Enteroendocrine cells couple nutrient sensing to nutrient absorption by regulating ion transport

Abstract

The ability to absorb ingested nutrients is an essential function of all metazoans and utilizes a wide array of nutrient transporters found on the absorptive enterocytes of the small intestine. A unique population of patients has previously been identified with severe congenital malabsorptive diarrhea upon ingestion of any enteral nutrition. The intestines of these patients are macroscopically normal, but lack enteroendocrine cells (EECs), suggesting an essential role for this rare population of nutrient-sensing cells in regulating macronutrient absorption. Here, we use human and mouse models of EEC deficiency to identify an unappreciated role for the EEC hormone peptide YY in regulating ion-coupled absorption of glucose and dipeptides. We find that peptide YY is required in the small intestine to maintain normal electrophysiology in the presence of vasoactive intestinal polypeptide, a potent stimulator of ion secretion classically produced by enteric neurons. Administration of peptide YY to EEC-deficient mice restores normal electrophysiology, improves glucose and peptide absorption, diminishes diarrhea and rescues postnatal survival. These data suggest that peptide YY is a key regulator of macronutrient absorption in the small intestine and may be a viable therapeutic option to treat patients with electrolyte imbalance and nutrient malabsorption.

Fig. 1. The PYY-VIP axis regulates ion and water transport in mouse and human small intestine.

a PYY and VIP regulate ion and water transport in HIO-derived small intestinal enteroids. VIP-induced ion and water transport as measured by enteroid swelling (****P < 0.0001) in a CFTR-dependent manner. EEC-deficient enteroids had an elevated response to VIP compared to wild-type enteroids (*P = 0.04), which was inhibited in both cultures upon addition of PYY. Chemical inhibition of the PYY receptor NPY1R with BIBO3304 resulted in swelling of wild-type enteroids to EEC-deficient levels (****P < 0.0001) and abolished the inhibitory effects of PYY in both genotypes (****P < 0.0001). Scale bars = 500 μm. n = 283 wild-type, n = 351 EEC-deficient enteroids over three biologically independent lines. Statistics calculated by two-way ANOVA with Sidak’s multiple comparisons test; upper row indicates comparison to vehicle; lower row indicates comparison between wild-type and EEC-deficient. b EEC-deficient enteroids displayed impaired NHE3 activity. Quantification is of initial rate of Na⁺-dependent intracellular pH recovery (red line) after acid load using the ratiometric pH indicator SNARF-4F. n = 16 wild-type, n = 18 EEC-deficient enteroids; *P = 0.01; statistics calculated by unpaired, two-tailed Student’s t test. c The localization of VIPR1 and NPY1R was comparable between wild-type and EEC-deficient human intestinal epithelium. PYY+ and CHGA+ cells were only found in wild-type HIOs. Scale bars = 50 μm. Representative images from four independent organoids are shown. d PYY modulates the stimulatory effects of VIP in mouse and human small intestine. Using an Ussing chamber we observed EEC-deficient small intestinal tissue displayed a greater response (ΔI_sc) to 10 nM VIP than wild-type (mouse, n = 20 wild-type, 8 mutant, ****P < 0.0001; human, n = 15 wild-type, 9 mutant, **P = 0.001). Inhibition of NPY1R in wild-type tissue with BIBO3304 resulted in an elevated response to VIP (mouse, n = 24, *P = 0.01; human, n = 7, *P = 0.04), whereas addition of exogenous PYY reduced the magnitude of EEC-deficient response to VIP (n = 8 mutant mice, ****P < 0.0001; n = 7 mutant HIOs, **P = 0.007) to wild-type levels. Electrogenic responses to VIP were blocked by the CFTR inhibitor CFTR-172 (dotted lines). One representative trace is shown (mouse), with baseline I_sc normalized to 0 μA/cm². Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. All error bars are + SEM.

Fig. 2. PYY restores normal glucose absorption in EEC-deficient human and mouse small intestine.

a Schematic depicting how the PYY-VIP paracrine axis might regulate ion, water, and nutrient transport in the small intestine. b In the absence of EECs, ion, water, and nutrient transport are dysregulated due to loss of one arm of the PYY-VIP axis. In EEC-deficient small intestine, loss of PYY results in increased chloride transport and increased water and sodium accumulation in the intestinal lumen. Reduced NHE3 transport activity would cause accumulation of cytosolic H⁺ and a decrease in pH. Subsequently, nutrient absorption would be dysregulated, with diminished di-/tri-peptide absorption due to increased intracellular proton accumulation and increased uptake of glucose due to an exaggerated Na⁺ gradient across the apical membrane. c Na⁺-coupled glucose transport is deranged in EEC-deficient human and mouse small intestine. Wild-type and EEC-deficient human and mouse intestinal tissues were treated with VIP prior to 25 mM D-glucose. EEC-deficient intestine had an elevated initial response to glucose (mouse, n = 28 wild-type, n = 9 mutant, **P = 0.001; HIO, n = 6 wild-type, n = 4 mutant, **P = 0.002) that was returned to wild-type levels by pre-treatment with 10 nM exogenous PYY (mouse, n = 7, *P = 0.04; HIO, n = 3). Inhibition of NPY1R in wild-type tissues using BIBO3304 caused an abnormal initial response to glucose that mimicked EEC-deficient tissues (mouse, n = 12, **P = 0.005; HIO, n = 6). Bar graphs represent the slope of the curve depicted within the boxed area. Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. d The subcellular distribution of glucose transporters SGLT1 and GLUT2 is normal in human intestinal tissue lacking EECs. Representative images from eight independent organoids are shown. Scale bars = 50 μm. All error bars are + SEM.

Fig. 3. H⁺-coupled dipeptide absorption is impaired in EEC-deficient small intestine.

a EEC-deficient human and mouse small intestine did not respond to the dipeptide Gly-Sar after exposure to VIP (mouse, n = 9 wild-type, n = 6 mutant, ****P < 0.0001; human, n = 11 wild-type, n = 5 mutant, **P = 0.006). Pre-treatment of EEC-deficient tissue with exogenous PYY (mouse, n = 6, ns; human, n = 5, *P = 0.03), or of wild-type tissue with BIBO3304 (mouse, n = 9; human, n = 6) did not improve the I_sc response to Gly-Sar. Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. b Expression and localization of peptide transporter PEPT1 is unchanged in EEC-deficient human small intestine. Representative images from eight independent organoids are shown. Scale bars = 50 μm. c The PYY-VIP axis regulates intracellular pH in human small intestinal cells. EEC-deficient enteroids differentiated with VIP for 5–7 days developed an H⁺ imbalance with an acidic cytoplasm whereas wild-type enteroids were able to maintain their intracellular pH (**P = 0.004). Concurrent treatment with 10 nM PYY normalized the pH in EEC-deficient enteroids and was dependent on NPY1R. pHrodo MFI was analyzed by flow cytometry and normalized to vehicle-treated wild-type. n = 3 independent experiments. Statistics calculated by mixed effects analysis using the Holm–Sidak method. d Small intestinal EECs regulate proton transport in a paracrine fashion. Using reporter animals with mosaic loss of EECs we found that regions of jejunal epithelium that escaped recombination had normal pH as measured by pHrodo MFI. Adjacent regions that expressed tdTomato exhibited increased pHrodo MFI, indicating elevated cytosolic H⁺ (n = 4 mice, ***P = 0.0002). There was no difference in pHrodo MFI between mosaic regions in wild-type reporter jejunum (n = 8 mice). Statistics calculated by two-way ANOVA with Sidak’s multiple comparisons test. All error bars are + SEM.

Fig. 4. Exogenous PYY rescues EEC-deficient mice from malabsorptive diarrhea and restores normal glucose and dipeptide transport.

a Survival curve of wild-type mice (n = 100), EEC-deficient mice (n = 34), EEC-deficient mice treated once daily with 10 μg PYY (n = 25) beginning at postnatal day 10 (P10), and vehicle-treated EEC-deficient mice (n = 18). ****P < 0.0001 comparison of survival curves to untreated mutant by log-rank Mantel–Cox test. b At postnatal day 28, PYY-treated mutant mice experienced significant weight gain compared to vehicle-treated mutant mice (*P = 0.01, n = 5 mice per condition). Statistics calculated by one-way ANOVA. c EEC-deficient mice have intractable watery diarrhea from birth (n = 34 mice; ****P < 0.0001 from wild-type littermates, n = 100 mice). Within 48 h of PYY treatment, EEC-deficient animals improved to slightly soft yet well-defined fecal pellets (n = 25 mice, ****P < 0.0001 from untreated mutant). Vehicle-treated mutant mice did not improve (n = 18 mice). Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. d PYY treatment of EEC-deficient animals restored a normal resting I_sc to small intestine (n = 6 mice, ****P < 0.0001) and a normal electrogenic response to VIP (n = 6 mice, ****P < 0.0001). Treatment of mutant mice with vehicle did not improve basal I_sc (n = 6 mice) or response to VIP (n = 4 mice). Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. e PYY treatment restored a normal glucose response in EEC-deficient mouse and human intestine (mouse, n = 6, **P = 0.003; HIO, n = 5, **P = 0.004). Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. f Proton transport, as measured by pHrodo MFI, was normalized between mosaic regions in EEC-deficient reporter animals following PYY treatment (n = 2 mice). Statistics calculated by two-way ANOVA with Sidak’s multiple comparisons test. g PYY improved dipeptide transport in EEC-deficient mouse and human intestine. Long-term treatment of EEC-deficient animals and transplanted HIOs with PYY resulted in improved I_sc response to Gly-Sar compared to untreated mutant tissue (mouse, n = 6, **P = 0.009; HIO, n = 5, ***P = 0.0001). Vehicle-treated mutant mice did not exhibit improvement in Gly-Sar response (n = 18 mice, **P = 0.002). Statistics calculated by one-way ANOVA with Tukey’s multiple comparisons test. All wild-type and untreated mutant mouse data points are the same as shown in Figs. 1–3, and S2. All error bars are + SEM.