Genetic and physiological activation of osmosensitive gene expression mimics transcriptional signatures of pathogen infection in C. elegans

Abstract

The soil-dwelling nematode C. elegans is a powerful system for comparative molecular analyses of environmental stress response mechanisms. Infection of worms with bacterial and fungal pathogens causes the activation of well-characterized innate immune transcriptional programs in pathogen-exposed hypodermal and intestinal tissues. However, the pathophysiological events that drive such transcriptional responses are not understood. Here, we show that infection-activated transcriptional responses are, in large part, recapitulated by either physiological or genetic activation of the osmotic stress response. Microarray profiling of wild type worms exposed to non-lethal hypertonicity identified a suite of genes that were also regulated by infection. Expression profiles of five different osmotic stress resistant (osr) mutants under isotonic conditions reiterated the wild type transcriptional response to osmotic stress and also showed substantial similarity to infection-induced gene expression under isotonic conditions. Computational, transgenic, and functional approaches revealed that two GATA transcription factors previously implicated in infection-induced transcriptional responses, elt-2 and elt-3, are also essential for coordinated tissue-specific activation of osmosensitive gene expression and promote survival under osmotically stressful conditions. Together, our data suggest infection and osmotic adaptation share previously unappreciated transcriptional similarities which might be controlled via regulation of tissue-specific GATA transcription factors.

Figure 1. Expression profiling of the hypertonic stress response in C. elegans.

(A) Hierarchical clustered heat map of gene expression changes associated with physiological activation of osmosensitive gene expression (200 mM NaCl for the indicated time interval) and the regulation of those genes in various osr mutants under normal conditions (50 mM NaCl). For filtering and statistical criteria, see ‘Methods’. (B) Venn diagram showing the overlap between genes differentially regulated by osmotic stress, osm mutants, and dpy mutants.

Figure 2. Osmotically regulated genes exhibit significant overlap with pathogen regulated genes.

(A) Venn diagram showing the genes in common between osmotic stress and a compilation of all pathogen microarray data. Pathogen data were compiled from existing microarray datasets , , , to obtain 5095 unique genes that were significantly regulated in at least one model of infection. (B) Enrichment of the indicated gene class among genes regulated by osmotic stress in wild type or under isotonic conditions in osm and dpy mutants. (C) Hierarchical clustering of antimicrobial cnc, nlp, and clec gene expression in osmotically stressed wild type or unstressed osm and dpy mutants.

Figure 3. GATA transcription factor binding sites are over-represented among upregulated ORGs.

(A) TRANSFAC-annotated transcription factor DNA binding motifs with enrichment scores >1.5 and p-values<0.05 in both C. elegans and C. briggsae ORGs. The number of ORGs with each motif is plotted. For details of enrichment analysis, see ‘Methods’. Only binding motifs with p-values of <0.05 in both C. elegans and C. briggsae datasets are shown. GATA - GATA binding protein 2; TATA - TATA sequence binding protein; USF - Upstream stimulatory factor (USF1 and USF2); Clock:BMAL - brain-muscle-ARNT-like protein 2 (ARNT-2); mtTFA - mitochondrial transcription factor A; HLF - hepatic leukemia factor; N-Myc - neuroblastoma MYC oncogene; c-Myc - c-myc proto-oncogene; Evi-1 - Ecotropic viral integration site 1 transcription factor. (B) Consensus GATA2 motif from 181 occurrences in the C. elegans ORG dataset. (C) Analysis of the gpdh-1 promoter from C. elegans, C. briggsae, and C. remanei. Regions containing >85% identify among all three promoters are indicated in black shading.

Figure 4. GATA binding sites are required in cis for physiological and genetic activation of gpdh-1 expression.

(A) Schematic diagram of the gpdh-1 promoter constructs, with the location of GATA binding sites and the mutagenized sequences indicated by arrows. (B) Animals expressing either the wild type (top) or ΔGATA gpdh-1p::GFP transgene (bottom) were exposed to 200 mM NaCl for 4 or 8 hours before being imaged (B) or quantified using the COPAS Biosort (C). For COPAS data, n>100 young adult animals for two independent wild type lines and six independent ΔGATA lines. (D) Animals carrying either the wild type (top) or the ΔGATA gpdh-1p::GFP transgene (bottom) were fed osm-7(RNAi) or osm-11(RNAi) for two generations. The RNAi treatment was highly effective in all lines, as 100% of osm-7(RNAi) or osm-11(RNAi) animals exhibited an osmotic stress resistance (osr) phenotype. (E) The induction of GFP following EV, osm-7, or osm-11 RNAi was measured on the COPAS Biosort; n>100 for all lines. Horizontal lines indicate data from independent transgenic lines.

Figure 5. The GATA factors elt-2 and elt-3 mediate cell-type specific activation of gpdh-1 expression following physiological or genetic activation.

(A) gpdh-1p::GFP animals exposed to the indicated RNAi treatments were exposed to 200 mM NaCl for 24 hours and GFP was imaged. Scale bar = 75 µ. (B) osm-7(n1515); gpdh-1p::GFP animals were exposed to the indicated RNAi treatments under isotonic conditions and GFP was imaged. Scale bar = 75 µ. Open arrows point to the intestine and closed arrowheads point to the hypodermis.

Figure 6. The GATA factors elt-2 and elt-3 are required to mediate physiological and genetic resistance to hyperosmotic stress in vivo.

(A) Wild type or elt-3(gk121) animals were exposed to either control (EV) or elt-2(RNAi) treatment as described in ‘Methods’. Young adult animals were transferred to 500 mM NaCl plates and survival was measured each 24 hour period. Animals that died from bagging or crawling up the edge of the plate were censored at the time of the event. N>50 animals per genotype. (B) osm-8(n1518) or osm-8(n1518); elt-3(gk121) animals were exposed to either control (EV) or elt-2(RNAi) treatment as described in ‘Methods’. Young adult animals were transferred to 500 mM NaCl and survival was measured each 24 hour period. Animals that died from bagging or crawling up the edge of the plate were censored at the time of the event. N>40 animals per genotype. * - p<0.05.