uPAR deficiency triggers TGFβ1-mediated fibrotic remodeling in a cardiac perivascular-like microenvironment
Yulia Goltseva, Zoya Tsokolaeva, Irina Beloglazova, Victoria Stepanova, Maria Boldyreva, Elizaveta Ratner, Andrew Mazar, Alexander Andreev, Andrey Shiryaev, Yelena Parfyonova, Konstantin Dergilev
TL;DR
This study shows that a lack of uPAR in heart tissue worsens fibrosis by increasing TGFβ1 activity and disrupting cell interactions.
Contribution
The study identifies uPAR as a novel regulator of TGFβ1-mediated fibrotic remodeling in the cardiac perivascular microenvironment.
Findings
uPAR deficiency increases TGFβ1 activity and ECM deposition in cardiac perivascular-like microenvironments.
Loss of uPAR disrupts cell-cell and cell-matrix interactions, leading to endothelial cell loss and fibrotic transformation.
uPAR deficiency amplifies Akt signaling in fibroblasts, exacerbating fibrotic responses.
Abstract
Cardiac fibrosis represents a significant health burden, with endothelial dysfunction and damaged perivascular microenvironment increasingly recognized as key contributors to fibrotic remodeling. The urokinase plasminogen activator receptor (uPAR), a critical component of the urokinase system, plays a pivotal role in vascular remodeling and fibrosis. While prior evidence indicates that uPAR deficiency leads to microvascular dysfunction and perivascular fibrosis, the underlying mechanisms remain poorly defined. This study investigates how uPAR deficiency contributes to fibrotic remodeling of the cardiac perivascular-like microenvironment. Single-cell RNA sequencing data analysis and immunofluorescence staining on mouse heart cryosections were performed to characterize uPAR expression within the cardiac perivascular microenvironment. To model this microenvironment in vitro, cardiospheres…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9- —https://doi.org/10.13039/501100006769Russian Science Foundation
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsCardiac Fibrosis and Remodeling · Protease and Inhibitor Mechanisms · Congenital heart defects research
Background
Cardiac fibrosis, characterized by excessive deposition of extracellular matrix (ECM) proteins, is a growing global health burden linked to poor outcomes in patients with heart disease [1, 2]. One of the mechanisms leading to cardiac fibrosis is an endothelial dysfunction and subsequent damage of a perivascular microenvironment resulting from latent inflammation, metabolic and neurohumoral disorders, oxidative stress, and hypertension [3–5]. The cardiac perivascular microenvironment comprises a complex interaction network of endothelial cells (ECs), pericytes (PCs), smooth muscle cells (SMCs), fibroblasts (FBs), cardiac progenitor cells (CPCs) and components of ECM [5, 6]. Impaired microvasculature integrity disrupts tissue homeostasis and impedes effective reparative mechanisms, fostering initial perivascular remodeling and the progressive spread of fibrosis into surrounding healthy tissue [7, 8]. Although the contribution of EC dysfunction in the pathogenesis of heart disease is well studied [9], the molecular mechanism governing the cardiac perivascular microenvironment and fibrosis progression remains to be elucidated.
The urokinase plasminogen activator system is involved in the regulation of vascular function and fibrosis progression [10–12]. This system includes a serine protease urokinase (uPA), urokinase receptor (uPAR) and plasminogen activator inhibitors (PAI-1 and PAI-2) [13]. uPAR and uPA promote angiogenesis and viability of ECs [14, 15], modulate SMC phenotype [16] and regulate transformation of FBs into myofibroblasts [17, 18]. Studies, including ours, have demonstrated that uPAR deficiency leads to age-dependent microvascular regression, EC apoptosis, endothelial-to-mesenchymal transition (EndMT) and fibrosis in murine lungs, skin and heart [19–21]. Clinically, elevated plasma level of soluble urokinase receptor (suPAR) is associated with incident heart failure risk and adverse prognosis [22, 23]. Nevertheless, the contribution of uPAR in the regulation of the perivascular microenvironment and cardiac fibrosis remains largely uninvestigated.
uPAR/Plaur is a glycosylphosphatidylinositol-anchored membrane protein. Despite lacking transmembrane or intracellular domains, uPAR modulates signaling pathways (e.g., ERK, Akt, and FAK) through its association with diverse co-receptors, including integrins, growth factor receptors, and low-density lipoprotein receptors, thereby regulating multiple cellular functions [13]. While uPAR is predominantly expressed by immune [24] and senescent [25] cells, or within the tissues undergoing extensive remodeling [18], it is also expressed by ECs and c-Kit-positive CPCs in an intact heart [26, 27]. This dual localization suggests a potential role for uPAR in regulating both perivascular microenvironment and stem cell niches.
This study aimed to define potential mechanisms by which uPAR deficiency contributes to the development of cardiac perivascular fibrosis. We characterized uPAR expression within the cardiac perivascular microenvironment by combining analysis of single-cell RNA sequencing dataset of non-myocyte cardiac cells with immunofluorescence staining of mouse hearts. Since 3D cell culture approaches, and cardiospheres (CSs) in particular, may imitate in vitro a niche-like microenvironment [28, 29], our complementary experiments utilized CSs derived from wild-type and uPAR-knockout mice to assess contribution of uPAR deficiency to fibrotic remodeling within the cardiac perivascular-like microenvironment.
Methods
Analysis of single-cell RNA-sequencing data of the cardiac vascular niche
We analyzed publicly available single-cell RNA-sequencing (scRNA-seq) dataset of vascular cell niche from control mouse hearts (GEO accession number GSE166403 [6]). Preprocessing and data analysis were performed using R version 4.3.2. The complete analysis workflow is available in Supplementary File 1.
Animals
All animal experiments in this study were approved and conducted in accordance with the Ethics Committee of the Institutional Animal Care and Use Committee of the National Medical Research Centre of Cardiology named after Academician E.I. Chazov. The work has been reported in line with the ARRIVE guidelines 2.0. All experiments were designed and performed in accordance with the 3R principles (Replacement, Reduction, Refinement). The minimal number of animals required to obtain sufficient cell yield and ensure adequate statistical power for biological replicates was used. All experiments were performed using 4-weeks old male C57BL/6 wild-type (Wt, N = 10 mice) or 4–6-weeks old male C57BL6/SV129 Wt and Plaur-knockout (uPAR-/-) mice (N = 12 mice per group, total N = 24) [21, 27]. No animals were excluded from the study or analysis. Mice were randomly distributed across experimental groups by an independent researcher not involved in outcome assessment. Animals received standard care in the vivarium of the National Medical Research Centre of Cardiology Named after Academician E.I. Chazov. uPAR-/- mice lacking exons 2–5 of the Plaur gene, which encode the DI and DII uPA-binding domains of uPAR [30], were originally developed by Flanders Institute of Biotechnology (Belgium) [31]. For organ harvesting, mice were anaesthetized intraperitoneally with 100 µg/mL avertin (2.5% 2,2,2-Tribromoethanol (Sigma-Aldrich) and 2.5% 2-Methyl-2-butanol (Sigma-Aldrich) dissolved in sterile water) and euthanized by cervical dislocation. Cells isolated from three animals were pooled to constitute one biological replicate (N = 3–4 independent replicates per group).
Patient samples
The study was approved by the Ethics Committee of the Institutional Animal Care and Use Committee of the National Medical Research Centre of Cardiology named after Academician E.I. Chazov. Written informed consent was obtained from all human participants prior to sample collection. Right atrial appendage samples were obtained from four male patients during aortocoronary bypass surgery at the Department of Cardiovascular Surgery, National Medical Research Center of Cardiology named after Academician E.I. Chazov. The tissue specimens were immediately transferred in Dulbecco’s Phosphate-Buffered Saline (DPBS, Paneco) and processed within 30–60 min. Cells isolated from an individual right atrial appendage sample was considered the experimental unit within the study.
Cell cultures
Mouse cardiac explant-derived cells (mEDCs)
Non-myocyte cardiac cells were isolated by explant culture method from hearts of Wt and uPAR-/- C57Bl6/SV129 mice [29, 32]. Hearts were harvested in the Krebs-Ringer buffer (Sigma) supplemented with 11.25 U/mL heparin and minced to small fragments (1–2 mm), washed with Ca^2+^,Mg^2+^-free DPBS (Paneco) and incubated with a 0.1 mg/mL collagenase A solution (Roche) for 10 min at 37 °C in rotation. The reaction was stopped by addition of 10% of fetal bovine serum (FBS, HyClone, Cytivia). The isolated tissue fragments (explants) were placed onto fibronectin (Imtek) coated 150 mm Petri dishes (one heart per dish) and cultured at 37 °C in 5% CO_2_ in Explant medium based on IMDM (Servicebio) supplemented with 20% FBS (Gibco), 0.1 mM β-mercaptoethanol (Sigma), and 100 U/mL penicillin-streptomycin (Gibco). After 10–14 days, mouse explant-derived cells (mEDCs) were trypsinazed, filtered through 100 μm cell strainer (BD) and expanded on fibronectin-coated dishes in Explant medium. mEDCs were passaged no more than five times. The culture medium was changed every 72 h.
To analyze cell migration from the explants, images were taken using Image ExFluorer AI microscope (LСI) and cell migration area was quantified by subtracting the area of the explants using ImageJ software (NIH).
Human cardiac explant-derived cells (hEDCs)
Right atrial appendage samples were minced and briefly digested with 0.25% trypsin (Gibco) for 5 min at 37 ℃. After inactivation with 10% FBS, cardiac explants were plated on fibronectin-coated dishes and maintained under identical culture conditions to mEDCs.
Cardiosphere formation
CSs were generated from mEDCs or hEDCs on poly-D-lysine (PDL)-coated plates as previously described [32, 33]. In brief, EDCs were seeded at 60,000 cells per cm^2^ in PDL-coated 12-well plates (Corning) and cultured in 1 mL of Full CS medium for 72 h at 37 °C with 5% CO_2_. Full CS medium contained 65% DMEM/F-12 (Gibco), 35% IMDM (Gibco), 3% FBS (Gibco), and 100 U/mL penicillin-streptomycin (Gibco) supplemented with 20 ng/mL EGF (Paneco), 20 ng/mL bFGF (Prospec), 8 ng/mL cardiotrophin-1 (PeproTech), 1 U/mL thrombin (Sigma), 1× NeuroBrew-21 w/o Vit. A (Miltenyi Biotec).
Mouse cardiosphere-derived cells (mCDCs)
Mouse cardiosphere-derived cells (mCDCs) was generated as previously described by spontaneous disassembly of spheroids to monolayer culture [34]. In brief, floated mouse cardiospheres (mCSs) were collected, washed with DPBS by centrifugation at 50 g for 3 min and cultured on fibronectin-coated Petri dishes in Explant medium at 37 °C with 5% CO_2_ for 1–2 weeks with medium change every 72 h. Cells were passaged no more than five times.
3T3 fibroblasts
3T3 FBs (ATCC) were used as model stromal cells. FBs were grown in DMEM low glucose (Servicebio), supplemented with 1 g/L glucose, 10% FBS (Gibco) and 100 U/mL penicillin-streptomycin (Gibco) at 37 °C with 5% CO_2_ with medium change every 48 h.
Mouse lung endothelial cells (MLECs)
ECs were isolated from lungs of Wt C57Bl/6 mice according to modified protocol [35]. Lungs were minced, washed by Hanks’ Balanced Salts Solutions (HBSS, Paneco) and centrifuged at 200 g for 5 min at 10 °C. Floated tissue pieces were dissociated by 1 mg/mL collagenase I (Worthington) 3 times for 20 min at 37 °C. The 2nd and 3rd harvest was filtered through a 40 μm cell strainer (BD Falcon) to get single-cell suspension, centrifuged at 300 g for 10 min at 10 °C and resuspended in ice-cold PBE buffer (0.5% bovine serum albumin (BSA), 200 mM EDTA, Ca^2+^,Mg^2+^-free DPBS). Mouse lung ECs (MLECs) were separated from total cell suspension by immunomagnetic selection with anti-CD31 beads (Miltenyi Biotec) according to manufacturer’s recommendations. MLECs were seeded at Petri dishes coated with 0.2% gelatin (Paneco) and cultivated in EGM-2MV Microvascular Endothelial Cell Growth Medium-2 (Lonza) at 5% CO_2_, 37 °C with medium change every 48 h. MLECs were passaged no more than two times. Prior to experiments, MLECs were reselected using anti-CD31 beads.
All brightfield images of cell cultures were taken using Image Exfluorer AI microscope (LCI).
Generation of Plaur knockout FBs via lentiviral transduction
Lentiviral particles generation, transfection of HEK293T cells and transduction were conducted as described previously [36]. In brief, calcium-phosphate transfection was performed with 4 µg/mL pMD2G (Addgene) and 8 µg/mL pCMV-R8.91 (Addgene) packaging plasmids and 10 µg/mL Plaur CRISPR/Cas9 knockout vectors with three sgRNAs (ABM Inc.) or 10 µg/mL Scrambled K010 plasmid (ABM Inc.). 3T3 FBs were transducted with lentiviral supernatants containing 8 µg/mL polybrene for 24 h. After 72 h cells were selected with 4 µg/mL puromycin (Sigma Aldrich) during 7 days with medium change every 24 h. To generate stable Plaur knockout (KO) cell line, we performed clonal selection.
Immunofluorescence (IF) staining
To prepare frozen sections, mouse hearts or mCSs were embedded in Tissue-Tek^®^ O.C.T. Compound (Sakura) and frozen in liquid nitrogen vapor. Cross-section slices with a 7 μm thickness were prepared. Otherwise, cells (MLECs or 3T3 FBs) were seeded on gelatin-coated coverslips. Thawed slides of frozen sections or cells on coverslips were washed with DPBS and fixed in 4% formaldehyde solution (PanReac AppliChem) for 10 min. To detect intracellular proteins, samples were permeabilized with 0.1% Triton X-100 (Bio-Rad) in PBS. The heart sections were blocked with 10% Normal Goat Serum (Vector Laboratories), while mCS sections and MLECs were blocked with 1% BSA, 0.3 M glycine for 1 h at room temperature (RT), then probed with the primary antibodies (see Supplementary Table 1) overnight at 4 °C followed by incubation with the secondary antibodies conjugated with AlexaFluor 488 (Invitrogen) or AlexaFluor 594 (Invitrogen) for 1 h at RT. Nuclei were stained with 0.1 µg/mL DAPI (Lumiprobe). Slides were mounted with Aqua-Poly/Mount (Polyscience). IF staining was examined with the Image Exfluorer AI microscope (LCI). Confocal images were acquired using Leica Stellaris 5 confocal microscope (Leica).
The number of CD31-, Sca-1-, and c-Kit-positive cells and nuclear localization of uPA within mCSs were counted routinely in ImageJ software (NIH) or automatically using NIS-Elements General Analysis 3 pipeline (Nikon) and were normalized to total nuclei number per mCS section. Nuclear uPA localization was confirmed when ≥ 60% of uPA fluorescence colocalized with DAPI-stained nuclei.
RNA isolation, reverse transcription, and quantitative real-time PCR
mEDCs and mCSs were cultured in the same Full CS medium for 3 days, then were washed with DPBS, lysed with RLT lysis buffer (Qiagen), and homogenized with a 29G insulin syringe (BD). RNA was isolated with RNeasy Mini Kit (Qiagen), treated with DNase I (Qiagen), and quantified using NanoDrop 2000 spectrometer (Thermo Scientific). 1 µg of total RNA was used in reverse transcription step using MMLV RT kit with random hexamer primers (Eurogene). cDNA was amplified for 40 cycles including 95 ℃ for 15 s, 60 ℃ for 1 min using qPCRmix-HS SYBR+HighROX kit (Eurogen) and StepOnePlus™ Real-Time PCR System (AppliedBiosystems). DNase/RNase-free water was added to the reaction mixture instead of the sample as a negative control. Primer sequences are listed in Supplementary Table 2. Gene expression was normalized per housekeeping gene Gapdh (glyceraldehyde-3-phosphate dehydrogenase) or Actb (β-actin). Fold changes of gene expression are calculated as 2^−ΔΔCT^. The same protocol was applied to the 3T3 cells.
Western blot
mEDCs and mCSs were cultured in the same Full CS medium for 3 days, then were washed with DPBS, lysed in RIPA buffer supplemented with protease and phosphatase inhibitors (Abcam), incubated on ice for 30 min, homogenized with a 29G insulin syringe and centrifuged at 13,200 g for 10 min at 4 °C. Samples were balanced based on protein concentration measured by the ProteOrange kit (Lumiprobe). Proteins were heated in Laemmli Sample Buffer supplemented with 4% (v/v) β-mercaptoethanol (Sigma-Aldrich) for 5 min at 95 °C and resolved on SDS-PAGE gels (10–15%) using Tris-Glycine SDS running buffer, then transferred to polyvinylidene difluoride (PVDF) membranes (Roche) overnight at 100 mA 4 °C. Membranes were washed, then blocked in 5% BSA in tris-buffered saline with 0.05% Tween-20 (TBS-T) for 1 h RT and incubated with primary antibodies (see Supplementary Table 1) diluted in 1% BSA/TBS-T overnight at 4 °C. After washing steps with PBS-T, the membranes were incubated with HRP-conjugated anti-rabbit antibodies (AS014, Abclonal) or anti-mouse antibodies from SuperSignal™ West Pico Complete Mouse IgG Detection kit (Thermo Scientific™) for 1 h RT, followed by washing steps. SuperSignal™ West Pico PLUS Chemiluminescent Substrate (Thermo Scientific™) was used to visualize protein bands with FUSION-SL system (Vilber Lourmat). Densitometric analysis was performed using ImageJ software (NIH) and protein levels were normalized to β-actin or GAPDH. The same protocol was applied to the 3T3 cells.
Flow cytometry
EDCs or 3T3 FBs were harvested using 0.025% trypsin (Gibco), inactivated with 10% FBS and washed with ice-cold fluorescence-activated cell sorting (FACS) buffer (5% FBS, 0.1% sodium azide in Ca^2+^,Mg^2+^-free DPBS). Cells (250,000 per sample) were incubated with primary antibodies listed in Supplementary Table 1 for 30 min at 4 °C. Isotype IgG or unstained samples were used as negative staining controls. If unconjugated primary antibodies were used, the cells were incubated with the secondary antibodies labeled with AlexaFluor 488 (A11006; Invitrogen) for 30 min at 4 °C. To exclude dead cells from analysis, mEDCs were stained with 0.2 µg DAPI (Servicebio) for 15 min. The cells were analyzed using FACSAria III cell cytometer (BD) with FACSDiva software (BD). The histograms were generated using floreada.io WASM Version: SIMD (accessed March 20, 2025).
Cell size analysis
Multi-well-plates (Corning) were coated for 2 h at 37 °C with 40 µg/mL fibronectin (Imtek), 10 µg/mL collagen I (Imtek), 2.5 µg/mL vitronectin (Imtec), or were left uncoated. mEDCs were seeded at 2,000 cells/cm^2^ in Explant medium. After 24 h, cells were washed with DPBS, stained with CellTracker Green CMFDA (Invitrogen) and 8 µM Hoechst 33,342 (Invitrogen), then fixed by 4% formaldehyde for 15 min RT. Cell size were analyzed by measuring cell area using NIS-Elements General Analysis 3 pipeline (Nikon).
Cell proliferation/viability assay
PrestoBlue Cell Viability Reagent (Invitrogen) was used to analyze cell proliferation/viability as described previously [37]. mEDCs or MLECs were deprived under serum-free conditions the day before and seeded into 96-well culture plates (Corning) at 5,000 cells per well. mEDCs were cultured in (1) serum-free DMEM (Gibco) supplemented with 1% BSA and 100 U/mL penicillin-streptomycin (Gibco), (2) DMEM supplemented 10% FBS (HyClone, Cytivia) and 100 U/mL penicillin-streptomycin, or (3) full CS medium. MLECs were cultured in EGM-2MV (Lonza) or combined medium 40% base EBM-2 (Lonza) supplemented with 2% FBS (HyClone, Cytivia) and 60% mCDC conditioned medium. Cells were cultured for 24 h, 48 h, and 72 h with medium change every 48 h (for mEDCs) or 24 h (for MLECs). At the specified time the cell medium was changed to serum-free medium supplemented with PrestoBlue reagent (1:10) for 1 h at 37 °C in the dark. The fluorescence signal proportional to the number of viable cells was detected on a Victor X3 plate reader (Perkin Elmer) at λ_ex_ = 520 nm and λ_em_ = 590 nm.
Adhesion assay
mEDCs were deprived under serum-free conditions overnight. 96-well-plates were coated with 40 µg/mL fibronectin, 10 µg/mL collagen I, 2.5 µg/mL vitronectin, or were left uncoated overnight at 4 °C, then washed with DPBS and incubated with 1% BSA/DMEM for 1 h at 37 °C. mEDCs were seeded at 10,000 cells per well in DMEM (Gibco) supplemented with 1% BSA and 100 U/mL penicillin-streptomycin (Gibco). Cells were led to adhere for 1 h, 2 h, and 4 h, non-adhered cells were washed with DPBS. Adhered cells were fixed with 4% formaldehyde for 5 min, then were stained with Crystal Violet dye for 20–30 min RT, washed with distilled water and dried at RT overnight. The dye was dissolved in 10% acetic acid for 10 min and the absorbance was detected at 590 nm and 650 nm (for background subtraction) using a Victor X3 plate reader (Perkin Elmer).
Wound healing assay
Wound healing assay was used to analyze migration ability of mEDCs. 24-well-plates were coated with 40 µg/mL fibronectin, 10 µg/mL collagen I, 2.5 µg/mL vitronectin, or were left uncoated overnight at 4 °C, then washed with DPBS. mEDCs were seeded at 800,000 cells per well in Explant medium for 24 h. Then cells were serum-starved for another 24 h. A sterile pipette tip was used to create a linear “wound” on a cell layer across the well diameter. Wounded cells were washed with DPBS and cultured in DMEM medium containing 1% BSA. Visualization of cell migration and wound closure was performed after 0 h, 3 h, and 6 h using Image Exfluorer AI microscope (LCI). Wound area was measured using a Wound_healing_size_tool ImageJ plugin [38].
Inhibitor assay
Inhibitor assay was performed to investigate the role of uPAR interaction with uPA and integrins during CS formation. mEDCs were seeded on PDL-coated plates to form mCSs and were treated with (1) 10 µM cilengitide (Selleck Chemical), a peptide antagonist of αvβ3, αvβ5 and α5β1 integrins; (2) 100 nM SB273005 (Selleck Chemical), a non-peptide antagonist of αvβ3 and αvβ5 integrins. hEDCs were seeded on PDL-coated plates to form hCSs and treated with (1) 100 µg/mL ATN-292 antibody [39, 40] blocking the uPAR binding site on uPA; (2) 100 µM amiloride (Calbiochem), which is an inhibitor of uPA proteolytic activity; (3) 75 µg/mL ATN-658 blocking antibody (Selleck Chemical), preventing uPAR interaction with αMβ2 and α5β1 integrins. Solvents (water or DMSO) were used as controls. Cells were cultured for 72 h without medium changing. Brightfield images were taken using Image ExFluorer AI (LCI). CS size (diameter) and number were analyzed using NIS-Elements Segment.ai and General Analysis 3 modules (Nikon).
uPA-induced cell signaling
mEDCs were plated at 10,000 cells/cm² in Explant medium and cultured for 48 h, followed by 2 h serum starvation. Then cells were treated with 10 nM mouse uPA [41] and incubated for 1, 5, 15, 30, or 60 min. For positive control, 20% FBS was added for 30 min. After treatment, cells were lysed in RIPA buffer supplemented with protease and phosphatase inhibitors. Activation of ERK1/2 and Akt pathways was assessed by immunoblotting using phospho/total protein ratios.
Live/dead assay
The presence of viable and dead cells inside the mCSs was assessed using LIVE/DEAD™ Viability/Cytotoxicity Kit, for mammalian cells (Invitrogen). According to the protocol [42], mCSs were treated with 2.5 µg/mL Hoechst 33,342 (Invitrogen), 2 µM Calcein-AM, and 4 µM ethidium homodimer-1 for 3 h at 37 ℃, 5% CO_2_ in the dark. mCSs were washed with DPBS and transferred on a plate with glass bottom in RPMI 1640 Medium without phenol red (Gibco) supplemented with 3% FBS (Gibco). Stained cells were detected using Image Exfluorer AI microscope (LCI) with life-maintaining module (37 °C, 5% CO_2_). Cell viability and cell death were quantified by measuring mean intensity of fluorescence signal per spheroid area using ImageJ software (NIH).
Measurement of spheroid stiffness
mCS stiffness was measured using micro-scale parallel-plate compression testing system (CellScale, Canada) as described in [43]. A single mCS was plated into PBS-filled measurement chamber and then was compressed between a static rigid substrate and a dynamical cantilever beam until 50% vertical compression of the spheroid was achieved. A compressive force and upper plate displacement were monitored via the CellScale software.
TGFβ1 treatment of 3T3 cells
FBs were serum-starved overnight. To analyze transforming growth factor beta 1 (TGFβ1)-induced signaling activation, cells were treated with 10 ng/mL mouse TGFβ1 (PKSM041167, Elabscience) in DMEM low glucose (Servicebio) for 30 min. Activation of SMAD2/SMAD3 and Akt pathways was assessed by immunoblotting using phospho/total protein ratios. To assess fibrotic transformation and ECM deposition, FBs were treated with 10 ng/mL mouse TGFβ1 in DMEM low glucose supplemented with 2% FBS (Gibco) for 48 h. An equivalent volume of TGFβ1 diluent was added as a negative control.
Secretome profiling of mCDCs
mCDCs were seeded on fibronectin-coated dishes at 10,000 cells/cm^2^ in Explant medium for 4 days. Then cells were washed with DPBS and cultivated in serum-free IMDM (Servicebio) medium supplemented with 100 U/mL penicillin-streptomycin (Gibco) for 48 h. Condition mediums (CMs) were centrifuged at 2,000 g 4 °C for 20 min. Supernatants were transferred to new tubes and stored at −70 °C. After harvesting CMs, cells were counted to balance CM concentration. mCDC secretome was analyzed using ProteomeProfiler antibody arrays (ARY015, R&D) following the manufacturer’s guidelines. Dot-blot images were taken using FUSION-SL system (Vilber Lourmat). Secreted protein levels were analyzed using ImageJ Protein Array Analyzer plugin (NIH).
Enzyme-linked Immuno sorbent assay (ELISA)
VEGF-A and MCP-1 concentrations in mCDC CMs were analyzed using ELISA for mouse VEGFA kit (SEA143Mu, Cloud-Clone Corp.) and Mini Samples ELISA Kit for MCP1 (MEA087Mu, Cloud-Clone Corp.). Optical density was measured on a Victor X3 plate reader (Perkin Elmer) at 450 nm.
Tube assay on Matrigel™
MLECs were seeded at 30,000 cells per well of 96-well plates (Corning) pre-coated with 50 µL of growth factor reduced Matrigel™ (Corning) and cultured for 6 h in base medium EBM-2 (Lonza) supplemented with 2% FBS (Gibco). 60% v/v Wt and uPAR-/- mCDC CMs were added to cells in base medium. Brightfield images were taken using Image ExFluorer AI (LCI). Tube length and number of tubes, meshes and nodes were quantified routinely in ImageJ software (NIH).
In vitro endothelial-to-mesenchymal transition (EndMT) assay
MLECs were seeded at 20,000 cells/cm^2^ at EBM-2 (Lonza) supplemented with 2% FBS. After adhesion, cells were treated with 60% v/v Wt and uPAR-/- mCDC CMs for 72 h. Then, cells were fixed with 4% formaldehyde, and IF staining was performed as described above. EndMT was assessed by quantifying of F-actin+/SMA+ myofibroblast-like cells in MLEC culture. F-actin was stained with using Alexa Fluor 488-conjugated phalloidin (A12379, Invitrogen).
Statistical analysis
All experiments included 3–4 biological replicates per condition. Statistical analysis and data visualization were performed using R (version 4.3.3) packages Tidyverse (v 2.0.0), rstatix (v 0.7.2), and ggpubr (v 0.6.0). Results are expressed as mean ± standard deviation. Data normality and homogeneity of variance were assessed using Shapiro-Wilk and Levene’s tests, respectively. For comparison between two groups, unpaired Student’s t-test or multiple t-tests with Bonferroni or Benjamini & Hochberg (for multiple gene expression comparisons) corrections were applied. For pairwise comparisons (2D vs. 3D cultures, control vs. treatment), paired t-test was used. For multi-group comparisons, one-way ANOVA with Tukey’s post-hoc correction was applied. For multifactorial analysis, two-way ANOVA followed by Tukey’s post hoc test was employed. P-value < 0.05 (or adjusted p-value < 0.05 for multiple testing) was considered statistically significant.
Results
uPAR expression in the cardiac perivascular microenvironment
To investigate Plaur expression in the cardiac perivascular cell microenvironment, we analyzed publicly available scRNA-seq dataset of vascular and stromal cells isolated from mouse hearts (GSE166403 [6]). Nine distinct cell clusters were identified (Fig. 1A) based on marker gene expression (Supplementary Figure S1A): Mfap4-fibroblasts (FB_1), Cxcl1-fibroblasts (FB_2), interferon fibroblasts (IFN FB), Schwann cells (Schwann), pericytes (PC), smooth muscle cells (SMC), endothelial cells (EC), proliferating endothelial cells (Prolif. EC) and lymphatic endothelial cells (Lymph. EC).
Fig. 1uPAR expression in the cardiac perivascular microenvironment. A UMAP visualization of stromal and vascular cells of mouse hearts. B UMAP visualization of Plaur expression. C Proportion of Plaur-expressing cells within each cell cluster (normalized to the specific cell type). D and E Immunofluorescence staining of mouse heart cryosections: uPAR (red), von Willebrand factor (vWF, green), α smooth muscle actin (SMA, green), nuclei were stained with DAPI. Arrows indicate uPAR-positive staining. The boxed areas represent the regions depicted at higher magnification
In mouse hearts, Plaur expression was low and restricted to a subset of FBs (3–4% of Plaur-expressing FBs per total FBs) and ECs (8% of Plaur-expressing ECs per total ECs) (Fig. 1B and C). Although there were a high proportion of Plaur-expressing cells within proliferating and lymphatic ECs, this result should be interpreted with caution due to the low number of cells analyzed in these clusters. In vivo few uPAR-expressing vascular and perivascular cells were observed within mouse heart cryosections (Fig. 1D and E). uPA/Plau expression was also comparatively low across all clusters (Supplementary Figure S1B), while PAI-1/Serpine1, an inhibitor of both uPA and tissue-type plasminogen activator (tPA) [44], was highly expressed in FBs (Supplementary Figure S1C).
Taken together, our findings demonstrate that uPAR is expressed by cardiac ECs and FBs – two cell populations that could promote perivascular fibrotic remodeling [3].
Modelling the cardiac perivascular-like microenvironment in vitro with cardiospheres
To model the cardiac perivascular-like microenvironment in vitro, we utilized a 3D cell culture approach to generate CSs from the mouse cardiac explant-derived cells (mEDCs) (Fig. 2A). mEDCs were grown on PDL-coated plates in growth factor-enriched medium, promoting mCS formation within 72 h (Fig. 2B). Mouse CSs (mCSs) recapitulated key stromal and vascular components, containing mesenchymal cells (CD105+, CD90+), FBs/SMCs (SMA+), ECs (CD31+), CPCs (c-Kit + and Sca-1+) and ECM-producing cells (fibronectin and collagen I) (Fig. 2C).
Fig. 2. Cardiospheres as an in vitro model of the cardiac perivascular-like cell microenvironment. A Key steps of mouse cardiosphere (mCS) formation. Created with Servier Medical Art, licensed under CC-BY 4.0. B Phase-contrast images of explant culture (2 and 10 days after initiation), mouse explant-derived cells (mEDCs), and mCSs. C Immunofluorescence images of mCS cryosections stained for mesenchymal markers CD105 (red) and CD90 (green), matrix proteins fibronectin (FN, green) and collagen I (COL I, green), α smooth muscle actin (SMA, green), endothelial marker CD31 (red), cardiac progenitor cell markers c-Kit (red) and Sca-1 (red), and urokinase system proteins urokinase (uPA, green) and urokinase receptor (uPAR, red). Nuclei were stained with DAPI. **D **qRT-PCR analysis quantified relative gene expression (fold change), normalized to Gapdh, in 3D mCSs versus 2D mEDCs. *adjusted p-value < 0.05, **adjusted p-value < 0.01, N = 3. E Protein levels of uPAR, single-chain uPA (sc-uPA) and two-chain uPA (tc-uPA) in 3D mCSs and 2D culture of mEDCs, normalized to β-actin. N = 3. F Representative images of western blot results for panel “D”. Full-length blots/gels are presented in Supplementary Figure S5
qRT-PCR analysis indicated that CS formation reproduced key aspects of cell microenvironment remodeling, including upregulation of ECM and endothelial-to-mesenchymal transition (EndMT)-related genes, altered vascular cell markers, and modulation of the urokinase system (Fig. 2D). Specifically, compared to 2D-cultured mEDCs, mCSs exhibited increased Plaur, unchanged Plau, and decreased Serpine1 expression (Fig. 2D). mCSs were positively stained for uPAR and uPA that were localized both in the central zone and in the periphery of spheroids (Fig. 2C). Although Plau gene expression was not altered, the proteolytically active form of uPA protein (two-chain uPA/tc-uPA) was increased (Fig. 2E and F), implying contribution of uPA-mediated proteolysis into mCS formation. According to the previously published data [33], we suggest that decrease of PAI-1 expression further facilitates uPA-mediated proteolysis during mCS formation. These results indicate that spheroid formation drives comprehensive remodeling of the cellular microenvironment, including upregulation of Plaur gene expression and tc-uPA.
Thus, mCSs can be considered as a simplified model of the cardiac perivascular-like microenvironment, composed of stromal and vascular cells, as well as ECM components.
uPAR interaction with integrins mediates cardiosphere assembly
To investigate how uPAR regulates the cardiac perivascular-like microenvironment, we generated mCSs from the cells isolated from young Wt and uPAR-/- mice hearts that have not yet developed the fibrosis [20]. While 60% of Wt mEDCs expressed surface uPAR, uPAR-/- cells showed complete absence (Fig. 3A and B). Despite comparable immunophenotypes (including CD31 + ECs and CD105 + or CD90 + mesenchymal cells), cell death rate, morphology, and proliferation, adhesion and migration capacity on fibronectin and vitronectin (Supplementary Figures S2 and S3), uPAR-/- cells exhibited enhanced adhesion to collagen I without increased motility (Supplementary Figure S3).
Despite similar characteristics, we found that uPAR deficiency altered CS assembly. uPAR-/- cells yielded a greater number of spheroids with reduced size compared to Wt cells (Fig. 3C and D). Since we have shown that tc-uPA levels and Plaur gene expression were elevated in mCSs compared to 2D culture (Fig. 2D-F), we have asked if uPA/uPAR and integrins (as uPAR coreceptors) concert in regulating CS formation. Integrins were shown to be crucial for spheroid assembly, especially on adhesive substrates, such as the poly-D-lysine [45–47].
Fig. 3. Regulation of mouse (m) and human (h) cardiosphere formation by uPA/uPAR-integrins interplay. A Flow cytometry analysis of Wt and uPAR-/- mouse explant-derived cells (mEDCs) stained with anti-uPAR antibodies (red), isotype IgG (blue), or unstained (gray). B Proportion of uPAR-positive Wt and uPAR-/- mEDCs measured by flow cytometry, N = 3. C Brightfield images of Wt and uPAR-/- mouse cardiospheres (mCSs) at 72 h after seeding. D Number and diameter of Wt and uPAR-/- mCSs. *p < 0.05, **p < 0.01, N = 3. E Schematic representation of integrin inhibitors cilengitide and SB273005 action; bold text indicates integrins that were shown to associate with uPAR. F Brightfield images of Wt mCSs treated with cilengitide and SB273005 or control treatment (H_2_O or DMSO). G Number and diameter of Wt mCSs treated with cilengitide or SB273005 and control. *p < 0.05, N = 3. H Immunophenotype of hEDCs measured by flow cytometry, N = 3. I Schematic representation of action of blocking antibodies ATN-658 and ATN-292, and non-specific uPA inhibitor amiloride. J Brightfield images of hCSs treated with ATN-658, ATN-292 and amiloride or control treatment. K Number and diameter of hCSs treated with ATN-658 or ATN-292 and control. **p < 0.01, N = 4. “E” and “I” were created with Servier Medical Art, licensed under CC-BY 4.0
When integrins (αvβ3, αvβ5 and α5β1) were blocked by cilengitide and SB273005 (Fig. 3E), cells formed CSs of smaller-size compared to untreated cells (Fig. 3F and G), suggesting uPAR cooperates with integrins to regulate 3D assembly. uPA and uPAR are known to promote integrin-dependent signaling pathways, such as ERK and PI3K/Akt, which regulate spheroid formation and properties [48, 49]. We found that uPA activated ERK1/2 phosphorylation in Wt but not in uPAR-/- cells, whereas Akt-S473 phosphorylation was not significantly changed between both cell origins (Supplementary Figure S4). This result suggested that spheroid formation may depend on the ERK signaling pathway activated by the association of uPA-occupied uPAR with co-receptors.
In order to delineate the significance of uPAR association with integrins in CS formation, we used ATN-658 to inhibit uPAR DIII domain, thus preventing its association with αMβ2 and α5β1 integrins but not with uPA [50] (Fig. 3I). Since mouse and human uPAR shares only 60% of sequence identity [51] and most antibodies against human uPAR does not bind mouse uPAR effectively, we isolated hEDCs from human heart samples. These cells had similar immunophenotype as mEDCs including high percent of uPAR-expressing cells (Fig. 3H) and formed CSs under the same culture conditions. Blocking uPAR-integrin association via ATN-658 abolished hCS formation, yielding only adhesive cell clusters and a limited number of spheroids (Fig. 3J and K). This result indicates that uPAR-integrin association is essential for the self-assembly of hCSs on PDL.
Since uPAR interaction with uPA enhanced the association of uPAR with its co-receptors we used anti-uPA antibody ATN-292 to block uPA binding to uPAR [39] (Fig. 3I) and found that it decreased hCS yield without the complete inhibition of spheroid assembly (Fig. 3J and K). Protease-dependent invasion is crucial for spheroid assembly [47]. Of interest, a specific uPA (among other proteases) inhibitor amiloride [52] totally suppressed spheroid and cell cluster formation (Fig. 3J).
In summary, urokinase system orchestrates CS formation by uPA/uPAR-integrins interplay and uPA-mediated proteolysis. These mechanisms likely extend to fibrotic remodeling, where aberrant cell-ECM interactions may exacerbate fibrosis [53, 54].
uPAR deficiency induces fibrotic remodeling and loss of ECs within mCSs
uPAR-/- cells were able to form mCSs (as demonstrated above) that recapitulate the cardiac perivascular-like microenvironment, although with some differences compared to Wt mCSs (Fig. 4). Despite the smaller size, uPAR-/- mCSs exhibited lower cell viability compared to Wt mCSs (Fig. 4A and B), and higher stiffness (Fig. 4C), which was attributed to increased deposition of collagens I and IV, and fibronectin (Fig. 4D-F), suggesting fibrotic remodeling. We also observed that uPAR-/- mCSs possessed higher protein levels of vimentin, SMA and SM22α (Fig. 4D-F) that suggests fibroblast-to-myofibroblast differentiation and/or smooth muscle-like cell enrichment. uPAR-/- mCSs demonstrated a significant reduction of ECs (Fig. 4D and G), while the number of vascular CPCs, showing positive staining for c-Kit and Sca-1, remained unchanged compared to Wt (Fig. 4D and G). These results are in accordance with previous in vivo studies, including our own, that have shown development of perivascular and interstitial cardiac fibrosis accompanied by endothelial dysfunction and apoptosis in aged uPAR-/- mice [20, 21].
Collectively, uPAR deficiency drives fibrotic remodeling, leading to ECM deposition, enrichment of myofibroblasts/smooth-muscle-like cells, and EC depletion, possibly reflecting fibrosis development and EndMT observed in the uPAR-/- hearts [21].
Fig. 4uPAR deficiency in mCSs leads to reduced viability, fibrotic remodeling and EC loss. A Live/dead staining of Wt and uPAR-/- mCSs. Live cells were stained with calcein-AM (green), dead cells were stained with ethidium homodimer-1 (red). B Viability of Wt and uPAR-/- mCSs measured as ratio of viable per dead cell signal, *p < 0.05, N = 3. C Stiffness (Young modulus) of Wt and uPAR-/- mCSs, **p < 0.01, N = 4. D Immunofluorescence images of Wt and uPAR-/- mCS cryosections stained for collagen I (COL I, red), fibronectin (FN, red), α-smooth muscle actin (SMA, green), CD31 (red), c-Kit (red), Sca-1 (red). Nuclei were stained with DAPI. E Protein levels of COL I, collagen IV (COL IV), FN, vimentin (VIM), αSMA and smooth muscle protein 22α/transgelin (SM22α) measured in Wt and uPAR-/- mCSs, * - p < 0.05, **p < 0.01, N = 4. F Representative images of western blot results. Full-length blots/gels are presented in Supplementary Figure S6. G Relative number of CD31-, c-Kit- and Sca-1-positive cells within Wt and uPAR-/- mCSs, **p < 0.01, N = 3
uPAR deficiency promotes TGFβ1 activation, uPA accumulation and SNAIL upregulation in mCSs
To investigate the underlying mechanisms of fibrotic remodeling induced by uPAR deficiency, we focused on TGFβ-dependent mechanisms, given the elevated TGFβ1 levels found in the cardiac tissue of uPAR-/- mice [21]. TGFβ1 is a major inducer of both fibrosis [55–57] and EndMT [58, 59]. We observed elevated production of active form of TGFβ1 in uPAR-/- mCSs compared to Wt, while no difference in Tgfb1 gene expression was observed (Fig. 5A and H). These results suggest that uPAR deficiency is associated with increased TGFβ1 activation.
Fig. 5uPAR deficiency promotes TGFβ1 activation, uPA nuclear accumulation and SNAIL upregulation. A Protein and gene expression levels of transforming growth factor β1 (TGFβ1) in Wt and uPAR-/- mCSs, *p < 0.05, N = 3–4. B Protein levels of single-chain urokinase (sc-uPA) and two-chain urokinase (tc-uPA) in Wt and uPAR-/- mCSs, **p < 0.01, N = 4. C uPA nuclear accumulation (percent of uPA-positive nuclei) within Wt and uPAR-/- mCSs, *p < 0.05, N = 3. D Confocal images of intracellular uPA staining (green) of Wt and uPAR-/- mCS cryosections. Nuclei were stained with DAPI. White segmented arrows indicate the direction of fluorescence intensity profiling presented on panel “E”. Straight arrows indicate colocalization of uPA and DAPI. E Intensity profile plots of the uPA and nuclei/DAPI fluorescence signal. Arrows indicate overlay of fluorescence signals. F Nuclear uPA expression (mean fluorescence intensity per nuclei) within Wt and uPAR-/- mCSs, *p < 0.05, N = 3. G Protein levels of SNAIL and TWIST1 in Wt and uPAR-/- mCSs, ****p < 0.0001, N = 4. H Representative images of western blot results for panels “A”, “B” and “G”. Full-length blots/gels are presented in Supplementary Figure S7. I Confocal images of double-staining for CD31 (red) and α smooth-muscle actin (SMA, green) in Wt and uPAR-/- mouse cardiospheres (mCSs). Nuclei were stained with DAPI. White arrows indicate colocalization of CD31 and SMA. J Intensity profile plots of the CD31 and SMA fluorescence signal. Arrows indicate overlay of fluorescence signals
TGFβ1 activation may occur through uPA-dependent mechanisms including proteolytic cleavage from the latent complex by plasmin, MMP-2 and MMP-9 [60–63]. Indeed, we observed a significant increase of single-chain uPA (sc-uPA) in uPAR-/- mCSs vs. Wt (Fig. 5B and H). Moreover, sc-uPA was shown to translocate into cell nuclei and upregulate SMA expression [64], indicating its possible role in myofibroblast transformation and EndMT. We revealed accumulation of nuclear uPA within uPAR-/- mCS cryosections compared to Wt (Fig. 5C, D, E and F).
Given that uPAR deficiency is associated with EndMT of dermal and cardiac vessels in vivo [21, 65], we investigated whether mCSs recapitulate this pathological process. Our findings revealed that uPAR deficiency led to a significant increase in SNAIL protein level and a trend towards higher TWIST1 in mCSs (Fig. 5G and H), consistent with EndMT progression. The pattern of EndMT activations is demonstrated in Fig. 5I and J, showing colocalization of CD31 and SMA within uPAR-/- mCSs.
Collectively, these results indicate that uPAR deficiency promotes TGFβ1 activation, uPA accumulation and upregulation of EndMT-related transcription factor SNAIL.
CSISPR/Cas9-based Plaur knockout in fibroblasts results in ECM deposition and TGFβ1 activation
Perivascular stromal cells, such as FBs, are primary targets of TGFβ1, which promotes ECM deposition and FB-to-myofibroblast differentiation [3, 56]. We next sought to determine whether the fibrotic remodeling induced by uPAR deficiency is mediated by FB activity. To address this, we generated Plaur knockout (Plaur KO) FBs using the CRISPR/Cas9 system (Fig. 6A). Knockout efficiency was confirmed by a significant reduction in Plaur mRNA levels (quantified by qRT-PCR, Fig. 6C) and a loss of surface uPAR protein expression (assessed by flow cytometry, Fig. 6B and D).
Fig. 6Plaur knockout in fibroblasts resulted in enhanced ECM deposition and TGFβ1 activation. A Schematic representation of Plaur knockout in fibroblasts using CRISPR/Cas9 system with three single guide RNAs (sgRNA). The illustration was created with Servier Medical Art, licensed under CC-BY 4.0. B Flow cytometry analysis of Wt, Scrambled (control plasmid) and Plaur knockout (Plaur KO) fibroblasts stained with anti-uPAR antibodies (red), isotype IgG (green), or unstained (blue). C Plaur gene expression in fibroblasts quantified by qPCR, **p < 0.01, ***p < 0.001, ****p < 0.0001, N = 3. D Analysis of surface uPAR expression on fibroblasts by flow cytometry. Data are presented as median fluorescence intensity (MFI), ***p < 0.001, ****p < 0.0001, N = 3. E Representative images of western blot results for panels “F” and “G”. Full-length blots/gels are presented in Supplementary Figure S8. F Protein levels of COL I (pro-collagen I and mature collagen I α-chain), FN, and αSMA measured in fibroblasts, *p < 0.05, **p < 0.01, N = 3. G Protein levels of latent and active (dimer and monomer) forms of TGFβ1 in fibroblasts, *p < 0.05, **p < 0.01, N = 3. H Tgfb1 gene expression in fibroblasts, *p < 0.05, **p < 0.01, N = 3
Indeed, Plaur KO FBs produced higher levels of ECM proteins, such as collagen I and fibronectin, compared to control Wt and Scrambled FBs (Fig. 6E and F), consistent with previously published data [18]. Although we observed upregulation of SMA protein level in Plaur KO FBs compared to Wt cells, there was no significant difference between Plaur KO and Scrambled FBs (Fig. 6E and F).
Similarly to uPAR-/- mCSs, Plaur KO in FBs resulted in increased levels of active TGFβ1 (both dimer and monomer forms, Fig. 6G and E), while Tgfb1 mRNA expression levels remained unchanged compared to Wt FBs (Fig. 6H). Notably, Scrambled FBs (with high uPAR expression, Fig. 6B-D) had lower TGFβ1 levels than Wt cells (Fig. 6E, G and H), suggesting that both uPAR deficiency and upregulation affect TGFβ1.
Taken together, these data demonstrate that uPAR deficiency induced by CRISPR/Cas9-based Plaur knockout promotes ECM deposition and active TGFβ1 production by FBs.
TGFβ1 stimulation further enhances Akt phosphorylation and ECM synthesis in Plaur knockout fibroblasts
Next, we investigated whether the pro-fibrotic effects of TGFβ1 are potentiated by uPAR deficiency. To test this, Wt, Scrambled and Plaur KO FBs were treated with TGFβ1 (Fig. 7A).
Following TGFβ1 treatment, Plaur KO FBs exhibited higher levels of pSMAD2/pSMAD3 (canonical pathway) and, especially, pAkt-S473 (non-canonical pathway) compared to Scrambled cells (Fig. 7B and C). Notably, basal Akt phosphorylation was increased in Plaur KO cells compared to Scrambled, even in the absence of TGFβ1 stimulation, suggesting that uPAR deficiency preferentially induces activation of non-canonical TGFβ1-signaling pathway.
According to the enhanced TGFβ1-signaling pathway, Plaur KO fibroblasts exhibited a further, more pronounced increase in ECM protein synthesis (collagen I and fibronectin) in response to TGFβ1 treatment (Fig. 7D, E and F). Notably, control FBs (Wt and Scrambled) required TGFβ1 stimulation to reach the baseline ECM production level of untreated Plaur KO FBs, suggesting that uPAR-deficient FBs exist in a constitutively activated, pro-fibrotic state. No significant differences in SMA protein levels after TGFβ1 treatment were observed (Fig. 7E and F).
Fig. 7TGFβ1 further enhances Akt phosphorylation and ECM synthesis in Plaur knockout fibroblasts. A Schematic representation of the experimental design for assessing TGFβ1 effects on Plaur knockout (Plaur KO) fibroblasts. B Analysis of SMAD2/SMAD3 and Akt phosphorylation in Scrambled and Plaur KO fibroblasts following 30 min TGFβ1 stimulation. Phospho-protein levels were normalized to total proteins levels, *p < 0.05, **p < 0.01, N = 3. C Representative images of western blot results for panel “B”. Full-length blots/gels are presented in Supplementary Figure S9. D Immunofluorescence images of Wt, Scrambled and Plaur KO fibroblasts after 48-hour TGFβ1 stimulation stained for collagen I (COL I, green) and EDA-fibronectin (EDA-FN, red). Nuclei were stained with DAPI. E Protein levels of COL I (pro-collagen I and mature collagen I α-chain), FN, and αSMA measured in fibroblasts after 48-hour TGFβ1 stimulation, *p < 0.05, N = 3. F Representative images of western blot results for panel “E”. Full-length blots/gels are presented in Supplementary Figure S10. G Schematic summary illustrating that uPAR deficiency in fibroblasts promotes TGFβ1 activation, enhances TGFβ1-dependent signal transduction, and stimulates ECM synthesis. “A” and “G” were created with Servier Medical Art, licensed under CC-BY 4.0
Our data from Plaur knockout FBs demonstrate that uPAR deficiency not only increases the level of active TGFβ1 but also potentiates its pro-fibrotic activity, including a non-canonical activation of Akt pathway [66] and excessive ECM protein synthesis (Fig. 7G). These results suggest that TGFβ1-activated FBs are primary effectors of the fibrotic remodeling observed in uPAR deficiency.
uPAR deficiency leads to anti-angiogenic and pro-fibrotic shift of mCDC sectretome
The perivascular microenvironment relies on paracrine signaling to maintain EC and perivascular cell homeostasis [67–70]. However, as demonstrated above, uPAR deficiency in stromal cells induced overproduction of active TGFβ1 and amplified TGFβ1-mediated excessive ECM synthesis, thereby creating a microenvironment that may disrupt EC function and promote EndMT [58, 59, 71]. Moreover, ERK signaling pathway that stimulates growth factor secretion by CSs [72] was downregulated in uPAR-/- mCSs (Supplementary Figure S4A). This led us to analyze the secretome of the uPAR-deficient mCSs and its functional effects on ECs.
Given that Wt and uPAR-/- mCSs differed significantly in size, which may affect the diffusion of secreted molecules out of the spheroids, we disassembled mCSs to 2D monolayer culture of mCS-derived cells (mCDCs) (Fig. 8A and B). A bidirectional shift of uPAR-/- mCDC secretory profile was observed compared to Wt (Fig. 8C). On the one hand, insulin-like growth factor binding protein-2 (IGFBP-2), thrombospondin-2, VEGF, Serpinf1/PEDF, and platelet-derived growth factors (PDGF-AA and PDGF-BB/AB) were increased in uPAR-/- mCDC CMs compared to Wt. On the other hand, MCP-1, IGFBP-3, pentraxin-3, endothelin-1, angiogenin, proliferin, angiopoetin-1, and PIGF-2 were decreased in uPAR-/- mCDC CMs. VEGF and MCP-1 concentrations in mCDC CMs were further evaluated by ELISA, given the established role of these factors in regulating EC function [73, 74] and their high secretion levels in mCDCs. We observed no statistically significant differences in VEGF concentrations (Fig. 8D), but reduced MCP-1 in uPAR-/- mCDC CMs compared to Wt (Fig. 8D).
Fig. 8uPAR deficiency leads to anti-angiogenic and pro-fibrotic shift of mCDC sectretome. A Phase-contrast images of Wt and uPAR-/- mouse cardiosphere-derived cells (mCDCs). B Schematic representation of experiment design. Mouse lung endothelial cells (MLECs) were stained for CD31 (green) and VE-cadherin (red), nuclei were stained with DAPI. C Secretome profile of Wt and uPAR-/- mCDCs analyzed with ProteomeProfiler antibody arrays (R&D). D VEGF and MCP-1 concentrations in Wt and uPAR-/- mCDC condition mediums (CMs) measured by ELISA. Concentrations were normalized per 10^6^ cells. **p < 0.01, N = 4. E Proliferation/viability of MLECs cultivated in the presence of Wt or uPAR-/- mCDC condition mediums. IMDM was used as a control, N = 4. F Tube assay on Matrigel™ of MLECs cultivated in the presence of Wt or uPAR-/- mCDC condition mediums during 6 h. G Bar plots shows average tube length, total tube length and number of nodes formed by MLECs stimulated with Wt or uPAR-/- mCDC condition mediums. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, N = 4. H Immunofluorescence images of MLECs stimulated with Wt or uPAR-/- mCDC condition mediums during 72 h. MLECs were stained for F-actin (phalloidin, green) and SMA (red). Nuclei were stained with DAPI. I Bar plot quantifying F-actin+/SMA+ myofibroblast-like cells within the MLEC culture after 72 h of stimulation with Wt or uPAR-/- mCDC condition mediums. *p < 0.05, ***p < 0.001, N = 4. J Schematic summary illustrating that secretome produced by uPAR-deficient mCDCs disrupts the angiogenic capacity of ECs and promotes their myofibroblast-like transformation. “B” and “J” were created with Servier Medical Art, licensed under CC-BY 4.0
To evaluate how this bidirectional shift in secretory profile affects EC function, we assessed the proliferation/viability of MLECs and their in vitro angiogenic capacity (tubulogenesis on Matrigel) in the presence of 60% Wt and uPAR-/- mCDC CMs (Fig. 8B). While no effect on MLEC viability/proliferation was observed (Fig. 8E), angiogenic capacity was impaired (Fig. 8F and G). CMs from Wt mCDC promoted MLEC tube formation on Matrigel, while CMs from uPAR-/- mCDC failed to stimulate or even suppressed tubulogenesis (Fig. 8G). This result indicates the prevalence of anti-angiogenic effects of the uPAR-/- cell secretome.
To investigate whether the uPAR-/- cell secretome drives EndMT, we stimulated MLECs with Wt and uPAR-/- mCDC CMs for 72 h and quantified the generation of myofibroblast-like cells based on the formation of F-actin+/SMA+ stress fibers. As demonstrated on Fig. 8 (H and I), stimulation with uPAR-/- mCDC CMs increased the proportion of myofibroblast-like cells within the EC culture compared to Wt mCDC CMs and control (IMDM).
These results demonstrated that uPAR deficiency alters cell secretome, and this aberrant paracrine communication consequently disrupts angiogenic capacity of ECs and induces their myofibroblast-like transformation (Fig. 8J).
Discussion
In this study, using a 3D in vitro model of the cardiac perivascular-like microenvironment, we demonstrated that uPAR deficiency leads to TGFβ1-mediated fibrotic remodeling. Despite its extremely low expression in the intact tissues and vessels [10, 26], uPAR deficiency in vivo leads to age-dependent fibrotic lesions and endothelial dysfunction in multiple organs and tissues, including the heart [19–21, 65]. Although we found that uPAR is expressed by minor cell populations within the heart, such as ECs and FBs (Fig. 1), it seems that its basal expression contributes to maintenance of the “normal” phenotype and function of these cells, as our data demonstrate that uPAR deficiency leads to fibrotic transformation and loss of ECs within the perivascular-like microenvironment modeled by mCSs.
uPAR expression is upregulated when cells engage in tissue remodeling during regeneration [75], fibrosis [76], ischemic injury [77], inflammation, aging, atherosclerosis and carcinogenesis [10, 25, 26]. This elevated uPAR expression promotes migration of SMCs [78] and FBs [79, 80], wound healing [75], and angiogenesis [81, 82] as part of the compensatory program attempting to support normal tissue integrity. Ultimately, pathological stimuli, such as altered protease expression and activity [83], trigger the proteolysis-dependent shedding of uPAR or its fragments from the cell surface [17, 84], thereby compromising its functions. Indeed, increased suPAR plasma level is associated with heart failure development and poor prognosis [23, 85]. Moreover, uPAR-/- mice develop spontaneous age-dependent fibrosis in the lungs, skin and heart [19–21]. However, uPAR overexpression could also contribute to pathological remodeling. Notably, uPAR expression is elevated in M1 macrophages and a subset of senescent cells, both of which produce pro-inflammatory factors [86, 87]. Selective elimination of these uPAR-expressing cells via CAR-T therapy has been shown to ameliorate liver fibrosis induced by non-alcoholic steatohepatitis [25]. These seemingly contradictory findings (uPAR deficiency promotes fibrosis vs. elimination of uPAR-expressing cells alleviates it) suggest that tissue homeostasis requires a precise balance of uPAR expression across stromal and immune cells. Increased uPAR expression is likely essential for ensuring stromal and vascular cell activation [78–80], while its downregulation restrains the pro-inflammatory phenotype of immune cells [88, 89].
In this work, we focused on uPAR’s role in regulating stromal and vascular cells within the cardiac perivascular-like microenvironment modeled in vitro with CSs (Fig. 2). We found that uPAR deficiency promoted loss of ECs and fibrotic remodeling, as demonstrated by elevated levels of SMA, SM22α and SNAIL, excess ECM deposition, and increased mCS stiffness (Figs. 4 and 5). The findings are consistent with the data provided by in vitro and in vivo studies [17, 18, 20, 21], showing that uPAR deficiency, knockdown or cleavage promotes FB-to-myofibroblast differentiation, ECM deposition, EC apoptosis, and cardiac fibrosis development.
Using Plaur KO FBs, we demonstrated that uPAR deficiency-induced ECM deposition is orchestrated by FB activity through TGFβ1-mediated mechanisms (Figs. 6 and 7). TGFβ1 plays a key role in cardiac fibrosis progression [71, 90], and its levels are elevated in the cardiac tissue of uPAR-/- mice [21]. In the current work, we found that both uPAR-/- mCSs and Plaur KO FBs produced higher levels of active TGFβ1 despite no difference in Tgfb1 gene expression compared to control groups (Figs. 5 and 6), suggesting that uPAR deficiency contributes to TGFβ1 activation. In turn, uPAR deficiency amplified TGFβ1-induced ECM deposition by FBs, potentially through activation of Akt signaling (Fig. 7), a pathway known to stimulate ECM synthesis [66]. Thus, we suggest that uPAR deficiency exacerbates TGFβ1-mediated pro-fibrotic effects.
uPA/uPAR system could regulate the release of active TGFβ1 from the latent complex by two different mechanisms: (1) proteolytic cleavage induced by uPA-dependent plasmin or activated MMPs generation, and (2) integrin-dependent mechanical release [61, 62, 91]. Proteolytically active form of uPA (tc-uPA) remained unchanged in uPAR-/- mCSs, while uPA precursor (sc-uPA) level was remarkably increased (Fig. 5B). Independent of its proteolytic activity, sc-uPA can also exert pro-fibrotic effects via nuclear translocation, which was upregulated in uPAR-/- mCSs (Fig. 5C-F), leading to subsequent activation of SMA expression [64] and epithelial-to-mesenchymal transition (EMT) [92]. The mechanical release of TGFβ1 depends on integrin function [93], which is impaired by uPAR and uPA deficiency [53]. Our experiments analyzing the impact of the uPAR/uPA/integrin (αvβ3, αvβ5, α5β1 and αMβ2) axis on spheroid assembly revealed that uPAR deficiency impairs integrin functions (Fig. 3). These results may partially explain the excess TGFβ1 activation we observed. TGFβ1, in turn, can subsequently enhance the expression and activation of integrins, thereby creating a pro-fibrotic positive feedback loop [93].
Moreover, impaired integrin functions, such as downregulated ERK signaling pathway and hyper-adhesion to collagen I (Supplementary Figures S3B and 4), could further promote pro-fibrotic cell activation. It was reported that knockdown of uPAR or uPA affects the activity of integrins β5 and β1 and thus leads to myofibroblast transformation [53]. We assume that uPAR deficiency-induced cell adhesion to collagen I (a β1 integrin ligand [94]) triggers transformation of FBs into myofibroblasts [54].
Overproduction of active TGFβ1, excessive ECM synthesis, stiffness and aberrant cell-ECM interactions (e.g., cell hyper-adhesion to collagen) create a pathological microenvironment that impairs EC function and promotes EndMT [58, 59, 71]. Indeed, our previous study showed that TGFβ1 disrupts endothelial capillary-like network within CSs [29]. According to the current paradigm, endothelial dysfunction is an initiating event in the pathogenesis of myocardial fibrosis and chronic heart failure [95]. While uPAR’s contribution into the regulation of EC survival and functions is poorly understood, our findings demonstrated that uPAR deficiency triggers EC loss (Fig. 4D, G) and upregulation of EndMT-related transcription factor SNAIL (Fig. 5G), as well as decreases cell viability within mCSs (Fig. 4A and B). Moreover, uPAR deficiency leads to the anti-angiogenic and pro-fibrotic shift of cell secrotome, resulting in suppressed angiogenic function and activation of myofibroblast-like transformation of ECs (Fig. 8). This aligns with prior reports linking uPAR dysfunction (deficiency or cleavage) to EndMT and apoptosis in cardiac and dermal vessels [20, 21, 65]. However, the role of uPAR in EMT or mesothelial-to-mesenchymal transition (MesoMT), which are similar processes to the EndMT, remains controversial and requires further investigation [92, 96, 97].
Altogether, our findings suggest that within the cardiac perivascular-like microenvironment, uPAR loss disrupts both normal cell-ECM interactions (i.e., integrin functions) and paracrine cell-to-cell communication (specifically through TGFβ1-mediated mechanisms), leading to fibrotic transformation and inhibition of blood vessel formation (summarized in Fig. 9). We can contemplate that preservation of basal membrane-bound uPAR expression within stromal and vascular cells could coordinate normal homeostasis of perivascular microenvironment to delay cardiac fibrosis progression. However, it is important to consider the main limitation of CS-based model, particularly the absence of immune cells, which play a crucial role in the development of fibrosis and endothelial dysfunction [95, 98]. Nevertheless, our key findings, including fibrotic remodeling, EC loss and TGFβ1 accumulation induced by uPAR deficiency in both mCSs and FBs, are in accordance with previously published studies [17, 18, 20, 21].
Fig. 9. Fibrotic remodeling of the cardiac perivascular microenvironment induced by uPAR deficiency. uPAR: urokinase receptor; uPA: urokinase; EC: endothelial cell; SM: smooth muscle; EndMT: endothelial-to-mesenchymal transition; ECM: extracellular matrix; COL: collagen; TGFβ1: transforming growth factor β1; Created with Servier Medical Art, licensed under CC-BY 4.0
Conclusion
Our findings provide novel insights into uPAR’s role in fibrotic remodeling within the cardiac perivascular-like microenvironment. We have shown for the first time using a more complex and multicomponent model based on CSs in contrast to 2D monocultures and at the same time a simpler model in contrast to in vivo tissues including inflammatory cells, that uPAR deficiency promotes fibrotic remodeling, accompanied with EC loss, altered cell-ECM and paracrine cell-cell interactions. Using Plaur KO FBs, we demonstrated that uPAR deficiency elevates active TGFβ1 levels and amplified TGFβ1-mediated signaling pathways, resulting in enhanced ECM deposition. We propose that the overproduction of active TGFβ1 likely represents a potential mechanism underlying the pro-fibrotic transformation induced by uPAR deficiency.
Supplementary Information
Supplementary Material 1.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Schaefer W, Hedemann N, Christmann T, Guledani A, Sebens S, Geisen R et al. Detection of Necrotic Cores in Tumor Spheroids Using a Live/Dead Assay and NYONE® Scientific. SYNENTEC Gmb H. 2023. 1–6.