Intestinal epithelial cell secretion of RELM-beta protects against gastrointestinal worm infection

Abstract

Th2 cells drive protective immunity against most parasitic helminths, but few mechanisms have been demonstrated that facilitate pathogen clearance. We show that IL-4 and IL-13 protect against intestinal lumen-dwelling worms primarily by inducing intestinal epithelial cells (IECs) to differentiate into goblet cells that secrete resistin-like molecule (RELM) beta. RELM-beta is essential for normal spontaneous expulsion and IL-4-induced expulsion of Nippostrongylus brasiliensis and Heligmosomoides polygyrus, which both live in the intestinal lumen, but it does not contribute to immunity against Trichinella spiralis, which lives within IEC. RELM-beta is nontoxic for H. polygyrus in vitro but directly inhibits the ability of worms to feed on host tissues during infection. This decreases H. polygyrus adenosine triphosphate content and fecundity. Importantly, RELM-beta-driven immunity does not require T or B cells, alternative macrophage activation, or increased gut permeability. Thus, we demonstrate a novel mechanism for host protection at the mucosal interface that explains how stimulation of epithelial cells by IL-4 and IL-13 contributes to protection against parasitic helminthes that dwell in the intestinal lumen.

Figure 1.

IEC IL-4Rα expression is required for expulsion of N. brasiliensis and RELM-β production. (A) Kinetics of fecal egg count in N. brasiliensis–inoculated IL-4Rα^flox/− (WT phenotype) and IL-4Rα^flox/−/Villin-Cre (selective IEC IL-4Rα deficiency) mice. (B) Intestinal worm burden 30 d after inoculation. Means ± SE of 8–10 mice/group are shown. The experiment was performed three times. (C) IL-4 and IFN-γ secretion measured by IVCCA 10 d after inoculation. (D) Serum levels of IL-13–soluble IL-13Rα2 complexes 10 d after inoculation. The error bars indicate the SE of 8–10 mice per group. ***, P < 0.001. (E) Quantitation of goblet cells per villus 10 d after inoculation. Mean ± SE is shown of 400 villi/group. (F) Immunofluorescence staining for RELM-β (red) and cell nuclei (blue) demonstrates colocalization of RELM-β and goblet cell mucus. 400× magnification. Yellow bars, 10 µm. A representative photo is shown from three independent experiments. **, P < 0.01; *, P < 0.05, compared with WT. N.D., none detected.

Figure 2.

Intestinal RELM-β expression is important for N. brasiliensis and H. polygyrus expulsion. (A and B) RELM-β (A) and RELM-α (B) mRNA were quantitated by real-time PCR in intestinal and lung tissue after N. brasiliensis infection. Results are representative of two independent experiments with mean ± SE. n = 6–8 mice/group. **, P < 0.01. (C) Lung worm burden in N. brasiliensis–inoculated WT and RELM-β^−/−. Mean ± SE is shown of five to six mice per group. Error bars are too small to be seen. The experiment was performed twice. (D) Intestinal worm burden in N. brasiliensis–inoculated WT and RELM-β^−/−. Mean ± SE is shown of 8–10 mice/group. The experiment was performed four times. (E) Real-time PCR analysis of IL-4 and IL-13 in gut mRNA 7 d after inoculation. Mean ± SE is shown of five to six mice per group. (F) Serum levels of IFN-γ and IL-4 measured by IVCCA 10 d after inoculation. **, P < 0.01. (G) Intestinal worm burden of WT and RELM-β^−/− mice 10 d after inoculation in mice treated with anti–IFN-γ (XMG-6) or isotype control mAb (GL113). (H) Intestinal worm burden 7, 10, and 14 d after second infection of WT or RELM-β^−/− with H. polygyrus. Dpi, days post inoculation. The error bars indicate the SE of five to six mice per group. The experiment was performed four times. ***, P < 0.001, compared with WT.

Figure 3.

RELM-β contribution to IL-4–mediated N. brasiliensis and H. polygyrus expulsion. (A) Worm counts 9 d after N. brasiliensis inoculation of WT or RELM-β^−/− treated with anti-CD4 mAb and vehicle or IL-4C. The experiment was performed three times. The error bars indicate the SE of five to six mice per group. (B) Number of goblet cells per villus 10 d after inoculation in the experiment shown in A. Data show mean ± SE of 400 villi/group. (C) FITC-dextran permeability of muscle-free jejunum segments from mice in the experiment shown in A. (D) Transepithelial resistance of muscle-free segments of jejunum from mice in the experiment shown in A. *, P < 0.05 compared with WT or WT vehicle-treated group; **, P < 0.01. The error bars in C and D indicate the SE of five to six mice per group. (E) Intestinal worm burdens 14 d after primary infection with H. polygyrus in WT or RELM-β^−/−. Mice were treated with vehicle or IL-4C on days 8, 10, and 12. ***, P < 0.001. Means ± SE are shown of 8–10 mice/group. The experiment was performed three times.

Figure 4.

RELM-β directly inhibits fecundity and survival of adult H. polygyrus. (A) Numbers of adult worms recovered 2 d after inoculation of WT and IL-4Rα–deficient (IL-4Rα^−/−) BALB/c mice with H. polygyrus adult worms treated in vitro with rIL-4 or 1–100 ng/ml of rRELM-β. The experiment was performed three times. (B) Fecal egg counts from the experiment shown in A. The experiment was performed three times. (C) Intestinal worm burden 2 d after inoculation of IL-4Rα^−/− with H. polygyrus adults treated in vitro with rRELM-α, rRELM-β, heat-treated (Δ°) rRELM-β, or rIL-4. The experiment was performed three times. (D) Intestinal worm burden of WT mice 2 d after oral challenge with equal numbers of rRELM-α– or rRELM-β–pretreated H. polygyrus adults. Worms were labeled with CFSE or vehicle (mock) to distinguish RELM-β– from RELM-α–pretreated worms. The experiment was performed three times, with selective CFSE labeling of RELM-β–treated worms in some experiments and RELM-α–treated worms in other experiments. The vertical bars indicate the mean of five to six mice per group. ***, P < 0.001. (E) Intestinal worm burden 2 d after inoculation of WT or RAG-2–deficient mice with untreated or RELM-β–treated H. polygyrus adult worms. The experiment was performed two times. The error bars indicate the SE of five to six mice per group.

Figure 5.

RELM-β impairs feeding and reduces ATP and protein content in H. polygyrus adult worms. (A) Representative photomicrographs of H. polygyrus adults recovered from WT mice 2 d after oral gavage with adult worms pretreated with rRELM-α or rRELM-β. 4× magnification. The experiment was performed four times. Bar, 2 mm. (B) ATP levels of individual adult worms recovered from WT mice 2 d after oral inoculation. Worms were treated with rIL-4, rRELM-β, Δ°rRELM-β, or rRELM-α before inoculation of mice. Data are expressed as relative luminescence units (RLU). Error bars indicate means for 15–17 worms/group. The experiment was performed three times. (C) Protein content of adult worms treated as in Fig. 4 A. (D) H. polygyrus adults were treated with rRELM-α or rRELM-β and inoculated into naive WT. Mice were injected i.v. 2 d later with rhodamine B and worms were recovered from mouse intestines 1 h afterward. 100× magnification. The experiment was performed three times. Bar, 100 µm. (E) Number of viable H. polygyrus adult worms and eggs produced per worm after 16 h in vitro culture in medium supplemented with saline and 1 µg/ml of rRELM-α or rRELM-β. The experiment was performed three times. (F) Fecundity of H. polygyrus adult worms cultured for 24, 48, or 72 h with rRELM-α or rRELM-β. Worms were moved to fresh culture wells every 24 h to allow determination of egg production for each 24-h period. (G) ATP levels of adult H. polygyrus worms cultured for 24, 48, or 72 h with rRELM-α or rRELM-β. (H) Protein content of adult H. polygyrus worms cultured for 24, 48, or 72 h with rRELM-α or rRELM-β. The experiments were performed three times. The error bars indicate the SE of 15–17 worms per group. *, P < 0.05; **, P < 0.01; ***, P < 0.001, compared with IL-4–treated worms.