HIC1 links retinoic acid signalling to group 3 innate lymphoid cell-dependent regulation of intestinal immunity and homeostasis

Abstract

The intestinal immune system must be able to respond to a wide variety of infectious organisms while maintaining tolerance to non-pathogenic microbes and food antigens. The Vitamin A metabolite all-trans-retinoic acid (atRA) has been implicated in the regulation of this balance, partially by regulating innate lymphoid cell (ILC) responses in the intestine. However, the molecular mechanisms of atRA-dependent intestinal immunity and homeostasis remain elusive. Here we define a role for the transcriptional repressor Hypermethylated in cancer 1 (HIC1, ZBTB29) in the regulation of ILC responses in the intestine. Intestinal ILCs express HIC1 in a vitamin A-dependent manner. In the absence of HIC1, group 3 ILCs (ILC3s) that produce IL-22 are lost, resulting in increased susceptibility to infection with the bacterial pathogen Citrobacter rodentium. Thus, atRA-dependent expression of HIC1 in ILC3s regulates intestinal homeostasis and protective immunity.

Fig 1. Hic1 is required for intestinal immune homeostasis.

Intestinal lamina propria (LP) cells from Hic1^Vav and Hic1^fl/fl mice at steady state were analyzed by flow cytometry to enumerate populations of: (A, B) CD45⁺ leukocytes, (C, D) TCRβ⁺ and TCRγσ⁺ T cells, (E, F) CD11c⁺ MHCII⁺ CD64⁺ macrophages, CD11c⁺ MHCII⁺ CD64^- DCs, (G, H) RORγt⁺ ILC3s, GATA3⁺ ILC2s. Data pooled from 2 independent experiments (n = 4 per group). *, P < 0.05; Mann-Whitney test. Error bars indicate SEM.

Fig 2. Hematopoietic deficiency of HIC1 results in susceptibility to Citrobacter rodentium infection.

Hic1^Vav and Hic1^fl/fl mice were orally inoculated with C. rodentium. (A) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. (B, C) Bacterial loads (CFU/g) from fecal pellets (B) and liver (C) were measured at 11 days post inoculation. (D) Quantitative RT-PCR was performed to determine expression of Il17a, Il22 and Reg3g from distal colon tissue 11 days post inoculation. Data are pooled from 2 independent experiments (n = 8–9 per group). *, P < 0.05; **, P < 0.01; Mann-Whitney test. Error bars indicate SEM. nd, none detected.

Fig 3. Hic1 expression in T cells and dendritic cells is not required for immunity to Citrobacter rodentium infection.

(A–C) Hic1^CD4 and Hic1^fl/fl mice were orally inoculated with C. rodentium. (A) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. (B) Bacterial loads (CFU/g) from fecal pellets were measured over course of infection. (C) Quantitative RT-PCR was performed to determine expression of Il17a, Il22 and Reg3g from distal colon tissue 14 days post inoculation. (D–F) Hic1^CD11c and Hic1^fl/fl mice were orally inoculated with C. rodentium. (D) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. (E) Bacterial loads (CFU/g) from fecal pellets were measured over course of infection. (F) Quantitative RT-PCR was performed to determine expression of Il17a, Il22 and Reg3g from distal colon tissue 11 days post inoculation. (A-C) Data are pooled from 3 independent experiments (n = 7–11 per group). (D-F) Data are pooled from 2 independent experiments (n = 4–5 per group) *, P < 0.05; Mann-Whitney test. Error bars indicate SEM. ns, not significant.

Fig 4. ILC3-intrinsic HIC1 is required for immunity to Citrobacter rodentium infection.

Hic1^Rorc and Hic1^fl/fl mice were orally inoculated with C. rodentium. (A) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. (B, C) Bacterial loads (CFU/g) from fecal pellets (B) and liver (C) were measured at 11 days post inoculation (p.i.). (D) Quantitative RT-PCR was performed to determine expression of Il17a, Il22 and Reg3g from distal colon tissue 11 days p.i. (E, F) H&E stained histological sections of colon (E) and liver (F) from 11 days p.i. Scale bar represents 100μm. Black arrows indicate inflammatory infiltrate. (A–D) Data are pooled from 3 independent experiments (n = 13–14 per group). *, P < 0.05; **, P < 0.01; Mann-Whitney test. Error bars indicate SEM. nd, none detected.

Fig 5. ILC3-intrinsic HIC1 is required for immunity to Citrobacter rodentium in T cell-depleted mice.

Hic1^Rorc and Hic1^fl/fl mice were treated with a depleting anti-CD4 antibody and then orally inoculated with C. rodentium. (A) Colonic mRNA expression of Cd4 in control and anti-CD4 antibody treated Hic1^Rorc and Hic1^fl/fl mice. (B) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. (C, D) Bacterial loads (CFU/g) from fecal pellets (C) and liver (D) were measured at 11 days post inoculation (p.i.). (E) Quantitative RT-PCR was performed to determine expression of Il17a, Il22 and Reg3g from distal colon tissue 11 days p.i. (F, G) H&E stained histological sections of colon (F) and liver (G) from 11 days p.i. Scale bar represents 100μm. Black arrows indicate inflammatory infiltrate. Data from one experiment (n = 4–6 per group). *, P < 0.05; Mann-Whitney test. Error bars indicate SEM. nd, none detected.

Fig 6. Hic1 expression in intestinal ILCs is vitamin A dependent and is required for intestinal immune homeostasis.

(A) ILCs (lin^neg CD90⁺ CD127⁺ cells) were analyzed by flow cytometry for Hic1^Citrine reporter expression from the intestinal lamina propria (LP). Data representative of 2 independent experiments (B) Hic1 reporter expression in intestinal LP ILCs (lin^neg CD90⁺ CD127⁺ cells) from Hic1^Citrine mice fed a control diet, Hic1^Citrine mice fed a vitamin A deficient (VAD) diet, and controls fed a control diet was analyzed by flow cytometry. Data are representative of 2 independent experiments (n = 4–5 per group). (C, D) Intestinal LP cells from Hic1^fl/fl and Hic1^Rorc mice at steady state were analyzed by flow cytometry to enumerate populations of ILC3s (RORγt⁺) and ILC2s (GATA3⁺). Data are from 3 independent experiments (n = 6–7 per group) *, P < 0.05; **, P < 0.01; Mann-Whitney test. Error bars indicate SEM.

Fig 7. Hic1 does not regulate ILC precursors in the bone marrow.

(A) Gating strategy and (B) cell numbers of CLPs (CD45⁺ lin^neg CD127⁺ Flt3⁺ α4β7⁻), α4β7⁺ lymphoid progenitors (αLP; CD45⁺ lin^neg CD127⁺ Flt3⁻ α4β7⁺), ChILPs (CD45⁺ lin^neg CD127⁺ Flt3⁻ α4β7⁺ CD25⁻ c-Kit⁺) and ILC2 progenitors (ILC2p; (CD45⁺ lin^neg CD127⁺ Flt3⁻ α4β7⁺ CD25⁺ c-Kit⁻) from bone marrow of Hic1^Vav and Hic1^fl/fl mice. Data are from two independent experiments (n = 4 per group). *, P < 0.05; Mann-Whitney test. Error bars indicate SEM.

Fig 8. Recombinant mouse IL-22 is sufficient to promote resistance to Citrobacter rodentium in Hic1^Rorc mice.

(A-D) Intestinal ILC3s from Hic1^Rorc and Hic1^fl/fl mice were analysed for intracellular IL-22 by flow cytometry at steady state (A, B) or at day 4 post infection with C. rodentium (C, D). (E–G) Hic1^Rorc mice were infected with C. rodentium and treated with or without rmIL-22. (E) Weight loss (percentage of initial weight) was calculated for each mouse over course of infection. H&E stained histological sections of colon (F) and liver (G) from 11 days post infection. Scale bar represents 100μm. Black arrows indicate inflammatory infiltrate. (A–D) Data are from 2 independent experiments (n = 4 per group) (E) Data are from 2 independent experiments (n = 5–6 mice per group). *, P < 0.05; Mann-Whitney test. Error bars indicate SEM.