Quantification of secreted hormones

Oxytocin in the tissue or organoid supernatants was measured using an ELISA (Enzo Life Sciences, USA, ADI-900-153A-0001) or Luminex panel (Millipore Sigma, USA, HNPMAG-35K). Secretin was similarly measured using an ELISA (Biomatik, USA, {"type":"entrez-protein","attrs":{"text":"EKU07226","term_id":"421950348","term_text":"EKU07226"}}EKU07226). For pig tissue, supernatants were concentrated 5× for the ELISA by lyophilizing. Extrapolated values from the assays less than the limit of detection were given the value of half the lowest standard in the standard curve for the ELISA. Data were analyzed using either a linear or a linear mixed model. Linear mixed models were used for data derived from human or pig intestinal tissue with the patient or pig variable added as a random effect, for the piglet data, sex was added as a random effect as there was only a single male and a single female piglet, and for organoids when multiple organoid batches or lines were used and these were included as a random effect. The random effect terms allowed for inclusion of replicate data at the patient/pig/organoid level without inflating the power of the model. For mouse data, linear models were used as there were no replicates per mouse and tissue segment. The specific models used are provided in Supplemental Table S2. Linear mixed models were run using the lmer function in lme4 package⁸³ with REML = FALSE and the control optimizer = “bobyqa” and linear models with the lm function in R.⁸¹ Least-squares means estimates were computed with the emmeans function in the emmeans package⁸² using a Benjamini-Hochberg multiple testing correction (see Supplemental Table S3).