Tissue-layer-resolved proteome landscape of Crohn’s disease strictures highlights potential drivers of fibrosis progression
Johannes Alfredsson, Carina Sihlbom Wallem, Maja Östling, Hanna de la Croix, Elinor Bexe-Lindskog, Mary Jo Wick
Johannes Alfredsson, Carina Sihlbom Wallem, Maja Östling, Hanna de la Croix, Elinor Bexe-Lindskog, Mary Jo Wick
View: Text | PDF
Research Article Cell biology Gastroenterology

Tissue-layer-resolved proteome landscape of Crohn’s disease strictures highlights potential drivers of fibrosis progression

Abstract

The chronic inflammation of Crohn’s disease frequently leads to fibrosis and muscular hypertrophy of the intestinal wall. This often culminates in strictures, a serious condition lacking directed therapy. Severe pathological changes occur in the submucosa and muscularis propria intestinal wall layers of strictures, yet stricture-associated proteome changes in these layers is unexplored. We perform unbiased proteomics on submucosa and muscularis propria microdissected from transmural sections of strictured and nonstrictured ileum. Proteome changes in strictured submucosa reflected a transition from homeostasis to tissue remodeling, inflammation, and smooth muscle changes. Top submucosal features included reduced vascular components and lipid metabolism proteins accompanied by increased proteins with immune-, ECM-, or stress-related functions, including CTHRC1, TNC, IL-16, MZB1, and TXNDC5. In parallel, predominant changes in strictured muscularis propria included increased ECM (POSTN) and immune (mast cell CPA3) proteins alongside decreased proteins with lipid metabolic, mitochondrial, or key muscle functions. Finally, trends of differentially expressed proteins along nonstrictured submucosa suggest progressive profibrotic tissue remodeling and muscle expansion as proximity to strictures increases. The comprehensive proteome map presented here offers tissue-layer-resolved insight into the stricture microenvironment and potential drivers of fibrotic disease, providing a valuable resource to fuel biomarker and therapeutic target research.

Authors

Johannes Alfredsson, Carina Sihlbom Wallem, Maja Östling, Hanna de la Croix, Elinor Bexe-Lindskog, Mary Jo Wick

×

Figure 1

Overview of the workflow.

Options: View larger image (or click on image) Download as PowerPoint
Overview of the workflow. The figure illustrates the workflow from FFPE ...
The figure illustrates the workflow from FFPE block preparation through (A) initial imaging analysis, (B) laser microdissection (LMD) of submucosa (SM) and muscularis propria (MP) tissue layers, and (C) layer-wise mass spectrometry (MS) analysis. (A) Reference slides from each FFPE block were stained and scanned for initial image analysis. Given the “dense” versus “loose” tissue density in STR and NSTR samples (right), we developed an image-based method to identify one SM and one MP region from each sample to standardize tissue content of 44 mm2 across the cohort. Preliminary region outlines were drawn, and the net foreground tissue coverage was estimated using image analysis. Region sizes were then iteratively adjusted until reanalysis showed tissue coverage of 44 mm2 was obtained (see Supplemental Methods). The percentages inside the detected tissue overlays to the right indicate the foreground-to-background area ratio (“density”) detected in the tissue scan that was subsequently adjusted for by our standardization. (B) After standardized regions were identified, LMD slides were prepared from additional serial sections, scanned for use in creating an LMD guide, and stored until LMD. The standardized regions were redrawn onto these prescanned LMD slide images in the analysis software and used as visual guides during subsequent LMD. At the LMD microscope, the outlines of the standardized 44 mm2 tissue regions were laser microdissected to extract SM and MP layers for proteomic analysis. As illustrated in the schematic, the final sample cohort comprised SM and MP dissected from both STR and NSTR regions of 12 patients with CD and 8 CTRL individuals, yielding a total of 32 samples per layer. (C) The 64 LMD samples were prepared and analyzed using TMT-labeled MS. The number of samples exceeded the unique barcodes (TMT labels) available and were thus analyzed as 2 separate sets of 16 samples per layer (SM1 and SM2; MP1 and MP2; see Supplemental Methods 6 and 7). For internal validation, one-tenth of the volume of each sample was set aside before TMT labeling and used for label-free proteomics using TIMS-TOF-MS. The results from the 2 MS methodologies were integrated for robust downstream analysis at the combined level. MT, Masson’s trichrome.