Biotin-Ahx-LPETGS-NH2

BALB/cj (7-12 weeks old, male and female, OVA-alum food allergy model)^[1]

20 mM stock; first injection 400 µL, injections 2-9 each 200 µL

i.p.; nine times at 20 min intervals

Enabled sortase-mediated covalent labeling of immune cells physically interacting with intestinal epithelial cells.
In Villin-uLIPSTIC mice, both epithelium-associated and lamina propria-resident mast cells exhibited substantial biotin labeling, and epithelium-associated eosinophils exhibited higher biotin labeling than their lamina propria counterparts.
Biotin labeling of eosinophils was minimal after the first OVA-OC and progressively increased after the fourth and sixth challenges, correlating with severity of allergic diarrhea (t=0.83).
Single-cell RNA sequencing of FACS-purified, epithelial-associated immune cells from Villin-uLIPSTIC mice after OVA-OC identified mast cells, T cells, and pDCs with higher biotin labeling than eosinophils and ILC3s.
Among four eosinophil subclusters, subcluster-0 and subcluster-1 displayed the highest biotin labeling, defined by wound repair (Alox15, Emilin2) and chemotaxis (Dock2, Elmo1) signatures, and were enriched for secreted effectors (IL-4 and Csf1) and cell surface adhesion molecules (Itgal and Alcam).
The majority of epithelial-associated eosinophils after the sixth OVA-OC were LFA-1hiCD43int and carried significantly more biotin than LFA-1loCD43lo eosinophils.
Eosinophils accounted for the vast majority of IL-13-expressing biotin-positive cells after the sixth OVA-OC.
In iEC subset-specific uLIPSTIC mice (Pou2f3GFP-CreERT2 for tuft cells, Defa6Cre for Paneth cells, Neurod1CreERT2 for EECs), under physiological conditions Paneth cells interacted with immune cells at a significantly higher rate than EECs or tuft cells after normalization to cell abundance, while tuft cell interactions were exceedingly rare.
Following OVA-OC, tuft cell-immune interactions remained minimal; EECs exhibited substantial and regionally variable interactions (mast cells preferentially labeled in duodenum, eosinophils in ileum and colon); Paneth cells displayed regionally variable interactions with ileal eosinophils and duodenal mast cells overrepresented (interaction score > 1).

C57Bl/6 (7-12 weeks old; Il5tdTom-iCre/+Rosa26uLIPSTIC/+ and PdgframGFP-CreERT2/+Rosa26uLIPSTIC/+; IL-33-induced inflammation model)^[2]

100 µL of 20 mM solution

i.p.; every 20 minutes; 2 hours

In PBS-treated control mice, pancreatic ILC2 did not label many Ly6c+ or Ly6c- fibroblasts or other cell-types.
Even during acute IL-33-driven inflammation, no direct ILC2-fibroblast interactions were observed, although CD172a+CD11b+ cDC2 acquired the biotin signal.
In Pdgfra-uLIPSTIC mice, fibroblasts did not biotinylate ILC2 but did biotinylate F4/80+MHCII+/low macrophages in the pancreas after IL-33 administration.
Biotin-tagged F4/80+ macrophages showed increased GFP signal, suggesting uptake of Pdgfra-GFP+ cells.