PAI-1 deficiency drives pulmonary vascular smooth muscle remodeling and pulmonary hypertension

Abstract

Pulmonary arterial hypertension (PAH) is a progressive disease characterized by vasoconstriction and remodeling of small pulmonary arteries (PAs). Central to the remodeling process is a switch of pulmonary vascular cells to a proliferative, apoptosis-resistant phenotype. Plasminogen activator inhibitors-1 and -2 (PAI-1 and PAI-2) are the primary physiological inhibitors of urokinase-type and tissue-type plasminogen activators (uPA and tPA), but their roles in PAH are unsettled. Here, we report that: 1) PAI-1, but not PAI-2, is deficient in remodeled small PAs and in early-passage PA smooth muscle and endothelial cells (PASMCs and PAECs) from subjects with PAH compared with controls; 2) PAI-1^−/− mice spontaneously develop pulmonary vascular remodeling associated with upregulation of mTORC1 signaling, pulmonary hypertension (PH), and right ventricle (RV) hypertrophy; and 3) pharmacological inhibition of uPA in human PAH PASMCs suppresses proproliferative mTORC1 and SMAD3 signaling, restores PAI-1 levels, reduces proliferation, and induces apoptosis in vitro, and prevents the development of SU5416/hypoxia-induced PH and RV hypertrophy in vivo in mice. These data strongly suggest that downregulation of PAI-1 in small PAs promotes vascular remodeling and PH due to unopposed activation of uPA and consequent upregulation of mTOR and transforming growth factor-β (TGF-β) signaling in PASMCs, and call for further studies to determine the potential benefits of targeting the PAI-1/uPA imbalance to attenuate and/or reverse pulmonary vascular remodeling and PH.

NEW & NOTEWORTHY This study identifies a novel role for the deficiency of plasminogen activator inhibitor (PAI)-1 and resultant unrestricted uPA activity in PASMC remodeling and PH in vitro and in vivo, provides novel mechanistic link from PAI-1 loss through uPA-induced Akt/mTOR and TGFβ-Smad3 upregulation to pulmonary vascular remodeling in PH, and suggests that inhibition of uPA to rebalance the uPA-PAI-1 tandem might provide a novel approach to complement current therapies used to mitigate this pulmonary vascular disease.

INTRODUCTION

Pulmonary arterial hypertension (PAH) is characterized by vasoconstriction and remodeling of small pulmonary arteries (PAs), leading to luminal narrowing, increased PA pressure, RV afterload, right heart failure, and premature death (1). Current treatment approaches, which predominantly target the vasoconstriction component, do not reverse pulmonary vascular remodeling or stop disease progression. Thus, there is a clear need for newer therapies based on molecular mechanisms focused on vascular remodeling to complement contemporary management (1).

Unchecked hyperproliferation and resistance to apoptosis of resident pulmonary vascular cells in small muscular PAs are important pathological components of pulmonary vascular remodeling. Dysregulation of the bone morphogenic protein receptor 2 (BMPR2)/transforming growth factor-β (TGF-β) network and overactivation of receptor tyrosine kinase and growth factor-mediated pro-proliferative signaling pathways, including Akt/mechanistic target of rapamycin (mTOR), play important roles in supporting PASMC hyperproliferation, pulmonary vascular remodeling, and PAH (1). The mechanistic link between these key molecular players in PAH is not fully established, and the strategies to reverse PASMC molecular reprogramming, remodeling, and PAH are not fully developed (1, 2).

The plasminogen activator system may play an important role in the remodeling program in PAH based on its roles in vascular postinjury restenosis, fibrosis, angiogenesis, and tumorigenesis. PAI-1 is the primary physiological inhibitor of uPA and tPA. PAI-1 deficiency leads to unopposed uPA activity. uPA promotes vascular remodeling (3) through plasmin-mediated proteolysis (4), activation of matrix metalloproteinases (5), and growth factors, including TGF-β1 (5). TGF-β1 is secreted in a biologically inactive (latent) form that must be proteolytically activated by uPA-generated plasmin to bind to its cell surface receptors (6). TGF-β1 is overproduced by human PAH PASMCs and promotes PASMC hyperproliferation, pulmonary vascular remodeling, and PH via canonical Smad2/3 and noncanonical PI3K-Akt-mTOR pathways (7, 8). However, both, proproliferative (9) and growth suppressor functions for PAI-1 in PASMCs have been reported (10), and the potential attractiveness of targeting PAI-1 signaling in PAH remains to be determined.

Here, we aimed to determine the role of PAI-1 in pulmonary vascular cell proliferation, remodeling, and PH using lung tissues and cells from the small muscular PAs of PAH and nondiseased subjects, PAI-1^−/− mice, and mouse SU5416-hypoxia (SuHx) PH model.

METHODS

The online supplement provides more detailed information [Materials and Methods, Antibodies specificity and validation (Supplemental Table S1 and Supplemental Fig. S1)].

Human Tissues and Cell Cultures

Human lung tissues from nondiseased (rejected transplant) and participants with idiopathic PAH (IPAH) (explant) and early-passage (3–8) PASMCs and PAECs were provided by UC Davis Lung Center Pulmonary Vascular Disease Program human specimens biobank (tissues, cells), UC San Francisco transplant program (tissues for cell isolation), University of Pittsburgh VMI Cell Processing Core, and the Pulmonary Hypertension Breakthrough Initiative (PHBI) (cells). Cell isolation, characterization, and maintenance were performed as described previously under protocols adopted by the PHBI (8, 11).

Animals

All animal studies were performed under the protocols approved by the University of California, Davis, and Ohio State University Animal Care and Use Committees. To study the effect of PAI-1 loss in spontaneous PH development, 10-mo-old PAI-1^−/− (Molecular Innovations Inc, MI) and WT mice (C57BL/6J, Jackson Laboratories, ME) were subjected to hemodynamic analysis as described before (11). To study the effect of uPA catalytic activity inhibition on PH development, the WT (C57BL/6J, Jackson Laboratories, ME) male mice were randomly assigned to control and experimental groups. Experimental groups were given daily intraperitoneal injections of 10 mg/kg BB2-30F or vehicle from days 1–21 of exposure to SU5416 (20 mg/kg sc once a week) and chronic hypoxia (10% O₂). Hemodynamic measurements and tissue harvest were made on day 21 as described before (11, 12). Controls were same-age same-sex mice maintained under normoxia.

Statistical Analysis

Statistical analysis was performed using StatView (SAS Institute, Cary, NC) and GraphPad Prism 9.02 (GraphPad Software, San Diego, CA) software. Statistical comparisons between two groups were performed by Mann–Whitney U test and among three groups by one-way ANOVA with Tukey’s multiple comparison test. Statistical significance was defined as P ≤ 0.05.

RESULTS

PAI-1 Is Deficient in Small Muscular PAs and Resident PA Cells from PAH Lungs

Using immunohistochemical and immunoblot analyses, we detected a marked decrease in PAI-1 protein in situ in smooth muscle α-actin (SMA)-positive areas in small remodeled PAs from patients with PAH (Fig. 1A) and in vitro in distal early-passage human PAH PAECs and PASMCs compared with controls (Fig. 1, B–E). PAI-1 deficiency in PAH PASMCs was accompanied by a significant increase in plasminogen activation (Fig. 1F) and cell proliferation (Fig. 1, G and H), suggesting the link between PAI-1 deficiency and pulmonary vascular remodeling. We also measured PAI-2 in the IPAH and Control PASMCs and PAEC sets that we used to analyze PAI-1. In contrast to PAI-1, our results did not reveal significant differences in PAI-2 content between control and PAH cells (Supplemental Fig. S2).

Figure 1.

PAI-1 is deficient in pulmonary vascular cells from small PAH PAs. A: IHC analysis of human lung tissue sections to detect PAI-1 (red). Green—SMA; blue—DAPI; yellow—red and green overlap. Images are representative from three subjects/group. Bar = 50 µm. B–E: immunoblot analysis of PAECs and PASMCs from human control and PAH lungs to detect indicated proteins. Gels for the detection of β-actin were ran separately from the same sample preparations. F: rates of the plasmin generation in control and PAH PASMCs normalized to the cell numbers. G and H: cell proliferation was measured by Ki67 staining (red); data represent the percentile of Ki67-positive cells/total number of cells detected by DAPI (blue). Bar = 50 µm. B–G: data are means ± SD; n = 4–5 human subjects/group. Mann–Whitney U test was used; closed black circles—cells from males; open circles—cells from females. PAs, pulmonary arteries; PAEC, pulmonary artery endothelial cells; PAI-1, plasminogen activator inhibitor-1; PAH, pulmonary arterial hypertension; PASMC, pulmonary artery smooth muscle cells.

Mice with PAI-1 Loss Develop Spontaneous PA Remodeling and PH

PAI-1 knockout mice (PAI-1^−/−) showed remodeling of the small muscular PAs assessed by an increase in medial wall thickness (PA MT) (Fig. 2, A and B), significant increase in systolic RV pressure (sRVP) (Fig. 2C), and RV hypertrophy (Fulton index) compared with age- and sex-matched wild-type (WT) mice on the same background (Fig. 2D). Importantly, small PAs of PAI-1^−/− mice showed smooth muscle-specific increase in phospho-Smad3 and phospho-S6 (Fig. 2, E and F), suggestive of activation of TGF-β/Smad3 and mTORC1. In agreement with these observations, siRNA PAI-1 significantly increased mTORC1-specific phosphorylation of ribosomal protein S6 in control PASMCs (Fig. 2G). Together, these data show that loss of PAI-1 results in spontaneous pulmonary vascular remodeling and PH and suggest that TGF-β/Smad3 and mTORC1 act as downstream effectors.

Figure 2.

Loss of PAI-1 in mice results in the development of spontaneous PH. A–F: morphological and hemodynamic analysis of 10-mo-old WT C57BL/6J and PAI-1^–/– mice. A and B: representative H&E images (bar = 30 µm) (A) and PA MT (B) from 10 PAs (<150 µm outer diameter) per mouse; C and D: sRVP (C) and RV/(LV+S) (Fulton index) (D). n = 6 mice/WT group (B–D); n = 6 (B and D) and n = 5 (C) mice/PAI-1^–/– group. E and F: IHC analysis of lung tissue sections from PAI-1^–/– and WT mice. Red: phospho-Smad3 (E) or phospho-S6 (F), green—SMA, blue—DAPI. Bar = 20 µm. Representative images from three animals/group. G: human control PASMCs were transfected with siRNA PAI-1 (+) or control scrambled siRNA (–); 48 h posttransfection, immunoblot analysis was performed to detect indicated proteins. Representative images (top) and statistical analysis of n = 5 subjects/group (bottom). Data are means ± SD; Mann–Whitney U test was used; closed black circles—cells from males; open circles—cells from females. PAs, pulmonary arteries; PAI-1, plasminogen activator inhibitor-1; PASMC, pulmonary artery smooth muscle cells; sRVP, systolic RV pressure.

Pharmacological Inhibition of uPA Restores PAI-1, Reduces Human PAH PASMC Proliferation, and Protects Mice from PH Development

Because PAI-1 activators are not currently available, we next asked whether the functional consequences of PAI-1 deficiency in pulmonary vasculature could be mitigated through the control of uPA. We found that inhibitor of human and mouse uPA BB2-30F (13) and WX671 and inhibitor of human uPA and some other trypsin-like proteases (14), reduced TGF-β-dependent Smad3 phosphorylation and inhibited Akt-mTORC1 in human PAH PASMCs, as evidenced by significant decreases in P-Smad3, P-S473 Akt, and P-S6 (Fig. 3, A and B). Furthermore, WX671 inhibited proliferation and induced apoptosis in human PAH PASMC (Fig. 3, C and D). Of interest, inhibition of uPA activity was accompanied by an increase in PAI-1 protein content (Fig. 3, A and B).

Figure 3.

uPA inhibition reduces proproliferative signaling, restores PAI-1, inhibits proliferation, and induces apoptosis in human PAH PASMCs. Human PAH PASMCs were incubated with 10 µM uPA inhibitors BB2-30F (A) or WX671 (B), or diluent for 48 h followed by immunoblot analysis to detect indicated proteins (A and B), cell proliferation (Ki67) (C), or apoptosis (In Situ Cell Death Detection Kit) (D) assessment. n = 3 subjects/group. Data are means ± SD from three independent experiments, each performed using cells from different human subjects. Bar = 100 µm. Mann–Whitney U test was used; open circles indicate cells from female patients. PAI-1, plasminogen activator inhibitor-1; PAH, pulmonary arterial hypertension; PASMC, pulmonary artery smooth muscle cells; uPA, urokinase-type plasminogen activator.

To test whether uPA mediates pulmonary vascular remodeling and PH in vivo, we examined the effect of BB2-30F on the development of SuHx PH in mice. We recognize the potential confounding effect of VEGF receptor inhibition; however, we used the mouse SuHx model as it induces more severe PH than chronic hypoxia exposure alone in mice, increasing the signal-to-noise ratio of the data. Vehicle-treated SuHx-exposed mice by day 21 of PH development showed upregulation of mTORC1 (assessed by phospho-S6) in small PAs (Fig. 4A), pulmonary vascular remodeling and PH, as evidenced by significant increase of PA MT (Fig. 4, B and C), sRVP (Fig. 4D), RV hypertrophy (Fig. 4E), and decrease in PA acceleration time/pulmonary ejection time (PAAT/PET) ratio (Fig. 4F) compared with age- and sex-matched normoxic controls. The SuHx group treated with BB2-30F (10 mg/kg, ip daily, days 1–21) was protected from all of these pathological changes (Fig. 4, A–F), showing that uPA inhibition protects mice from the development of SuHx PH.

Figure 4.

Inhibition of uPA prevents the development of SuHx-induced PH in mice. WT (C57BL/6J) male mice were given daily intraperitoneal injections of 10 mg/kg uPA inhibitor BB2-30F or vehicle from days 1–21 of exposure to VEGFR inhibitor SU5416 (20 mg/kg sc once a week) and hypoxia ($\text{F}{\text{I}}_{{\text{O}}_2}$ 10%). Hemodynamic measurements and tissue harvest were made on day 21. Controls (con) were same-age mice kept under normoxia. A: IHC of lung tissue sections. Red—phospho-S6, green—SMA, blue—DAPI. Bar = 50 µm. Images are representative from three animals/group. B and C: representative H&E images (bar = 30 µm) (B) and PA MT (C); n = 6 mice/group, 10 PAs with outer diameter <150 μm per mouse. D–F: sRVP (D), RV/(LV+S) (Fulton index) (E), and PA acceleration time/ejection time (PAAT/PET) ratio (F); C, E, and F: n = 6 mice/group. D: n = 6/4/5 mice/groups Con/SuHx-Veh/SuHx-BB2. C, D, E, and F: one-way ANOVA with Tukey’s multiple comparison test was used; Data are means ± SD. G: schematic representation of the mechanism by which PAI-1 deficiency promotes PA remodeling and PH. PAs, pulmonary arteries; PAI-1, plasminogen activator inhibitor-1; PASMC, pulmonary artery smooth muscle cells; sRVP, systolic RV pressure; uPA, urokinase-type plasminogen activator. [Figure created with BioRender.com.]

DISCUSSION

Multiple factors including genetic predisposition, epigenetic mechanisms, oxidative stress, metabolic reprogramming, extracellular matrix (ECM) remodeling, and inflammation contribute to PASMC hyperproliferation, remodeling, and PAH (2). The plasminogen activator system plays key roles in modulating physiological and pathological vascular remodeling, but the role of its key component PAI-1 in PAH is unclear. In this study, we report several novel findings with respect to the plasminogen activator system and PH. First, a systemic loss of PAI-1 results in spontaneous pulmonary vascular remodeling and PH in mice. Second, PAI-1 protein, but not PAI-2, is significantly reduced in PAECs and PASMCs from the small muscular PAs of human PAH lungs, resulting in uncontested activation of uPA and generation of plasmin. Third, pharmacological inhibition of uPA retards pulmonary vascular remodeling in vivo and experimental PH in mice. This link is supported by in vitro observations that PAI-1 is restored by small molecular weight inhibitors of uPA that downregulate TGF-β/Smad3 and Akt/mTOR, suppress proliferation, and induce apoptosis of human PAH PASMCs (Fig. 4G).

PAI-1 levels vary widely in vivo in response to diverse external stimuli. The synthesis and function of PAI-1 in the pulmonary vasculature and elsewhere are subject to complex regulatory pathways not currently amenable to regulation in the clinical setting. PAI-1 has the potential to modulate vascular remodeling by multiple mechanisms, such as promoting PA intimal hyperplasia by stabilizing intravascular fibrin, which often forms in response to endothelial injury and supports the invasion of VSMCs and other cell types to form a neointima (15, 16), and promoting VSMC proliferation and inhibiting apoptosis (10, 17). PAI-1 may also reduce pathological vascular remodeling by 1) inhibiting uPA and, consequently, cell-associated plasmin formation; 2) binding to the ECM protein vitronectin, thereby competing for vitronectin receptors on VSMC (i.e., α_Vβ₃ and u-PAR) (18–21), which inhibits VSMC migration through ECM; and 3) prevention of TGF-β1 activation, which exerts pleiotropic effects on VSMC proliferation and migration (22).

Studies of the role of PAI-1 in PAH have also yielded discordant results. Elevation in PAI-1 serum levels correlated with disease severity (23) and lungs of patients with IPAH and rodents with hypoxia-induced PH were shown to have markedly higher levels of PAI-1 (24). In contrast, Kouri et al. reported reduced levels of PAI-1 protein in PASMC from IPAH lungs and linked this with increased PASMC proliferation (10), which is consistent with our observations. Chen et al. also found a trend of reduced PAI-1 protein levels in PASMCs from human subjects with IPAH and rats with SuHx PH (9). They also reported that microRNA miR-19a/b from the miR-17 ∼ 92 cluster can directly suppress PAI-1 to regulate PASMC proliferation through binding to the 3′-untranslated region (UTR) of PAI-1 mRNA (9). PAI-1 3′-UTR Hind III polymorphism associated with the Hd2 allele was detected in patients with PAH (25).

One consequence of PAI-1 loss is uncontested activation of uPA and plasmin generation. Previous work showing that mice deficient in uPA (plau^−/− mice) are protected from hypoxia-induced PH (26) and neointima formation after vascular injury (27, 28) suggests a key role for uPA in PA vascular remodeling. Indeed, our studies demonstrate that amiloride analog BB2-30F, which inhibits the catalytic activity of both human and mouse uPA (13) and suppresses metastasis in mouse models of pancreatic cancer (29, 30), impedes pulmonary vascular remodeling and experimental PH in mice. The lack of drug-only controls is a potential limitation for our study. However, drug toxicity studies performed in our prior studies demonstrated that 10 mg/kg is a well-tolerated dose in mice (29, 30). The in vivo data are supported by our in vitro observations that both BB2-30F and WX671 (upamostat), an inhibitor of enzymatic activity of human uPA and some other trypsin-like serine proteases, which was used in clinical trials for patients with pancreatic cancer (14), restore PAI-1, downregulate TGF-β/Smad3 and Akt/mTOR, suppress PASMC proliferation, and induce apoptosis in human PAH PASMCs. This suggests that excessive enzymatically active uPA further depletes PAI-1 in a feedforward pathogenic cycle and that inhibition of uPA may spare residual PAI-1 from further depletion. Although WX671 treatment also can inhibit tPA in PASMCs in vitro, our in vivo data using a specific inhibitor of uPA (BB2-30F) in the SuHx model of PH strongly suggest that namely uPA contributes to the pathogenesis of PH while PAI-1 is deficient. However, we will assess the role of tPA in this and other models when the specific tPA inhibitor becomes available.

Although we observed that PAI-1/uPA promotes remodeling and PH via TGF-β/Smad3 and Akt/mTOR, our findings from mice with global PAI-1 knockout cannot exclude the possibility that other cell types and other uPA-induced signaling pathways may also be involved. For example, PAH PASMCs accumulate higher amounts of uPA in their nuclei than controls (not shown), which could also support PASMC proliferation by suppressing p53 expression (31, 32).

In conclusion, this study identifies an important role for PAI-1 deficiency and unchecked uPA activity in pulmonary vascular remodeling and PH in vitro and in vivo, provides a novel mechanistic link from PAI-1 through uPA-induced Akt/mTOR and TGFβ-Smad3 upregulation to PASMC hyperproliferation, remodeling, and PH (Fig. 4G). The results also suggest that inhibition of uPA to rebalance the uPA-PAI-1 tandem as a therapeutic intervention might provide a novel approach to complement current therapies used to mitigate this severe and progressive pulmonary vascular disease. These findings call for further investigation to establish this molecular pathway as a potential novel target to attenuate or even reverse established PAH.

DATA AVAILABILITY

Data will be made available upon reasonable request.

SUPPLEMENTAL MATERIAL

Supplemental Methods, Supplemental Table S1, and Supplemental Figs. S1 and S2: https://doi.org/10.6084/m9.figshare.26371195.

GRANTS

This work is supported by National Institutes of Health (NIH)/National Heart, Lung, and Blood Institute (NHLBI) R01HL166932 (to T.V.K.), R01HL130261 (to E.A.G.), R01HL150638 (to E.A.G), R01HL172488 (to E.A.G.), RO1HL141462 (to V.S.), R01HL139881 (to L.F.), RO1HL159256 (to D.B.C.), Nina Ireland Program for Lung Health (to P.J.W.), VA ORD CSR&D CX002011 (to J.R.G.), R35HL150698 (to M.P.), The LAM Foundation LAM0139P07-19 (to V.S.), Department of the Army TS150032 (to V.S.), NHMRC Ideas grant APP1181179 (to M.R. and M.J.K.). Pulmonary Hypertension Breakthrough Initiative is supported by NIH/NHLBI R24 HL123767 and by the Cardiovascular Medical Research and Education Fund (CMREF).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

M.R., M.J.K., M.P., E.A.G., L.F., and V.S. conceived and designed research; T.V.K., B.P., S.B., A.S., L.T., A.P., S.V.Z., L.F., V.S., L.J., B.J.B., J.P.M., D.G., I.Z., and G.N. performed experiments; T.V.K., B.P., S.B., A.S., L.T., A.P., J.R.G., P.J.W., S.V.Z., E.A.G., V.S., L.J., B.J.B., J.P.M., D.G., I.Z., D.L., and G.N. analyzed data; T.V.K., B.P., S.B., A.S., L.T., J.R.G., P.J.W., M.J.K., M.P., S.V.Z., H.M.D., D.B.C., E.A.G., L.F., V.S., B.J.B., J.P.M., D.G., and I.Z. interpreted results of experiments; T.V.K., A.P., S.V.Z., E.A.G., L.F., V.S., D.G., I.Z., D.L., and G.N. prepared figures; E.A.G., L.F., and V.S. drafted manuscript; T.V.K., M.R., J.R.G., P.J.W., M.J.K., S.V.Z., D.B.C., E.A.G., L.F., V.S., L.J., and B.J.B. edited and revised manuscript; T.V.K., B.P., S.B., A.S., L.T., A.P., M.R., J.R.G., P.J.W., M.J.K., M.P., S.V.Z., H.M.D., D.B.C., E.A.G., L.F., V.S., L.J., B.J.B., J.P.M., D.G., I.Z., D.L., and G.N. approved final version of manuscript.