Extended Data Figure 9: Phenotypic consequences of Xbp1 deletion in Paneth cells and characterization of murine norovirus status.

a, b, TUNEL-labelled ileal sections (a) with quantification of TUNEL⁺ cells per cm gut shown in b (n = 7/11; unpaired Student’s t-test; mean ± s.e.m.). Scale bars, 50 μm. c, d, Ki67 immunohistochemical staining of ileal sections (c) with quantification of Ki67⁺ cells per total intestinal epithelial cell count along the crypt–villus axis shown in d (n = 5; unpaired Student’s t-test; mean ± s.e.m.). Scale bars, 20 μm. e, f, BrdU-labelled ileal sections (e) after 24 h with quantification of BrdU⁺ cells per total intestinal epithelial cell count along the crypt–villus axis in f (n = 5/3; unpaired Student’s t-test; mean ± s.e.m.). Scale bars, 50 μm. g, h, PAS-stained ileal sections (g) with quantification of PAS⁺ cells per total intestinal epithelial cell count along the crypt–villus axis shown in h (n = 5; unpaired Student’s t-test; mean ± s.e.m.). Scale bars, 50 μm. i, Enteritis histology score of indicated genotypes (n = 3; median shown; Kruskal–Wallis). j, Taqman qRT–PCR for MNV of faecal samples in three different animal facilities (n = 3; mean ± s.e.m.). Note detectable MNV genome copy numbers in Boston (Atg7/Xbp1^ΔIEC, Fig. 2c) and Innsbruck (Atg16l1/Xbp1^ΔIEC, Extended Data Fig. 5g) colonies, and absence of MNV after re-derivation of colonies into the MNV-free Cambridge facility (Atg16l1/Xbp1^ΔIEC, Fig. 2e), yet similar levels of enteritis in Xbp1^ΔIEC and Atg16l1/Xbp1^ΔIEC mice maintained in Innsbruck (MNV⁺) and Cambridge (MNV⁻) (Fig. 2b, e and Extended Data Fig. 5g) and relatively more severe disease in Xbp1^ΔIEC mice from Boston (MNV⁺) (Fig. 2c). Sentinel mice from the Cambridge colony also tested negative for MNV by serological analysis (data not shown). These observations suggest that the inflammatory phenotypes observed are not necessarily dependent on MNV. *P < 0.05.