NF-κB determines Paneth versus goblet cell fate decision in the small intestine

Abstract

Although the role of the transcription factor NF-κB in intestinal inflammation and tumor formation has been investigated extensively, a physiological function of NF-κB in sustaining intestinal epithelial homeostasis beyond inflammation has not been demonstrated. Using NF-κB reporter mice, we detected strong NF-κB activity in Paneth cells, in '+4/+5' secretory progenitors and in scattered Lgr5+ crypt base columnar stem cells of small intestinal (SI) crypts. To examine NF-κB functions in SI epithelial self-renewal, mice or SI crypt organoids ('mini-guts') with ubiquitously suppressed NF-κB activity were used. We show that NF-κB activity is dispensable for maintaining SI epithelial proliferation, but is essential for ex vivo organoid growth. Furthermore, we demonstrate a dramatic reduction of Paneth cells in the absence of NF-κB activity, concomitant with a significant increase in goblet cells and immature intermediate cells. This indicates that NF-κB is required for proper Paneth versus goblet cell differentiation and for SI epithelial homeostasis, which occurs via regulation of Wnt signaling and Sox9 expression downstream of NF-κB. The current study thus presents evidence for an important role for NF-κB in intestinal epithelial self-renewal.

Keywords: Epithelial self-renewal; Goblet cells; Intestine; Mouse; NF-κB; Paneth cells; Stem cells.

Fig. 1.

Cell type-specific NF-κB activity in crypts of the small intestine. (A) Technovit section of X-gal-stained PSIs of κ-Gal reporter mice (n=3). Inset shows magnification with X-Gal-stained cells in the bottom of the crypts and in +4/+5 position (black arrowheads). Arrowheads indicate specific X-Gal staining in crypts, asterisks diffuse, nonspecific staining. Scale bars: 50 μm (main panel); 10 μm (inset). (B) Indirect immunofluorescence using an anti-EGFP antibody on PSI sections of κ-EGFP and κ-EGFP;ΔN mice (representative of n=3 mice per group). Yellow arrowheads indicate EGFP expression/NF-κB activity, asterisk marks nonspecific staining in villi. Scale bars: 50 μm (far left and far right); 20 μm (middle). (C) EGFP IHC [DAB (brown) alone and DAB plus Alcian Blue staining; left panels] and immunofluorescence co-staining of EGFP with anti-lysozyme or anti-Olfm4 antibodies on PSI sections of κ-EGFP mice (n=3). Black arrows and yellow arrowheads point to Paneth cells; black or red arrowheads indicate EGFP expression/NF-κB activity in ISCs, mainly in +4 and +5 positions. Scale bars: 20 μm. Nuclear counterstain: DAPI (blue).

Fig. 2.

Unaltered IEC proliferation and cell death rate in ΔN mice compared with controls. (A) Anti-Ki67 antibody staining on PSI sections of ΔN and control mice indicating proliferative cells. n=3 per group. (B) Quantification of average numbers of Ki67⁺ cells as a percentage of total DAPI-stained cells. n=3 per group, n>20 crypts per mouse. (C) Anti-BrdU antibody staining on PSI sections of ΔN and control mice indicating BrdU incorporation 4 or 24 h after injection. Insets show magnified images of crypts. In A and C, in crypts of controls proliferating ISCs (black arrowheads) are intermingled with Paneth cells (red arrowheads). (D) Quantification of BrdU⁺ cells per crypt (4 h) and per crypt-villus axis (24 h) in control or ΔN mice. n=3 per group, n≥20 crypts per mouse. (E) Representative IHC images of anti-cleaved caspase 3 and Alcian Blue staining on PSI sections of ΔN and control (n=3 per group) mice, and of a tissue sample of γ-irradiated control mice (see Kolesnichenko et al., 2021). Black arrows point to caspase 3-positive (apoptotic) cells in villi of irradiated controls. Asterisk indicates nonspecific staining due to handling. (F) Analysis of cleaved caspase-3 protein expression. Total PSI and PSI crypt whole-cell extracts of control and ΔN mice (n=3 per group) were used for SDS-PAGE western blotting. β-Actin was used as loading control. Scale bars: 50 µm (A,C); 200 μm (E). In B,D, error bars represent s.e.m.; n.s., not significant (unpaired Welch's t-test).

Fig. 3.

Increased numbers of goblet cells in ΔN mice. (A) IHC on PSI sections of control, ΔN and Villin-ΔN mice (n=3 per group) stained with either Alcian Blue alone or together with an anti-lysozyme antibody. Scale bars: 50 µm. Insets show magnified images of crypts. (B) Quantification of the average number of Alcian Blue-positive cells per crypt. **P<0.01, ***P<0.001 (unpaired Welch's t-test). Error bars represent s.e.m. n=5 per group, n=50 crypts per sample. (C) qRT-PCR for mature goblet cell markers Gob5, Muc2 and Klf4 using RNA from PSIs of ΔN and control mice (n=5 per group). *P<0.05, **P<0.01 (multiple t-test with Bonferroni correction). (D) Immunofluorescence using anti-Muc2 and anti-E-cadherin (cadherin 1) antibody staining on sections of PSI of control and ΔN mice (n=3 per group). Nuclear counterstain: DAPI (blue). Scale bars: 50 µm (A,D, main panels); 10 µm (insets).

Fig. 4.

Strongly reduced numbers of Paneth cells in crypts of ΔN mice. (A) Top: Immunofluorescence using anti-lysozyme antibody staining on sections of PSI of control, ΔN and Villin-ΔN mice (n=3 per group). White arrows indicate remaining Paneth cells in ΔN and Villin-ΔN crypts. Scale bars: 20 µm. Nuclear counterstain: DAPI (blue). Bottom: Hematoxylin & Eosin-stained high-magnification images of single crypts to illustrate Paneth cells (red arrows) in PSIs of ΔN, Villin-ΔN and control mice. Scale bars: 50 µm. (B) Quantification of lysozyme-positive crypts per field of view (FOV) in control and ΔN mice (n=3 per group), n≥150 crypts per sample. **P<0.01 (unpaired Welch's t-test). (C) Quantification of the average number of Paneth cells per crypt in PSIs of control, ΔN or Villin-ΔN mice (n=3 per group). **P<0.01 (unpaired Welch's t-test). (D) qRT-PCR for Paneth cell markers lysozyme and cryptdin-1 using RNA from PSIs of ΔN and control mice (n=6 per group). ****P<0.0001 (multiple t-test with Bonferroni correction). (E) qRT-PCR for secretory lineage progenitor markers Gfi1, Math1, Dll1 and Dll4, as well as for Hes1 and Spdef using RNA from PSI crypts of ΔN and control mice (n=3 per group). **P<0.01 (multiple t-test with Bonferroni correction). Error bars represent s.e.m.

Fig. 5.

NF-κB is required for differentiation and maturation of Paneth cells. (A) ISH on PSI sections of control and ΔN mice (n=3/group) at P9, P15 and 8 weeks of age using a riboprobe for cryptdin-1. Blue arrows (P9 panels) point to cryptdin-1-positive cells. Scale bars: 200 µm. (B) Quantification of cryptdin-positive crypts in controls and ΔN mice at the indicated time points (n=3 per group). *P<0.05; ****P<0.0001 (two-way ANOVA). Error bars represent s.e.m. FOV, field of view. (C) TEM analysis of PSI crypts of ΔN and control mice (n=4 per group). Yellow arrowheads point to mature Paneth cells in control and immature intermediate cells in ΔN mice. Bottom left panel shows a ΔN crypt lacking Paneth cells. Scale bars: 10 µm (left, control); 5 µm (left, ΔN), 2 µm (middle, ΔN). (D) PTAB staining on PSI sections of ΔN and control mice (n=3 per group). Black arrowheads point to mature, lysozyme-containing Paneth cells (red granules) in controls, blue arrowheads to immature intermediate cells (blue-greenish) that lack lysozyme. Scale bars: 20 µm. (E) Quantification of immature intermediate cells versus mature Paneth cells in ΔN and control mice (n=3 per group). *P<0.05, **P<0.01 (multiple t-test with Bonferroni correction). Error bars represent s.e.m.

Fig. 6.

In crypt organoids, suppression of NF-κB activity results in strongly reduced growth, but also an increase in goblet cells and a loss of Paneth cells. (A) Quantification of viable organoid formation at days 4 and 8 after isolation of PSI crypts from control or ΔN mice (n=4 per group). ****P<0.0001 (unpaired t-test). Error bars represent s.e.m. (B) Quantification of organoid growth in the presence (WENR) or absence (ENR) of Wnt3 in culture medium at days 5 and 8 after isolation of PSI crypts from control or ΔN mice (n=3 per group). *P<0.05 (unpaired t-test). n.s., not significant. Error bars represent s.d. (C) Upper panels: Representative images of crypt organoids cultured in ENR medium (without Wnt) at day 1 (d1) and day 4 (d4) after isolation of PSI crypts from control and ΔN mice (n=4 per group). Black arrows indicate viable organoids. Lower panels: Representative images of crypt organoids cultured in WENR medium (plus Wnt) at day 1 (d1) and day 5 (d5) after isolation of PSI crypts from control and ΔN mice (n=3 per group). (D) Representative immunofluorescence staining using an anti-lysozyme antibody (green) and DAPI (blue) of control and ΔN crypt organoids at d2 or d7 in ENR medium. (E) Representative images of immunofluorescence staining using E-cadherin (E-Cad, red) and Muc2 (green) antibodies on control and ΔN crypt organoids at d2 in ENR or WENR medium, or of Alcian Blue staining of Lgr5-EGFP (control) and Lgr5-EGFP;ΔN (ΔN) organoids, generated from single Lgr5⁺ cells isolated by FACS, at d7 grown in WENR medium. Scale bars: 50 µm. (F,G) qRT-PCR for the Paneth cell marker lysozyme and the goblet cell marker Muc2 using RNA from organoids at d7 in ENR medium (F) or WENR medium (G). *P<0.05, ****P<0.0001 (multiple t-test with Bonferroni correction). Error bars represent s.e.m.

Fig. 7.

Wnt3, Wnt10a and Sox9 expression is significantly reduced or absent in SI crypts and crypt organoids of ΔN mice. (A) qRT-PCR for Wnt3 using RNA from isolated PSI crypts of ΔN and control mice (n=3 per group). ***P<0.001 (unpaired Welch's t-test). Error bars represent s.e.m. (B) Left: ISH using antisense riboprobes for Wnt10a on PSI sections of control and ΔN mice (n=3 per group). Black arrows point to mRNA expression in crypts. Insets show magnified images of crypts. Right: Sense riboprobe for Wnt10a on PSI sections of controls (negative control; top), and Wnt10a antisense riboprobe on sagittal skin sections showing Wnt10a mRNA expression in the inner root sheath of an anagen hair follicle (black arrows; positive control; bottom). (C) qRT-PCR for Wnt3, Wnt10a and Axin2 using RNA from bulk crypt organoids of ΔN and control mice (n=3 per group) cultured without Wnt (ENR). **P<0.01, ***P<0.001 (multiple t-test with Bonferroni correction). Error bars represent s.e.m. (D) ISH using an antisense riboprobe for Sox9 on PSI sections of control and ΔN mice. Black arrows point to mRNA expression in crypts. Insets show magnified images of crypts. (E) qRT-PCR for Sox9 using RNA from PSIs of ΔN and IκBα^IEC-KO mice (n=3 per group). Expression level of control mice set to one. **P<0.01 (unpaired Welch's t-test). Error bars represent s.e.m. (F) IHC using a Sox9 antibody on PSI sections of control and ΔN mice (n=3 per group) shows strongly reduced Sox9 expression in ΔN mice. Insets show magnified images of crypts. Scale bars: 50 µm (main panels); 20 µm (insets). (G) PSI cytoplasmic and nuclear extracts of control and ΔN mice (n=3 per group) were used for SDS-PAGE western blotting to confirm the presence of nuclear Sox9 (arrowhead). Lactate dehydrogenase A (Ldha) was used as positive control for cytoplasmic extracts. Parp1 and active nuclear non-phosphorylated β-catenin were used as positive controls for nuclear extracts.