Microengineered 3D pulmonary interstitial mimetics highlight a critical role for matrix degradation in myofibroblast differentiation

Abstract

Fibrosis, characterized by aberrant tissue scarring from activated myofibroblasts, is often untreatable. Although the extracellular matrix becomes increasingly stiff and fibrous during disease progression, how these physical cues affect myofibroblast differentiation in 3D is poorly understood. Here, we describe a multicomponent hydrogel that recapitulates the 3D fibrous structure of interstitial tissue regions where idiopathic pulmonary fibrosis (IPF) initiates. In contrast to findings on 2D hydrogels, myofibroblast differentiation in 3D was inversely correlated with hydrogel stiffness but positively correlated with matrix fibers. Using a multistep bioinformatics analysis of IPF patient transcriptomes and in vitro pharmacologic screening, we identify matrix metalloproteinase activity to be essential for 3D but not 2D myofibroblast differentiation. Given our observation that compliant degradable 3D matrices amply support fibrogenesis, these studies demonstrate a departure from the established relationship between stiffness and myofibroblast differentiation in 2D, and provide a new 3D model for studying fibrosis and identifying antifibrotic therapeutics.

Fig. 1. Matrix cross-linking and stiffness have opposing effects on myofibroblastic differentiation of fibroblasts plated on 2D versus in 3D hydrogels.

(A) Histological preparations of healthy control and bleomycin-treated murine lung tissue (n = 3 mice per group) stained for collagen by picrosirius red (scale bar, 100 μm). (B) Young’s modulus of mouse lung tissue as measured by AFM nanoindentation, with data fit to the Hertz contact model to determine Young’s modulus (n = 3 mice per group, n = 50 indentations per group on n = 9 tissue sections). (C) Schematic of proteolytically sensitive, cell-adhesive DexVS-VPMS bulk hydrogels. (D) Young’s modulus determined by AFM nanoindentation of DexVS-VPMS hydrogels formed with different concentrations of VPMS cross-linker (n = 4 samples per group, n = 20 total indentations per group). (E and F) Representative images of F-actin (cyan), nuclei (yellow), and α-SMA (magenta); image-based quantification of α-SMA expression (left axis, magenta bars, day 9) and nuclear Ki67 (right axis, gray bars, day 5) in 2D and 3D (n = 4 samples per group, n = 10 fields of view per group, n > 50 cells per field of view; scale bars, 200 μm). All data presented are means ± SDs with superimposed data points; asterisk denotes significance with P < 0.05 determined by one-way analysis of variance (ANOVA). AU, arbitrary units.

Fig. 2. Microengineered fibrous hydrogel composites to model the lung interstitium.

(A) SHG imaging of collagen microstructure within healthy and bleomycin-treated lungs on day 14, with quantification of average signal intensity (arrows indicate interstitial tissue regions adjacent to alveoli; n = 3 mice per group, n = 10 fields of view per group; scale bar, 100 μm). (B) Schematic depicting polymer cross-linking and functionalization for generating fibrous DexVS hydrogel composites to model changes in fiber density within lung interstitial tissue ECM. (C) Images and intensity quantification of fluorophore-labeled fibers within composites varying in fiber density (n = 4 samples per group, n = 10 fields of view per group; scale bar, 100 μm). Young’s modulus determined by AFM nanoindentation of fibrous composites formed with different concentrations of VPMS cross-linker (n = 4 samples per group, n = 20 measurements per group). (D) Representative high-resolution images of NHLFs on day 1 in fibrous composites formed with bulk hydrogels (12.5 mM VPMS) functionalized with integrin ligand arginylglycylaspartic acid (RGD) or heparin-binding peptide (HBP) [F-actin (cyan), nuclei (yellow), and DexVS fibers (magenta); scale bar, 50 μm]. Quantification of fiber recruitment as measured by contact between cells and DexVS fibers (n = 10 fields of view per group, n > 25 cells analyzed). (E) Representative high-resolution images of NHLF on day 1 fibrous composites formed with bulk hydrogels functionalized with integrin ligand RGD or HBP [F-actin (cyan), fibronectin (yellow), and DexVS fibers (magenta); scale bar, 5 μm]. Quantification of fibronectin deposition into tshe hydrogel matrix as measured by immunostain intensity (n = 10 fields of view per group, n > 25 cells analyzed). All data presented are means ± SDs with superimposed data points; asterisk denotes significance with P < 0.05 determined by one-way ANOVA or Student’s t test, where appropriate; NS denotes nonsignificant comparison.

Fig. 3. Increasing matrix fiber density in HBP-presenting 3D fibrous hydrogels primes fibroblasts for activation into MFs.

(A) Immunofluorescence images of NHLFs in hydrogel composites over a range of fiber densities after 3 days of culture [F-actin (cyan), fibronectin (FN, yellow), YAP (magenta), Ki67 (white), and nuclei (blue); scale bars, 100 μm (F-actin), 20 μm (FN), 20 μm (YAP), and 100 μm (Ki67/nuclei)]. (B) Corresponding image-based quantification of cell area, deposited FN, YAP nuclear to cytosolic ratio, and % of proliferating cells (n = 4 samples per group; for cell spread area analysis, n > 50 cells per group; for FN, YAP, and Ki67 analyses, n = 10 fields of view per group and n > 25 cells per field of view). (C) Cytokine secretion into culture medium on day 3 (all data were normalized to background levels in control medium, n = 4 samples per condition). (D) Expression of MF-related genes in NHLFs stimulated with TGF-β1 on day 3, in either highly fibrous (FD 5.0%) or nonfibrous (FD 0.0%) hydrogels (data presented are GAPDH-normalized fold changes relative to NHLFs within an FD 0% hydrogel lacking TGF-β1 supplementation). All data presented are means ± SDs with superimposed data points; asterisk denotes significance with P < 0.05 determined by one-way ANOVA or Student’s t test where appropriate.

Fig. 4. Profibrotic soluble and physical cues promote MF differentiation in 3D and initiate a progression of fibrosis-associated changes over long-term culture.

(A) Representative immunofluorescence images of NHLFs in microenvironmental conditions leading to low (top row) or high (bottom row) MF differentiation after 9 days in culture [α-SMA (magenta) and nuclei (cyan); n = 4 samples per group, n = 10 fields of view per group, and n > 50 cells per field of view; scale bar, 200 μm], with corresponding image-based quantification in (B) and (C). Insets depict representative fiber densities. (D) Measurement of total cytokine secretion over time as a function of fiber density (n = 4 samples per condition; * indicates significant differences between FD 5.0% and all other groups at a given time point; NS denotes nonsignificant comparison). (E) Secretion of specific cytokines and chemoattractants as a function of fiber density on day 9 (n = 4 samples per condition). (F) Representative images and quantification of tissue contraction within day 14 fibroblast-laden hydrogels of varying fiber density (n = 4 samples per group, dashed line indicates initial diameter of 5 mm). Photo credit: Daniel Matera, University of Michigan. (G) AFM measurements of day 14 fibroblast-laden hydrogels of varying fiber density (n = 20 measurements from n = 4 samples per group). Dashed line indicates original hydrogel stiffness. (H) SHG images of fibrous collagen within fibroblast-laden hydrogels after 21 days of culture in medium supplemented with ascorbic acid (scale bar, 100 μm). (I) Measurement of total collagen content within digested DexVS hydrogels at day 21 as measured by biochemical assay (n = 4 samples per group). All data presented are means ± SDs with superimposed data points; asterisk denotes significance with P < 0.05 determined by one-way ANOVA; NS denotes nonsignificant comparison.

Fig. 5. Pharmacologic treatment of NHLFs plated on tissue culture plastic in 2D compared to encapsulated within 3D fibrous hydrogels reveals differential drug responses.

(A) Representative confocal images stained for α-SMA (magenta), F-actin (cyan), and nuclei (yellow) of NHLFs after 9 days of culture on tissue culture plastic (TCP) (top row) or 3D fibrotic matrices (bottom row) with pharmacologic treatment indicated from days 3 to 9 (scale bar, 100 μm). Imaged regions were selected to maximize the number of α-SMA⁺ cells per field of view within each sample. (B) Quantification of α-SMA and (C) total cell count within 2D NHLF cultures. (D) Quantification of α-SMA and (E) total cell count within 3D fibrotic matrices (n = 4 samples per group, n = 10 fields of view per group, and n > 50 cells per field of view). All data presented are means ± SDs with superimposed data points; asterisk denotes significance with P < 0.05 determined by one-way ANOVA; NS denotes nonsignificant comparison.

Fig. 6. Bioinformatics analysis of IPF patient transcriptomes reveals matrix remodeling as a key signaling node in human disease.

(A) Schematic representation of bioinformatics workflow: Whole-genome transcriptomes from 91 healthy and 140 patients with lung fibrosis were fetched from the NCBI GEO. Differential expression analysis was used to assemble an uncurated list of the top 1000 differentially expressed genes. GO enrichment of choice biological pathways was used to assemble a curated list of genes associated with MF differentiation. Datasets were fed through a previous knowledge–based analysis pipeline to identify enriched signaling pathways (Reactome) and key protein signaling nodes (STRING) within patients with IPF. (B) Heatmaps of the top 20 differentially expressed genes within specified GO categories, which were manually selected for curated analysis. C_N values indicate a high degree of interaction between proteins selected for curated analysis. Colors are based on differential expression values that were not log-normalized. (C) Summary of the top 5 significantly enriched pathways in the curated and uncurated gene set. (D) Representative STRING diagram depicting protein interactions within the curated dataset, with summary of the top 5 signaling nodes in the uncurated and curated gene set. Blue nodes and edges represent interactions within the top 5 signaling nodes for the curated dataset.