In C. elegans, 100 mM pyruvate restored the post-fasting pattern of INS-7 secretion seen with OP50 refeeding, while 2% glucose did not. The result came alongside evidence that INT1, one intestinal region, carries a translating-mRNA program that differs from INT2-9 during acute fasting. It points to a nutrient-linked signal under the tested conditions, but it does not identify the pyruvate sensor or show that pyruvate is the only bacterial nutrient involved.
A map of the gut’s active messages
The study asked whether discrete intestinal segments have distinct food-sensing capabilities and fasting responses. Researchers used TRAP-Seq to isolate polysomal RNA, RNA associated with the machinery actively translating messages into proteins, from INT1 and INT2-9 in fed, fasted and refed states. Each TRAP-Seq sample used 100,000 day-one adult INT1::TRAP or INT2-9::TRAP animals, and the experiments were conducted in triplicate.
The methods also specify wild-type or matched vector-RNAi animals as controls, random selection within genotype and treatment, and blinding whenever feasible.
The resulting profiles separated clearly by promoter identity in a principal-component analysis, which groups samples according to their overall molecular pattern. One gene module was enriched for stress-response categories, including C-type lectin, pathogen and heat genes, and the analysis identified an INT1-favored stress-response program. In plain terms, the two groups were carrying different sets of actively translated messages. That is not the same as measuring the complete protein output of the cells.
Some of the INT1 markers held up across nutritional conditions. ShKT-domain proteins were reported as robust INT1-specific markers in both acute and chronic comparisons. Eleven genes shared across the acute conditions also appeared in a 12-gene shared chronic set, a pattern used to identify a stable INT1 signature.
The split was especially clear during acute fasting: 270 genes were upregulated in INT1 but downregulated in INT2-9. This describes opposite changes in translation profiles, not direct proof that the same proteins rose or fell by matching amounts.
A mixed picture for pathogen defense
To test whether the molecular differences had a functional counterpart, the researchers examined selected INT1 genes in animals exposed to PA14. INT1-specific knockdown of clec-86 or clec-160 impaired avoidance of PA14 but significantly increased survival on PA14. The result is not a general explanation of pathogen defense: the functional tests covered selected genes, and the study did not show that every immune or stress-related gene in the profiles has the same effect.
Other INT1-specific RNAi experiments changed the refeeding pattern of INS-7. After aakg-4 knockdown, secretion did not return to basal levels after refeeding. fmo-2 knockdown animals retained normal fasting-induced secretion, but had lower basal secretion and an impaired refeeding response. These selected-gene results show altered secretion patterns, not a complete pathway.
The signal changed with energy handling
Timing mattered in the pyruvate tests. In fasted animals, UV-killed OP50 restored basal INS-7 secretion, while 2% glucose failed to normalize it, and 100 mM pyruvate fully recapitulated OP50 refeeding. But supplying 100 mM pyruvate during fasting was associated with no INS-7 induction. RNAi targeting mitochondrial pyruvate transport or metabolism also altered the response: mpc-1/mpc-2 animals lacked pyruvate-induced restoration, while pyc-1, pdha-1 and pdhb-1 conditions showed suppressed fasting-induced secretion and absent secretion dynamics. The experiments therefore point to nutrient timing and mitochondrial handling, while leaving open whether INT1 senses pyruvate directly or responds to a downstream metabolite.
What the experiment can tell us
The study’s method also brings caveats. Elevated HA-tagged RPL22 expression could bias transcript utilization, and random insertion of multiple transgene copies could affect adjacent-gene expression. The authors reported that growth, fertility and fecundity were unchanged in the TRAP-Seq lines. Even so, TRAP-Seq measures actively translating mRNAs, so it does not directly establish protein abundance or function.
The work appears as a peer-reviewed research article in eLife reporting original C. elegans experiments. It is therefore evidence from an animal model, not direct evidence that the same nutrient-to-INS-7 mechanism operates in humans or mammals. Raw RNA-seq FASTQ files and a gene-count matrix are available in GEO under GSE306302.
Paper data and sources
Original title: RiboTRAP-seq identifies spatially distinct functions for the anterior and posterior intestine in immune and metabolic regulation in Caenorhabditis elegans
Authors: Liu C, Seban N, Srinivasan S
Journal/Repository: eLife
Status: Peer-reviewed
First online: 2026-10-19
DOI: Not available
Original paper · Full text