposure to a mixed lawn of Ochrobactrum BH3 and E. coli OP50 reduced accumulation of Ochrobactrum BH3 at four hours compared to a pure Ochrobactrum BH3 treatment, but this difference was minimal by 24 hours (Fig 7B and 7C, Supp Fig. 11C-F).

Figure 7:

These data suggested that the presence of Ochrobactrum in the intestine may be important for its ability to promote Orsay virus infection. We therefore fixed C. elegans exposed to O. vermis MYb71 or E. coli OP50 for examination via electron microscopy and imaged cross sections of the intestine. As expected from our fluorescence experiments, O. vermis MYb71 was readily observed in the intestinal lumen at 4 or 24 h of exposure (Fig. 7D and 7E). The glycocalyx is the mucus-like layer covering C. elegans intestinal cells and that recent work has suggested may play a role in defense against Orsay virus^63,64. The glycocalyx of animals exposed to O. vermis MYb71 was observed with regions of variable thickness and potential instances of O. vermis MYb71 deforming the microvilli brush border as opposed to animals exposed to E. coli OP50 where the glycocalyx appeared uniform and undisturbed (Fig. 7D and 7E). A recent report observed that Serratia marcescens promoted infection of the mosquito Aedes aegypti by Dengue, Zika, and Sindbis viruses by secreting a protein, enhancin, that degrades the mucus layer covering epithelial cells¹⁴. These observations lead us to speculate that O. vermis MYb71 mediated disruption of the brush border and glycocalyx promotes Orsay virus infection although this hypothesis has yet to be tested.

Discussion

In our study, we quantitatively define the effect of bacteria on the transmission of Orsay virus in the C. elegans host. While the presence of O. vermis MYb71 enhanced the transmission of Orsay virus, the presence of P. lurida MYb11 or P. aeruginosa strains PA01 and PA14 attenuated Orsay virus transmission. The enhancement of Orsay virus transmission by O. vermis MYb71 and reduction of transmission by P. lurida MYb11 and P. aeruginosa PA01 and PA14 was mirrored in assays assessing infection rates by exogenous virus. Our results using exogenous virus demonstrate that host susceptibility to Orsay virus infection may vary by over three orders of magnitude in the presence of O. vermis MYb71 versus P. lurida MYb11 which are found in the natural environment in association with C. elegans^36,39. Moreover, we observe that pathogenic P. aeruginosa, which may be found in natural environments, can further attenuate Orsay virus transmission³¹. Our work is also consistent with a recent study by González and Félix that also reported that monoaxenic cultures of other bacteria from the environment of C. elegans impact host susceptibility⁶⁵.

Our observations that two Ochrobactrum species promoted transmission of Orsay virus are intriguing given that other Ochrobactrum species have also been linked to viral infections. Ochrobactrum intermedium promoted poliovirus infection in mice and enhanced poliovirus stability in vitro¹³. Ochrobactrum anthropi, an opportunistic human pathogen, was identified using a random forest analysis as one of the most predictive features differentiating the upper respiratory tract of human patients recovering from influenza infection versus healthy controls⁶⁶. Our observations in the C. elegans host raise the speculative possibility that Ochrobactrum colonization may have roles in evolutionarily diverse hosts in modulating viral infection. Further, our electron micrographs point to an intriguing hypothesis that Ochrobactrum disrupts the brush border region of epithelial cells thereby promoting viral infection. Such a mechanism would be consistent with recent work identifying that Serratia marcescens produces the protein enhancin, which promotes viral infection by degrading the mucus layer covering epithelial cells in the mosquito Aedes aegypti¹⁴. Additional work will need to be performed to explicitly test whether such a mechanism is at play in Ochrobactrum mediated enhancement of Orsay virus infection and transmission.

P. lurida MYb11 and P. aeruginosa PA01 and PA14 shared the ability to attenuate Orsay virus infection and transmission. P. lurida MYb11 promotes developmental rate and fitness at a cost to overall lifespan^39,46. In contrast, P. aeruginosa rapidly kills C. elegans and is detrimental to host fitness^30,44. Under our experimental conditions exposure to all three shortened lifespan compared to exposure to E. coli OP50. Given that O. vermis MYb71 shortened lifespan to the same extent as P. aeruginosa PA14 but had the opposite effect on host susceptibility to Orsay virus, we do not believe that general virulence by any bacterium is responsible for the observed effects on host susceptibility. Rather, unique virulence processes specific to each bacterium may contribute to the unique effects of each bacterium on host susceptibility to Orsay virus.

We identified that gacA regulates the attenuation of Orsay virus by P. aeruginosa PA01 and PA14 and P. lurida MYb11. In Pseudomonads, gacA/gacS regulate processes related to quorum sensing and virulence⁶⁷. Quorum signaling influences additional aspects of P. aeruginosa physiology, including swarming behaviors, biofilm formation, secondary metabolism rates, and overall transcription patterns⁴⁸. To gain additional insights into the P. aeruginosa PA14 effects on Orsay virus transmission we conducted a candidate-based screen of genes influenced by quorum signaling and genes that regulate virulence towards C. elegans^51,59,60. We identified six additional genes, ptsP, prpC, kinB, clpA, glnK, and fabF1, that when mutated suppressed P. aeruginosa PA14 attenuation of Orsay virus infection. Each of these genes, in addition to gacA, rhlR, and lasI has been shown to be required for full P. aeruginosa PA14 virulence in C. elegans⁶⁰. However, we observed that only these six out of the total 41 virulence-related genes identified in Feinbaum et al. influenced Orsay virus infection, arguing further against a role for general virulence in the attenuation of Orsay virus transmission and infection caused by P. aeruginosa PA14 and P. aeruginosa PA01⁶⁰. Additionally, while mutation of the ptsP, prpC, and kinB orthologs of P. lurida MYb11 failed to suppress the attenuation of Orsay virus, it is unclear whether these genes influence the effect of P. lurida MYb11 on host lifespan or even whether P. lurida MYb11 pathogenicity mediated by other factors leads to attenuation of Orsay virus infection.

C. elegans is unlikely to associate with a single bacterium in its natural environment. However, C. elegans shows clear behavioral preferences for grazing on certain bacteria from its environment and may eat monoxenic lawns in the wild⁶⁸. Our data demonstrate that individual bacteria can have a profound impact on Orsay virus transmission rates in a species-specific manner. Additionally, the tractability of the C. elegans-Orsay virus experimental system allowed us to identify molecular determinants of viral transmission and will be useful for identifying additional biotic factors that influence viral transmission. Our work builds on the expanding body of knowledge showing that the microbiota can influence the interactions between viruses and their animal hosts.

Materials and methods

C. elegans strains and growth conditions

C. elegans were maintained on NGM agar plates (17 g agar, 2.5 g peptone, 3 g NaCl per 1 L water) containing E. coli OP50⁴¹. Strains bearing the glp-4(bn2) temperature sensitive mutation were maintained at 16°C, while all other were maintained at 20°C. All assays were performed on SKA assay plates at 20°C (17 g agar, 3.5 g peptone, 3 g NaCl per 1 L water)³⁰. A full list of C. elegans strains used in this study is contained within the Supplement (Supplement Table 1). For all assays relying upon pals-5p::GFP, induction was manually assessed using a Nikon SMZ18 stereofluorescent microscope.

Bacterial strains and growth conditions

All bacteria were grown in Luria broth (10 g tryptone, 5 g yeast, 10 g NaCl per 1 L water). E. coli OP50 and P. aeruginosa strains were grown at 37°C with shaking, while all other strains were grown at 27°C with shaking. A full list of bacterial strains used in this study is contained within the Supplemental Information (Supplement Table 2).

Orsay virus isolation and batch testing

Orsay virus was isolated from infected WUM31(rde-1(ne219);jyIs9[pals-5p::gfp;myo-2p::mCherry]) individuals. Plates with gravid WUM31 were bleached to obtain eggs. Eggs were hatched overnight in M9 solution rotating at 20°C and the L1 larvae were arrested to synchronize the population. L1 larvae were combined with 100 μL of 6x concentrated E. coli OP50 liquid culture and 50 μL of Orsay virus filtrate, plated on SKA plates, and once dried, placed at 20°C for 48 h. Four GFP positive individuals were then transferred to a new 6cm NGM plate with E. coli OP50 and maintained until just starved. Twenty such 6cm plates were washed with 10 mL of M9 and the resulting suspension was Dounce homogenized. Alternatively, two 3.5 cm SKA plates containing infected WUM31 animals were allowed to starve and then equally chunked onto eight 10 cm NGM plates with E coli OP50. Once these plates were just starved, the plates were washed with 10 mL of M9 and homogenized as above. The homogenized suspension was then centrifuged at 13,200xg for five min. The supernatant was passed through a 0.22 μm filter and aliquoted. Each batch of virus was tested for potency and an ID₅₀ calculated for ZD2611 populations in the presence of E. coli OP50 (Supp. Fig. 2). On average, 3.6 μL of Orsay virus stock prepared as indicated in the Methods was required to infect 50% of a population of ZD2611 animals in the presence of E. coli OP50 after 24 hours (Supp. Fig. 2). We observed variability in the measured ID₅₀ even when using the same batch of Orsay virus (e.g. Fig. 1D vs Supp. Fig. 3A vs. Supp. Fig. 3B). We therefore chose to control all our experiments internally rather than attempt to normalize between Orsay virus batches and we report doses used in arbitrary units (a.u.), however 1 a.u. corresponds to 1 μL of Orsay virus filtrate.

Preparation of Uninfected Individuals for transmission and susceptibility assays

Prior to all assays, plates with fecund ZD2611(glp-4(bn-2);jyIs8[pals-5p::gfp;myo-2p::mCherry]) were bleached to obtain eggs. Eggs were hatched overnight in M9 solution rotating at 20°C and the L1 larvae were arrested to synchronize the population. ZD2611 L1s were dropped onto plates containing E. coli OP50 and placed at 20°C for 64-72 h. At this temperature reproduction is delayed, but not eliminated, and no L1s are observed during transmission or susceptibility assays. The resulting young adults were washed off the plates in M9 and centrifuged at 1000xg for one minute. After removing the supernatant, the young adults were once again washed in M9, then centrifuged at 1000 x g for one minute. Young adults were directly dropped onto prepared plates described below.

Preparation of Bacteria for transmission and susceptibility assays

Unless otherwise indicated, bacteria were cultured overnight and added to cover the assay plates. Plates containing E. coli OP50 or P. aeruginosa strains were placed at 37°C for 24 h. Plates containing O. vermis MYb71 or P. lurida MYb11 strains were placed at 25°C for 24 h. All plates were then moved to room temperature for an additional 24 h before the addition of virus.

For the natural isolate transmission screen and early transmission assay cultures were grown to late stationary phase. Cultures were then spun down for six minutes at 4000xg and reconstituted to 25 mg/mL in their own supernatant. 150 μL of this suspension was combined with 50 μL of a mix of M9 for transmission assays or 50 μL of a combination of Orsay virus filtrate and M9 for the early transmission assay. This mixture was combined with young adult ZD2611 individuals and directly added to cover assay plates before drying.

Transmission assays

To obtain infected spreader animals, fecund ZD2610 (glp-4(bn-20); rde-1(ne219); jyIs8[pals-5p::gfp;myo-2p::mCherry]) were bleached to obtain eggs. Eggs were hatched overnight in M9 solution rotating at 20°C and the L1 larvae were arrested to synchronize the population. Arrested ZD2610 L1s were dropped onto NGM plates containing E. coli OP50 and placed at 20°C for 48 h. Animals were then washed off the plates in M9 and centrifuged at 1000xg for one min. After removing the supernatant, the young adults were washed with M9 again then centrifuged at 1000xg for one min. Supernatant was once again removed, and individuals were immediately mixed with 100 μL of 6x overnight E. coli OP50 culture and 50 μL of Orsay virus for a minimum of 18-20 h. pals-5p::GFP-positive ZD2610 young adults were identified and picked onto transfer plates containing the appropriate bacteria prepared as indicated above for 4-6 h. After this period, five spreaders were transferred to assay plates containing the same bacteria and approximately 100 uninfected ZD2611 individuals to start the assay. Transmission assays were scored 24 h after adding spreader individuals.

For the natural isolate transmission screen, arrested ZD2610 L1s were combined with 100 μL of 6x concentrated E. coli OP50 and 50 μL of Orsay virus filtrate and once dried, placed at 20°C for 64-72 h. On the day of the assay, pals-5p::GFP-positive ZD2610 young adults were identified and picked onto transfer plates containing the appropriate bacteria prepared as indicated above for 4-6 h. After this period, five spreaders were transferred to assay plates containing the same bacteria and approximately 100 uninfected ZD2611 individuals to start the assay.

Incidence proportion

Incidence proportion was calculated by dividing the number of newly infected animals by the total non-spreader population per plate. Individual plates for which the total number of pals-5p::GFP positive individuals after 24 h was less than the initial number of spreaders placed on the plate were excluded to eliminate potential differences in incidence proportion caused by unequal numbers of infected spreader animals. For Fig. 1C a single plate was censored from P. lurida MYb11, for Fig. 2A six plates were censored from P. aeruginosa PA14, and for Fig. 4E 12 plates were censored: seven from P. aeruginosa PA14, three from P. aeruginosa PA01, one from P. aeruginosa PA14ΔgacA, and one from P. aeruginosa PA14ΔlasI.

Susceptibility assays

Unless otherwise indicated, Orsay virus filtrate at the indicated doses was diluted in filtered M9 solution. 200 μL total was applied to each assay plate and swirled to cover the entire bacterial lawn. Plates were then dried before the addition of approximately 100 uninfected ZD2611 individuals prepared as indicated above. Assays were scored 24 h later using a Nikon SMZ18 stereo fluorescent microscope and the fraction of pals-5p::GFP-positive individuals was calculated.

Early Transmission Assay

For early transmission assay, young ZD2611 adults were exposed to 0 a.u. or 5 a.u. exogenous Orsay virus in the presence of O. vermis MYb71. 8 hours post infection five infected individuals from the 5 a.u. plate were transferred to the 0 a.u. plate to assess whether transmission from these individuals would occur. Both plates were scored 16 hours after the transfer.

Supernatant Transfer Assay

Assay plates were prepared with E. coli OP50 for susceptibility assays as indicated above. Bacteria were inoculated into 6 mL of LB and allowed to grow for 48 h. Cultures were then spun at 4,000xg for 6 minutes and the spent media supernatant was passed through a 0.22 μm filter. 200 μL of the filtered supernatant was added to the E. coli OP50 assay plates just prior to the addition of Orsay virus and ZD2611 young adults.

Fluorescence in situ hybridization

Previously published probes were obtained to target the RNA1 segment of Orsay virus²⁵. Animals were infected according to the methods described above for susceptibility assays. Animals were processed according to the Stellaris RNA FISH Protocol for C. elegans (LGC Biosearch Technologies) with minor modifications. Briefly, young adults were washed off the plate and rinsed twice in M9. Animals were then fixed for 30 min rotating in a microcentrifuge tube at 20°C. After washing twice with 1 mL of phosphate buffered saline animals were permeabilized in 70% of ethanol and stored at 4°C for 1-7 d. Animals were washed with Wash Buffer A before addition of 100 μL of hybridization buffer containing 3 μL of RNA1 probe mix. Probe was hybridized overnight at 46°C. Animals were then washed with Wash Buffer A alone once, and then again with Wash Buffer A containing 5 ng/mL DAPI. Lastly 100 μL of Wash Buffer B was added before mounting animals on slides with 25 μL of Vectashield Mounting Medium. The fraction of individuals with RNA1 staining was then quantified using a Nikon SMZ18 stereofluorescent microscope or a Zeiss AxioImager Z1 compound fluorescent microscope.

Viral Stability

Assay plates with each bacterium were prepared as described above. 50 a.u. Orsay virus was added to each plate as in a susceptibility assay however no C. elegans were added. Plates were washed with 1 mL of ddH₂0 after 30 minutes or 24 h. A spike-in control was also included where 50 a.u. of Orsay virus was directly added to 1ml H₂0 and processed the same as all other samples. The resulting mixture was spun at 15000 rpm (max speed) at 4°C for 10 min. 1 μl of the supernatant was used to make cDNA (Promega GoScript reverse transcriptase using random primers (A2801)). Orsay virus RNA1 level was quantified by qPCR using 1 μl of 1/5 diluted cDNA (GoTaq Promega A6001) and run on QuantStudio 3 Real Time PCR system (RNA1 qPCR primers GW194 and GW195)²³.

Viral Shedding

Infected ZD2610 animals were prepared as indicated above with infection occurring at the L4 stage. 24 h later as young adults, 20 pals-5p::GFP positive animals were transferred to assay plates prepared with bacteria as indicated above for a transmission assay. These animals were allowed to shed virus onto the plate for 24 h at 20°C. Spreaders were then removed and the assay plates were washed with 1 mL of M9. The resulting mixture was spun at 15000 rpm (max speed) at 4°C for 10 min. 1 μl of the supernatant was used to make cDNA (Promega GoScript reverse transcriptase using random primers (A2801)). Orsay Virus RNA1 level was quantified by qPCR using 1 μl of 1/5 diluted cDNA (GoTaq Promega A6001) and run on QuantStudio 3 Real Time PCR system. (RNA1 qPCR primers GW194 and GW195)²³

RNA Extraction

Animals were washed 5 times in M9 and collected in TRIzol reagent (Invitrogen) and stored at – 80°C before extraction. RNA extraction was performed using Direct-zol RNA microprep kits (Zymo Research) following the manufacturer’s instructions.

qPCR

cDNA was made using 1000 ng of RNA as template (Promega GoScript Reverse Transcriptase using Random Primers). cDNA was diluted at 1/80 and qPCR were performed using 1 μl of diluted cDNA (GoTaq Promega) and run on QuantStudio 3 Real Time PCR system. RNA1 levels were then quantified via qPCR using previously published primers²³. Orsay virus RNA levels were normalized to the internal control host gene snb-1 (primers GCTCAGGTTGATGAAGTCGTC and GGTGGCCGCAGATTTCTC).

Plasmid-based replication experiments

Adult animals carrying a transgene containing the wild-type RNA1 or RNA1D601A Orsay virus genome segment under the control of a heat-inducible promoter were placed on the indicated bacteria for 4 h before heat-shock at 33°C for 2 h⁴⁷. Animals then recovered at 20°C for 20 h before harvesting for RNA extraction and qPCR as detailed above. RNA1 levels were normalized to the values obtained from animals bearing the wild-type RNA1 and exposed to E. coli OP50.

Orsay virus replication in the presence of P. lurida MYb11 and P. aeruginosa

Adult ERT54 (jyIs8[pals-5p::gfp;myo-2p::mCherry]) or WUM31 (rde-1(ne219); jyIs8[pals-5p::gfp;myo-2p::mCherry)] were exposed to exogenous Orsay virus in the presence of P. lurida MYb11, P. aeruginosa PA01, or P. aeruginosa PA14 prepared as in a susceptibility assay. After 2 h or 24 h, animals were harvested for RNA extraction and qPCR as detailed above. RNA1 levels were quantified via qPCR. Within each genotype, the data for each experiment were normalized to the 2 h timepoint.

P. lurida MYb11 mutant construction

A protocol developed for allelic exchange in P. aeruginosa was modified for use in P. lurida MYb11⁶⁹. Briefly, homology arms flanking the region to be deleted were obtained using polymerase chain reaction (PCR) and cloned into the pExG2-KanR suicide vector using Hi-Fi Assembly (New England Biolabs)⁷⁰. DH5α E. coli were transformed using a standard heat shock protocol. Successful transformants were selected for on LB+Kanamycin (50 μg/mL) plates and colony PCR was performed to check for proper insert size in the transformants. E. coli bearing the desired plasmid were grown overnight in LB+Kanamycin (50 μg/mL). Plasmids were then obtained using a Qiagen MiniPrep Kit (Qiagen). Plasmids were assessed for the desired sequence by Sanger sequencing and transformed into P. lurida MYb11 using the following electroporation procedure. P. lurida MYb11 was grown overnight then placed on ice for 30 min. The culture was spun down at 4°C and washed twice with ice-cold water. After reconstitution in 100 μL of ice-cold water the suspension was transferred to a pre-chilled cuvette and transformed at 630 kV using an Eporator (Eppendorf). 900 μL of LB was added, and the suspension was transferred to a microcentrifuge tube and incubated at 27°C for 2 h with shaking. The suspension was then plated on LB+Kanamycin (50 μg/mL) plates and grown for 48 h at 25 °C. Colonies were picked and grown overnight in LB. Cultures were streaked onto sucrose plates (15 g agar, 10 g tryptone, 5 g yeast, 60 g sucrose per 1 L water) to perform sucrose-based counter selection⁶⁹. Colonies that survived were genotyped for the expected deletion.

Lifespan Analysis

Assay plates, bacteria, and ZD2611 animals were prepared as indicated above, however plates were prepared with the addition of 50 μg of 5-fluorodeoxyuridine (FUdR) to eliminate the need to transfer animals during the assay by preventing reproduction. 30 young adult ZD2611 animals were added to each assay plate and assessed for survival every 24 h. Three replicate plates were prepared per experiment and three experiments were performed in total.

Pumping Rate

Assay plates, bacteria, and ZD2611 animals were prepared as indicated above. 20 young adult ZD2611 animals were transferred to each assay plate. At 6 h and 24 h, animals were recorded with slow-motion video capture (1/4x) while feeding. Cycles of pharyngeal pumping were counted and used to calculate the number of pharyngeal pumps per minute.

Mixing Susceptibility Assays and Colonization Assays

Bacteria were grown in LB for 48 h before being spun down at 4,000xg for 6 min. The resulting cell pellets were weighed and reconstituted to 25 mg/mL in their own culture supernatant. Concentrated bacteria were then combined with filtered M9 for colonization experiments or filtered M9 plus 1 μL of Orsay filtrate for mixing susceptibility assays. For mixing susceptibility assays young adult ZD2611 individuals prepared as above were added to the mixture before addition to plates such that the bacterial lawn covered the entire plate. For colonization assays N2 animals were prepared in the same manner to adulthood. Mixing susceptibility assays were scored as a susceptibility assay as indicated above. Colonization assays were scored at 4 h or 24 h. Briefly, animals were picked off of the plate and anaesthetized in sodium azide before mounting on a microscope slide. Fluorescent images of the region of the intestinal lumen behind the pharynx were imaged using a Zeiss AxioImager Z.1 compound microscope. Images were then imported into FIJI for analysis^71,72. An irregular polygon tracing the region behind the pharynx was drawn and the median fluorescence intensity measured.

Electron Microscopy

Animals were prepared for electron microscopy as indicated above for susceptibility assays. Animals were washed off their plates in M9 and allowed to pellet by gravity. The supernatant was removed and animals were anesthetized with sodium azide. Animals were then pipetted into a type A 6mm Cu/Au carrier (Leica) and frozen in a high-pressure freezer (EM ICE, Leica). This was followed by the following freeze-substitution protocol: (EM AFS2, Leica)

Freeze substitution protocol

Cocktail 1: 2% osmium tetroxide, 0.1% uranyl acetate in anhydrous acetone

Program

After the washes in acetone, samples were incubated in propylene oxide for 30 min, then slowly infiltrated in TAAB Epon (TAAB Laboratories Equipment Ltd, https://taab.co.uk) as follows:

3:1 Propylene Oxide:TAAB Epon – 3hrs at RT

1:1 Propylene Oxide:TAAB Epon – ON at 4C

1:3 Propylene Oxide:TAAB Epon – 4 hrs at RT

100% TAAB Epon – 2 hrs at RT

Samples were embedded in fresh TAAB Epon and polymerized at 60 degrees C for 48 hrs.

Ultrathin sections (about 80nm) were cut on a Reichert Ultracut-S microtome, picked up on to formvar/carbon coated copper grids, sections were stained with 0.2% lead citrate. The sections were examined in a JEOL 1200EX Transmission electron microscope and images were recorded with an AMT 2k CCD camera.

Data visualization and statistics

All experiments were performed three times. For transmission, qPCR, FISH-based susceptibility assays, and lifespan assays data from each experiment are combined. For susceptibility assays and colonization assays a single representative experiment is shown and replicates are shown in the supplementary material. Data were analyzed in R Studio⁷³. When comparing all the means of more than two groups p-values were calculated using one-way ANOVA followed by the Tukey HSD test. When comparing multiple experimental groups to a control group p-values were calculated using one-way ANOVA followed by Dunnett’s test. P-values for assays comparing only two groups were calculated using Student’s t-test or Welch’s t-test as indicated. E. coli OP50 is included in all experiments as a reference but was not included for statistical comparison unless explicitly noted. Susceptibility assay curves were modeled using the drc⁷⁴ package in R. A two-parameter log-logistic function was used to model the curve and ID₅₀ values were calculated using the ED function. Kaplan-Meier survival curves were made using the survival⁷⁵ package and curves were compared using the log-rank test within the survdiff function. Plots were made using the gdata⁷⁶, scales⁷⁷, drc⁷⁴, Rmisc⁷⁸, multcomp⁷⁹, survival⁷⁵, ggplot2⁸⁰, ggsignif⁸¹, and cowplot⁸² packages.

Acknowledgements

We would like to thank members of the Kim/Fischer laboratory for helpful comments during the preparation of this work. We would like to thank David Wang, Emily Troemel, Marie-Anne Félix, Eliana Drenkard, Simon Dove, E. Peter Greenberg, Matthew Parsek, Jon Paczkowski, and Read Pukkila-Worley for kindly providing strains. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). We would like to further thank Brendan O’Hara, Michael Gebhardt, and Simon Dove for assistance developing the P. lurida MYb11 transformation protocol. Electron microscopy consultation, sample fixing, and sample sectioning were performed in the HMS Electron Microscopy Facilty by Maria Ericsson and Anja Nordstrom. We additionally thank Rita Droste (MIT) for assistance with electron microscopy. B.G.V. was partially supported by the MIT Department of Biology Graduate Program. B.G.V., N.S., S.E.J.F., and D.H.K were supported by NIH Grant R35GM141794.

Supplementary Information

Supplementary Figure 1:

Supplementary Figure 2:

Supplementary Figure 3:

Supplementary Figure 4:

Supplementary Figure 5:

Supplementary Figure 6:

Supplementary Figure 7:

Supplementary Figure 8:

Supplementary Figure 9:

Supplementary Figure 10:

Supplementary Figure 11:

Supplementary Figure 12:

Supplementary Table 1:

Supplementary Table 2:

Supplementary Table 3: