Fig. 5: Docking of Gal-GalNAc-Thr to Fap2-ECD and validation of binding site.

a Surface representation of Fap2 colored according to Coulomb potential (red: negative, blue: positive charges) with docked Gal-GalNAc-Thr. A representative frame of the MD simulation of the best docking pose is shown (Supplementary Movie 4). The Fap2 model has two clefts adjacent to the docking sites, which is in agreement with the polypeptide chain of the Gal-GalNAc-containing O-glycosylated receptor going through (indicated by gray dashed line). b Hydrogen bonds between the docked Gal-GalNAc-Thr and R613, E619, E713, and F818 of Fap2, which remain most consistent in MD simulation. c Representative fluorescence micrographs that show HT-29 cell binding of Fap2-expressing E. coli (analogous as in Fig. 3b), in which Fap2 WT and Fap2 mutant R613P_E619A_F818P are compared. Red: E. coli labeled with CellBrite Fix 555, green: Actin labeled with Phalloidin-iFluor 488. Scale bar: 50 µm. The images are representatives of data sets of 267 micrographs each for Fap2 WT-expressing E. coli and control E. coli, and 273 micrographs for Fap2 mutant-expressing E. coli. d Quantification of experiments in (c), with average fluorescence signal of E. coli bound to HT-29 determined from the indicated number of micrographs each. The mean is shown as red dot. The boxes represent the interquartile range, with the median shown as horizontal line within the box. The distribution of data points is visually represented by violin plots. Statistical evaluation was done using the Games-Howell post-hoc test from the ggstatsplot package⁸² in R. e Pulldown of Fap2-ECD with hTIGIT-ECD in the absence (left; n = 3) and presence (right; n = 1) of 25 mM GalNAc. The arrow indicates the position of Fap2-ECD, the arrow with asterisk indicates the position of hTIGIT-ECD-Fc.