Maternal Epidermal Growth Factor Promotes Neonatal Claudin-2 Dependent Increases in Small Intestinal Calcium Permeability

Abstract

A higher concentration of calcium in breast milk than blood favors paracellular calcium absorption enabling growth during postnatal development. We aimed to determine whether suckling animals have greater intestinal calcium permeability to maximize absorption and to identify the underlying molecular mechanism. We examined intestinal claudin expression at different ages in mice and in human intestinal epithelial (Caco-2) cells in response to hormones or human milk. We also measured intestinal calcium permeability in wildtype, Cldn2 and Cldn12 KO mice and Caco-2 cells in response to hormones or human milk. Bone mineralization in mice was assessed by μCT. Calcium permeability across the jejunum and ileum of mice were 2-fold greater at 2 wk than 2 mo postnatal age. At 2 wk, Cldn2 and Cldn12 expression were greater, but only Cldn2 KO mice had decreased calcium permeability compared to wildtype. This translated to decreased bone volume, cross-sectional thickness, and tissue mineral density of femurs. Weaning from breast milk led to a 50% decrease in Cldn2 expression in the jejunum and ileum. Epidermal growth factor (EGF) in breast milk specifically increased only CLDN2 expression and calcium permeability in Caco-2 cells. These data support intestinal permeability to calcium, conferred by claudin-2, being greater in suckling mice and being driven by EGF in breast milk. Loss of the CLDN2 pathway leads to suboptimal bone mineralization at 2 wk of life. Overall, EGF-mediated control of intestinal claudin-2 expression contributes to maximal intestinal calcium absorption in suckling animals.

Keywords: EGF; calcium; development; intestine; mineral; paracellular; permeability.

Graphical Abstract

Epidermal growth factor in milk increases the paracellular permeability to calcium across the small intestine via increasing the intestinal claudin-2 abundance in suckling mammals.

Figure 1.

Calcium permeability across the small intestine is greater in young mice. (A) Duodenum (P = .0152), (B) jejunum (P = .0022), (C) ileum (P = .0007). Values presented as median and interquartile range. Analysis by Mann–Whitney test. P_Ca, Calcium permeability; P14, 2 wk old; 2 mo, 2 months old.

Figure 2.

Cldn2 and Cldn12 expression varies with age along the small intestine. Relative expression of (A–C) Cldn2 and (J–L) Cldn12 from P1 to 6 mo of age in the duodenum (A and J), jejunum (B and K), and ileum (C and L). Expression is normalized to Gapdh and relative to 1 mo. Data presented as mean ± SD, *P < .05 vs 1 mo by ANOVA with Dunnett’s multiple comparisons test. Immunoreactivity for CLDN2 was strong along both the villi and crypts of segments from (D) the duodenum, (E) the jejunum, and (F) the ileum at P14, whereas strong immunoreactivity was restricted to the crypts at 6 wk (G–I). Images were taken at 40× magnification and size bar indicates 50 µm. Representative images are obtained from evaluation of n = 3–4 mice per group.

Figure 3.

Claudin-2 confers greater P_Ca across jejunum and ileum in P14 mice. Calcium permeability across the duodenum (A, D), jejunum (B, E), and ileum (C, F) from Cldn2 (A–C) and Cldn12 (D–F) WT and KO mice at P14 and 2 mo. Data presented as mean ± SD and compared by one-way ANOVA with Tukey’s multiple comparisons test, *P < .05, ***P < .001, ^****P < .0001.

Figure 4.

Cldn2 KO pups at P14 display renal and intestinal compensation. (A) Urinary Ca²⁺ normalized to urinary creatinine (P = .011), (B) plasma PTH (P = .1680), (C) plasma calcitriol (P = .0323), and (D) plasma-ionized Ca²⁺ (P = .8182) in Cldn2 WT and KO mice at P14. Data presented as mean ± SD and compared by unpaired t-test. Relative abundance of genes expressing (E) TRPV6 and (F) Ca_v1.3 in the intestine of WT and Cldn2 KO mice at P14. Expression is normalized to β-Actin and presented relative to WT for each intestinal segment. (G) Net apical to basolateral ⁴⁵Ca²⁺ flux across ex vivo segments of jejunum from P14 mice (P = .0289). Each data point per segment represents an individual animal. Data presented as median ± IQR. Comparison of KO to WT for each intestinal segment by Mann–Whitney test. *P < .05. Duo, duodenum; Jej, jejunum; Ile, ileum; Col, colon. ND = not detected.

Figure 5.

Breast milk increases intestinal expression of claudin-2. mRNA abundance of Cldn2 in the (A) jejunum (P = .0019) and (B) ileum (P = .0003) of mice at P15 who remain with the dam (suckling) or who were weaned to solid chow at P12 (weaned). Expression is normalized to β-Actin and presented relative to suckling group as median ± IQR and compared by Mann–Whitney test. (C) Representative blots from 6 independent experiments of Caco-2 cells grown in media with increasing concentration as % by volume of human breast milk. (D) Semi-quantification of CLDN2 abundance from panel C normalized to β-actin and relative to control from the same blot. Data presented as mean ± SD and compared by one-way ANOVA with Tukey’s multiple comparisons test. *P < .05, **P < .01, ***P < .001. Con, control.

Figure 6.

EGF in breast milk increases claudin-2 expression. (A) Representative blot from 7 independent experiments of Caco-2 cells grown in media with additional hormones as indicated. (B) Semi-quantification of CLDN2 abundance from panel (A) normalized to β-actin and relative to control from the same blot. (C) Representative blots from 3 to 6 independent experiments of Caco-2 cells grown with or without human breast milk and the EGFR inhibitor erlotinib (0.1 µm). (D) Semi-quantification of CLDN2 abundance from panel C normalized to β-actin and relative to control from the same blot. (E) Calcium permeability across Caco-2 monolayers grown without and with 100 ng mL⁻¹ EGF normalized to control conditions. (F) Calcium permeability across Caco-2 monolayers in media containing 2% human milk without or with 0.1 µm erlotinib normalized to individual milk samples. *P < .05, **P < .01, ***P < .001, ^****P < .0001. Con, control; VitD, calcitriol; EGF, epidermal growth factor; Prl, prolactin; milk, human milk; P_Ca, calcium permeability.