Reduced Cathepsin L expression and secretion into the extracellular milieu contribute to lung fibrosis in systemic sclerosis

Abstract

Objectives: Lung fibrosis is the leading cause of death in SSc, with no cure currently available. Antifibrotic Endostatin (ES) production does not reach therapeutic levels in SSc patients, suggesting a deficit in its release from Collagen XVIII by the main cleavage enzyme, Cathepsin L (CTSL). Thus, elucidating a potential deficit in CTSL expression and activity unravels an underlying molecular cause for SSc-driven lung fibrosis.

Methods: Fibrosis was induced experimentally using TGF-β in vitro, in primary human lung fibroblasts (pLFs), and ex vivo, in human lung tissues. ES and CTSL expression was quantified using ELISA, RT-qPCR, immunoblotting or immunofluorescence. Recombinant NC1-FLAG peptide was used to assess CTSL cleavage activity. CTSL expression was also compared between SSc vs normal (NL)-derived pLFs and lung tissues.

Results: ES levels were significantly reduced in media conditioned by TGF-β-induced pLFs. TGF-β-stimulated pLFs significantly reduced expression and secretion of CTSL into the extracellular matrix (ECM). CTSL was also sequestered in its inactive form into extracellular vesicles, further reducing its availability in the ECM. Media conditioned by TGF-β-induced pLFs showed reduced cleavage of NC1-Flag and reduced release of the antifibrotic ES fragment. SSc-derived pLFs and lung tissues expressed significantly lower levels of CTSL compared with NL.

Conclusions: Our findings identify CTSL as a protein protective against lung fibrosis via its activation of antifibrotic ES, and whose expression in SSc pLFs and lung tissues is suppressed. Identifying strategies to boost CTSL endogenous levels in SSc patients could serve as a viable therapeutic strategy.

Keywords: Cathepsin L; SSc; TGF-β; endostatin; extracellular vesicles; fibroblasts; fibrosis; lung; scleroderma.

Fig. 1

Activated pLFs release reduced levels of ES and downregulate CTSL expression and secretion pLFs from different donors were treated with vehicle control (VC) or 5 ng/ml recombinant TGF-β. Supernatants and RNA were collected after 24, 48 and 72 h of stimulation. (A) ES concentrations in the supernatants were measured using ELISA, and fold-change estimates were calculated to compare TGF-β with vehicle (n = 5, three donors). (B) COL18A1 gene expression levels were quantified using qPCR, and fold-change estimates were calculated to compare TGF-β to vehicle (normalized to B2M) (n = 7, four donors). (C) CTSL gene expression levels were quantified using qPCR, and fold-change estimates were calculated to compare TGF-β to vehicle (normalized to B2M) (n = 8, four donors). (D) CTSL was detected in conditioned media and cellular lysates. GAPDH was used as a loading control for cellular lysates to which the band density quantification was normalized. (n = 3, three donors). Graphs analysed by 1-way analysis of variance (Dunnett correction). A dotted line at a fold-change of 1.0 (i.e. no change) is provided for reference. Bars represent mean (s.e.m.). *P < 0.05, **P < 0.01, ****P < 0.0001. pLF: primary human lung fibroblast; ES: Endostatin; CTSL: Cathepsin L; COL18A1: collagen type XVIII alpha 1 chain; B2M: β2-microglobulin; GAPDH: glyceraldehyde-3-phosphate dehydrogenase.

Fig. 2

CTSL expression in pLF is specifically inhibited by TGF-β’s canonical pathway NL fibroblasts from different donors were treated with vehicle control, recombinant TGF-β and/or SIS3 (4 μM), PI3K inhibitor (10 μM), or JNK inhibitor (10 μM). Lysates and RNA were collected after 48 h. (A) Gene expression levels of CTSL were quantified using qPCR (n = 5, three donors), and fold-change estimates were calculated to compare treated samples with the untreated one (normalized to B2M). (B) Protein expression levels of CTSL were analysed by immunoblotting, with GAPDH used as loading control (n = 3, three donors). Graphs analysed by 1-Way analysis of variance (Dunnett correction). Bars represent mean (s.e.m.). *P < 0.05, **P < 0.01. pLF: primary human lung fibroblast; NL: normal lung; SIS3: smad3 inhibitor; CTSL: Cathepsin L; GAPDH: glyceraldehyde-3-phosphate dehydrogenase.

Fig. 3

TGF-β reduces cleavage of NC1-Flag via downregulation of CTSL (A) Expression of recombinant flag-tagged NC1 peptide. NC1 coding sequence was inserted in a flag tag overexpression plasmid and transfected into HEK 293 cells for 96 h. Empty plasmid transfection was used as control. (B) NL fibroblasts from different donors were treated with vehicle control or recombinant TGF-β. Supernatants were collected after 72 h of stimulation, inoculated with recombinant NC1-Flag, and incubated for 1, 3, 6 and 24 h at 37°C. A selective CTSL inhibitor (CTSLi, 20 µM) was used in parallel on a VC-treated sample for 24 h. Cleavage of NC1-Flag was detected via immunoblotting. (C) Generated ES was detected via immunoblotting at 6 and 24 h, using ES specific antibody (n = 3, three donors). Graphs analysed by (B) 1-way analysis of variance (Dunnett correction) and (C) 2-way analysis of variance (Šídák correction). Bars represent mean (s.e.m.). *P < 0.05, **P < 0.01, ****P < 0.0001, ns: not significant. CTSL: Cathepsin L; ES: endostatin; VC: vehicle control.

Fig. 4

TGF-β increases packaging CTSL into EV, where it remains inactive NL fibroblasts from different donors were treated with vehicle control or recombinant TGF-β. Supernatants and lysates were collected after 48 h of stimulation. EV were isolated from supernatants and lysed. (A) EV size and concentration were quantified using NTA. (B) EV content of CTSL, CD81 and Calnexin (negative control) were analysed by immunoblotting upon loading equal EV number in each lane. Cellular lysates were analysed as well for proof of successful EV isolation (n = 3, thee donors). Densitometry analysis was performed. The CTSL/CD81 ratio in the EV fractions was calculated as an indicator of the extent of CTSL packaging in EV. Graphs (right panels) analysed by paired sample t-test. Bars represent mean (s.e.m.). (C) NC1-Flag was localized to the EVs of MRC-5 cells via overexpression, then cells were treated with vehicle control or recombinant TGF-β. EVs and lysates were collected after 48 h of stimulation. NC1-Flag, cleavage product ES-Flag and CTSL (pro and mature) were detected via immunoblotting in cellular and EV lysates (n = 3). GAPDH was used as a loading control for cellular lysates, and CD81 (EV marker) was used as a reference for EV lysates. EV: extracellular vesicles; ES: endostatin; CTSL: Cathepsin L; NL: normal lung; NTA: nanoparticle tracking analysis; GAPDH: glyceraldehyde-3-phosphate dehydrogenase.

Fig. 5

SSc-derived pLFs express lower levels of CTSL and sequester more CTSL in EVs than normal pLFs (A) RNA-seq data comparing RNA expression of pro and antifibrotic genes between five independent SSc-derived pLFs vs five independent NL-derived pLFs. CTSL indicated in red box. (B) Fold change in baseline CTSL gene expression levels were quantified using qPCR comparing NL (n = 16) vs SSc (n = 8) derived pLFs (normalized to B2M). (C) CTSL protein levels in cell lysates of NL (n = 3) and SSc (n = 3) pLFs in passage 3 were analysed by immunoblotting. GAPDH was used as a loading control for densitometry analysis. (D) Fold change in baseline COL18A1 gene expression levels were quantified using qPCR comparing NL (n = 7) vs SSc (n = 8) derived pLFs (normalized to B2M). (E) EV content of CTSL and CD81 in NL (n = 3) and SSc (n = 3) was analysed by immunoblotting (representative blot shown), and CTSL/CD81 densitometry ratio was measured. Gene expression levels of (F) CTSL and (G) COL1A1 were quantified using qPCR (n = 5, 3 independent SSc donors), and fold-change estimates were calculated to compare SIS3-treated pLFs with the DMSO-treated pLFs (normalized to B2M). (H) CTSL protein levels were analysed by immunoblotting in CM, cell lysates and EV lysates. GAPDH was used as a loading control for cell lysates, and CD81 as a reference for EV lysates, to which relative densitometry analysis was performed (n = 3, two independent SSc donors). Graphical presentation of the data analysed by paired sample t-test for SIS3- vs DMSO-treated pLFs, and unpaired sample t-test for NL vs SSc pLFs. Bars represent mean (s.e.m.). *P < 0.05, ***P < 0.001. pLF: primary human lung fibroblast; CTSL: Cathepsin L; EV: extracellular vesicles; COL18A1: Collagen type XVIII alpha 1 chain; B2M: β2-microglobulin; CM: conditioned media; NL: normal lung; GAPDH: glyceraldehyde-3-phosphate dehydrogenase.

Fig. 6

SSc-derived lung tissues have lower levels of CTSL than normal lung tissues (A) CTSL gene expression levels in TGF-β-treated NL tissues were quantified using qPCR, and fold-change estimates were calculated to compare TGF-β to vehicle (normalized to PPIB). (n = 4, four donors). (B) Representative images of immunofluorescence staining of NL vs SSc sections showing CTSL expression (green) and nuclear DAPI staining (blue) (scale = 200 μm). (C) Normalization of CTSL intensity to DAPI count as a representation of CTSL expression relative to tissue amount per field (fold change). Each colour represents a donor. Three fields were used from each of n = 5 independent NL donors and n = 5 independent SSc donors. (D) Representative images of co-immunofluorescence staining of NL and SSc sections showing CTSL expression (green), α-SMA (red) or (E) pro-SPC (red) and nuclear DAPI staining (blue) (Scale = 20 μm). Bars in panels (A) and (C) represent mean (s.e.m.). **P < 0.01, ****P < 0.0001. CTSL: Cathepsin L; COL18A1: Collagen type XVIII alpha 1 chain; PPIB: Peptidylprolyl Isomerase B; NL: normal lung.