ADAMTS14 cleaves collagen V to keep matrix stable; its loss weakens YAP nuclear translocation and IPF fibroblast activation

An unbiased siRNA screen (7,638 siRNA pools) identified ADAMTS14 as a regulator of YAP nuclear translocation and confirmed in deconvolution that it also modulates the fibroblast TGFβ response. ADAMTS14 was previously linked mainly to collagen I maturation; this work connects it to YAP-mediated mechanoactivation. High-throughput screening plus deconvolution validation, with positives defined as a translocation score ≥21.5 (FDR 1%) in ≥2 independent siRNA duplexes, narrowing 230 validated hits to 23 priority targets and selecting ADAMTS14.

In IPF patient lung single-cell and spatial transcriptomics, ADAMTS14 was enriched in CTHRC1-high myofibroblasts and localized to fibroblastic foci and other actively fibrotic niches, correlating with ECM production and matrix stiffness gene sets. It anchors the in vitro screen result to a specific cell subpopulation and spatial location in human IPF tissue. Based on a published scRNA-seq atlas (620,644 cells, 49 annotated cell types) and a spatial transcriptomics dataset (13 ILD, 11 non-ILD controls, ~50k Visium spots, 30,988 manually annotated); in ADAMTS14-positive CTHRC1-high myofibroblasts COL1A1 (log2FC 94.8) and COL1A2 (log2FC 63.1) were the top upregulated genes, with enrichment of YAP/TAZ, matrisome, and IPF signature gene sets.

In lung fibroblasts from five IPF donors, ADAMTS14 knockdown reduced CYR61, CTGF, COL1A1 and TGFβ-induced ACTA2 expression and attenuated TGFβ pathway activity; in the Scar-in-a-Jar model TGFβ raised collagen I and αSMA 25- and 41-fold in controls, and knockdown reduced both by 77% and 78%. It shows ADAMTS14's collagen-processing function directly determines fibroblast susceptibility to TGFβ, rather than acting only through non-collagenous modulation of TGFβ bioavailability. siRNA knockdown in five IPF donors with qPCR and bulk RNA-seq (DESeq2, 20,095 genes with multiple-testing correction), plus macromolecular crowding assays in five additional donors, analyzed by two-way ANOVA with Tukey/Dunnett correction.

Co-immunoprecipitation mass spectrometry identified collagen V (COL5A1, COL5A2) as the most enriched ADAMTS14 substrate; ADAMTS14 deficiency reduced cleavage at ADAMTS14-specific sites (PANQ(438).(439)DTIY in COL5A1 and GPPG(906).(907)ATGF in COL5A2), causing accumulation of immature collagen V, shifting the COL1:COL5 ratio from about 50:1 to about 20:1, with more numerous but disorganized focal adhesions, reduced FAK and AKT1 phosphorylation, rounder cell shape, and about 50% reduced durotaxis. It proposes an ADAMTS14–collagen V–focal adhesion axis as a feed-forward circuit linking extracellular proteolytic remodeling to focal adhesion dynamics and YAP mechanoactivation. coIP-MS (4 samples from 2 independent experiments) plus cell-derived matrix proteomics (4,208 protein hits, 208 matrisome proteins, PCA explaining 82% of variance) corroborated by Brillouin microscopy, SHG two-photon imaging, gel contraction, and durotaxis functional assays.

The results apply to human IPF-derived lung fibroblasts and patient lung tissue samples, providing mechanistic support for an anti-fibrotic strategy targeting ADAMTS14 to partially rather than fully inhibit YAP activation; for researchers interested in matrix-cell mechanical coupling, focal adhesion signaling, or post-translational collagen processing, it serves as a mechanistic framework and candidate target.

The authors note they have not demonstrated whether ADAMTS14 has pro-fibrotic effects independent of YAP, nor directly tested its effect on the YAP paralog TAZ; in mice ADAMTS14 is functionally redundant with ADAMTS2, so ADAMTS14-deficient mice showed no overt phenotype in lung fibrosis models, suggesting species differences or compensation; additionally, the inference that the altered COL1:COL5 ratio reduces matrix mechanical properties rests mainly on prior literature rather than direct measurement in this study.