Serotonergic chemosensory neurons modify the C. elegans immune response by regulating G-protein signaling in epithelial cells

Abstract

The nervous and immune systems influence each other, allowing animals to rapidly protect themselves from changes in their internal and external environment. However, the complex nature of these systems in mammals makes it difficult to determine how neuronal signaling influences the immune response. Here we show that serotonin, synthesized in Caenorhabditis elegans chemosensory neurons, modulates the immune response. Serotonin released from these cells acts, directly or indirectly, to regulate G-protein signaling in epithelial cells. Signaling in these cells is required for the immune response to infection by the natural pathogen Microbacterium nematophilum. Here we show that serotonin signaling suppresses the innate immune response and limits the rate of pathogen clearance. We show that C. elegans uses classical neurotransmitters to alter the immune response. Serotonin released from sensory neurons may function to modify the immune system in response to changes in the animal's external environment such as the availability, or quality, of food.

Figure 1. Exogenous serotonin inhibits the Dar phenotype and decreases pathogen clearance rates.

Adult wild type animals were exposed to M. nematophilum on plates containing exogenous serotonin and the Dar phenotype was scored in their progeny. Treatment with 3.8 mg/ml 5-HT caused a 35% decrease in the number of Dar animals following M. nematophilum infection using standard assay conditions (small lawn) (A). A similar decrease was observed when assay conditions were modified so that animals were unable to avoid the pathogen (big lawn) (A). The Dar phenotype was still decreased when wild type animals were infected in the presence of exogenous serotonin during development, at L1 or L2/3 stage, or 10–18 hours prior to adulthood, at L3/4 stage (B). Similar amounts of M. nematophilum bacteria, labelled using the nucleic acid stain SYTO13, still attached to the anal opening following serotonin treatment (C, D and E) (the rectal opening is indicated with an arrow head in C and D). SYTO13 labeled M. nematophilum was cleared from the anal opening of wild type animals and less than 50% of animals were colonized 90 minutes after transfer to plates without food (F). Treatment of infected animals with 3.8 mg/ml 5-HT significantly decreased the clearance of labeled pathogen from the anal opening (F).

Figure 2. The serotonin biosynthetic enzyme TPH-1 is required in chemosensory neurons to inhibit the Dar phenotype and decrease pathogen clearance rates.

Adult tph-1(mg280) or tph-1(n4622) animals lacking the serotonin biosynthetic enzyme TPH-1 were infected on lawns contaminated with 10% M. nematophilum. The percentage of tph-1(mg280) and tph-1(n4622) progeny with the Dar phenotype was indistinguishable from wild type (A). When animals were infected on lawns contaminated with 0.05% M. nematophilum the Dar phenotype was increased from 60.3% in wild type animals to 92.1% in tph-1(mg280) and 83.0% in tph-1(n4622) (A). This increase could be rescued by treatment with exogenous 5-HT (A) or expression of TPH-1 cDNA in ADF, but not NSM, neurons (B). tph-1(mg280) and tph-1(n4622) animals cleared SYTO13 labeled pathogen more quickly than wild type animals consistent with a role for TPH-1 in suppressing the immune response (C). This phenotype was rescued by 5-HT treatment (C) or expression of a TPH-1 cDNA in both ADF and NSM neurons or ADF neurons alone but not by expression in NSM alone (D). * indicates significance relative to wild type. § indicates significance relative to untreated mutant control (C) or tph-1(mg280) (D) (see materials and methods for details of statistical analysis).

Figure 3. Increased expression of TPH-1 in ADF chemosensory neurons is caused by reduced contact with contaminated bacterial lawns.

Wild type and egl-30(ad805) animals carrying an integrated tph-1p::DSRED transgene were infected with M. nematophilum or an avirulent form of M. nematophilum using standard (small lawn) or “big lawn” assay conditions. The mean tph-1p::DSRED fluorescence in ADF neurons was quantified. Expression of tph-1p::DSRED was significantly increased when wild type animals were grown on small lawns contaminated with virulent M. nematophilum. This increase in expression was not observed under conditions when animals were unable to leave the bacterial lawn; in egl-30(ad805) animals or when wild type animals were infected on “big lawns”.

Figure 4. The Gαo RGS EGL-10 is required in rectal epithelial cells to regulate the immune response and affect pathogen clearance.

egl-10 mutants were infected with M. nematophilum using either standard assay conditions (small lawn) or conditions where animals were unable to avoid the pathogen (big lawn) and the percentage of Dar progeny scored. egl-10(n692) and egl-10(md176) significantly decreased the percentage of Dar animals (A). Expression of EGL-10 cDNA, using a heatshock-inducible promoter, at L1, L2/L3 and L3/L4 stage (hs::EGL-10), or a rectal epithelial promoter (egl-5p::EGL-10), rescued the Dar phenotype in egl-10(n692) animals (A). Although egl-10(n692) animals failed to produce a wild type Dar response similar amounts of M. nematophilum, labeled using the nucleic acid stain SYTO13, still attached to the anal opening (indicated with an arrow in B) (B and C). The rate of clearance of SYTO13 labeled pathogen was significantly decreased in egl-10(n692) animals and this was rescued by expression of EGL-10 cDNA in the rectal epithelium which cleared labeled pathogen more rapidly than wild type animals (D). In D * indicates significance relative to wild type. § indicates significance relative to egl-10(n692).

Figure 5. Rectal epithelial EGL-10 acts downstream of serotonin to modify the immune response and affect pathogen clearance.

Treatment of wild type animals with 3.8/ml 5-HT caused a decrease in the number of Dar animals following infection with M. nematophilum (A) and decreased the clearance of SYTO13 labeled pathogen from the rectal opening (B). Serotonin was unable to decrease the percentage of Dar animals (A) or the rate of pathogen clearance (B) when EGL-10 cDNA was overexpressed in the rectal epithelium of wild type animals suggesting that GOA-1(Gáo) signaling in the rectal epithelium is required for serotonin to suppress the immune response. Animals lacking tph-1 have wild type levels of Dar response on lawns contaminated with 10% M. nematophilum (A) but are more able to clear pathogen infections than wild type (C). Conversely activation of GOA-1(Gáo) using egl-10 loss-of-function mutants results in a decrease in the percentage of Dar animals (A) and infections clear more slowly than wild type animals (C). To determine whether GOA-1(Gáo) acts downstream of serotonin we combined egl-10(n692) with tph-1(mg280) or tph-1(n4622). The percentage of Dar animals (A) and the rate of pathogen clearance was indistinguishable between egl-10(n692) and these double mutants (C). Thus GOA-1(Gáo) signaling acts downstream of serotonin synthesis to suppress the immune response to M. nematophilum infection.

Figure 6. Serotonin synthesis in chemosensory neurons inhibits the immune response by altering rectal epithelial G-protein signaling.

In response to environmental cues, such as the presence or absence of food, serotonin, released from ADF chemosensory neurons acts, directly or indirectly, to regulate GOA-1(Gαo) signaling in the rectal epithelium. This signaling suppresses the Dar phenotype that forms part of the innate immune response and limits the rate of pathogen clearance from the rectal opening.