Perivascular Matrix Densification Dysregulates Angiogenesis and Activates Pro‐Inflammatory Endothelial Cells

Figures

• 

  FIGURE 1. Emergence of aberrant ECs in the bleomycin lung injury model. (a) Breeding strategy and study timeline for EC lineage tracing during bleomycin injury‐induced lung fibrogenesis. (b) Pipeline for 3D imaging of lung tissue. Representative low magnification whole‐lobe tile‐scan images and high magnification single z‐slice images of peribronchial vasculature in (c) uninjured control lung and (d) bleomycin‐treated lung tissues 21 days post‐injury. Movie S1: same image set presented as a slice‐through movieshowing multiple vessels and abundant single ECs. (e) Representative high magnification single z‐slice images (left, middle) and 3D rendering (right) depicting cross‐sectional and longitudinal views of bronchial vasculature from day 21 bleomycin treated mice. Yellow arrow heads mark individualized Tie2‐Cre‐GFP positive cells expressing αSMA. BA: bronchial arterioles, BV: bronchial venules, BE: bronchial epithelium. Movie S2, S3: same image sets presented as slice‐through movies.

• 

  FIGURE 2. Heightened matrix fiber density induces spontaneous tip cell formation in the absence of soluble cues. (a) Schematic of microvessel microphysiological model and tip cell formation assay. (b) S1P‐induced tip cell formation over 4‐days within 10 mg/ml fibrin hydrogels as a function of S1P concentration. Nuclei (magenta), F‐actin (cyan), yellow dashed lines indicate parent vessel edge. (c) Quantification of number of tip cells (data representative of N>3 replicate studies with n>5 fields of view analyzed per condition) and (d) tip cell invasion depth as a function of [S1P] (data representative of N>3 biological replicates with n>7 invading tip cells per group). (e) 3D rendering of fiber‐embedded fibrin hydrogels at varying fiber density. Synthetic DexVS fiber segments (cyan), fluorescently labeled fibrin hydrogel (magenta). (f) Quantification of fiber density and Young's modulus as a function of input fiber stock volume. N > 3 individual hydrogels per group. (g) Flow‐induced fiber alignment perpendicular (0°) or parallel (90°) to the long axis of the parent vessel. (h) Fiber‐induced tip cell formation (without S1P added to the chemokine channel) over 4‐days within 10 mg/ml fibrin…

• 

  FIGURE 3. Fiber‐mediated ATEC activation and invasion share signaling and cellular functions previously implicated in EndMT and angiogenic tip cells. (a–c) Representative confocal images (maximum intensity projections) depicting the differential localization of immunostained VE‐cadherin (a), YAP (b), and SNAI1 (c) invading ECs and ECs within the parent microvessel (red arrow) after 4 days of culture. Top row: nuclei (magenta), F‐actin (cyan), DexVS‐RGD fibers (grey), yellow dashed lines indicate parent vessel edge. Bottom row: inverted grayscale images, magenta dashed lines indicate boundaries of nuclei, yellow dashed lines indicate parent microvessel edge. (d–f) Representative images (d) of vehicle control (DMSO 1:1000 dilution), marimastat (1 µm), blebbistatin (30 µm), or Y27632 (25 µm) treated microvessels (treated over days 1–4 of culture) within fibrin + FD 2% hydrogels after 4 days of culture, along with corresponding quantification of the number of invading ECs (e: N=3 replicate studies with n>6 fields of view analyzed per condition) and EC invasion depth (f: N=3 biological replicates with n>9 invading cells analyzed per condition. (g–i) Representative images (g) of vehicle control…

• 

  FIGURE 4. VE‐cadherin destabilization increases vessel permeability and ATEC activation. (a,b) Representative maximum intensity projection images of microvessels with or without FD 2% cell‐adhesive fibers after one day of culture (a) along with corresponding quantification of vessel wall cell density and microvessel diameter (b). Data representative of N=4 replicate studies with n>24 fields of view analyzed per condition. (c) Representative single z‐slice (at mid‐plane of microvessel) following fluorescent dextran perfusion on day 4 of culture. (d) Quantification of vessel wall diffusive permeability of microvessels cultured for 1 or 4 days in FD 0% or 2% hydrogels. N = 4 biological replicates with n>12 fields of view analyzed per condition. (e,f) Representative images (e) of EC monolayers cultured for 2 days on glass coverslips or electrospun DexVS fiber matrices along with corresponding quantifications of VE‐cadherin signal per cell, pulsed VE‐cadherin signal per cell, and adherens junction (AJ) width as determined by segmentation of VE‐cadherin+ cell‐cell junctions (f). N=3 replicate studies with n>20 fields of view analyzed per condition for. Data representative of N=3 replicate…

• 

  FIGURE 5. Transcriptomic analysis of ATEC suggests an inflammatory and pro‐fibrotic secretory phenotype. (a) Representative maximum intensity projection images of HUVECs embedded within control fibrin (FD 0%) or fibrous (fibrin + FD 2%) hydrogels cultured over 3 days to model ATECs at sufficient numbers for transcriptomic analysis. Corresponding quantification of EC spread area. Images and quantification representative of N=3 replicate experiments. (b) Schematic depiction of conditions for transcriptomic analysis. Experiment was performed once with n>2 technical replicates. (c) Expression of several curated gene categories relative to parent microvessels after 3 days of culture. (d) Volcano plot visualizing differentially expressed genes of HUVECs in fibrin + FD 2% hydrogels compared to parent microvessel HUVECs after 3 days of culture. (e) p‐value and number of differentially regulated genes for Gene Ontology (GO) terms comparing HUVECs in Fibrin + FD 2% to parent microvessel HUVECs. (f) Top 10 genes due based on differentially expressed genes predicted to be upstream regulators of these identified genes. (g) Top 10 differentially regulated GO terms based on impacted pathways. (h)…

• 

  FIGURE 6. EC engagement to fibrous topography modulates TGF‐β signaling via VE‐cadherin–TGF‐βR2 dependent interactions. (a) Diagram depicting VE‐cadherin BioID construct. hMVECs stably expressing VE‐cadherin‐BioID were incubated in media containing biotin and then stained for VE‐cadherin and biotinylated proteins (Streptavidin‐AF568). (b) Schematic of VE‐cadherin‐BioID mass spectrometry screen to identify proteins interacting with VE‐cadherin. (c) Western blot micrograph of TGF‐βR2/VE‐cadherin co‐immunoprecipitation from ECs cultured on flat glass or DexVS‐RGD fibrous substrates. Blot representative of N=3 replicate experiments. (d) Representative images of hMVECs expressing TGF‐βR2‐mEmerald cultured on flat substrates or fibrous topographies immunostained for VE‐cadherin along with quantification TGF‐βR2 localization to VE‐cadherin AJs. (e) Western blot of scrambled control or VECKO HUVEC lysates treated with or without TGF‐β2 ligand (10 ng/mL). Blot representative of N=3 replicate experiments. (f–h) Representative images (f) and quantifications of total VE‐cadherin intensity (g) and nuclear SMAD3 (h) as a function of varied ratios of scrambled control HUVECs and VECKO HUVECs. (i–k)…

• 

  FIGURE 7. EC engagement to a matrix with heightened fiber density potentiates apoptosis in vitro and in vivo. (a) Schematic of study conditions and timeline for assessing TGF‐β2 secretion from ECs encapsulated in fibrin with (FD 2%) or without (FD 0%) heightened fiber density. (b) Quantification of TGF‐β2 secretion in culture media was performed via microfluidic ELISA. Data presented is from a single experiment with n=3 biological replicates. (c) Schematic of study conditions and timeline to assess the impact of heightened perivascular fiber density on TGF‐β2‐mediated apoptosis. (d) Representative images of parent vessel EC apoptosis in response to TGF‐β2 (10 ng/ml) via caspase3 immunostaining and corresponding quantification (e). (f) Representative images of large airways, alveolar space, arterioles, and venules from saline control and bleomycin treated mouse lungs at 3 weeks with TUNEL+ apoptotic ECs shown in white.

• 

  FIGURE 8. Summary schematic of key findings and outstanding questions (denoted with question mark).