The microbiota is a key determinant of the physiology and antiviral immunity of animal hosts. The factors governing the transmissibility of viruses between susceptible hosts are incompletely understood. Bacteria serve as food for Caenorhabditis elegans and represent an integral part of the natural environment of C. elegans. We determined the effects of bacteria isolated with C. elegans from its natural environment on the transmission of Orsay virus in C. elegans using quantitative virus transmission and host susceptibility assays. We observed that Ochrobactrum species promoted Orsay virus transmission, whereas Pseudomonas lurida MYb11 attenuated virus transmission relative to the standard laboratory bacterial food Escherichia coli OP50. We found that pathogenic Pseudomonas aeruginosa strains PA01 and PA14 further attenuated virus transmission. We determined that the amount of Orsay virus required to infect 50% of a C. elegans population on P. lurida MYb11 compared with Ochrobactrum vermis MYb71 was dramatically increased, over three orders of magnitude. Host susceptibility was attenuated even further in presence of P. aeruginosa PA14. Genetic analysis of the determinants of P. aeruginosa required for attenuation of C. elegans susceptibility to Orsay virus infection revealed a role for regulators of quorum sensing. Our data suggest that distinct constituents of the C. elegans microbiota and potential pathogens can have widely divergent effects on Orsay virus transmission, such that associated bacteria can effectively determine host susceptibility versus resistance to viral infection. Our study provides quantitative evidence for a critical role for tripartite host-virus-bacteria interactions in determining the transmissibility of viruses among susceptible hosts.
This important study identifies differential Orsay virus infection of C. elegans when animals are fed on different bacteria. The evidence for this is however, incomplete, as experiments to control for feeding rate and bacterial pathogenicity are needed as well as direct quantification of viral load.
Viruses are ubiquitous and abundant1,2. 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 transmission3–5. Abiotic factors such as temperature6 and humidity7 and biotic factors such as viral load8,9 and host immune status10,11 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 organisms12–14. 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 vitro13,16. 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 nutrients17–21. 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 fish23,24. 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 cells23,25. Viral infection activates host defense mechanisms including the RNA-interference response and a transcriptional program known as the intracellular pathogen response23,26–29.
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. elegans30–32. Infection of C. elegans with P. aeruginosa activates innate immunity and stress responses, as well as behavioral avoidance responses33–35. Recently, the bacteria resident in the natural environment of C. elegans in the wild have been of increasing interest22,36–38. The intestinal lumen of free-dwelling C. elegans is occupied by taxonomically and functionally diverse bacteria that can affect C. elegans fitness and physiology36,37,39.
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.
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 be observed spreading 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 OP5036,37,39,41. We set up a transmission assay by placing infected animals (“spreaders”) together with uninfected animals on a monoaxenic lawn of each bacteria and examining the incidence proportion, or, 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 and 1B). 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.
Of further note, we observed a substantial amount of variability in the incidence proportion in the presence of most bacteria (Fig. 1C). Considering both the spreaders and uninfected animals are isogenic populations, the variability we observe between technical and experimental replicates may be due to the inherently stochastic elements of virus transmission. It is striking that the two Ochrobactrum species and P. lurida MYb11 were exceptions to this rule, as it suggests these bacteria somehow mitigate the semi-random elements of virus transmission under these conditions.
We reasoned that increased transmission of Orsay virus could be due to bacterial effects on the shedding of virus by infected spreader animals or bacterial modulation of host susceptibility to viral infection. 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 a fixed amount of Orsay virus to plates of each bacterial species and monitored infection using the pals-5p::GFP reporter. Specifically, we added Orsay virus in doses that ranged across four orders of magnitude and then quantified the fraction of individuals that became infected after 24 h (Fig. 1A and 1D). 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). 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). We further quantified these impacts on host susceptibility by calculating the dose of Orsay virus required to infect 50% of the population after 24 h of exposure, which we defined as the ID50 (Fig. 1D and 1E). On average we observed that the ID50 in the presence of E. coli OP50 was 17-fold higher than the ID50 in the presence of O. vermis MYb71 (Fig. 1E). The ID50 observed in the presence of P. lurida was 120-fold and 2300-fold higher than the ID50 in the presence of E. coli OP50 or O. vermis MYb71 respectively (Fig. 1E).
We corroborated our observations from the pals-5p::GFP reporter by scoring a susceptibility assay 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.
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 extensively30,42. 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). Once more we observed substantial variation in the incidence proportion in the presence of E. coli OP50, while the incidence proportions in the presence of either O. vermis MYb71, P. aeruginosa PA01, or P. aeruginosa PA14 had minimal variation, suggesting a potent overall effect on transmission under these conditions (Fig 2A).
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. 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). 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 33% and 62% of the population respectively compared to E. coli OP50 which supported infection of 85% 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 the presence of P. lurida and P. aeruginosa species to sharply reduce and even effectively block, host susceptibility to infection with Orsay virus.
Attenuation of Orsay virus transmission by P. lurida and P. aeruginosa is not due to effects on Orsay virus replication
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 host43. 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 conditions43. 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 efficiency43. 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.
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 following rare events of natural infection of C. elegans in the presence of P. aeruginosa PA14. We exposed young adult wild-type (N2) or RNAi defective animals (rde-1(ne219)) to exogenous Orsay virus while in the presence of P. aeruginosa PA14 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. In the N2 background viral RNA1 levels increased by 3.5-fold at 24 h relative to levels at 2 h post-exposure (Fig. 3B). Orsay virus is more effective at replicating in the rde-1 background compared to N2 and in the rde-1 background viral RNA1 levels increased by 9000-fold at 24 h relative to levels at 2 h post-exposure (Fig. 3B)23. Together, these results suggest that Orsay virus RNA1 can also replicate once natural infection is achieved in both the wild-type and rde-1- defective background while animals are in the presence of P. aeruginosa PA14. These data confirm our findings that bacteria induced differences in viral replication do not explain the substantial attenuation of Orsay virus transmission in the presence of P. aeruginosa PA14.
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 behavior in response to growing population density44. 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)44. 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)45–47. P. aeruginosa possesses three exopolysaccharide biosynthetic clusters that each can 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 Pel48–52. 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.
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). Knockout of gacA in P. aeruginosa PA01 also suppressed the attenuation of Orsay virus infection (Fig. 4B). On the other hand, mutation of any of the three exopolysaccharide production pathways had no effect on infection (Fig. 4B). These data support a role for quorum sensing regulated processes in reducing Orsay virus infection rates but do not implicate a role for exopolysaccharide biosynthesis.
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). 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)53. Mutation of lasI or lasR suppressed the attenuation of Orsay virus infection to a lesser extent (Fig. 4C). Independent mutation of two genes responsible for pel or alg exopolysaccharide production had no effect on infection rates (Fig. 4C). 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 effects of wild-type P. aeruginosa on C. elegans infection in the presence of exogenous virus, we next confirmed that these P. aeruginosa mutants similarly affected not just host susceptibility, but also 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. 4D). 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. 4D). 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 specifically in P. aeruginosa PA14 to attenuate Orsay virus transmission.
It is additionally interesting to note that similar to the incidence proportion in the presence of E. coli OP50, the variation in the incidence proportion in the presence of all P. aeruginosa mutants increased (Fig 4D). This was true even for the incidence proportion observed in the presence of P. aeruginosa PA14 lasI mutant, which while not statistically different from wild-type P. aeruginosa PA14, had a qualitatively different distribution nonetheless (Fig 4D).
P. lurida 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, a well-characterized pathogen of C. elegans, could inform us further regarding the mechanisms underlying the attenuation of Orsay virus transmission in the presence of P. lurida MYb11, a non-pathogenic constituent of the C. elegans microbiota39. In particular, we identified a gacA ortholog using Orthovenn2 and generated a putative knockout allele by removing the central 194 out of 214 amino acids (Fig. 5A)54. Following exposure to exogenous Orsay virus, knockout of gacA in P. lurida MYb11 led to infection of 80% of the population compared to infection of only 14% of the population in the presence of wild-type P. lurida MYb11 from exogenous Orsay virus (Fig. 5B). 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. 5C). These data suggest that gacA has a conserved role across distant Pseudomonas species in the attenuation of Orsay virus transmission and infection of C. elegans.
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. 6A)55. 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 6A)47,56,57. Of the 201 genes tested, 15 putative hits 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 these 15 candidate genes, we performed an additional set of susceptibility assays which confirmed six of the hits (Fig. 6B, 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 only 4.2% of the population in the presence of wild-type P. aeruginosa PA14. 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. 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 again 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. 6C). However, similar to the effect observed with lasI mutation, the distributions of the incidence proportion observed in the presence of the prpC, kinB, or glnK mutants were qualitatively different from the nearly uniform incidence proportion observed in the presence of wild-type P. aeruginosa PA14 suggesting that mutation of these genes was impacting the attenuation of Orsay virus transmission to some extent (Fig. 6C).
All six of the hits originated from the set of genes required for full P. aeruginosa PA14 virulence in C. elegans. We noted that 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 virulence57. While these six genes have orthologs within P. lurida MYb11, 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(Supp. Fig. 2A and 2B)57,58. One explanation of these results is that these genes play different roles within P. lurida MYb11 and therefore their mutation did not have the same consequences for Orsay virus attenuation. Alternatively, these results might further suggest that ptsP, prpC, and kinB act to attenuate Orsay virus infection via some specific effect on P. aeruginosa PA14 virulence.
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. 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 susceptibility59.
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 controls60. 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.
P. lurida MYb11 and P. aeruginosa PA01 and PA14 shared the ability to attenuate Orsay virus infection and transmission, even as P. lurida and P. aeruginosa have markedly distinct effects on the C. elegans host. P. lurida MYb11 promotes developmental rate and fitness, although may do so at a minor cost to overall lifespan39,61. In contrast, P. aeruginosa is a pathogen of C. elegans and numerous other organisms30,62. We identified that gacA regulates the attenuation of Orsay virus by P. aeruginosa PA01 and PA14 and P. lurida MYb11. The gacS/gacA system regulates many phenotypes in a variety of ψ-proteobacteria63. In Pseudomonads, gacA/gacS regulate processes related to quorum sensing, virulence, and biocontrol in a population density dependent manner63. These data raise the possibility that Pseudomonas quorum sensing pathways are required for the effects on Orsay virus transmission and infection. Quorum sensing also influences additional aspects of P. aeruginosa physiology, including swarming behaviors, biofilm formation, secondary metabolism rates, and overall transcription patterns44.To gain additional insights into the P. aeruginosa effects on Orsay virus transmission we conducted a candidate-based screen including genes influenced by quorum sensing and genes that regulate virulence towards C. elegans47,56,57. We identified six genes, ptsP, prpC, kinB, clpA, glnK, and fabF1, that when mutated suppressed P. aeruginosa PA14 attenuation of Orsay virus infection. Each of these genes has been shown to be required for full P. aeruginosa PA14 virulence in C. elegans57. 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 against a role for general virulence in the attenuation of Orsay virus transmission and infection caused by P. aeruginosa PA14 and P. aeruginosa PA0157.
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 wild64. 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 dose 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 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, 3.5 cm SKA plates containing infected WUM31 animals was allowed to starve and then equally chunked onto four 10 cm NGM plates with E coli OP50. Once these plates were just starved, eight 10 cm plates were washed and homogenized as above. The homogenized suspension was then centrifuged at 13.2xg 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 E. coli OP50. Doses referenced throughout the manuscript refer to actual microliters applied to each plate. However, no method was used to normalize doses between batches of virus and so we have chosen to refer to doses in terms of arbitrary units (a.u.). The average ID50 while on E. coli OP50 for all virus batches used in this study was 3.6 suggesting an average dose of 1 a.u. roughly corresponds to 0.28x the ID50 on Escherichia coli.
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. 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 assay plates before drying.
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. 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, 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 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 GFP positive individuals after 24 h was less than the initial number of spreaders placed on the plate were excluded.
Unless otherwise indicated, Orsay virus filtrate at the indicated doses was diluted in filtered M9 solution. 200 uL 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 GFP-positive individuals was calculated.
Early Transmission Assay
For early transmission assay, young ZD2611 adults were exposed to 0a.u. or 5a.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.
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.
Animals were washed 5 times in M9 and collected in TRIzol reagent (Invitrogen) and stored at - 80C before extraction. RNA extraction was performed using Direct-zol RNA microprep kits (Zymo Research) following the manufacturer’s instructions.
cDNA was made using 500ng of RNA as template (Promega GoScript Reverse Transcriptase/Random Primers). cDNA was diluted at 1/40 and qPCR were performed using 1ul 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.
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 h43. 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. aeruginosa PA14
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. aeruginosa PA14. 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.
Pseudomonas lurida MYb11 mutant construction
A protocol developed for allelic exchange in P. aeruginosa was modified for use in P. lurida MYb1165. 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)66. 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 selection65. Colonies that survived were genotyped for the expected deletion.
Data visualization and statistics
All experiments were performed three times. For transmission, qPCR, and FISH-based susceptibility assays data from each experiment are combined. For susceptibility assays a single representative experiment is shown. Data were analyzed in R Studio67. 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 explicitely noted. Susceptibility assay curves were modeled using the drc68 package in R. A two-parameter log-logistic function was used to model the curve and ID50 values were calculated using the ED function. Plots were made using the gdata69, scales70, drc68, Rmisc71, multcomp72, ggplot273, ggsignif74, and cowplot75 packages.
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. 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.
- 1.Viral metagenomicsNat Rev Microbiol 3:504–510
- 2.Phages across the biosphere: contrasts of viruses in soil and aquatic environmentsRes Microbiol 159:349–357
- 3.Environmental factors affecting the transmission of respiratory virusesCurr Opin Virol 2:90–95
- 4.Animal models for influenza virus pathogenesis, transmission, and immunologyJ Immunol Methods 410:60–79
- 5.Animal models of SARS-CoV-2 transmissionCurr Opin Virol 50:8–16
- 6.Influenza Virus Transmission Is Dependent on Relative Humidity and TemperaturePLoS Pathog 3
- 7.Airborne Transmission of Influenza Virus Infection in MiceNature 195:1129–1130
- 8.Viral Load and Heterosexual Transmission of Human Immunodeficiency Virus Type 1New England Journal of Medicine 342:921–929
- 9.A Mouse Model for the Study of Contact-Dependent Transmission of Influenza A Virus and the Factors That Govern TransmissibilityJ Virol 86:12544–12551
- 10.Mucosal Immunization with a Candidate Universal Influenza Vaccine Reduces Virus Transmission in a Mouse ModelJ Virol 88:6019–6030
- 11.Inactivated Influenza Vaccine That Provides Rapid, Innate-Immune-System-Mediated Protection and Subsequent Long-Term Adaptive ImmunitymBio 6
- 12.Successful Transmission of a Retrovirus Depends on the Commensal MicrobiotaScience (1979) 334:245–249
- 13.Intestinal Microbiota Promote Enteric Virus Replication and Systemic PathogenesisScience (1979) 334:249–252
- 14.A Gut Commensal Bacterium Promotes Mosquito Permissiveness to ArbovirusesCell Host Microbe 25:101–112
- 15.Interaction between microbiota and immunity in health and diseaseCell Res 30:492–506
- 16.Bacteria and bacterial envelope components enhance mammalian reovirus thermostabilityhttps://doi.org/10.1371/journal.ppat.1006768
- 17.Wolbachia and Virus Protection in InsectsScience (1979) 322:702–702
- 18.The Bacterial Symbiont Wolbachia Induces Resistance to RNA Viral Infections in Drosophila melanogasterPLoS Biol 6
- 19.A Wolbachia Symbiont in Aedes aegypti Limits Infection with Dengue, Chikungunya, and PlasmodiumCell 139:1268–1278
- 20.The Endosymbiotic Bacterium Wolbachia Induces Resistance to Dengue Virus in Aedes aegyptiPLoS Pathog 6
- 21.The Antiviral Effects of the Symbiont Bacteria Wolbachia in InsectsFront Immunol 11
- 22.The Natural Biotic Environment of Caenorhabditis elegansGenetics 206:55–86
- 23.Natural and Experimental Infection of Caenorhabditis Nematodes by Novel Viruses Related to NodavirusesPLoS Biol 9
- 24.Natural Viruses of Caenorhabditis NematodesAnnu Rev Genet 53:313–326
- 25.Orsay, Santeuil and Le Blanc viruses primarily infect intestinal cells in Caenorhabditis nematodesVirology 448:255–264
- 26.Competition between virus-derived and endogenous small RNAs regulates gene expression in Caenorhabditis elegansGenome Res 23:1258–1270
- 27.Ubiquitin-Mediated Response to Microsporidia and Virus Infection in C. elegansPLoS Pathog 10
- 28.An evolutionarily conserved transcriptional response to viral infection in Caenorhabditis nematodesBMC Genomics 18
- 29.Terminal uridylyltransferases target RNA viruses as part of the innate immune systemNat Struct Mol Biol 25:778–786
- 30.Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesisProceedings of the National Academy of Sciences 96:715–720
- 31.The environmental occurrence of Pseudomonas aeruginosaAPMIS 128:220–231
- 32.Pseudomonas aeruginosa: a clinical and genomics updateFEMS Microbiol Rev 45
- 33.A Conserved p38 MAP Kinase Pathway in Caenorhabditis elegans Innate ImmunityScience (1979) 297:623–626
- 34.Signaling in the innate immune responseWormBook :1–35https://doi.org/10.1895/wormbook.1.83.2
- 35.Pathogenic bacteria induce aversive olfactory learning in Caenorhabditis elegansNature 438:179–184
- 36.The native microbiome of the nematode Caenorhabditis elegans: gateway to a new host-microbiome modelBMC Biol 14
- 37.Caenorhabditis elegans responses to bacteria from its natural habitatsProceedings of the National Academy of Sciences 113
- 38.Assembly of the Caenorhabditis elegans gut microbiota from diverse soil microbial environmentsISME J 10:1998–2009
- 39.CeMbio - The Caenorhabditis elegans Microbiome ResourceG3 Genes|Genomes|Genetics 10:3025–3039
- 40.Orsay δ Protein Is Required for Nonlytic Viral EgressJ Virol 92
- 41.THE GENETICS OF CAENORHABDITIS ELEGANSGenetics 77:71–94
- 42.Molecular Mechanisms of Bacterial Virulence Elucidated Using a Pseudomonas aeruginosa– Caenorhabditis elegans Pathogenesis ModelCell 96:47–56
- 43.An Evolutionarily Conserved Pathway Essential for Orsay Virus Infection of Caenorhabditis elegansmBio 8
- 44.The hierarchy quorum sensing network in Pseudomonas aeruginosaProtein Cell 6:26–41
- 45.The global activator GacA of Pseudomonas aeruginosa PAO positively controls the production of the autoinducer N-butyryl-homoserine lactone and the formation of the virulence factors pyocyanin, cyanide, and lipaseMol Microbiol 24:309–319
- 46.Pseudomonas aeruginosa GacA, a factor in multihost virulence, is also essential for biofilm formationMol Microbiol 40:1215–1226
- 47.The GacS/GacA signal transduction system of Pseudomonas aeruginosa acts exclusively through its control over the transcription of the RsmY and RsmZ regulatory small RNAsMol Microbiol 73:434–445
- 48.Two Genetic Loci Produce Distinct Carbohydrate-Rich Structural Components of the Pseudomonas aeruginosa Biofilm MatrixJ Bacteriol 186:4457–4465
- 49.Genes involved in matrix formation in Pseudomonas aeruginosa PA14 biofilmsMol Microbiol 51:675–690
- 50.Biosynthesis of the Pseudomonas aeruginosa Extracellular Polysaccharides, Alginate, Pel, and PslFront Microbiol 2
- 51.Identification of psl, a Locus Encoding a Potential Exopolysaccharide That Is Essential for Pseudomonas aeruginosa PAO1 Biofilm FormationJ Bacteriol 186:4466–4475
- 52.Putative Exopolysaccharide Synthesis Genes Influence Pseudomonas aeruginosa Biofilm DevelopmentJ Bacteriol 186:4449–4456
- 53.The PqsE and RhlR proteins are an autoinducer synthase–receptor pair that control virulence and biofilm development in Pseudomonas aeruginosaProceedings of the National Academy of Sciences 115
- 54.OrthoVenn2: a web server for whole-genome comparison and annotation of orthologous clusters across multiple speciesNucleic Acids Res 47:W52–W58
- 55.An ordered, nonredundant library of Pseudomonas aeruginosa strain PA14 transposon insertion mutantsProceedings of the National Academy of Sciences 103:2833–2838
- 56.The PqsE-RhlR Interaction Regulates RhlR DNA Binding to Control Virulence Factor Production in Pseudomonas aeruginosaMicrobiol Spectr 10
- 57.Genome-Wide Identification of Pseudomonas aeruginosa Virulence-Related Genes Using a Caenorhabditis elegans Infection ModelPLoS Pathog 8
- 58.OrthoVenn3: an integrated platform for exploring and visualizing orthologous data across genomesNucleic Acids Res 51:W397–W403
- 59.Natural monobacterial environments modulate viral infection in Caenorhabditis elegansbioRxiv
- 60.Microbiome disturbance and resilience dynamics of the upper respiratory tract during influenza A virus infectionNat Commun 11
- 61.Exploring Effects of C. elegans Protective Natural Microbiota on Host PhysiologyFront Cell Infect Microbiol 12
- 62.Common Virulence Factors for Bacterial Pathogenicity in Plants and AnimalsScience (1979) 268:1899–1902
- 63.Gac/Rsm signal transduction pathway of γ-proteobacteria: from RNA recognition to regulation of social behaviourMol Microbiol 67:241–253
- 64.Dietary choice behavior in Caenorhabditis elegansJournal of Experimental Biology 209:89–102
- 65.Precision-engineering the Pseudomonas aeruginosa genome with two-step allelic exchangeNat Protoc 10:1820–1841
- 66.ExsE, a secreted regulator of type III secretion genes in Pseudomonas aeruginosaProceedings of the National Academy of Sciences 102:8006–8011
- 67.R: A language and environment for statistical computing
- 68.Dose-Response Analysis Using RPLOS ONE 10
- 69.gdata: Various R Programming Tools for Data Manipulation
- 70.scales: Scale Functions for Visualization
- 71.Rmisc: Ryan Miscellaneous
- 72.Simultaneous Inference in General Parametric ModelsBiometrical Journal 50:346–363
- 73.ggplot2: Elegant Graphics for Data Analysis
- 74.ggsignif: R Package for Displaying Significance Brackets for ‘ggplot2’PsyArxiv
- 75.cowplot: Streamlined Plot Theme and Plot Annotations for ‘ggplot2’