Abstract
The intestine integrates food-derived cues to coordinate organismal physiology, yet the molecular specialization of discrete regions along the intestinal epithelium remains unclear. Here, we generate cell type-specific translatomes of the Caenorhabditis elegans intestine during fasting and refeeding using discrete promoters for the anterior quartet of intestinal cells (INT1) and the remaining eight pairs of intestinal cells (INT2-9). We found that the anterior-most INT1 cells are a translationally distinct intestinal subcompartment that is particularly enriched for immune- and stress response genes. Functional assays using novel INT1-specific genes emerging from this study reveal that these specialized cells play a previously unappreciated role in pathogen avoidance and organismal survival. A second critical function of INT1 cells is their role in sensing and responding to the contents of the gut lumen. We show that luminal pyruvate is the key signal linking bacterial nutrients in the gut to secretion of the gut insulin antagonist INS-7. These findings establish INT1 as sentinel enteroendocrine cells that integrate metabolic and immune cues to couple food status with immune and endocrine responses. Our studies also provide a rich resource for dissecting segment-specific intestinal biology, an overlooked and fertile area for future research.
Research organism: C. elegans
Introduction
The intestine is a critical organ for food digestion and nutrient absorption in animals. Caenorhabditis elegans, as a bacterivore, constantly ingests a variety of bacteria with an intestinal transit time as short as 2 min in adults (Ghafouri and McGhee, 2007). This rapid process implies the existence of food-sensing mechanisms within the intestine that can continuously monitor and respond to the acute fluctuation in food availability. When food is absent, a transcriptional metabolic response is evoked via the induction of the fasting response genes, including lipases, peroxisomal and mitochondrial β-oxidation genes, and some fatty acid desaturases in the intestine to maintain energy homeostasis (Van Gilst et al., 2005; O’Rourke and Ruvkun, 2013; Jo et al., 2009). However, it has been unclear whether discrete segments within the intestine have specific food-sensing capabilities or exhibit different responses to fasting in comparison to the organ as a whole. To our knowledge, no studies to date have addressed this question with high spatial and temporal resolution. The C. elegans intestine is comprised of 20 cells designated INT1-INT9 (Sulston et al., 1983). The anterior ring comprises four cells called INT1, which are the most proximal intestinal cells adjacent to the pharynx and represent the initial location of contact with ingested food, which for C. elegans is digested bacterial components. The remaining 16 cells are arranged in pairs to form the latter 8 posterior rings (INT2-9) arranged in a stereotypical structure. A number of years ago, Sulston et al., 1983 had observed that the microvilli of INT1 cells are shorter than those in the rest of the intestine, which implies that INT1 cells may not have the same food absorption capacity that INT2-9 cells have.
In addition to the morphological differences, we recently reported that INT1 functions as an enteroendocrine cell and secretes the gut peptide INS-7 shortly after food withdrawal. INS-7 secretion levels increase over the duration of fasting, which are reset by the re-introduction of food. Gut INS-7 acts in the nervous system to suppress the fat loss-stimulating peptide FLP-7, functional ortholog of the C. elegans tachykinin peptide (Liu et al., 2024). This mechanism allows the stimulation of fat loss to be regulated, in part, by incoming food signals from the gut. Furthermore, Quintin et al. showed that worms treated with H2O2 have an induction of the antioxidant enzyme PRDX-2 specifically in INT1 cells, suggesting that INT1 cells may have additional oxidative stress-sensing functions (Quintin et al., 2022). Although these lines of evidence have suggested that INT1 cells may be a group of specialized intestinal cells that sense the quality of the ingested food and have distinct fasting responses, many questions remain. First, the molecular features of INT1 cells and the molecular distinctions between INT1 and the rest of the intestine (INT2-9) have not been defined. Second, the role of INT1 cells in modulating physiological responses to changes in food availability remains undetermined. Finally, the differential effects of metabolic challenge via fasting and re-feeding in the subsets of intestinal cells are unclear.
In this study, we aim to address these questions using spatial translatomics. In short, the approach is as follows: actively translating mRNAs are bound to clusters of ribosomes, referred to as polysomes (Warner and Knopf, 2002; Warner et al., 1963). Translating ribosome affinity purification (TRAP) was developed to isolate cell type-specific polysome-mRNA complexes (Heiman et al., 2014; Heiman et al., 2008), allowing quantitative analysis of translational regulation (Zhao et al., 2019). Here, we performed TRAP-Seq using a modified protocol based on McLachlan and Flavell, 2019, in which RPL-22 was tagged with HA to enable affinity purification of polysome-mRNA complexes, which is followed by RNA sequencing. This approach provided a comprehensive spatially distinct translatomic profile of actively translating mRNAs in INT1 and INT2-9 cells. To discern novel temporal features in response to metabolic challenge, we examined the translational responses of these intestinal subsets to acute and chronic fasting and refeeding paradigms. Our datasets reveal that INT1 cells exhibit a distinct molecular identity, characterized by the enrichment of immune- and stress-responsive genes. Furthermore, fasting and refeeding elicit divergent regulatory programs in INT1 and INT2-9 cells. To our knowledge, this represents the first study to determine the molecular identities of discrete intestinal cell subsets in response to fasting and refeeding in C. elegans.
Results
Isolation and sequencing of translating RNA from INT1 and INT2-9 cells
To identify the molecular differences between INT1 and INT2-9 cells in fed, fasted, and refed states, we employed TRAP (Heiman et al., 2014; Heiman et al., 2008) to isolate polysomal RNA from INT1 and INT2-9 cells followed by RNA sequencing. This approach captures actively translating mRNAs, thereby providing a cell type-specific snapshot of the translational landscape under fasting and refeeding conditions. To selectively target INT1 and INT2-9 cells, we expressed the HA-tagged ribosomal protein (RPL-22-3xHA) under the control of the Pges-1ΔB (Liu et al., 2024; Schroeder and McGhee, 1998) and Ppho-1 (Beh et al., 1991) promoters, respectively. As published (Liu et al., 2024), we noted that Pges-1ΔB is strictly expressed in INT1 cells, whereas Ppho-1 is expressed in INT2-9 cells with detectable, albeit lower expression in INT2 cells (Figure 1A).
Figure 1. Conditions and workflow for INT1- and INT2-9-specific translatomics.
(A) Representative images of wild-type animals bearing integrated INT1::GFP and INT2-9::mCherry transgenes. Upper panel, GFP expression in INT1 cells and mCherry expression in INT2-9 cells; lower panel, merge with DIC. Scale bar, 100 μm. (B) Representative images of wild-type animals in fed, 30 min fasted, and 30 min refed states fixed and stained with Oil Red O. (C) Fat content was quantified for each condition and expressed as a percentage of wild-type animals in fed state ± SEM (n=20). nsp>0.05 by one-way ANOVA (Tukey). (D) Representative images of wild-type animals in fed, 180 min fasted, 30 min refed, 90 min refed, and 180 min refed states fixed and stained with Oil Red O. (E) Fat content was quantified for each condition and expressed as a percentage of wild-type animals in fed state ± SEM (n=18–20). ***p<0.001, nsp>0.05 by one-way ANOVA (Sidak). (F) 100000 C. elegans expressing INT1::RPL-22::3xHA or INT2-9::RPL-22::3xHA were grown to day 1 adults. Worms were fasted and refed for the time indicated and harvested for TRAP-Seq. Panel F created with BioRender.com.
We operationally defined the ‘acute’ conditions at 30 min fasting followed by 30 min refeeding, because under these conditions, intestinal fat content remains unchanged (Figure 1B and C). However, INS-7 secretion was induced within the 30 min acute fasting period and restored to baseline levels within 30 min upon refeeding (Liu et al., 2024), indicating that intestinal responses to food withdrawal can occur before detectable changes in lipid stores. Chronic conditions were established as 3 hr fasting followed by 90 min refeeding. In this paradigm, animals lost approximately half of intestinal lipid stores after chronic fasting, which were subsequently replenished upon refeeding for 90 min (Figure 1D and E). Thus, chronic conditions represent a more pronounced metabolic challenge. For each TRAP-Seq sample preparation, we cultured 100,000 INT1::TRAP or INT2-9::TRAP animals to the day 1 adult stage, subjected them to acute or chronic conditions (Figure 1F), and performed RNA isolation following a modified TRAP protocol (McLachlan and Flavell, 2019). These experiments were conducted in triplicate.
Stress response genes distinguish INT1 from INT2-9 translatomes across acute and chronic conditions
Principal component analysis (PCA) of all TRAP-Seq datasets revealed clearly distinct clustering by promoter identity, indicating that INT1 and INT2-9 cells maintain fundamentally different molecular identities independent of any and all acute or chronic conditions (Figure 2A). We found PC2 to be significantly associated with promoter identity (FDR = 1.2e-10, Pearson’s correlation) and PC1 with metabolic stress (FDR = 1.4e-6, ANOVA). To systematically define these differences, we performed weighted correlation network analysis (WGCNA) (Langfelder and Horvath, 2008), which unbiasedly identified 12 gene modules associated with condition-specific responses across the INT1 and INT2-9 datasets (Figure 2B and C, Figure 2—figure supplement 1A). Notably, the expression of genes in module red was positively correlated to all INT1 conditions but negatively correlated to all INT2-9 conditions, suggesting a transcriptional program preferentially translated in INT1 cells (Figure 2C, Figure 2—figure supplement 1A). Examination of normalized counts confirmed uniformly higher expression of module red genes across all INT1 conditions compared with INT2-9 (Figure 2D). To identify the categories and functions of module red genes, we used WormCat (Higgins et al., 2022; Holdorf et al., 2020), an annotation and category enrichment tool with near-complete annotation of the C. elegans genome. Among the 100 genes from module red, we found genes involved in “Stress Response”, including C-type lectin, pathogen, and heat categories (Figure 2E). In addition, Gene Ontology (GO) term enrichment analysis of module red genes further highlighted the enrichment for the terms “Immune system process” and “Response to biotic stimulus” (Figure 2—figure supplement 1B).
Figure 2. Identification of the fundamental differences between INT1 and INT2-9 translatomes.
(A) Principal component analysis showing translatomic variability across the first two principal components of TRAP-Seq data from INT1 and INT2-9 promoters under both acute and chronic conditions (fed, fasted, and refed). Ellipse was computed using the mean and covariance of PC1 and PC2 for each line. (B) Cluster dendrogram produced by average linkage hierarchical clustering in WGCNA. The color row under the dendrogram indicates the module assignment. (C) Pearson’s correlations of TRAP-Seq conditions with WGCNA modules. Each row represents a WGCNA module and its mean correlation to the indicated trait across replicates. The color of each cell indicates the mean Pearson’s correlation coefficient between the module and the condition; red represents a positive correlation and blue represents a negative correlation. (D) Average expression of red module genes in the TRAP-Seq conditions. The y-axis shows mean variance stabilized expression across all genes and replicates per condition with standard deviation indicated by the error bars. (E) WormCat visualization of categories enriched in genes of module red. Categories 1 are all bold uppercase; Categories 2 are capitalized; Categories 3 are gray italics. The size of the bubbles indicates the gene counts in the category and the color of the bubble represents the adjusted p-value. (F) Volcano plot of differentially expressed genes for INT1 fed versus INT2-9 fed. High-confidence differentially expressed genes are labeled. The horizontal dashed line indicates an adjusted p-value cutoff of 0.05. (G) WormCat visualization of categories enriched in genes upregulated in INT1 cells versus INT2-9 cells in the fed state. Categories 1 are all bold uppercase; Categories 2 are capitalized; Categories 3 are gray italics. The size of the bubbles indicates the gene counts in the category and the color of the bubble represents the adjusted p-value. (H) A hierarchical clustering tree of significantly enriched GO terms (Biological Process, BP) in the upregulated genes for INT1 fed versus INT2-9 fed by ShinyGO. GO terms with many shared genes are clustered together. Larger dots indicate more significant adjusted p-values. (I) Fluorescent images of transgenic animals bearing PB0024.4::mNeonGreen, Pclec-86::mNeonGreen, Pclec-160::mNeonGreen, PF01D5.5::mNeonGreen or PF01D5.1::mNeonGreen transgenes. The white arrowhead indicates mNeonGreen expression in INT1 cells. Scale bar, 100 μm. (J) Fluorescent image of a transgenic animal bearing Pnpr-28::mNeonGreen transgene. Scale bar, 100 μm.
Figure 2—figure supplement 1. The genes differentially expressed in INT1 cells.
Figure 2—figure supplement 2. InterPro annotations of differentially expressed genes for INT1 fed versus INT2-9 fed.
Recently, Ghaddar et al. presented a gene expression atlas of adult C. elegans with single-cell RNA sequencing (Ghaddar et al., 2023). They categorized the intestinal cells into multiple subtypes: intestine anterior (two), intestine middle, and intestine posterior (two). To place our findings in the context of previously defined intestinal cell subtypes, we compared the results of our datasets to theirs. INT1-enriched genes were upregulated in the anterior intestine cluster relative to other intestinal subtypes, supporting our conclusion that INT1 represents a translationally distinct intestinal subset (Figure 2—figure supplement 1C).
Next, we decided to characterize the translatomic differences between INT1 and INT2-9 cells. We began with the differentially expressed genes (DEGs) in INT1 cells relative to INT2-9 cells in the fed state. There were 153 upregulated and 58 downregulated genes (Figure 2F), and we found ins-7 among the upregulated genes, consistent with our prior report of its restricted expression in INT1 (Liu et al., 2024). To identify the categories of the DEGs in INT1 cells under the fed condition, we performed WormCat category enrichment analysis and also used ShinyGO (Ge et al., 2020) to conduct GO term enrichment analysis with a hierarchical clustering tree, allowing us to visualize the overlapping relationships among enriched GO terms. In the upregulated genes with WormCat analysis, we found genes involved in “Stress Response”, including pathogen, C-type lectin, and heat-shock categories (Figure 2G). Further, the results from ShinyGO demonstrated a cluster of GO terms related to “Immune Response” and “Unfolded Protein Response” (Figure 2H). In contrast, no functional enrichment was detected among the INT1-downregulated genes. Together, these findings show that INT1 cells are characterized by a stress- and immune-responsive translational program in the fed state, marking a strong and clear distinction from the rest of the intestine.
C-type lectins and ShKT domain-containing proteins are enriched in INT1 cells
We noticed that many INT1-differentially expressed genes were not assigned to a category by WormCat or annotated by ShinyGO. To further explore their molecular functions, we used the InterPro database to annotate the protein domains for each gene and assign the domains into several main categories, such as “Enzyme”, “Membrane Protein”, and “Transcription Factor”. Genes encoding domains that did not fall into these categories, as well as those lacking identifiable domains, were grouped as “Genes with Unknown Functions” (Figure 2—figure supplement 2A and B). Among the 149 upregulated genes in INT1, we found 11 encoding C-type lectins (Figure 2—figure supplement 2A). C-type-lectin-domain-containing proteins are notable for pathogen recognition and the initiation of immune responses in vertebrates (Mayer et al., 2017). In C. elegans, the expression of many distinct C-type lectins has previously been found to be induced in a pathogen-specific exposure (Schulenburg et al., 2008; Pees et al., 2021).
In addition, we discovered a subset of INT1 upregulated genes encoding ShKT domains within the genes with unknown functions (Figure 2—figure supplement 2A). The ShK toxin (ShKT) is a potassium (K+) channel blocker originally isolated from the sea anemone Stichodactyla helianthus (Castañeda et al., 1995). ShKT domain-containing proteins have been reported in parasitic nematode secretomes (Loukas et al., 2000) and shown to modulate the activity of human effector memory T cells by blocking voltage-gated potassium channels (Chhabra et al., 2014). Their biological functions in C. elegans, however, remain unclear. Our findings that specific ShKT domain-containing proteins are specifically expressed in INT1 cells are intriguing. Notably, several INT1-upregulated ShKT proteins, including F01D5.1, F01D5.5, and SYSM-1, were previously found to be upregulated by PMK-1 (p38 MAPK) (Troemel et al., 2006), a key regulator of pathogen-induced immune responses in C. elegans (Troemel et al., 2006; Shivers et al., 2008), suggesting that these ShKT proteins may contribute to intestinal immune functions.
To validate our TRAP-Seq findings and define the spatial expression of INT1-enriched genes, we chose key candidate genes with high fold change and strong significance (Figure 2I, Figure 2—figure supplement 1D-H) to generate transcriptional reporters using their endogenous promoters and 3′ UTRs. We started with B0024.4, a membrane raft protein-coding gene with the lowest adjusted p-value in the INT1-enriched genes, which revealed tight expression restricted to INT1 cells. We next decided to examine reporters for two C-type lectin genes, clec-86 and clec-160, and two ShKT protein-coding genes, F01D5.1 and F01D5.5. Reporters for clec-86, clec-160, and F01D5.5 also exhibited highly specific expression in INT1, whereas F01D5.1 was expressed in both INT1 and INT2. Out of curiosity, we also generated a reporter line for the most highly downregulated gene npr-28 in INT1, and interestingly, found that it was restricted to the posterior intestinal cells INT7-9, with no expression in the anterior intestine (Figure 2J). Collectively, these analyses corroborate the spatial restriction of our TRAP-Seq method and demonstrate that INT1-upregulated and INT1-downregulated genes exhibit distinct spatial expression patterns along the intestine.
Comparisons of enriched categories under acute and chronic conditions identify ShKT domain-containing proteins as stable INT1-specific markers
We next wished to evaluate whether there are specific differences between INT1 and INT2-9 cells under acute fasting and refeeding. In the acute fasted state, INT1 displayed 90 upregulated and 12 downregulated genes (Figure 3A). WormCat analysis of upregulated genes revealed those involved in “Stress Response”, including C-type lectin and pathogen categories (Figure 3B), while ShinyGO identified a cluster of GO terms related to “Defense Response” (Figure 3C). In the acute refed state, we observed 52 INT1-upregulated and 35 INT1-downregulated genes (Figure 3D). Enriched categories of the upregulated genes again included “Stress Response”, particularly pathogen and C-type lectins by WormCat (Figure 3E), with GO terms related to “Defense Response” and “Immune Response” by ShinyGO (Figure 3F). InterPro annotations of INT1-upregulated genes revealed consistent representation of “Enzyme”, “C-type Lectin”, and “Membrane Protein” categories across fed, acute fasted, and acute refed states (Figure 3—figure supplement 1A and C). Although enriched functional categories were similar between the 2 intestinal sub-compartments, a Venn diagram analysis of the “Stress Response” genes demonstrated only partial overlap across states (Figure 4A), suggesting that translational differences between INT1 and INT2-9 vary depending on physiological states.
Figure 3. Identification of the differences between INT1 and INT2-9 translatomes under acute and chronic conditions.
(A, D, G, J) Volcano plot of differentially expressed genes for INT1 versus INT2-9 in acute fasted, acute refed, chronic fasted, and chronic refed states. High-confidence differentially expressed genes are labeled. The horizontal dashed line indicates an adjusted p-value cutoff of 0.05. (B, E, H, K) WormCat visualization of categories enriched in genes upregulated in INT1 cells versus INT2-9 cells in acute fasted, acute refed, chronic fasted, and chronic refed states. Categories 1 are all bold uppercase; Categories 2 are capitalized; Categories 3 are gray italics. The size of the bubbles indicates the gene counts in the category and the color of the bubble represents the adjusted p-value. (C, F, I, L) A hierarchical clustering tree of significantly enriched GO terms (Biological Process, BP) in the upregulated genes for INT1 versus INT2-9 in acute fasted, acute refed, chronic fasted, and chronic refed states by ShinyGO. GO terms with many shared genes are clustered together. Larger dots indicate more significant adjusted p-values.
Figure 3—figure supplement 1. InterPro annotations of differentially expressed genes for INT1 versus INT2-9 under acute conditions.
Figure 3—figure supplement 2. InterPro annotations of differentially expressed genes for INT1 versus INT2-9 under chronic conditions.
Figure 4. Comparisons of InterPro categories for INT1 versus INT2-9 under acute and chronic conditions.
(A–H) Venn diagrams showing the number of upregulated or downregulated genes in specific categories for INT1 versus INT2-9 under the three acute conditions (fed, fasted, and refed). (I–Q) Venn diagrams showing the number of genes in specific categories for INT1 versus INT2-9 under the three chronic conditions (fed, fasted, and refed). (R) Venn diagram showing the overlap between overlapped upregulated ShKT proteins for INT1 versus INT2-9 under acute and chronic conditions.
To further test this possibility, we compared gene lists from four major InterPro categories (“Enzyme”, “Membrane Protein”, “C-type Lectin”, and “ShKT Protein”) across fed, acute fasted, and acute refed states (Figure 4B–H). While “Enzyme”, “Membrane Protein”, and “C-type Lectin” categories exhibited only partial overlap, the “ShKT Protein” category showed near-complete overlap across conditions (Figure 4E), suggesting that ShKT proteins represent stable INT1-specific markers under acute conditions.
We next extended our analysis to chronic fasting and refeeding, which elicit more pronounced metabolic shifts in the intestine. In the chronic fasted state, INT1 exhibited 111 upregulated and 77 downregulated genes (Figure 3G). WormCat analysis of the upregulated genes again identified enrichment for “Stress Response” categories, including C-type lectins and pathogen-related genes (Figure 3H), while ShinyGO revealed a cluster of GO terms related to “Immune Response” (Figure 3I). Similar enrichments were observed in the chronic refed state (Figure 3K and L). InterPro annotations of INT1 upregulated genes revealed categories consistent with acute conditions, including “Enzyme”, “Membrane Protein”, and “C-type Lectin” (Figure 3—figure supplement 2A and C). Comparison of gene lists demonstrated partial overlap within these categories (Figure 4J–L and N–P). Strikingly, however, upregulated ShKT genes showed near-complete overlap across chronic conditions (Figure 4M).
Finally, we compared the INT1-upregulated ShKT genes across both acute and chronic conditions. Venn diagram analysis revealed that the 11 ShKT genes shared across acute conditions were also present in the 12-gene set shared across chronic conditions (Figure 4R). These findings demonstrate that ShKT domain-containing proteins represent robust INT1-specific markers, stable across both acute and chronic states. Thus, pathogen-sensing, detoxification, and mounting an organismal response to incoming information from the gut lumen are the principal functions of INT1.
INT1 and INT2-9 cells exhibit distinct fasting and refeeding responses under acute and chronic conditions
Given the fundamentally distinct translatomes of the INT1 and INT2-9 cells, we next investigated whether these differences extend to their responses to fasting and refeeding. Specifically, we sought to determine if the translatomic profiles of INT1 and INT2-9 cells exhibit distinct regulatory patterns in response to metabolic shifts induced by fasting and subsequent refeeding. Looking at the concordance of DEGs during chronic and acute fasting between INT1 and INT2-9 cells, we identified fasting-response genes with opposing regulation between the two intestinal subsets (Figure 5A and B). For instance, 270 genes were upregulated in INT1 following acute fasting but downregulated in INT2-9 under the same condition (Figure 5A), indicating divergent translational responses to metabolic shifts.
Figure 5. Comparison of fasting and refeeding responses between INT1 and INT2-9 cells under acute and chronic conditions.
(A, B) Scatterplots comparing the acute and chronic fasting response genes in INT1 versus INT2-9 differential expression. The top 1000 DEGs by significance were selected for each line and fasting response to be compared, with a regression line fitted to compare INT1 and INT2-9 responses. (C) Table of differentially expressed genes (FDR <0.1) under acute and chronic conditions for INT1 and INT2-9 datasets. (D) Venn diagrams showing the number of differentially expressed genes (FDR <0.1) in fasting and refeeding conditions for INT1 and INT2-9 datasets. (E, F) WormCat visualization of categories enriched in upregulated genes in INT1 cells and INT2-9 cells under chronic fasting or refeeding conditions. Categories 1 are all bold uppercase; Categories 2 are capitalized; Categories 3 are gray italics. The size of the bubbles indicates the gene counts in the category and the color of the bubble represents the adjusted p-value. (E) is part of Figure 5—figure supplement 2A, and (F) is part of Figure 5—figure supplement 3A.
Figure 5—figure supplement 1. WormCat annotations of differentially expressed genes under acute fasting and refeeding conditions for INT1 and INT2-9 cells.
Figure 5—figure supplement 2. WormCat annotations of differentially expressed genes under chronic fasting condition for INT1 and INT2-9 cells.
Figure 5—figure supplement 3. WormCat annotations of differentially expressed genes under chronic refeeding condition for INT1 and INT2-9 cells.
We next listed all DEGs in INT1 and INT2-9 under acute and chronic conditions. Overall, INT1 exhibited fewer DEGs than INT2-9 (Figure 5C). To rule out the possibility that INT1 DEGs represented a subset of INT2-9 DEGs, we used Venn diagrams to compare all the INT1 and INT2-9 DEGs from both acute and chronic conditions (Figure 5D). We found two major features of these fasting and refeeding response genes. Notably, INT1 DEGs were not a subset of INT2-9 DEGs regardless of acute or chronic conditions, but rather represented distinct gene sets (Figure 5D). Another key distinction is that for both INT1 and INT2-9, the acute response genes did not overlap substantially with chronic response genes, indicating that distinct translational programs are engaged during acute versus chronic conditions (Figure 5D).
To identify the enriched categories of these INT1 and INT2-9 DEGs under each condition, we performed WormCat analysis and found that INT1 datasets generally had fewer enriched categories than INT2-9 datasets under each condition (Figure 5—figure supplement 1A and B, Figure 5—figure supplement 2A and B). In the acute fasted state, upregulated genes in both INT1 and INT2-9 were enriched for “Metabolism”, whereas categories including “Stress Response”, “Transcription Factor”, and “Lysosome” were specific to INT2-9 (Figure 5—figure supplement 1A). Among the downregulated genes in both cell types in the acute fasted state, we found genes involved in “Ribosome”, including ribosomal subunit genes (Figure 5—figure supplement 1B). Interestingly, in the acute refed state, genes involved in “Ribosome” were upregulated in INT2-9 (Figure 5—figure supplement 1C), suggesting the sensitivity of ribosomal gene expression to acute fasting and refeeding. Additional categories, such as “Transcription Factor” and “Proteolysis Proteasome”, were enriched exclusively among INT2-9 downregulated genes in the acute refed state (Figure 5—figure supplement 1D).
Further, we observed that chronic conditions elicited broader and more diverse enrichment patterns than acute states in both INT1 and INT2-9 (Figure 5—figure supplements 1–3). Although this is an expected result, the category term “Stress Response” was surprisingly enriched only among INT1-upregulated genes in the chronic fasted state (Figure 5E), whereas “Pathogen” was uniquely enriched among INT1-upregulated genes in the chronic refed state (Figure 5F), underscoring an important potential specialization of INT1 in stress and immune responses. Conversely, both INT1 and INT2-9 datasets shared enrichment for “Ribosome” and “Metabolism” among downregulated genes in the chronic fasted state and upregulated genes in the chronic refed state (Figure 5—figure supplements 2B and 3A), suggesting that protein translation and metabolic regulation are highly responsive to chronic fasting and refeeding.
Taken together, these results demonstrate that INT1 and INT2-9 cells engage distinct translational programs during fasting and refeeding. The divergence is particularly evident under chronic conditions, where INT1 uniquely activates stress- and pathogen-related programs, suggesting specialized roles for INT1 in maintaining homeostasis under chronic conditions. However, because a large portion of the INT1 DEGs remain unassigned by WormCat, the full extent of INT1-specific regulatory programs may not be fully captured by gene set enrichment analyses.
INT1 cells modulate pathogen avoidance, survival, and peptide secretion
To begin to annotate and assign functions to the genes emerging from our datasets, we examined the pathogen response and peptide secretion functions of the INT1 cells, and the roles of the most highly regulated genes therein. To study pathogen response, we noted that previous studies had implicated INT1-enriched genes in immune responses. For example, expression of B0024.4 is induced upon exposure to pathogenic S. maltophilia JV3 strain (Radeke and Herman, 2020), and clec-42, one of our INT1-upregulated genes, is exclusively expressed in INT1 at all larval stages and mediates resistance to pathogen Bt247 (Pees et al., 2021). To test whether INT1-upregulated genes contribute to immune responses, we used an INT1-specific RNAi strain (sid-1; Pges-1ΔB::sid-1) to knock down clec-86 and clec-160, and then performed PA14 avoidance and PA14 slow-killing assays (Tan et al., 1999). Interestingly, animals with INT1-specific knockdown of either gene exhibited impaired PA14 avoidance (Figure 6A and B) and significantly increased survival on PA14 (Figure 6C). These findings suggest that INT1 cells execute specialized, gene-dependent immune functions during pathogen exposure.
Figure 6. INT1-specific genes modulate behavior and metabolism.
(A) The vector, clec-86, and clec-160 RNAi-treated sid-1;INT1::sid-1 animals after 8 hr exposure to PA14 lawns are shown. (B) The distribution of indicated RNAi-treated sid-1;INT1::sid-1 animals at hour 8 after being transferred to PA14 plates. At hour 0, a total of 30 animals were placed near the PA14 lawn. Data are expressed as a percentage of the animals staying inside PA14 lawn at hour 8± SEM (n=3, 3 plates per RNAi treatment, 30 animals on each plate). **p<0.01 by one-way ANOVA (Dunnett). (C) The lifespan of each indicated RNAi-treated animal after being transferred to PA14 lawn was determined by counting the number of live and dead animals every 6 hr until all animals had died. Data are plotted as the percentage of animals that survived on any given hour relative to the number of live animals at hour 0. The number of animals for each RNAi treatment is indicated in the figure panels. ***p<0.001 by log-rank test. The median survival is 44 hr for all the RNAi treatments. (D–F) Average expression of brown, magenta, and turquoise module genes in the TRAP-Seq conditions. The y-axis shows mean variance stabilized expression across all genes and replicates per condition and module. (G) Representative images of sid-1;INT1::sid-1 animals bearing integrated INT1::INS-7mCherry and CLM::GFP transgenes treated with vector RNAi at the indicated time points. Scale bar, 5 μm. (H, J, L) sid-1;INT1::sid-1 animals bearing integrated INT1::INS-7mCherry and CLM::GFP transgenes were treated with vector, aakg-4 or fmo-2 RNAi. INS-7 secretion levels were measured after 30 min fasting and after 30 min refeeding. The intensity of INS-7mCherry fluorescence within a single coelomocyte was quantified and normalized to the area of CLM::GFP expression for each RNAi treatment. Data are expressed as a percentage of the normalized INS-7mCherry fluorescence intensity of fed animals treated with vector RNAi ± SEM (n=21–27). *p<0.05, **p<0.01, ***p<0.001, nsp>0.05 by one-way ANOVA (Dunnett T3). (I, K) Normalized expression plots of aakg-4 and fmo-2 under all conditions in INT1 and INT2-9 cells by replicate.
Figure 6—figure supplement 1. WormCat annotations of undulating genes in INT1 and INT2-9 cells.
From the WGCNA, we noted that some modules (brown, magenta, turquoise) contained genes with an undulating change in expression levels in response to fasting and refeeding under both acute and chronic conditions (Figure 6D–F). These genes could be classified as “fasted up/refed down” or “fasted down/refed up”. WormCat analysis of brown module genes identified enrichment for “Ribosome” and “Metabolism”, consistent with categories previously linked to ribosomal subunit genes (Figure 5—figure supplements 2B and 3A and Figure 6—figure supplement 1A). Notably, modules magenta and turquoise contained “fasted up/refed down” genes. To determine the strength of the induction across lines we used a Wilcoxon test comparing the distribution of the normalized expression of module genes between fed and fasted conditions. We found the induction in INT2-9 cells (p=5.6e-4, average LFC = 0.07) after acute fasting was weaker in the turquoise module when compared to INT1 cells (p=5.6e-50, average LFC = 0.12), suggesting these undulating genes were more INT1-specific (Figure 6E and F).
Given our prior finding that the gut peptide INS-7 is secreted from INT1 cells in response to acute fasting (Liu et al., 2024), we next investigated whether INT1-specific undulating genes could modulate this response. Using the INT1-specific RNAi strain (sid-1; Pges-1ΔB::sid-1) crossed into an INS-7 secretion reporter, we confirmed that vector RNAi-treated animals displayed the expected undulating INS-7 secretion levels upon acute fasting and refeeding (Figure 6G and H), as previously described (Liu et al., 2024). We next targeted two INT1-specific undulating genes, aakg-4 and fmo-2, which showed the largest absolute fold changes across acute fasting and refeeding for INT1-specific RNAi (Figure 6I and K). aakg-4 is the gene encoding one of the five γ subunits of the AMP-activated protein kinase (AMPK) in C. elegans (Tullet et al., 2014). However, the physiological role of aakg-4 in the intestine is unknown. Remarkably, INT1-specific aakg-4 knockdown prevented INS-7 secretion from returning to basal levels after refeeding, indicating a requirement for aakg-4 in resetting INS-7 secretion and suggesting that INS-7 release from INT1 cells is subject to energy-dependent regulation, consistent with its modulation by fasting and refeeding (Figure 6J). The second gene fmo-2, encoding flavin-containing monooxygenase-2, is known to promote longevity under hypoxia and dietary restriction (Leiser et al., 2015). FMOs contain the prosthetic group that binds FAD+ and utilize the cofactor NADPH together with O2 to catalyze substrate oxygenation, generating NADP+ as a by-product (Ziegler, 2002). Animals subjected to INT1-specific fmo-2 knockdown displayed a normal fasting-induced INS-7 secretion, but exhibited reduced basal INS-7 secretion and impaired refeeding response (Figure 6L). Given that the NADPH/NADP+ ratio is a central determinant of cellular redox state and energy metabolism (Xiao et al., 2018), this result further supports the notion that regulation of INS-7 secretion in INT1 cells is coupled to energy-sensing metabolic pathways.
Together, these results demonstrate that INT1 cells integrate immune and metabolic cues. By regulating pathogen avoidance and survival through lectin genes and modulating INS-7 secretion via energy-responsive genes such as aakg-4 and fmo-2, INT1 cells function as a specialized intestinal subset that senses energy states and coordinates gut peptide secretion to maintain organismal homeostasis.
Pyruvate is crucial for regulating INS-7 peptide secretion from INT1 cells
We were intrigued by the findings with aakg-4 and fmo-2, which suggested that cellular energy status may govern the refeeding response of INS-7 secretion. Notably, the effect does not appear to be mediated by changes in intestinal fat content, as lipid levels remain unaltered during acute fasting and refeeding (Figure 1B and C). These observations led us to hypothesize that INT1 cells may integrate specific sensory cues to reset INS-7 secretion in response to food availability. To explore this possibility, we first evaluated the contribution of mechanosensory inputs. Fasted worms refed with 0.5 μm latex beads, comparable in size to Escherichia coli (Raizen et al., 2012; Fukushige et al., 1998), failed to restore INS-7 secretion to basal levels, whereas refeeding with OP50 fully normalized INS-7 secretion (Figure 7A), suggesting that mechano-stimulus alone was not sufficient to induce the refeeding response of INS-7 secretion. Because C. elegans can detect secreted molecules from bacteria and trigger a stress response (Samuel et al., 2016), we next exposed fasted animals to cell-free OP50 culture supernatant (Samuel et al., 2016; Zaslaver et al., 2015). Supernatant alone likewise failed to reset INS-7 secretion (Figure 7A). A combination of beads and supernatant was similarly ineffective (Figure 7A). In contrast, UV-killed OP50 completely reinstated basal INS-7 secretion (Figure 7A), indicating that chemicals or nutrients from bacteria, rather than mechano-stimulus or secreted molecules, mediate the refeeding response.
Figure 7. Pyruvate is critical for maintaining INS-7 secretion dynamics in INT1 cells.
(A) INS-7 secretion dynamics during fasting and re-feeding with OP50, latex beads, day 1 OP50 supernatant (D1SN), day 1 OP50 supernatant with latex beads (D1SN + Beads), and UV-killed OP50 were determined at the indicated time points. The intensity of INS-7mCherry fluorescence within a single coelomocyte was quantified and normalized to the area of CLM::GFP expression for each time point. Data are expressed as a percentage of the normalized INS-7mCherry fluorescence intensity of wild-type fed animals ± SEM (n=20–26). **p<0.01, ***p<0.001, nsp>0.05 by one-way ANOVA (Dunnett T3). (B) INS-7 secretion dynamics during fasting and re-feeding with OP50, 2% glucose, and 100 mM pyruvate were determined at the indicated time points. The intensity of INS-7mCherry fluorescence within a single coelomocyte was quantified and normalized to the area of CLM::GFP expression for each time point. Data are expressed as a percentage of the normalized INS-7mCherry fluorescence intensity of wild-type fed animals ± SEM (n=14–26). **p<0.01, ***p<0.001, nsp>0.05 by one-way ANOVA (Dunnett T3). (C–F) INS-7 secretion dynamics in INT1-specific vector RNAi, mpc-1 and mpc-2 RNAi, pyc-1 RNAi, pdha-1 and pdhb-1 RNAi-treated animals during fasting and re-feeding with 100 mM pyruvate were determined at the indicated time points. The intensity of INS-7mCherry fluorescence within a single coelomocyte was quantified and normalized to the area of CLM::GFP expression for each time point. Data are expressed as a percentage of the normalized INS-7mCherry fluorescence intensity of vector RNAi fed animals ± SEM (n=13–24). *p<0.05, **p<0.01, ***p<0.001, nsp>0.05 by one-way ANOVA (Dunnett T3). (G) INS-7 secretion dynamics during fasting and fasting with the presence of 100 mM pyruvate were determined at the indicated time points. The intensity of INS-7mCherry fluorescence within a single coelomocyte was quantified and normalized to the area of CLM::GFP expression for each time point. Data are expressed as a percentage of the normalized INS-7mCherry fluorescence intensity of fed animals ± SEM (n=16–19). ***p<0.001, nsp>0.05 by one-way ANOVA (Dunnett T3).
Figure 7—figure supplement 1. The role of pyruvate in refeeding responses of INS-7 secretion from INT1 cells.
Next, we set out to pinpoint the bacterial nutrient or chemical most critical for regulating INS-7 secretion during refeeding. Based on the requirement of aakg-4 and fmo-2 for INS-7 secretion dynamics, we tested whether restoring cellular energy states after fasting could return INS-7 secretion to baseline. Glucose serves as a primary cellular energy source, undergoing glycolysis to generate pyruvate, which in turn fuels downstream metabolic pathways to produce energy. However, refeeding fasted worms with 2% glucose, a concentration considered a high-glucose diet for C. elegans (Lee et al., 2009), failed to normalize INS-7 secretion, which remained elevated (Figure 7B). In contrast, remarkably, refeeding with 100 mM pyruvate, a dose effective in prior studies (Tauffenberger et al., 2019), fully recapitulated the refeeding response observed with OP50 (Figure 7B). This effect was dose-dependent (Figure 7—figure supplement 1A). Thus, pyruvate, a key molecule derived from glycolysis and a precursor metabolite for numerous biochemical processes, is sufficient to substitute for bacterial food in restoring INS-7 secretion to baseline after fasting.
After entering the cell, pyruvate can be converted to lactate or alanine in the cytosol or transported into the mitochondrial matrix via the mitochondrial pyruvate carrier (MPC) (Bricker et al., 2012; Herzig et al., 2012) to enter the citric acid cycle, which generates energy (Gray et al., 2014). INT1-specific RNAi of the two C. elegans MPC genes (mpc-1 and mpc-2) abolished the pyruvate-induced restoration of INS-7 secretion (Figure 7C and D), demonstrating that mitochondrial import of pyruvate is required. Within the matrix, pyruvate is converted to acetyl-CoA by pyruvate dehydrogenase (pdha-1 and pdhb-1 in C. elegans) or to oxaloacetate by pyruvate carboxylase (pyc-1 in C. elegans). INT1-specific knockdown of either pathway suppressed fasting-induced INS-7 secretion and abolished secretion dynamics (Figure 7E and F). Since chronic pdha-1 and pdhb-1 knockdown lowered basal INS-7 secretion level, we acutely inhibited pyruvate dehydrogenase during refeeding with the specific inhibitor 3-fluoropyruvate (3-FP); this intervention prevented INS-7 secretion from returning to basal levels (Figure 7—figure supplement 1B), suggesting that the conversion of pyruvate to acetyl-CoA is required for the refeeding response of INS-7 secretion.
If intracellular pyruvate is the crucial determinant of INS-7 secretion, we predicted that maintaining pyruvate levels in the INT1 cells during fasting should suppress the fasting-induced INS-7 secretion. Indeed, worms fasted on NGM agar supplemented with 100 mM pyruvate showed no induction of INS-7 secretion (Figure 7G), suggesting pyruvate alone is sufficient to suppress fasting-induced INS-7 secretion. Collectively, these findings identify mitochondrial pyruvate metabolism as the key signal sensed by INT1 enteroendocrine cells to couple nutritional availability with the fluctuation in INS-7 peptide secretion during fasting and refeeding.
Discussion
In the human intestine, both epithelial and immune cell compositions shift dramatically along the anterior–posterior axis, with region-specific cellular neighborhoods for epithelial, stromal, and immune cells contributing to distinct physiological functions (Hickey et al., 2023). In contrast, the C. elegans intestine comprises only 20 epithelial cells and lacks resident immune cells, raising the question of whether spatially defined molecular and functional heterogeneity also exists in this simplified but powerful model system. Our findings, together with prior observations, support the view that C. elegans intestinal subsets are molecularly and functionally distinct. The data show that the most pronounced difference between INT1 and the rest of the intestine at the RNAseq level is the expression of stress response genes in INT1. Although there are some nuanced differences, the prevalence of stress response genes persists across feeding and fasting conditions. This difference, combined with the evidence that INT1 cells secrete the enteroendocrine peptide INS-7 (Liu et al., 2024; Figure 6), is strongly reminiscent of the mammalian enteroendocrine cells, which also secrete peptides and show strong expression of stress response genes (Bany Bakar et al., 2023; Zietek and Rath, 2016). We suggest that this category term reflects not only a canonical stress response but also a broader response to shifts in the luminal environment, which INT1 cells are anatomically poised to detect well before the absorption of nutrients has begun further down the intestine (INT2-9). Thus, we believe INT1 cells are a newly defined enteroendocrine cell-type within the C. elegans intestine.
In the present study, we utilized INT1- and INT2-9-specific promoters to perform TRAP-Seq analyses under fasting and refeeding conditions. While comparisons between INT1 and INT2-9 cells revealed marked differences, this approach likely underrepresents heterogeneity among the INT2-9 cells themselves. Prior studies demonstrated that posterior intestinal cells initiate the rhythmic defecation motor program (Mahoney et al., 2008; Beg et al., 2008), and our dataset identified npr-28, which is specifically expressed in the posterior intestinal cells INT7-9 (Figure 2J), further underscoring molecular diversity across intestinal segments. Future identification of promoters with more restricted expression within the INT2-9 subset will be crucial for functionally resolving the spatial map of the C. elegans intestine.
We previously identified INT1 as enteroendocrine cells that secrete the gut peptide INS-7 preferentially during acute food withdrawal to suppress neuronally stimulated fat loss (Liu et al., 2024). Additional lines of evidence also point to the specialization of the anterior-most INT1 cells. First, acid phosphatase staining for intestinal brush border could not be detected in INT1 cells (Beh et al., 1991; Gt et al., 1995). Second, methyl red staining showed strong red staining of acidic vacuoles in INT1 cells. From INT2 to INT9 cells, vacuoles were stained gradually from orange-yellow (neutral-alkaline) to yellow (alkaline) with methyl red (Gt et al., 1995), suggesting the pH of intracellular vacuoles is lower in INT1 cells. Third, ultrastructural analysis highlighted enrichment of rough ER, Golgi, and mitochondria in INT1, consistent with a secretory function, whereas INT2-9 cells preferentially accumulate yolk and lipid vacuoles for storage (Borgonie et al., 1995). Fourth, components of a cullin-RING ubiquitin ligase complex critical for thermotolerance and the intracellular pathogen response (IPR), including CUL-6 and SKR-5, were found to be specifically expressed in INT1 cells (Panek et al., 2020), suggesting that INT1 cells may play an integral role in proteostasis responses to intracellular pathogen infection. Collectively, these findings establish INT1 as a specialized cell-type that integrates endocrine, stress, and immune functions within the C. elegans intestine.
We uncovered pyruvate as a crucial metabolite that functions as a true surrogate for bacterial food to regulate the secretion of the endogenous insulin antagonist INS-7 (Liu et al., 2024). Pyruvate must enter the TCA cycle to elicit this effect, suggesting that either actual metabolism of pyruvate is required or that downstream metabolites may also act as mediators of this process. In other systems, TCA cycle intermediates such as acetyl-CoA, itaconate, and L-2-hydroxyglutarate (L-2-HG) modulate immune responses (Martínez-Reyes and Chandel, 2020), while succinate and fumarate regulate hypoxia sensing upon cytosolic release (Picard and Shirihai, 2022). A similar mechanism may underpin the communication between mitochondria and the nucleus in C. elegans, exemplified by the translocation of ATFS-1 during the activation of the mitochondrial unfolded protein response (Nargund et al., 2015). Thus, pyruvate metabolism may provide a metabolic checkpoint that couples nutrient availability to enteroendocrine signaling.
Many bacterial species, including Yersinia enterocolitica, Citrobacter freundii, Enterobacter aerogenes, and E. coli, excrete pyruvate into the extracellular environment during nutrient-replete conditions (Behr et al., 2017; Fried et al., 2013) and reabsorb extracellular pyruvate during nutrient insufficiency (Paczia et al., 2012; Yasid et al., 2016). Thus, the presence of extracellular pyruvate may serve as an indicator of nutrient abundance in bacterial environments, a sensing mechanism that C. elegans may have co-evolved by leveraging the INT1 cells to couple environmental conditions with host physiology. It was reported that C. elegans exhibits chemotaxis to volatile pyruvic acid via AWC neurons (Zhang et al., 1997). However, the molecular mechanisms underlying both INT1 and neuronal sensing of pyruvate remain uncharacterized. Thus, both intracellular and extracellular bacterial pyruvate appear to regulate the metabolism or behavior of C. elegans, indicating that pyruvate is a crucial signaling molecule.
In summary, we expect our TRAP-Seq datasets presented here to provide the scientific community with valuable resources for investigating the dynamic translatomes of specific subsets of intestinal cells in response to fasting and refeeding. More broadly, the discovery of cell-type heterogeneity within the C. elegans intestine highlights the evolutionary conservation of spatially organized intestinal biology and lays a foundation for future mechanistic studies of metabolic and immune regulation in the intestine.
Limitations of the TRAP-Seq in the current study
Not all previous studies have demonstrated a strong correlation between mRNA expression levels and protein abundance (Harvald et al., 2017; Maier et al., 2009; Vogel and Marcotte, 2012). TRAP-Seq provides a more precise assessment of gene expression by detecting actively translating mRNAs. Specifically, TRAP-Seq better reflects the translational responses under various conditions, such as temperature fluctuations or fasting. Moreover, by employing tagged ribosomal proteins, TRAP-Seq enables the selective purification of polysomes without the need for cell dissociation or sorting, thereby reducing RNA contamination from neighboring cells or tissues.
To obtain sufficient ribosome-mRNA complexes for RNA sequencing, we generated strains with an integrated extrachromosomal array expressing HA-tagged RPL22 in either INT1 or INT2-9 cells. While this strategy enables robust affinity purification, the elevated expression of HA-tagged RPL22 in this system could potentially introduce unanticipated effects on transcript utilization or translational bias. Nevertheless, because all analyses rely on internal referencing, the differential gene expression signatures remain valid. In addition, the random integration of multi-copy transgenes into unknown genomic loci may alter the expression of adjacent genes, leading to additional variability. Importantly, we did not observe any changes in growth rate, fertility, or fecundity in the TRAP-Seq transgenic lines, suggesting that potential confounding effects are minimal.
TRAP has traditionally tagged proteins of the 60S large ribosomal subunit, including RPL10A (Heiman et al., 2014; Heiman et al., 2008; Thomas et al., 2012; Huang et al., 2013; Zhang et al., 2016; Gracida and Calarco, 2017) and RPL22 (McLachlan and Flavell, 2019; Sanz et al., 2009; Wester et al., 2023; Rhoades et al., 2019), to enable tissue- or cell type-specific capture of polysome-mRNA complexes followed by RNA sequencing. A key distinction among previous studies lies in the choice of ribosomal protein for tagging. Both RPL10A and RPL22 are evolutionarily conserved ribosomal proteins. In C. elegans, RPL10A (Gracida and Calarco, 2017; Gracida et al., 2017; Koterniak et al., 2020) and RPL22 McLachlan and Flavell, 2019; Rhoades et al., 2019; Nousch, 2020 have been successfully used for tagging in TRAP-Seq experiments, yielding robust results. However, recent findings in mammalian systems indicate that RPL10A is significantly sub-stoichiometric in polysomes of mouse embryonic stem cells (Shi et al., 2017), suggesting it is not associated with all actively translating mRNAs. Similarly, in C. elegans, RPL10A shows limited expression in the germline and weak association with ribosomes, impairing the co-purification of translating mRNAs from the germline (Nousch, 2020). These observations suggest that tagging RPL10A in TRAP experiments may not capture the complete translatome in specific cell types or tissues. Further, Nousch recently demonstrated that RPL-4 and RPL-9 are more effective than RPL10A and RPL22 for purifying polysome-mRNA complexes from germ cells in C. elegans (Nousch, 2020), suggesting that the choice of ribosomal proteins for tagging in TRAP-Seq experiments may vary depending on the tissue of interest. Specifically, certain ribosomal proteins may offer improved specificity and efficiency for isolating tissue-specific mRNAs. Importantly, the RPL22-based TRAP in our experiments achieved robust immunoprecipitation efficiency for intestinal polysomal mRNAs (Supplementary file 4), thereby validating our approach. Future studies may further investigate whether alternative ribosomal proteins could provide enhanced isolation of translating mRNAs in different tissues, including the intestine.
Materials and methods
C. elegans maintenance and strains
Worms were cultured on nematode growth medium (NGM) agar plates with E. coli strain OP50 at 20°C as described (Brenner, 1974). Bristol strain N2 was obtained from the Caenorhabditis Genetics Center (CGC) and used as the wild-type strain. All transgenic strains used in this study are listed in Supplementary file 1. Worms were synchronized for Oil Red O staining by hypochlorite treatment, after which hatched L1 larvae were seeded on plates with the appropriate bacteria; worms were synchronized for INS-7 peptide secretion assay, PA14 avoidance, and slow killing assays by letting gravid adult worms lay eggs on plates with the appropriate bacteria for 1.5 hr. All experiments were performed on day 1 adults.
Acute and chronic fasting
For the fasting experiments, synchronized day 1 adult worms were removed from the OP50 seeded NGM plates with M9 buffer, washed three times to remove remaining OP50, then placed on the unseeded NGM plates for 30 min (acute fast) or 180 min (chronic fast). For the refeeding experiments, fasted worms were removed from the fasting plate with M9 buffer, then placed onto OP50 seeded plates for 30 min (acute fast/refeeding) or 90 min (chronic fast/refeeding).
Oil Red O staining
Oil Red O staining was performed as described (Noble et al., 2013). Worms were harvested with phosphate-buffered saline (PBS) and incubated on ice for 10 min. Worms were fixed as described (Noble et al., 2013). Following fixation, worms were stained in filtered Oil Red O (Thermo Scientific) working solution (60% Oil Red O in isopropanol; 40% water) overnight. Approximately 2000 worms were fixed and stained for all conditions within one experiment. We visually examined about 100 worms on slides and then blindly chose 20 worms for imaging. All experiments were repeated at least three times.
Image acquisition and quantitation
Oil Red O-stained worms were imaged using a ×10 objective on a Zeiss Axio Imager microscope. Images were acquired with the software AxioVision (Zeiss). Lipid droplet staining in the first four pairs of intestinal cells was quantified using ImageJ software (NIH). All reported results were consistent across biological replicates. Fluorescent images of mNeonGreen reporters were acquired with software AxioVision (Zeiss) using a ×10 objective on a Zeiss Axio Imager microscope. Fluorescent images of reporters for INS-7 secretion were acquired with software AxioVision (Zeiss) using a ×20 objective on a Zeiss Axio Imager microscope. The images were taken at the indicated time point or immediately after treatment, such as refeeding. The first pair of coelomocytes was imaged. mCherry fluorescence intensity in one of the two imaged coelomocytes was quantified and normalized to the surface area of the coelomocyte (Punc-122::GFP) as previously described and validated (Palamiuc et al., 2017). Within each experiment, at least 15 worms from each condition were quantified using ImageJ software (NIH).
Cloning and transgenic strain construction
For generating Pges-1∆B::rpl-22-3xHA and Ppho-1::rpl-22-3xHA constructs, promoters and rpl-22-3xHA were amplified with standard PCR techniques and subcloned into donor vectors using Gateway Cloning Technology (Life Technologies). All plasmids for mNeonGreen reporters were generated by Gibson Assembly. Primers used for cloning in this study are listed in Supplementary file 3. Transgenic strains were generated by injecting plasmids into the germline of wild-type worms followed by visual selection of co-injection markers (Pmyo-2::mCherry, Pges-1∆B::GFP, or Ppho-1::mCherry) under the fluorescence microscope. The INS-7 secretion line was previously developed and validated (Liu et al., 2024). For INT1::TRAP strain, N2 worms were injected with 10 ng/μL of the Pges-1∆B::rpl-22-3xHA plasmid, 20 ng/μL of the Pges-1∆B::GFP plasmid, and 70 ng/μL of an empty vector to bring the final concentration of injection mix to 100 ng/μL. For INT2-9::TRAP strain, N2 worms were injected with 25 ng/μL of the Ppho-1::rpl-22-3xHA plasmid, 50 ng/μL of the Ppho-1::mCherry plasmid, and 25 ng/μL of an empty vector to bring the final concentration of injection mix to 100 ng/μL. A transgenic line with high transmission rate and consistent expression was integrated using the Stratalinker UV Crosslinker 2400 (Stratagene) and backcrossed four times before experimentation. For other microinjections, we injected the animals with 30 ng/μL of the desired plasmid, 10 ng/μL of the Pmyo-2::mCherry, and 60 ng/μL empty vector to maintain a final injection mix concentration of 100 ng/μL. Two lines were selected for experimentation based on the transmission rate and consistency of expression.
RNAi
RNAi experiments were performed as described (Kamath and Ahringer, 2003; Srinivasan et al., 2008). Carbenicillin-IPTG plates were seeded with HT115 bacteria containing the empty vector or the relevant RNAi clone four days before seeding larvae.
PA14 avoidance assay
The PA14 avoidance assay was performed as previously described (Filipowicz et al., 2022). Pseudomonas aeruginosa strain PA14 was cultured by inoculating a single colony into 2 mL of LB broth and incubating at 37°C for 6 hr. 20 μL of the culture was then seeded onto 3.5 cm slow-killing (SK) plates and incubated at 37°C for 15 hr. For each assay, 30 synchronized day 1 gravid adults (containing 6–10 eggs) grown on RNAi bacteria were cleaned and transferred to the agar surface outside the PA14 lawn. The number of animals on and off the PA14 lawns was scored after 8 hr. Three plates were used per RNAi treatment in each experiment.
PA14 slow killing assay
PA14 slow killing assay was performed as described (Tan et al., 1999). Briefly, overnight cultures of P. aeruginosa strain PA14 were seeded onto slow-killing (SK) plates and incubated at 37°C for 24 hr, followed by 25°C for an additional 24 hr. Synchronized animals at the closed vulva stage subjected to RNAi treatment were transferred to 3.5 cm SK plates seeded with PA14 (30 animals per plate), with three plates per RNAi treatment. Plates were maintained at 25°C, and worm survival was monitored until all animals had died. Survival curves were analyzed using GraphPad Prism, and statistical significance was determined by the log-rank test.
Translating ribosome affinity purification
The protocol was adapted from reference (McLachlan and Flavell, 2019) and described in detail in Supplementary file 8. Briefly, transgenic worms containing the Ribo-TRAP constructs were made to drive expression of rpl-22-3xHA in INT1 and INT2-9. Age-matched transgenic worms for each experimental condition were washed, flash-frozen, and lysed to release ribosomes and cellular components in the presence of cycloheximide to pause translation as well as RNase and protease inhibitors. Immunoprecipitation of ribosomes from our desired tissue was performed with a monoclonal anti-HA antibody (Sigma-Aldrich, catalog number: H3663-200UL, RRID:AB_262051) and protein G-beads. Nascent RNA was eluted from ribosomes with beta mercaptoethanol (BME). Percent IP RNA compared to input, on average less than 1%, was used to determine if the IP was successful.
RNA purification
RNA from the IP, flowthrough, and input was purified with Agilent Absolutely RNA Nanoprep Kit according to the manufacturer’s instructions including the on-column DNase treatment. Final RNA isolates were eluted in 15 μL and used to make diluted aliquots for sequencing, validation, and QC.
RNA quality assessment
We tested for RNA degradation and quality by running the input, flowthrough, and IP samples on an Agilent Bioanalyzer chip (nano or pico, depending on concentration). Because IP samples had a small concentration, it was difficult to accurately measure the RNA integrity number (RIN) in all conditions, but the quality of RNA for these unknown samples could be extrapolated from the flowthrough and input samples. In all conditions, the average RIN across sample types was greater than 7.5, indicating RNA integrity sufficient for sequencing, and in the majority of conditions it was greater than 9, indicative of high-quality samples.
RNAseq library prep, sequencing, and alignment
Total RNA samples were prepared into RNAseq libraries using the NEBNext Ultra Directional RNA Library Prep Kit for Illumina following the manufacturer’s recommended protocol. Briefly, for each sample 100 ng total RNA was polyA-selected, converted to double-stranded cDNA, followed by fragmentation and ligation of sequencing adapters. The libraries were then PCR amplified for 12 cycles using barcoded PCR primers, purified, and size selected using AMPure XP Beads before loading onto an Illumina NextSeq500 for 75 base single-read sequencing. Reads were aligned to C. elegans genome (WBcel235/r67) using STAR (version 2.6.1d) (Dobin et al., 2013), and gene counts were produced using featureCounts (version 1.6.4) (Liao et al., 2014). On average, 10.6 million unique single-end (SE) reads aligned per INT1 sample, and 19.5 million unique SE reads per INT2-9 sample. These differences were due to the larger RNA input for INT2-9 samples. Data quality was assessed using FastQC (version 0.11.8) (Andrews, 2010). Raw RNA-seq FASTQ files and a gene count matrix are available on GEO (GSE306302). Bioinformatic analysis was performed using R programming language (version 4.4.2) (R Development Core Team, 2024) and RStudio integrated development environment (version 2024.12.1) (Posit Team, 2024). Plots were generated using ggplot2 R package (version 3.5.2).
Differential expression analysis
Differential expression analysis was performed using the DESeq2 R package (version 1.46.0) (Love et al., 2014), which models the gene expression counts with a negative binomial distribution while adjusting for library size and dispersion. DEGs were identified as having an adjusted p-value less than 0.05 using a Benjamini–Hochberg false discovery rate correction. DESeq2 variance stabilized counts were used to generate normalized count matrices that were used for plotting expression data of individual genes and those averaged across WGCNA modules. PCA was performed using the top 500 variable genes from variance stabilized counts output in DESeq2. Statistical significance of PCA correlation was determined with a Pearson’s correlation for two-factor comparisons (line) or an ANOVA for multi-factor comparisons (metabolic stress). Plot ellipses comparing groups were drawn with ggplot2 stat_ellipse() function, which computes the mean and covariance of the PC1 and PC2 for each line to build the ellipse at a 95% confidence level.
WGCNA
WGCNA was performed with the WGCNA R package (version 1.73) (Langfelder and Horvath, 2008) to identify modules of co-expressed genes and their associations with external traits in the analysis. Prior to analysis, samples with low expression values were removed to reduce variation associated with batch. One 30 min fasting sample and one 30 min refeeding sample were removed from the Pges-1∆B experiment based on these criteria. DESeq2 variance stabilized counts were used as the input and only genes in the 95th percentile of variance were kept to reduce noise (n=2338). A soft-thresholding power of 10 was selected using the scale-free topology criterion to approximate scale-free network topology. A signed adjacency matrix was constructed and transformed into a topological overlap matrix (TOM), followed by hierarchical clustering to detect modules of highly correlated genes. Modules were identified using dynamic tree cutting with a minimum module size of 30 genes and subsequently merged based on eigengene similarity (cut height = 0.25). Module eigengenes (MEs) were correlated with external traits (e.g., line and fasting condition), and Pearson correlation coefficients along with associated p-values were computed. Modules showing significant trait associations were further investigated for biological relevance and functional enrichment. Module GO enrichment analysis was performed using the ClusterProfiler R package (version 4.14.6) (Wu et al., 2021). For GO analysis, we reported terms with FDR <0.1. The test set of module genes was compared against the background set of genes in the WGCNA analysis passing the variance criteria.
Statistics
Sample sizes were determined by relevant published studies using comparable assays and by pilot experiments assessing phenotype significance. Wild-type animals or matched vector RNAi-treated animals served as controls, as indicated in the corresponding figure legends. Within each genotype and treatment, animals were randomly selected for assays. Investigators performed blinded scoring whenever feasible. Error bars represent the standard error of the mean (SEM), unless otherwise specified in the figure legends. Student’s t-test and one-way ANOVA were used as indicated in the figure legends. Statistical analyses of phenotypic assays were performed in GraphPad Prism 9.3 (GraphPad Software); sequencing-data analyses were performed in R as described above. Appropriate multiple comparison corrections were used for ANOVAs.
Materials availability statement
The worm strains generated in this study are available from the corresponding author upon reasonable request.
Acknowledgements
This work was supported by research grants to SS from the NIH/NIDDK (R01 DK124706) and NIH/NIA (R01 AG056648). This work was supported by the resources and staff at the Scripps Research Genomics Core Facility. CL was supported by a Dorris Scholar Award from the Dorris Neuroscience Center, The Scripps Research Institute.
Funding Statement
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Contributor Information
Supriya Srinivasan, Email: supriya@scripps.edu.
Sylvia Lee, Cornell University, United States.
Sofia J Araújo, Universitat de Barcelona, Spain.
Funding Information
This paper was supported by the following grants:
National Institutes of Health R01 DK124706 to Supriya Srinivasan.
National Institutes of Health R01 AG056648 to Supriya Srinivasan.
Additional information
Competing interests
No competing interests declared.
Author contributions
Conceptualization, Resources, Data curation, Formal analysis, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing.
Conceptualization, Resources, Data curation, Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing – original draft, Writing – review and editing.
Conceptualization, Resources, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing – original draft, Project administration, Writing – review and editing.
Additional files
Genotypes, sources, and identifiers of the C. elegans strains used in this study. Integration and backcrossing information is provided where applicable.
Construct descriptions, sources, and identifiers of plasmids used for translating ribosome affinity purification (TRAP), fluorescent marking of intestinal cells, and generation of mNeonGreen reporters. mNG, mNeonGreen; 3′ UTR, 3′ untranslated region.
Oligonucleotide sequences, primer functions, and supplier information for Gibson assembly of the clec-160, clec-86, B0024.4, F01D5.5, F01D5.1, and npr-28 reporter constructs. Primer pairs were used to amplify the corresponding promoters and 3′ untranslated regions, the mNeonGreen coding sequence, and the pUC19 vector backbone.
Quality-control measurements and analysis results for TRAP-Seq profiling of the four anterior intestinal cells (INT1) and the remaining 16 intestinal cells (INT2-9). The acute paradigm comprised 30 min of fasting followed by 30 min of refeeding; the chronic paradigm comprised 180 min of fasting followed by 90 min of refeeding, with separate fed controls for each paradigm. Worksheets provide RNA concentrations, immunoprecipitation recovery relative to input (%IP), mean RNA integrity numbers (RIN), sample-exclusion notes, DESeq2 differential-expression results, weighted gene co-expression network analysis (WGCNA) module assignments, Gene Ontology enrichment analysis results, and module eigengenes. Differential-expression tables include gene identifiers, mean normalized counts, log₂ fold changes, standard errors, test statistics, and unadjusted and Benjamini–Hochberg-adjusted p-values, including nonsignificant results. In comparison codes, G and P denote INT1 and INT2-9; F, S, and R denote fed, fasted, and refed states; and 30 and 180 identify the acute and chronic paradigms. Positive log₂ fold changes indicate higher expression in the first member of each comparison. The ‘DE’ labels indicate fold-change direction; statistical significance is reported by the adjusted p-values.
Gene lists comparing INT1 with INT2-9 in fed, fasted, and refed states within the acute and chronic paradigms. Acute experiments comprised 30 min of fasting followed by 30 min of refeeding, and chronic experiments comprised 180 min of fasting followed by 90 min of refeeding, with separate fed controls for each paradigm. The summary worksheet contains genes selected at Benjamini–Hochberg-adjusted p<0.05. Additional worksheets separate genes with higher (‘Up’) or lower (‘Down’) expression in INT1 relative to INT2-9 for each condition. Annotations include WormBase identifiers, gene names, log₂ fold changes, adjusted p-values, WormCat functional categories and descriptions, UniProt identifiers, and InterPro protein-family and domain annotations.
TRAP-Seq differential-expression results for INT1 during acute fasting (30 min) and subsequent refeeding (30 min), and chronic fasting (180 min) and subsequent refeeding (90 min). The merged worksheet contains DESeq2 results for fasted versus fed, refed versus fasted, and refed versus fed comparisons within each paradigm, including nonsignificant results. Additional worksheets provide gene lists selected at Benjamini–Hochberg-adjusted p<0.1, with WormBase identifiers, gene names, log₂ fold changes, adjusted p-values, and WormCat functional annotations. ‘Fasted’ worksheets compare fasted with fed animals; ‘Refed’ worksheets compare refed with fasted animals. ‘Up’ and ‘Down’ indicate positive and negative log₂ fold changes, respectively.
TRAP-Seq differential-expression results for INT2-9 during acute fasting (30 min) and subsequent refeeding (30 min), and chronic fasting (180 min) and subsequent refeeding (90 min). The merged worksheet contains DESeq2 results for fasted versus fed, refed versus fasted, and refed versus fed comparisons within each paradigm, including nonsignificant results. Additional worksheets provide gene lists selected at Benjamini–Hochberg-adjusted p<0.1, with WormBase identifiers, gene names, log₂ fold changes, adjusted p-values, and WormCat functional annotations. ‘Fasted’ worksheets compare fasted with fed animals; ‘Refed’ worksheets compare refed with fasted animals. ‘Up’ and ‘Down’ indicate positive and negative log₂ fold changes, respectively.
Detailed protocol used to isolate intestinal ribosome-associated RNA from C. elegans expressing HA-tagged RPL-22 in INT1 or INT2-9. Procedures cover animal synchronization and culture, harvesting in the presence of cycloheximide, flash-freezing and cryogenic grinding, lysate preparation and pre-clearing, anti-HA immunoprecipitation, washing, and RNA elution. Adapted from McLachlan and Flavell, 2019, Bio-protocol 9 (15) (doi:10.21769/BioProtoc.3328).
Numerical source data are provided in a single Excel file, with separate worksheets labeled according to the corresponding figure and panel. Each worksheet contains the associated data and experimental group labels.
Data availability
Numerical source data and processed sequencing results are provided in the accompanying Source Data file and Supplementary files 4–7. Raw TRAP-seq reads (FASTQ files) and the gene count matrix generated in this study have been deposited in the Gene Expression Omnibus (GEO) under accession number GSE306302. The published adult Caenorhabditis elegans single-cell RNA-sequencing dataset used for comparison of intestinal gene-expression patterns (Ghaddar et al., 2023) is available through GEO under accession number GSE226715.
The following dataset was generated:
Liu CC, Seban N, Srinivasan S. 2025. Ribo-TRAPseq identifies spacially distinct roles for the anterior and posterior intestine for metabolic regulation in C. elegans. NCBI Gene Expression Omnibus. GSE306302
The following previously published dataset was used:
Ghaddar A, Armingol E, Huynh C, Gevirztman L, Lewis NE, Waterston R, O'Rourke EJ. 2023. Whole-body gene expression atlas of an adult metazoan. NCBI Gene Expression Omnibus. GSE226715
References
- Andrews S. FastQC: A Quality Control Tool for High Throughput Sequence Data. 2010. [May 29, 2021]. http://www.bioinformatics.babraham.ac.uk/projects/fastqc
- Bany Bakar R, Reimann F, Gribble FM. The intestine as an endocrine organ and the role of gut hormones in metabolic regulation. Nature Reviews. Gastroenterology & Hepatology. 2023;20:784–796. doi: 10.1038/s41575-023-00830-y. [DOI] [PubMed] [Google Scholar]
- Beg AA, Ernstrom GG, Nix P, Davis MW, Jorgensen EM. Protons act as a transmitter for muscle contraction in C. elegans. Cell. 2008;132:149–160. doi: 10.1016/j.cell.2007.10.058. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Beh CT, Ferrari DC, Chung MA, McGhee JD. An acid phosphatase as a biochemical marker for intestinal development in the nematode Caenorhabditis elegans. Developmental Biology. 1991;147:133–143. doi: 10.1016/s0012-1606(05)80013-2. [DOI] [PubMed] [Google Scholar]
- Behr S, Brameyer S, Witting M, Schmitt-Kopplin P, Jung K. Comparative analysis of LytS/LytTR-type histidine kinase/response regulator systems in γ-proteobacteria. PLOS ONE. 2017;12:e0182993. doi: 10.1371/journal.pone.0182993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Borgonie G, Claeys M, Dewaele D, Coomans A. Ultrastructure of the intestine of the bacteriophagous nematodes Caenorhabditis-Elegans, panagrolaimus superbus and acrobeloides maximus (nematoda, rhabditida) Fundamental and Applied Nematology. 1995;18:123–133. [Google Scholar]
- Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94. doi: 10.1093/genetics/77.1.71. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bricker DK, Taylor EB, Schell JC, Orsak T, Boutron A, Chen Y-C, Cox JE, Cardon CM, Van Vranken JG, Dephoure N, Redin C, Boudina S, Gygi SP, Brivet M, Thummel CS, Rutter J. A mitochondrial pyruvate carrier required for pyruvate uptake in yeast, Drosophila, and humans. Science. 2012;337:96–100. doi: 10.1126/science.1218099. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Castañeda O, Sotolongo V, Amor AM, Stöcklin R, Anderson AJ, Harvey AL, Engström A, Wernstedt C, Karlsson E. Characterization of a potassium channel toxin from the Caribbean Sea anemone Stichodactyla helianthus. Toxicon. 1995;33:603–613. doi: 10.1016/0041-0101(95)00013-c. [DOI] [PubMed] [Google Scholar]
- Chhabra S, Chang SC, Nguyen HM, Huq R, Tanner MR, Londono LM, Estrada R, Dhawan V, Chauhan S, Upadhyay SK, Gindin M, Hotez PJ, Valenzuela JG, Mohanty B, Swarbrick JD, Wulff H, Iadonato SP, Gutman GA, Beeton C, Pennington MW, Norton RS, Chandy KG. Kv1.3 channel-blocking immunomodulatory peptides from parasitic worms: implications for autoimmune diseases. FASEB Journal. 2014;28:3952–3964. doi: 10.1096/fj.14-251967. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29:15–21. doi: 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Filipowicz A, Lalsiamthara J, Aballay A. Dissection of a sensorimotor circuit underlying pathogen aversion in C. elegans. BMC Biology. 2022;20:229. doi: 10.1186/s12915-022-01424-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fried L, Behr S, Jung K. Identification of a target gene and activating stimulus for the YpdA/YpdB histidine kinase/response regulator system in Escherichia coli. Journal of Bacteriology. 2013;195:807–815. doi: 10.1128/JB.02051-12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fukushige T, Hawkins MG, McGhee JD. The GATA-factor elt-2 is essential for formation of the Caenorhabditis elegans intestine. Developmental Biology. 1998;198:286–302. [PubMed] [Google Scholar]
- Ge SX, Jung D, Yao R. ShinyGO: a graphical gene-set enrichment tool for animals and plants. Bioinformatics. 2020;36:2628–2629. doi: 10.1093/bioinformatics/btz931. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghaddar A, Armingol E, Huynh C, Gevirtzman L, Lewis NE, Waterston R, O’Rourke EJ. Whole-body gene expression atlas of an adult metazoan. Science Advances. 2023;9:eadg0506. doi: 10.1126/sciadv.adg0506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ghafouri S, McGhee JD. Bacterial residence time in the intestine of Caenorhabditis elegans. Nematology. 2007;9:87–91. doi: 10.1163/156854107779969718. [DOI] [Google Scholar]
- Gracida X, Calarco JA. Cell type-specific transcriptome profiling in C. elegans using the translating ribosome affinity purification technique. Methods. 2017;126:130–137. doi: 10.1016/j.ymeth.2017.06.023. [DOI] [PubMed] [Google Scholar]
- Gracida X, Dion MF, Harris G, Zhang Y, Calarco JA. An Elongin-Cullin-SOCS box complex regulates stress-induced serotonergic neuromodulation. Cell Reports. 2017;21:3089–3101. doi: 10.1016/j.celrep.2017.11.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gray LR, Tompkins SC, Taylor EB. Regulation of pyruvate metabolism and human disease. Cellular and Molecular Life Sciences. 2014;71:2577–2604. doi: 10.1007/s00018-013-1539-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gt B, Claeys M, Waele DD, Coomans A. In-vivo and in-vitro characterization of the intestine of fifteen bacteriophagous nematodes (Nematoda: Rhabditida) Fundamental and Applied Nematology. 1995;18:115–122. [Google Scholar]
- Harvald EB, Sprenger RR, Dall KB, Ejsing CS, Nielsen R, Mandrup S, Murillo AB, Larance M, Gartner A, Lamond AI, Færgeman NJ. Multi-omics analyses of starvation responses reveal a central role for lipoprotein metabolism in acute starvation survival in C. elegans. Cell Systems. 2017;5:38–52. doi: 10.1016/j.cels.2017.06.004. [DOI] [PubMed] [Google Scholar]
- Heiman M, Schaefer A, Gong S, Peterson JD, Day M, Ramsey KE, Suárez-Fariñas M, Schwarz C, Stephan DA, Surmeier DJ, Greengard P, Heintz N. A translational profiling approach for the molecular characterization of CNS cell types. Cell. 2008;135:738–748. doi: 10.1016/j.cell.2008.10.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Heiman M, Kulicke R, Fenster RJ, Greengard P, Heintz N. Cell type-specific mRNA purification by translating ribosome affinity purification (TRAP) Nature Protocols. 2014;9:1282–1291. doi: 10.1038/nprot.2014.085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Herzig S, Raemy E, Montessuit S, Veuthey J-L, Zamboni N, Westermann B, Kunji ERS, Martinou J-C. Identification and functional expression of the mitochondrial pyruvate carrier. Science. 2012;337:93–96. doi: 10.1126/science.1218530. [DOI] [PubMed] [Google Scholar]
- Hickey JW, Becker WR, Nevins SA, Horning A, Perez AE, Zhu C, Zhu B, Wei B, Chiu R, Chen DC, Cotter DL, Esplin ED, Weimer AK, Caraccio C, Venkataraaman V, Schürch CM, Black S, Brbić M, Cao K, Chen S, Zhang W, Monte E, Zhang NR, Ma Z, Leskovec J, Zhang Z, Lin S, Longacre T, Plevritis SK, Lin Y, Nolan GP, Greenleaf WJ, Snyder M. Organization of the human intestine at single-cell resolution. Nature. 2023;619:572–584. doi: 10.1038/s41586-023-05915-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Higgins DP, Weisman CM, Lui DS, D’Agostino FA, Walker AK. Defining characteristics and conservation of poorly annotated genes in Caenorhabditis elegans using WormCat 2.0. Genetics. 2022;221:iyac085. doi: 10.1093/genetics/iyac085. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Holdorf AD, Higgins DP, Hart AC, Boag PR, Pazour GJ, Walhout AJM, Walker AK. WormCat: an online tool for annotation and visualization of Caenorhabditis elegans genome-scale data. Genetics. 2020;214:279–294. doi: 10.1534/genetics.119.302919. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Huang Y, Ainsley JA, Reijmers LG, Jackson FR. Translational profiling of clock cells reveals circadianly synchronized protein synthesis. PLOS Biology. 2013;11:e1001703. doi: 10.1371/journal.pbio.1001703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jo H, Shim J, Lee JH, Lee J, Kim JB. IRE-1 and HSP-4 contribute to energy homeostasis via fasting-induced lipases in C. elegans. Cell Metabolism. 2009;9:440–448. doi: 10.1016/j.cmet.2009.04.004. [DOI] [PubMed] [Google Scholar]
- Kamath RS, Ahringer J. Genome-wide RNAi screening in Caenorhabditis elegans. Methods. 2003;30:313–321. doi: 10.1016/s1046-2023(03)00050-1. [DOI] [PubMed] [Google Scholar]
- Koterniak B, Pilaka PP, Gracida X, Schneider L-M, Pritišanac I, Zhang Y, Calarco JA. Global regulatory features of alternative splicing across tissues and within the nervous system of C. elegans. Genome Research. 2020;30:1766–1780. doi: 10.1101/gr.267328.120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee SJ, Murphy CT, Kenyon C. Glucose shortens the life span of C. elegans by downregulating DAF-16/FOXO activity and aquaporin gene expression. Cell Metabolism. 2009;10:379–391. doi: 10.1016/j.cmet.2009.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Leiser SF, Miller H, Rossner R, Fletcher M, Leonard A, Primitivo M, Rintala N, Ramos FJ, Miller DL, Kaeberlein M. Cell nonautonomous activation of flavin-containing monooxygenase promotes longevity and health span. Science. 2015;350:1375–1378. doi: 10.1126/science.aac9257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30:923–930. doi: 10.1093/bioinformatics/btt656. [DOI] [PubMed] [Google Scholar]
- Liu CC, Khan A, Seban N, Littlejohn N, Shah A, Srinivasan S. A homeostatic gut-to-brain insulin antagonist restrains neuronally stimulated fat loss. Nature Communications. 2024;15:6869. doi: 10.1038/s41467-024-51077-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Loukas A, Hintz M, Linder D, Mullin NP, Parkinson J, Tetteh KK, Maizels RM. A family of secreted mucins from the parasitic nematode Toxocara canis bears diverse mucin domains but shares similar flanking six-cysteine repeat motifs. The Journal of Biological Chemistry. 2000;275:39600–39607. doi: 10.1074/jbc.M005632200. [DOI] [PubMed] [Google Scholar]
- Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biology. 2014;15:550. doi: 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mahoney TR, Luo S, Round EK, Brauner M, Gottschalk A, Thomas JH, Nonet ML. Intestinal signaling to GABAergic neurons regulates a rhythmic behavior in Caenorhabditis elegans. PNAS. 2008;105:16350–16355. doi: 10.1073/pnas.0803617105. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Maier T, Güell M, Serrano L. Correlation of mRNA and protein in complex biological samples. FEBS Letters. 2009;583:3966–3973. doi: 10.1016/j.febslet.2009.10.036. [DOI] [PubMed] [Google Scholar]
- Martínez-Reyes I, Chandel NS. Mitochondrial TCA cycle metabolites control physiology and disease. Nature Communications. 2020;11:102. doi: 10.1038/s41467-019-13668-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Mayer S, Raulf MK, Lepenies B. C-type lectins: their network and roles in pathogen recognition and immunity. Histochemistry and Cell Biology. 2017;147:223–237. doi: 10.1007/s00418-016-1523-7. [DOI] [PubMed] [Google Scholar]
- McLachlan IG, Flavell SW. Cell type-specific mRNA purification in Caenorhabditis elegans via translating ribosome affinity purification. Bio-Protocol. 2019;9:e3328. doi: 10.21769/BioProtoc.3328. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nargund AM, Fiorese CJ, Pellegrino MW, Deng P, Haynes CM. Mitochondrial and nuclear accumulation of the transcription factor ATFS-1 promotes OXPHOS recovery during the UPR(mt) Molecular Cell. 2015;58:123–133. doi: 10.1016/j.molcel.2015.02.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Noble T, Stieglitz J, Srinivasan S. An integrated serotonin and octopamine neuronal circuit directs the release of an endocrine signal to control C. elegans body fat. Cell Metabolism. 2013;18:672–684. doi: 10.1016/j.cmet.2013.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nousch M. RPL-4 and RPL-9 –mediated ribosome purifications facilitate the efficient analysis of gene expression in Caenorhabditis elegans germ cells. G3: Genes, Genomes, Genetics. 2020;10:4063–4069. doi: 10.1534/g3.120.401644. [DOI] [PMC free article] [PubMed] [Google Scholar]
- O’Rourke EJ, Ruvkun G. MXL-3 and HLH-30 transcriptionally link lipolysis and autophagy to nutrient availability. Nature Cell Biology. 2013;15:668–676. doi: 10.1038/ncb2741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paczia N, Nilgen A, Lehmann T, Gätgens J, Wiechert W, Noack S. Extensive exometabolome analysis reveals extended overflow metabolism in various microorganisms. Microbial Cell Factories. 2012;11:122. doi: 10.1186/1475-2859-11-122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Palamiuc L, Noble T, Witham E, Ratanpal H, Vaughan M, Srinivasan S. A tachykinin-like neuroendocrine signalling axis couples central serotonin action and nutrient sensing with peripheral lipid metabolism. Nature Communications. 2017;8:14237. doi: 10.1038/ncomms14237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Panek J, Gang SS, Reddy KC, Luallen RJ, Fulzele A, Bennett EJ, Troemel ER. A cullin-RING ubiquitin ligase promotes thermotolerance as part of the intracellular pathogen response in Caenorhabditis elegans. PNAS. 2020;117:7950–7960. doi: 10.1073/pnas.1918417117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pees B, Yang W, Kloock A, Petersen C, Peters L, Fan L, Friedrichsen M, Butze S, Zárate-Potes A, Schulenburg H, Dierking K. Effector and regulator: Diverse functions of C. elegans C-type lectin-like domain proteins. PLOS Pathogens. 2021;17:e1009454. doi: 10.1371/journal.ppat.1009454. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Picard M, Shirihai OS. Mitochondrial signal transduction. Cell Metabolism. 2022;34:1620–1653. doi: 10.1016/j.cmet.2022.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Posit Team . Boston, MA, USA: Posit Software, PBC; 2024. https://posit.co/ [Google Scholar]
- Quintin S, Aspert T, Ye T, Charvin G. Distinct mechanisms underlie H2O2 sensing in C. elegans head and tail. PLOS ONE. 2022;17:e0274226. doi: 10.1371/journal.pone.0274226. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radeke LJ, Herman MA. Identification and characterization of differentially expressed genes in Caenorhabditis elegans in response to pathogenic and nonpathogenic Stenotrophomonas maltophilia. BMC Microbiology. 2020;20:170. doi: 10.1186/s12866-020-01771-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Raizen D, Song BM, Trojanowski N, You YJ. Methods for measuring pharyngeal behaviors. WormBook. 2012;1:1–13. doi: 10.1895/wormbook.1.154.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- R Development Core Team . Vienna, Austria: R Foundation for Statistical Computing; 2024. https://www.R-project.org/ [Google Scholar]
- Rhoades JL, Nelson JC, Nwabudike I, Yu SK, McLachlan IG, Madan GK, Abebe E, Powers JR, Colón-Ramos DA, Flavell SW. ASICs mediate food responses in an enteric serotonergic neuron that controls foraging behaviors. Cell. 2019;176:85–97. doi: 10.1016/j.cell.2018.11.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Samuel BS, Rowedder H, Braendle C, Félix M-A, Ruvkun G. Caenorhabditis elegans responses to bacteria from its natural habitats. PNAS. 2016;113:E3941–E3949. doi: 10.1073/pnas.1607183113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sanz E, Yang L, Su T, Morris DR, McKnight GS, Amieux PS. Cell-type-specific isolation of ribosome-associated mRNA from complex tissues. PNAS. 2009;106:13939–13944. doi: 10.1073/pnas.0907143106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schroeder DF, McGhee JD. Anterior-posterior patterning within the Caenorhabditis elegans endoderm. Development. 1998;125:4877–4887. doi: 10.1242/dev.125.24.4877. [DOI] [PubMed] [Google Scholar]
- Schulenburg H, Hoeppner MP, Weiner J, III, Bornberg-Bauer E. Specificity of the innate immune system and diversity of C-type lectin domain (CTLD) proteins in the nematode Caenorhabditis elegans. Immunobiology. 2008;213:237–250. doi: 10.1016/j.imbio.2007.12.004. [DOI] [PubMed] [Google Scholar]
- Shi Z, Fujii K, Kovary KM, Genuth NR, Röst HL, Teruel MN, Barna M. Heterogeneous ribosomes preferentially translate distinct subpools of mRNAs genome-wide. Molecular Cell. 2017;67:71–83. doi: 10.1016/j.molcel.2017.05.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shivers RP, Youngman MJ, Kim DH. Transcriptional responses to pathogens in Caenorhabditis elegans. Current Opinion in Microbiology. 2008;11:251–256. doi: 10.1016/j.mib.2008.05.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Srinivasan S, Sadegh L, Elle IC, Christensen AGL, Faergeman NJ, Ashrafi K. Serotonin regulates C. elegans fat and feeding through independent molecular mechanisms. Cell Metabolism. 2008;7:533–544. doi: 10.1016/j.cmet.2008.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sulston JE, Schierenberg E, White JG, Thomson JN. The embryonic cell lineage of the nematode Caenorhabditis elegans. Developmental Biology. 1983;100:64–119. doi: 10.1016/0012-1606(83)90201-4. [DOI] [PubMed] [Google Scholar]
- Tan MW, Mahajan-Miklos S, Ausubel FM. Killing of Caenorhabditis elegans by Pseudomonas aeruginosa used to model mammalian bacterial pathogenesis. PNAS. 1999;96:715–720. doi: 10.1073/pnas.96.2.715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tauffenberger A, Fiumelli H, Almustafa S, Magistretti PJ. Lactate and pyruvate promote oxidative stress resistance through hormetic ROS signaling. Cell Death & Disease. 2019;10:653. doi: 10.1038/s41419-019-1877-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Thomas A, Lee P-J, Dalton JE, Nomie KJ, Stoica L, Costa-Mattioli M, Chang P, Nuzhdin S, Arbeitman MN, Dierick HA. A versatile method for cell-specific profiling of translated mRNAs in Drosophila. PLOS ONE. 2012;7:e40276. doi: 10.1371/journal.pone.0040276. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Troemel ER, Chu SW, Reinke V, Lee SS, Ausubel FM, Kim DH. p38 MAPK regulates expression of immune response genes and contributes to longevity in C. elegans. PLOS Genetics. 2006;2:e183. doi: 10.1371/journal.pgen.0020183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tullet JMA, Araiz C, Sanders MJ, Au C, Benedetto A, Papatheodorou I, Clark E, Schmeisser K, Jones D, Schuster EF, Thornton JM, Gems D. DAF-16/FoxO directly regulates an atypical AMP-activated protein kinase gamma isoform to mediate the effects of insulin/IGF-1 signaling on aging in Caenorhabditis elegans. PLOS Genetics. 2014;10:e1004109. doi: 10.1371/journal.pgen.1004109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Van Gilst MR, Hadjivassiliou H, Yamamoto KR. A Caenorhabditis elegans nutrient response system partially dependent on nuclear receptor NHR-49. PNAS. 2005;102:13496–13501. doi: 10.1073/pnas.0506234102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vogel C, Marcotte EM. Insights into the regulation of protein abundance from proteomic and transcriptomic analyses. Nature Reviews. Genetics. 2012;13:227–232. doi: 10.1038/nrg3185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Warner JR, Knopf PM, Rich A. A multiple ribosomal structure in protein synthesis. PNAS. 1963;49:122–129. doi: 10.1073/pnas.49.1.122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Warner JR, Knopf PM. The discovery of polyribosomes. Trends in Biochemical Sciences. 2002;27:376–380. doi: 10.1016/S0968-0004(02)02126-6. [DOI] [PubMed] [Google Scholar]
- Wester LE, Lanjuin A, Bruckisch EHW, Perez-Matos MC, Stine PG, Heintz C, Denzel MS, Mair WB. A single-copy knockin translating ribosome immunoprecipitation toolkit for tissue-specific profiling of actively translated mRNAs in C. elegans. Cell Reports Methods. 2023;3:100433. doi: 10.1016/j.crmeth.2023.100433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, Feng T, Zhou L, Tang W, Zhan L, Fu X, Liu S, Bo X, Yu G. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. Innovation. 2021;2:100141. doi: 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xiao W, Wang R-S, Handy DE, Loscalzo J. NAD(H) and NADP(H) Redox couples and cellular energy metabolism. Antioxidants & Redox Signaling. 2018;28:251–272. doi: 10.1089/ars.2017.7216. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yasid NA, Rolfe MD, Green J, Williamson MP. Homeostasis of metabolites in Escherichia coli on transition from anaerobic to aerobic conditions and the transient secretion of pyruvate. Royal Society Open Science. 2016;3:160187. doi: 10.1098/rsos.160187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zaslaver A, Liani I, Shtangel O, Ginzburg S, Yee L, Sternberg PW. Hierarchical sparse coding in the sensory system of Caenorhabditis elegans. PNAS. 2015;112:1185–1189. doi: 10.1073/pnas.1423656112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Chou JH, Bradley J, Bargmann CI, Zinn K. The Caenorhabditis elegans seven-transmembrane protein ODR-10 functions as an odorant receptor in mammalian cells. PNAS. 1997;94:12162–12167. doi: 10.1073/pnas.94.22.12162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang KX, Tan L, Pellegrini M, Zipursky SL, McEwen JM. Rapid changes in the translatome during the conversion of growth cones to synaptic terminals. Cell Reports. 2016;14:1258–1271. doi: 10.1016/j.celrep.2015.12.102. [DOI] [PubMed] [Google Scholar]
- Zhao J, Qin B, Nikolay R, Spahn CMT, Zhang G. Translatomics: The global view of translation. International Journal of Molecular Sciences. 2019;20:212. doi: 10.3390/ijms20010212. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ziegler DM. An overview of the mechanism, substrate specificities, and structure of FMOs. Drug Metabolism Reviews. 2002;34:503–511. doi: 10.1081/dmr-120005650. [DOI] [PubMed] [Google Scholar]
- Zietek T, Rath E. Inflammation meets metabolic disease: gut feeling mediated by GLP-1. Frontiers in Immunology. 2016;7:154. doi: 10.3389/fimmu.2016.00154. [DOI] [PMC free article] [PubMed] [Google Scholar]