Fig 2: Immunostaining of sections from distal colon from Agr2+/+ (+/+) n = 5, Agr2+/- (+/-) n = 2, and Agr2-/- (-/-) n = 3 mice, two sections of each mouse were stained and examined.Two sections each of all Agr2-/- animals are shown in Fig. S2. The two upper panels display two sections each for the mature Muc2 (detected with anti-MUC2C3 antiserum, green) using a glycosylation independent antiserum (the gap between the inner mucus layer and the epithelium is due to shrinking during processing). A decreased amount of mature Muc2 stored in the mucin granules can be observed in the Agr2-/- mice. The middle panel with the DAPI channel alone most likely show bacteria stained as the smallest dots close to the epithelium in the Agr2-/- mice as indicated by arrows. The larger blue dots further away from the epithelium are shed cells or fecal content. The two lower panels show sections at two magnifications stained with an antiserum against the nonO-glycosylated Muc2 precursor (anti-gpA antiserum, green) found in the ER. A more fragmented ER with accumulation of immature Muc2 is observed in the Agr2-/- mice. DNA stained with DAPI (blue). Scale bar 25 µm for the top three panels, scale bar 10 µm for the lowest panel.

Fig 3: Western blot analysis of CHO-K1 cells stably expressing MUC2N and stably variable levels of AGR2.(A) Spent culture media and lysates normalized to similar number of cells were immunoblotted using an anti-AGR2 antibody after reducing SDS-PAGE. Notice that bands blighted in the center have the highest concentration of AGR2. The stable clones showed varying amounts of AGR2 (#9<#16<#4). Band intensities of AGR2 normalized to the strongest band (clone #4) in the blots of three replicates are ahown as a histogram in Fig. S1. (B) Immunoblot with an anti-GFP antibody detecting the MUC2N protein using the same media and lysates as in (A). In the lysates MUC2 was largely found as the not fully glycosylated form (lower band, slightly below 250 kDa kDa). The mature glycosylated form of MUC2 (higher band, above 250 kDa kDa) was found in the media from the control sample (MUC2N) and clone #9 (MUC2N AGR2 #9). Figure is representative of at least 3 replicates.

Fig 4: Agr2 is a component of the protective mucus layer in the whole gastrointestinal tract.(A) Boxplot of the relative levels (logarithm of the normalized intensities) of Agr2 in the mucus along the mouse gastrointestinal tract as revealed by proteomics. Sto: stomach; Duo: duodenum; Jej: jejunum; Ile: Ileum; PC: proximal colon; DC: distal colon. (B) Immunoblot detecting AGR2 in mucus from human colonic biopsies mounted in a horizontal perfusion hamber. The first lane (mucus) contains the collected secreted mucus. The second lane (lysate) shows the AGR2 remaining in the biopsy after mucus collection (diluted 1:100). (C) Molar ratios between Agr2, Muc2, Fcgbp and Clca1 in proximal and distal colonic murine mucus.

Fig 5: Western blot analysis detecting AGR2 in stably MUC2N- or MUC2C-expressing CHO-K1 cells.The same number of CHO-K1 cells were transiently transfected with equal amounts of expression plasmids coding for human AGR2 and its mutated variants. Equal amount lysates of MUC2N-expressing CHO-K1 cells (A) or MUC2C-expressing CHO-K1 cells (B) had been transfected with the different forms of AGR2, were separated by non-reducing SDS-PAGE and immunoblotted for AGR2. Only a single band at the expected size of AGR2 was observed, and no higher molecular bands were visible indicating that AGR2 did not bind to the MUC2N or MUC2C proteins expressed in the same cell line. Lysates of MUC2N- or MUC2C-expressing CHO-K1 cells immunostained for GFP are shown to the left and right, respectively. Figures are representative of at least 3 replicates. Band intensities of AGR2 normalized to the strongest band (MUC2N-C81S) in the blots of three replicates are plotted as a histogram (C) for MUC2N (correpsonds to gel in A) and (D) for MUC2C (correppnds to gel in B). Error bars represent standard deviation.