YAP in epithelium senses gut barrier loss to deploy defenses against pathogens

Pathogens commonly disrupt the intestinal epithelial barrier; however, how the epithelial immune system senses the loss of intestinal barrier as a danger signal to activate self-defense is unclear. Through an unbiased approach in the model nematode Caenorhabditis elegans, we found that the EGL-44/TEAD transcription factor and its transcriptional activator YAP-1/YAP (Yes-associated protein) were activated when the intestinal barrier was disrupted by infections with the pathogenic bacterium Pseudomonas aeruginosa PA14. Gene Ontology enrichment analysis of the genes containing the TEAD-binding sites revealed that “innate immune response” and “defense response to Gram-negative bacterium” were two top significantly overrepresented terms. Genetic inactivation of yap-1 and egl-44 significantly reduced the survival rate and promoted bacterial accumulation in worms after bacterial infections. Furthermore, we found that disturbance of the E-cadherin-based adherens junction triggered the nuclear translocation and activation of YAP-1/YAP in the gut of worms. Although YAP is a major downstream effector of the Hippo signaling, our study revealed that the activation of YAP-1/YAP was independent of the Hippo pathway during disruption of intestinal barrier. After screening 10 serine/threonine phosphatases, we identified that PP2A phosphatase was involved in the activation of YAP-1/YAP after intestinal barrier loss induced by bacterial infections. Additionally, our study demonstrated that the function of YAP was evolutionarily conserved in mice. Our study highlights how the intestinal epithelium recognizes the loss of the epithelial barrier as a danger signal to deploy defenses against pathogens, uncovering an immune surveillance program in the intestinal epithelium.


Introduction
The intestinal epithelium forms a physical barrier to limit entry of pathogens, microbial antigens, and toxins from the luminal environment into the mucosal tissues [1][2][3]. The epithelial barrier function is regulated by an apical junctional complex, consisting of the apical tight junctions (TJs), the subjacent adherens junctions (AJs), and the underlying desmosomes [3,4]. TJs are composed of multiple integral transmembrane proteins, such as claudins, occludins, junctional adhesion molecules, and tricellulin, and cytosolic scaffold proteins, such as cingulin and zonulae occludens (ZOs) [3]. AJs are composed of a cadherin (E-cadherin)-catenin complex and its associated proteins [4,5]. The cytoplasmic tail of E-cadherin interacts with the Armadillo repeat protein β-catenin, which in turn binds α-catenin. TJs and AJs link epithelial cells to form an extremely effective barrier to macromolecules and bacteria [6].
In addition to their primary function as a physically protective barrier, the intestinal epithelial cells can deploy a variety of mechanisms to activate inflammatory and immune responses against microbial pathogens [6][7][8][9]. For example, secretory intestinal epithelial cells are capable of producing and secreting mucins and antimicrobial peptides [7,9]. In addition, intestinal epithelia express various innate pattern recognition receptors (PRPs), such as Toll-like receptors (TLRs) and nucleotide-binding and oligomerization domain (NOD)-like receptors (NLRs), [6,7,9]. These receptors recognize a variety of microbial ligands or endogenous signals associated with pathogenesis, triggering transcriptional expression of inflammatory factors and antimicrobial peptides that coordinate the elimination of pathogens. Persistent bacterial infections cause disruption of the intestinal barrier [10][11][12][13]. Disruption of junction proteins occurs through degradation by bacterial proteases or by biochemical alterations, such as phosphorylation or dephosphorylation [14].
Accumulating evidence has demonstrated that in Drosophila, intestinal epithelial damage caused by toxins and bacterial infection induces proliferation and differentiation of intestine stem cells to replenish damaged cells [15,16]. During epithelial cell damage, Yorkie is activated, and induces such proliferation of intestine stem cells [17,18]. YAP (ortholog of Yorkie in Drosophila) is a major downstream effector of the Hippo signaling pathway [19][20][21]. The core components of Hippo-YAP signaling, initially discovered in Drosophila, are highly conserved in mammals [22]. During activation of this pathway, the MST1/2 (mammalian Ste20-like kinases 1/2, orthologs of Hippo in Drosophila) activates the downstream kinase LATS1/2 (large tumor suppressor 1/2, orthologs of Warts in Drosophila). LATS1/2 then phosphorylates YAP, resulting in its cytoplasmic retention. If Hippo signaling is reduced, the dephosphorylated YAP is translocated into the nucleus, where it binds to and activates the transcription factor TEAD (TEA Domain). Accumulating evidence indicates that YAP interacts with some of the components of the apical junctional complex, such as angiomotin (AMOT) family proteins [23,24] and α-catenin [25,26]. Together, these proteins in TJs and AJs modulate cell proliferation through regulating YAP activity [23][24][25][26].
Recent studies have demonstrated that YAP is involved in innate immune responses to a variety of pathogens. For instance, verteporfin, an inhibitor of YAP/TEAD interactions, renders mice susceptible to gut infection by the helminth parasite Trichuris muris, suggesting a protective role of the YAP-TEAD complex in epithelial cells [27]. In contrast, flies with overexpression of Yorkie in fat bodies exhibit susceptibility to infection by Gram-positive bacteria [28]. In the fat body of flies, Yorkie mediates expression of the Drosophila IkB homolog, Cactus, which in turn suppresses the NF-kB family transcription factors Dorsal (Dl) and Dorsalrelated immune factor (Dif) by retaining the latter in the cytoplasm. Likewise, an inhibitory role of YAP in the NF-kB signaling has also been demonstrated in mammals [29,30]. In addition, YAP has been identified as a negative regulator of host antiviral responses in mammalian cells [31,32]. YAP prevents the phosphorylation and nuclear translocation of the transcription factor IRF3, thus inhibiting IRF3-mediated production of type I interferon and interferonstimulated genes in response to virus infection.
The model organism Caenorhabditis elegans is capable of detecting infection and protecting itself from diverse pathogens [33][34][35]. In this report, we took advantage of the simple intestine of C. elegans to elucidate the role of the intestinal epithelium in host defense against bacterial infection when the epithelial barrier is disrupted. Here we reported that the YAP-TEAD complex was crucial for host defense against bacterial infections. YAP-1/YAP was activated when the E-cadherin complex was disturbed and the intestinal barrier was disrupted during bacterial infections. Mutations in egl-44 and yap-1 led to a defect in host defense against pathogen infections in C. elegans. Similarly, we showed that YAP was also activated in the gut of mice during loss of the intestinal barrier integrity and required for resistance to bacterial infections. Thus, EGL-44/TEAD and YAP-1/YAP play an important role in innate immune responses to bacterial infections.

EGL-44/TEAD transcription factor is involved in host response to disruption of barrier during infection
We first used two dyes, a food dye FD&C Blue No. 1 and the fluorescein isothiocyanate (FITC)-dextran of 40 kD, to evaluate intestinal epithelial barrier disruption [36]. A significant increase in permeability of the barrier was detected in worms at 24 and 48 hours, but not 12 hours, after infection with the Gram-negative bacterium P. aeruginosa PA14 (Fig 1A and 1B). In contrast, the barrier integrity remained intact in worms fed the standard laboratory food bacterium Escherichia coli OP50 (Fig 1A and 1B). Similar results were obtained from worms infected with another Gram-negative bacterium, Salmonella enterica serovar Typhimurium 468 (S1A and S1B Fig).
Two previous studies using whole-genome microarrays revealed that a set of immune response genes, such as C-type lectins, lysozymes, and CUB-like genes, were significantly induced during P. aeruginosa PA14 infection [37,38]. To clarify the mechanisms of innate immune responses mediated by epithelial barrier disruption, we used RNA-Seq analysis to study the transcriptional profiles of worms at 24 hours after P. aeruginosa PA14 infection versus worms fed E. coli OP50. Among the 832 differentially expressed genes, 593 were upregulated, and 239 were downregulated in wild type (WT) animals (P < 0.05 and fold change cutoff > 2.0) (Fig 1C; S1 and S2 Tables). These results revealed the host's broad transcriptional response to P. aeruginosa infection during disruption of the intestinal barrier. To identify transcription factors for host-response induction during intestinal barrier disruption, we investigated which transcription factor binding sites were overrepresented in the promoters of these 832 genes using the BioProspector program [39]. After analyzing 2 kb upstream of the transcription start sites (TSS), we identified an overrepresentation of the CATTCC DNA motif (P <1.58E-05) (Fig 1D; S3 Table). Of these P. aeruginosa-regulated genes, the promoters of 138 genes contain the TEAD-binding sites (Fig 1E; S3 Table). The CATTCC motifs were preferentially localized within the first 1000 bp upstream of potential target TSS ( Fig 1D). This DNA sequence is a consensus motif recognized by TEAD transcription factors, the downstream effectors of the Hippo signaling [19][20][21]. In the C. elegans genome, the gene egl-44 encodes the TEAD homolog [40]. We compared these 138 differentially expressed genes containing the TEAD-binding site with P. aeruginosa PA14 transcriptional responses at 4 hours from a previous study [41]. Of these 138 genes, 36 genes were upregulated and 19 genes downregulated at both 4 hours and 24 hours, whereas 56 genes were upregulated and 27 genes downregulated only at 24 hours, but unchanged at 4 hours (S2 Fig). Therefore, the 138 genes were classified as two clusters: the genes of Cluster I are regulated probably by other transcription factors at 4 hours and by the TEAD/EGL-44-YAP-1/YAP complex at 24 hours, while the genes of Cluster II are regulated mainly by the TEAD/EGL-44-YAP-1/YAP complex at 24 hours during P. aeruginosa PA14 infection. These results suggest that the TEAD/EGL-44-YAP-1/YAP complex is important for host-response induction during disruption of intestinal barrier.

EGL-44/TEAD and YAP-1/ YAP regulates immune response-related genes
Gene Ontology enrichment of the 138 genes containing the TEAD-binding via DAVID revealed that "Innate immune response" (GO: 0045087, P < 3.8E-39) and "Defense response to Gram-negative bacterium" (GO: 0050829, P < 2.5E-19) were two top significantly overrepresented terms (Fig 2A). Of these 39 "innate immune response" genes, 35 genes were upregulated and 4 genes were downregulated after P. aeruginosa PA14 infection. Of these 39 genes in the GO term category "Innate immune response", 26 and 13 belonged to Clusters I and II, respectively. Using qRT-PCR, we tested the expression of 6 upregulated genes of Cluster I (clec-67, clec-85, tsp-1, C49C8.5, T25D10.1, and Y46D2A.2) and 12 upregulated genes of Cluster II (clec-83, clec-86, fbxa-30, fbxa-60, prx-11, zip-10, F35E12.2, F39G3.5, F54E2.1, M04C3.2, valv-1, and clec-264) 24 h after P. aeruginosa PA14 infection. We found that the mRNA levels of these immune response-related genes were significantly increased in WT worms exposed to P. aeruginosa PA14 for 24 hours, compared with those in worms fed E. coli OP50 (Fig 2B). However, the expression of these genes was downregulated by mutations in yap-1(tm1416) and egl-44(mt2247) mutants after P. aeruginosa PA14 infection. In contrast, mutations in yap-1(tm1416) and egl-44(mt2247) did not influence their expression in worms fed E. coli OP50 ( Fig 2B). Meanwhile, we found that knockdown of yap-1 and egl-44 reduced the expression of clec-85p::gfp after P. aeruginosa PA14 infection (Fig 2C and 2D). Finally, we randomly selected 5 upregulated genes without a TEAD binding site (dod-21, pgp-7, F35E12.5, clec-45, and  ) and FITC-dextran staining in wild type (WT) worms exposed to P. aeruginosa PA14 or fed E. coli OP50 at 0 h, 12 h, and 24 h, respectively. (B) Quantification of body-cavity leakages was measured in animals over time. These results are means ± SD of three independent experiments (n � 50 worms per experiment). �� P < 0.01 relative to 0 h (One-way ANOVA followed by a Student-Newman-Keuls test). (C) Volcano Plot for differential gene expression in worms exposed to P. aeruginosa. The red dots represent the significantly upregulated genes and the green dots represent the significantly downregulated genes. The black and blue dots represent the 92 upregulated genes and 46 downregulated genes whose promoters contain the TEAD-binding sites, respectively. (D) Distribution of the identified CATTCC DNA motif in the promoter regions of genes in C. elegans responsive to P. aeruginosa infection. Insert shows logo representation of identified CATTCC DNA motif. (E) A Venn diagram comparing the overlap in P. aeruginosa-regulated genes and the genes whose promoters contain the TEAD-binding sites. Y51B9A.8) and tested their expression after P. aeruginosa PA14 infection. We found that expression of these five genes was upregulated, which was independent of the EGL-44/TEA-D-YAP-1/YAP complex ( Fig 2B). These results demonstrate that the EGL-44/TEAD-YAP-1/ YAP complex specifically controls a set of immune response-related genes in C. elegans.

Infection induces nuclear accumulation of YAP-1/YAP in worms during disruption of intestinal epithelial barrier
The transcriptional activity of TEAD requires YAP as a transcriptional coactivator [42,43]. Thus, the upregulation of the TEAD/EGL-44-mediated immune response is probably due to activation of YAP-1, the C. elegans orthologue of YAP [40]. To test this hypothesis, we examined the effect of pathogen infections on the nuclear translocation of YAP-1/YAP. Using the transgenic worms expressing yap-1p::yap-1::gfp, we observed that, under standard growth conditions, YAP-1/YAP was mainly located in the cytoplasm in all tissues. However, infections by P. aeruginosa PA14 and S. Typhimurium all induced significant nuclear translocation of YAP-1/YAP in the intestine at 24 hours but not 12 hours after infection (Fig 3A-3C). Western blotting analysis revealed that the levels of phosphorylated YAP-1/YAP were markedly reduced at 24 hours but not 12 hours after infections by P. aeruginosa PA14 (Fig 3D and 3F) and S. Typhimurium (Fig 3E and 3F). These results demonstrate that pathogen infections activate YAP-1/ YAP in C. elegans during disruption of intestinal barrier.

EGL-44/TEAD and YAP-1/YAP are required for resistance to pathogens
We next addressed the question of whether YAP-1/YAP and EGL-44/TEAD were involved in resistance to pathogenic bacteria. We found that egl-44(mt2247) and yap-1(tm1416) mutants were more sensitive to killing by P. aeruginosa PA14 ( Fig 4A) and S. Typhimurium ( Fig 4B) than WT worms. Likewise, egl-44 and yap-1 RNAi recapitulated the mutant phenotype ( Fig 4C  and 4D). To determine whether the susceptibility was attributable to accelerated bacterial infections in yap-1 and egl-44 mutants, we tested bacterial accumulation in the intestine, and found that the accumulation of P. aeruginosa PA14 or S. Typhimurium expressing GFP in yap-1(tm1416) and egl-44(mt2247) worms was significantly higher than that in WT worms ( Fig 4E). Furthermore, mutations in yap-1(tm1416) and egl-44(mt2247) markedly increased the colony forming units (CFU) of P. aeruginosa or S. Typhimurium in worms ( Fig 4F).  [40]. Thus, these results indicate that lifespan is not a major factor to determine pathogen susceptibility in yap-1(tm1416) mutants.

Fig 2. EGL-44/TEAD and YAP-1/YAP regulate the expression of immune-related genes after P. aeruginosa infection. (A)
Enrichment analysis of the GO biological process of the P. aeruginosa-regulated genes containing CATTCC DNA motifs is shown. The enrichment P value of each term was transformed to a -log 10 (P-value) . (B) Unsupervised hierarchical clustering of expression levels (qRT-PCR) of YAP-1-EGL-44 complex-dependent immune-related genes as well as YAP-1-EGL-44 complexindependent genes in worms exposed to P. aeruginosa (24 h) by using Origin 2019b. Each column represents an independent replicate. (C) Expression of clec-85p::gfp was upregulated in worms infected with P. aeruginosa compared to worms fed E. coli OP50. Knockdown of egl-44 and yap-1 by RNAi reduced such an increase of clec-85p::gfp expression. (D) Quantification of GFP levels. These results are means ± SD of three independent experiments. �� P < 0.01 relative to OP50 (n � 50 worms per experiment) (One-way ANOVA followed by a Student-Newman-Keuls test). https://doi.org/10.1371/journal.ppat.1008766.g002

Disturbance of the E-cadherin-catenin complex leads to activation of YAP-1/YAP
Accumulating evidence indicates that the apical junctional complex modulates epidermal cells or epithelial proliferation through regulating YAP activity [23][24][25][26]. A strong temporal relationship between the permeability of intestinal barrier and the nuclear translocation of YAP-1/ YAP raises the possibility that disruption of the epithelial barrier by bacterial infection could activate YAP-1/YAP. To test this hypothesis, we used RNAi to silence the major genes encoding TJ proteins (LAD-1/LiCAM and AJM-1), and the genes encoding AJ proteins (HMR-1/Ecadherin, HMP-1/α-catenin, and HMP-2/β-catenin) [44], in worms fed E. coli OP50. We found that genetic inactivation of hmr-1, hmp-1, and hmp-2, but not lad-1 and ajm-1, significantly enhanced the nuclear translocation of YAP-1/YAP ( Fig 5A and 5B), and reduced the levels of phosphorylated YAP-1/YAP in worms fed E. coli OP50 (Fig 5C and 5D). To further confirm the activation of YAP-1/YAP, we determined the expression of the 6 Cluster I genes and 12 Cluster II genes mentioned above in worms subjected to hmr-1 RNAi under normal growth conditions. We found that knockdown of hmr-1 upregulated their expression ( Fig 5E).

PP2A is crucial for the activation of YAP-1/YAP during infection-induced barrier loss
Activation of the Hippo pathway induces the phosphorylation of YAP by the serine/threonine kinase LATS1/2, leading to cytoplasmic sequestration of YAP [22,42]. We hypothesized that P. aeruginosa could promote nuclear translocation of YAP by inactivating the Hippo signaling. However, we found that the phosphorylation levels of WTS-1, the C. elegans orthologue of LATS1/2, were not altered after P. aeruginosa PA14 infection (S6A Fig), suggesting that there was no significant change in the activation status of WTS-1/LATS1/2. Meanwhile, although knockdown of hmr-1, hmp-1, and hmp-2 activated YAP-1/YAP in worms fed E. coli OP50 ( Fig  5A-5C), it failed to influence the phosphorylation levels of WTS-1/LATS1/2 (S6B Fig). These results suggested that the activation of YAP-1/YAP might be independent of the Hippo pathway during disruption of intestinal barrier.
In murine epidermal cells, loss of α-catenin leads to an efficient association of YAP and PP2Ac, the catalytic subunit of the protein phosphatase 2A (PP2A) [25]. PP2Ac mediates the dephosphorylation of YAP, thereby triggering its activation. We thus tested the role of serine/ Quantification of YAP-1 nuclear accumulation (n = 3 independent experiments, n = 100/condition). ��� P < 0.001 relative to 0 h (One-way ANOVA followed by a Student-Newman-Keuls test). (D and E) The levels of phosphorylated YAP-1 were reduced in wild type (WT) worms after P. aeruginosa (D) or S. Typhimurium (E) infection. The blot is typical of three independent experiments. yap-1, yap-1(tm1416) mutants. (F) Quantification of the ratio of p-YAP-1 to total YAP-1. These results are means ± SD of three experiments. �� P < 0.01 relative to 0 h (One-way ANOVA followed by a Student-Newman-Keuls test). https://doi.org/10.1371/journal.ppat.1008766.g003
Next, we tested the role of PP2A in the activation of YAP-1/YAP in worms subjected to genetic inactivation of hmr-1 under normal growth conditions. Knockdowns of let-92, paa-1, and sur-6 by RNAi markedly inhibited the nuclear translocation of YAP-1/YAP (Fig 7A and  7B). Again, knockdowns of let-92, paa-1, and sur-6 substantially restored the phosphorylation levels of YAP-1/YAP in hmr-1 RNAi worms (Fig 7C and 7D). These results suggest that PP2A is required for the activation of YAP-1/YAP during disruption of intestinal barrier.

YAP is required for resistance to P. aeruginosa PA14 or S. Typhimurium in mice
To test whether a similar phenomenon exists in the gut of mice, we first analyzed intestinal permeability in mice given P. aeruginosa PA14 or S. Typhimurium by gavage. Both pathogens dramatically induced intestinal barrier permeability, as measured by the dissemination of FITC-dextran 40 kD from the intestine into the bloodstream at 48 hours, but not 12 hours, after infection (S8A Fig). To evaluate the protein levels of E-cadherin, we performed immunofluorescence staining of the intestinal tissues with antibodies against E-cadherin. Immunofluorescence staining revealed that the protein levels of E-cadherin in the colon of mice were markedly decreased at 48 hours, but not 12 hours, after P. aeruginosa or S. Typhimurium infection (S8B Fig). Immunofluorescence staining also showed a significant increase in the nuclear translocation of YAP in the colon of mice infected with P. aeruginosa PA14 or S. Typhimurium at 48 hours (Fig 8A). Meanwhile, the levels of phosphorylated YAP in the colon were markedly decreased 48 hours after P. aeruginosa PA14 or S. Typhimurium infection ( Fig  8B and 8C). To test the role of YAP in host defense against bacterial infections in vivo, we knocked out the allele of Yap in the epithelium of the small intestine and colon by crossing floxed YAP mice with Villin-Cre (Vil-Cre) mice. Furthermore, we found that Vil-Cre;Yap fl/fl mice showed a significant reduction in survival rate (Fig 8D and 8E) and body weight (Fig 8F   Fig 4. EGL-44/TEAD and YAP-1/YAP are required for host defense against pathogen infections. (A and B) egl-44 and yap-1 mutants were more susceptible to killing by P. aeruginosa (PA) (A) or S. Typhimurium (ST) (B) than wild type worms (WT). P < 0.01 relative to WT (Log-rank test). (C and D) Knockdown of yap-1 and egl-44 led to enhanced susceptibility to P. aeruginosa (C) or S. Typhimurium (D) infection. P < 0.01 relative to empty vector (EV) (Log-rank test). (E and F) Mutations in egl-44 and yap-1 promoted accumulation of pathogens. Fluorescence of worms exposed to P. aeruginosa for 48 h or S. Typhimurium expressing GFP for 72 h is shown (E). The images are representative of three independent experiments. Colony-forming units (CFU) of P. aeruginosa and S. Typhimurium were measured in worms (F). These results are mean ± SD of six independent experiments (n � 50 worms per experiment). � P < 0.05, �� P < 0.01 relative to WT (Two-sample t-test). https://doi.org/10.1371/journal.ppat.1008766.g004

PLOS PATHOGENS
YAP-TEAD in innate immunity and 8G) after P. aeruginosa PA14 or S. Typhimurium infection. These results indicate that YAP is sufficient to confer protection against bacterial pathogenesis in mice.

Discussion
In the Drosophila, Yorkie is activated in response to intestinal epithelial cell damage, and mediates a nonautonomous increase in proliferation and differentiation of intestine stem cells, thereby promoting intestinal regeneration [15][16][17][18]. In this study, we characterized a mechanism by which the intestinal epithelium detects danger and triggers innate immune responses to bacterial infection during a breach of the intestinal epithelial barrier. The E-cadherin-catenin complex, the functional unit of AJs, interacts with YAP-1/YAP in the intestine of worms under normal conditions. When the intestinal barrier is infected, disturbances of the HMR-1/E-cadherin, HMP-1/α-catenin, and HMP-2/β-catenin result in the loss of inhibition of YAP-1/YAP in a PP2A dependent manner. Upon activation, YAP-1/YAP is translocated into the nucleus, where it activates EGL-44/TEAD-mediated transcription. Both yap-1 and egl-44 are required for resistance to bacterial infection in worms and mice. Importantly, conservation of such delicate danger-detecting machinery embedded within the adhesion architecture may represent an ancient immune surveillance program in the intestinal epithelium of multi-cellular organisms.
The YAP-TEAD complex is widely involved in cell proliferation, differentiation, organ size, tissue regeneration, and stem cell self-renewal [42,43]. Recent studies indicate that YAP plays sophisticated roles in innate immune responses to different types of pathogens [27,28,31,32]. YAP is involved in innate immune responses to pathogen infections mainly by influencing multiple immune signaling pathways. Our work here demonstrates that the complex of YAP-1/YAP and EGL-44/TEAD are required for resistance to bacterial infection in worms and mice. The complex function as a transcription factor to regulate the expression of immune response-related genes in worms during bacterial infection. In human umbilical vein endothelial cells, overexpression of YAP induces expression of a set of pro-inflammatory genes, such as IL-6, VCAM1, and ICAM1 [45]. Likewise, hepatocyte-specific overexpression of YAP in mice upregulates the expression of inflammatory factors, including IL-1β and TNFα [46]. Thus, YAP may directly regulate a set of genes related to immune and inflammatory responses. The PMK-1-ATF-7 signaling plays a crucial role in innate immunity in C. elegans [47,48]. A recent study has demonstrated that the PMK-1-ATF-7 signaling causes broad transcriptional changes including the upregulation of many immune-related genes after P. aeruginosa infection [41]. However, our results demonstrate that the expression of genes regulated by the EGL-44/TEAD-YAP-1/YAP complex are independent of the PMK-1-ATF-7 signaling (S9 Fig). Previous studies indicate that pmk-1 mutation leads to increased accumulation of P. aeruginosa PA14 in the gut of worms at 20-24 hours of infection [47,49]. In contrast, significant accumulation of P. aeruginosa or S. Typhimurium in yap-1(tm1416) and egl-44(mt2247) mutants only occurred at 48 hours, but not 24 hours. These data implicate that the EGL-44/TEAD-YAP-1/ YAP signaling exhibits immune function only after the complex is activated during disruption of intestinal barrier at a late stage of bacterial infection. Thus, these two pathways function in parallel to regulate immune-related genes during pathogen infection.

YAP-TEAD in innate immunity
It has been shown that YAP interacts with proteins found in TJs [23,24] and AJs [25,26]. AJs and the E-cadherin-catenin complex have been found to inhibit epithelial proliferation through regulating YAP activity [50,51]. For instance, E-cadherin mediates contact inhibition of proliferation by promoting the cytoplasmic localization of YAP in a Hippo signaling-dependent manner [50]. Conversely, overexpression of YAP blocks the E-cadherin-mediated contact inhibition of proliferation. Moreover, disruption of AJs by calcium depletion, and Ecadherin or α-catenin RNAi in confluent Caco-2 cells promotes nuclear translocation of YAP [51]. Our data showed that disturbance of E-cadherin, α-catenin, or β-catenin by bacterial infections led to the nuclear translocation and activation of YAP. Our discoveries regarding the function of the YAP-TEAD pathway provide additional insights into the role of the E-cadherin-catenin complex in immunity and inflammation. In mammals, the AMOT family proteins of TJs also regulate YAP activity by recruiting YAP to TJs, thus inhibiting its nuclear translocation [23,24]. However, we found that genetic ablation of several components of TJs failed to induce the nuclear translocation of YAP-1/YAP in C. elegans. One possible explanation is that unlike the components of AJs, the TJ components in C. elegans are not conserved.
It has been well-established that the Hippo pathway kinases LATS1 and LATS2 mediate the degradation of YAP [22,42]. In this study, bacterial infections reduced the phosphorylation of YAP-1/YAP in C. elegans during the disruption of intestine barrier. The fact that the phosphorylation of WST-1/LATS1 is not altered implicates that the Hippo signaling itself is unlikely to be involved in the dephosphorylation of YAP-1/YAP during loss of the epithelial barrier. A previous study has demonstrated that α-catenin binding prevents YAP dephosphorylation in murine epidermal cells [25]. Loss of α-catenin leads to an efficient association of YAP and PP2Ac, which potentially dephosphorylates YAP. Consistent with the observation, our data showed that knockdown of hmr-1/E-cadherin-mediated dephosphorylation of YAP-1/YAP was also required for PP2A in the intestine of worms. Meanwhile, PP2A is involved in the activation of YAP-1/YAP in worms after bacterial infections. Thus, PP2A is a critical regulator of YAP activation during loss of epithelial barrier.
Interestingly, compromise of the intestinal epithelial barrier occurs commonly in patients with inflammatory bowel disease (IBD) and celiac disease [3,[52][53][54]. These diseases are characterized by excessive activation of the immune and inflammatory responses. Most current studies attribute the link between intestinal barrier disruption and inflammation to the loss of the physical barrier function and subsequent permeation of luminal bacterial products and antigens and neutrophil transepithelial migration [3,52]. Our findings suggest a central role of the epithelial cells in triggering inflammatory responses after the intestinal barrier is damaged. Thus, our results implicate that the YAP-TEAD pathway likely contributes to the pathogenesis of IBD and celiac disease that involves chronic inflammation of the digestive tract. Therefore, interventions that suppress the YAP-TEAD pathway might provide new therapeutic avenues for IBD and other inflammatory intestinal diseases associated with barrier loss.

PLOS PATHOGENS
YAP-TEAD in innate immunity yuncare20190034). The investigation conforms to the Guide for the Care and Use of Laboratory Animals, published by the US National Institutes of Health. All mice were weighed and monitored daily to ensure animal welfare.

Nematode strains
Mutations and transgenic strains used in this study include egl-44(mt2247) and strains expressing clec-85p::gfp, hmr-1p::hmr-1::gfp, and hmp-1p::hmp-1::gfp, kindly provided by the Caenorhabditis Genetics Center (CGC; http://www.cbs.umn.edu/CGC), which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). yap-1(tm1416) was kindly provided by the National BioResource Project at Tokyo Women's Medical University School of Medicine. The strain that expresses yap-1p::yap-1::gfp was kindly provided by Dr. Y Hata (Tokyo Medical and Dental University, Japan). Mutants were backcrossed three times into the N2 strain used in the laboratory. All strains were maintained on nematode growth media (NGM) and fed with E. coli OP50 at 20˚C [55].

Animals
Kunming mice were obtained from the Animal Center, Kunming Medical University, and used at the ages of 3-4 months. Yap flox/flox and Villin-Cre mice were purchased from the Jackson Laboratory and National Resource Center of Model Mice (NRCMM) respectively. To generate a mouse in which Yap was conditionally deleted in the epithelium of the small intestine and colon, Yap flox/flox mice were bred to Villin-Cre mice.These mice were kept under a constant 12-hour light-dark cycle, and were allowed to eat and drink ad libitum.

Infection of worms with bacteria
Synchronized populations of worms were cultivated at 20˚C until they reached maturity. For survival assays of worms, 75 μg/ml of 5'-fluoro-2'-deoxyuridine (FUdR) was added to the assay plates to abolish the growth of progeny [56]. P. aeruginosa PA14 (a gift from Dr. K Zhu, Institute of Microbiology, CAS), S. enterica Typhimurium 468, and S. aureus ATCC 25923 (a gift from Dr. WH Lee, Kunming Institute of Zoology, CAS) were cultured in Luria broth (LB), then seeded on slow-killing plates as described previously [48]. P. aeruginosa or S. Typhimurium was incubated first for 24 h at 37˚C and then for 24 h at 25˚C. 50-60 young adult animals were transferred onto the plates at 25˚C. The numbers of living worms were measured at 12 h intervals. Immobile worms unresponsive to touch were scored as dead. Worms that crawled off the plate were censored. Three plates of each genotype were used per assay and all experiments were performed four times. The sizes of sample for each assay are provided in S4 Table.

RNA-Seq analysis
After 24 h of infection by P. aeruginosa PA14, worms were collected and sent to Novogene Corporation (Beijing, China) for RNA-sequencing analysis. P < 0.05 and fold change > 2 was used as a threshold for differential expression. The GEO accession number is GSE 121091.

Identification of CATTCC Motif in the Promoters of P. aeruginosa-Induced genes expressed in intestine
We extracted upstream sequences for 832 genes responsive to P. aeruginosa PA14 infection from Ensembl. Shared upstream motifs were identified by using the BioProspector program [39]. To compare the prevalence of identified motifs in promoters of the genes, we searched these promoters using MAST [58]. A motif occurrence is defined as a position in the sequence whose match to the motif has position p-value less than 0.0001. The CATTCC motif logo was prepared using WebLogo (http://weblogo.berkeley.edu).

RNA interference
RNAi bacterial strains containing targeting genes were obtained from the Ahringer RNAi library [59]. In the current study, all clones were verified by sequencing. RNAi feeding experiments were performed on synchronized L1 larvae at 20˚C. Briefly, E. coli strain HT115(DE3) expressing dsRNA was grown overnight in LB broth containing 100 μg/ml ampicillin at 37˚C, then spread onto NGM plates containing 100 μg/ml ampicillin and 5 mM isopropyl 1-thio-β-D-galactopyranoside (IPTG). The RNAi-expressing bacteria were grown overnight at 25˚C. Synchronized L1 larvae were placed on RNAi plates at 20˚C until they reached maturity. Young adult worms were used for further assays.

Construction of hmp-1 RNAi
To generate a clone directed against hmp-1, a 1075 bp DNA fragment of hmp-1 was chemically synthesized (Generay Biotech Co., Shanghai, China), and subcloned into a NotI-KpnI linearized L440 feeding vector [59]. The resulting vector containing hmp-1 was confirmed by DNA sequencing.

YAP-1/YAP Nuclear localization assay
After bacterial infections or treatment with RNAi, the worms expressing yap-1p::yap-1::gfp were immediately mounted in M9 onto microscope slides. The slides were viewed using a Zeiss Axioskop 2 Plus fluorescence microscope. The status of YAP-1/YAP localization was categorized as cytosolic localization or nuclear localization when localization was observed throughout the body from head to tail. When nuclear localization was visible, but not completely throughout the body, the status of YAP-1/YAP localization was characterized as intermediate localization. The numbers of worms with each category of YAP-1/YAP localization were counted. About 100 nematodes were counted in each experiment.

Infection of Mice with P. aeruginosa or S. Typhimurium
To induce infections, SPF mice were infected by gavage with the doses of 1×10 7 CFU of P. aeruginosa or S. Typhimurium expressing GFP for 12 to 72 hours. Then the mice were injected by gavage with FITC-dextran 40 KD (600 mg/kg body weight; Sigma) dissolved in sterile PBS buffer 4 hours before they were anesthetized and sacrificed. Then~0.5 ml of blood was collected by cardiac puncture for measurement of intestinal permeability. Meanwhile, the colon tissues and feces were collected and washed with PBS buffer three times. The tissues were excised for immunofluorescence staining and western blotting.

Measurement of Intestinal Permeability in Mice
Intestine barrier integrity was determined by FITC-dextran leakage assay. 10 μl of serum was transferred into 96-well plates in a Spectra Max M5 fluorescent microplate reader (Molecular Devices, Sunnyvale, CA) for quantification of FITC fluorescence with excitation at 488 nm and emission 518 nm. The levels of serum FITC-dextran were determined using a standard curve.

Quantitative RT-PCR
Total RNA was extracted from worms with TRIzol Reagent (Invitrogen, Carlsbad, CA). Random-primed cDNAs were generated by reverse transcription of the total RNA samples with SuperScript II (Invitrogen). A quantitative real-time PCR analysis was performed using SYBR Premix-Ex TagTM (Takara, Dalian, China) on a Roche LightCycler 480 System (Roche Applied Science, Penzberg, Germany).The relative amount of each mRNA to mRNA of ama-1 was calculated using the method described previously [60]. Primer pair sequences for qPCR amplification were designed using Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/ primer-blast/) and are available upon request. The values of mRNA were normalized against the control gene ama-1.

Immunofluorescence staining
Mouse colon tissues were fixed in acetone for 30 min then frozen at optimal cutting temperature. Standard immunofluorescence staining was performed on 6 μm frozen sections. After blocking with the mixture of 10% normal goat serum overnight, the sections were incubated with primary antibodies specific for YAP (#4912, rabbit polyclonal antibodies,1:200 dilution; Cell Signaling Technology) and E-cadherin (#14472, mouse mAb, 1:50 dilution; Cell Signaling Technology) overnight at 4˚C. The sections were washed three times in PBS with 0.1% Tween-20 (PBST) then incubated with FTIC-conjugated secondary antibodies (#111-095-003, goat anti-rabbit IgG, dilution 1:50, Jackson ImmunoResearch Laboratories, West Grove, PA) for 1 hour, respectively. After washing three times with PBST, the slides were stained with 1 μg/ml of 4,6-diamidino-2-phenylindole (DAPI) for 30 min to detect nuclei. Then the antiquenching agent was used to cover the slices. Images were acquired on a Zeiss Axioskop 2 Plus fluorescence microscope.

Microscopy
For imaging fluorescence in worms, clec-85p::gfp, hmr-1p::hmr-1::gfp, and hmp-1p::hmp-1::gfp animals or worms stained by FITC-dextran were mounted in M9 onto microscope slides. The slides were viewed using a Zeiss Axioskop 2 Plus fluorescence microscope. Averages and standard errors were calculated based on more than 100 worms for each condition. The intensity of GFP was analyzed using the ImageJ software (NIH). Imaging of immunofluorescence staining in mouse tissues was carried out using a Zeiss Axioskop 2 Plus fluorescence microscope.

Detection of bacterial accumulation in worms
Colony forming units (CFU) of P. aeruginosa or S. Typhimurium in the intestine of worms were determined according to the method described previously [61,62]. For measurement of P. aeruginosa or S. Typhimurium CFU, synchronized young adult worms were exposed to plates seeded with P. aeruginosa expressing GFP (a gift from Dr. K Zhu) or S. Typhimurium expressing GFP (the vector for GFP expression is a gift from Dr. XP Qi, Kunming Institute of Zoology, CAS) for 24 to 72 h at 25˚C. Then worms were transferred to NGM plates seeded with E. coli OP50 for 15 min to eliminate P. aeruginosa or S. Typhimurium stuck to the worm bodies. Animals were twice transferred to new NGM plates seeded with E. coli to further eliminate external P. aeruginosa or S. Typhimurium. Then worms were collected and soaked in M9 buffer containing 25 mM levamisole hydrochloride (Sangon Biotech Co., Shanghai, China) for 30 min at room temperature. For surface sterilization, the worms were transferred to M9 buffer containing 25 mM levamisole hydrochloride and 100 μg/ml kanamycin for 1 h at room temperature. Then worms were washed three times with M9 buffer. 50 nematodes per condition were transferred into 50 μl PBS plus 0.1% Triton and ground. The lysates were diluted by PLOS PATHOGENS YAP-TEAD in innate immunity 10-fold serial dilutions in sterilized water and spread over LB agar plates. After one day of incubation at 37˚C, colonies of GFP-P. aeruginosa, or GFP-S. Typhimurium were counted.

Statistics
All experiments were performed at least three times. Differences in survival rates were analyzed using the log-rank test. Differences in mRNA and protein levels, the percentage of body cavity-leakage, and CFU were assessed by performing a one-way ANOVA followed by a Student-Newman-Keuls test. Differences in YAP nuclear accumulation were analyzed using twosample t-test. Data were analyzed using GraphPad Prism 7.0.