ADAMTSL3 knock-out mice develop cardiac dysfunction and dilatation with increased TGFβ signalling after pressure overload

Figures

• 

  Fig. 1. ADAMTSL3 is upregulated in patients with heart failure and in wild-type mice with left ventricular pressure overload, and is produced by cardiac fibroblasts. a-b ADAMTSL3/RPL32 mRNA levels in left ventricular (LV) biopsies from patients with a aortic stenosis (AS, n = 11) and b ischemic dilated cardiomyopathy (iDCM, n = 9) vs. respective controls. c Western blot and quantification of ADAMTSL3 protein/total protein levels in LVs of patients with iDCM (n = 8) vs. controls (n = 7). d Adamtsl3/Rpl32 mRNA levels in LVs of wild-type (WT) mice 1 and 6 weeks after aortic banding (AB) or sham surgery (n = 5 sham and n = 7 AB at 1 week, n = 8 sham and n = 12 AB at 6 weeks). e In situ hybridization of Adamtsl3 mRNA (red staining) in the left atrium of AB and sham-operated mice, and in the LV wall of AB mice, 4 weeks post-AB, with hematoxylin nuclear counterstain (purple). f RNA sequencing data from the Genotype-Tissue Expression (GTEx) project portal. In GTEx, a total of 17382 samples across 54 non-diseased tissues, from 948 deceased human organ donors, represents a control population. Donors are 20-70 years old, 33% female, 85% White and 13% African-American. ADAMTSL3 expression is shown…

• 

  Fig. 2. Adamtsl3 inactivation results in contractile dysfunction and cardiac dilatation with high mortality in response to pressure overload. a–b Generation of the Adamtsl3 mutant allele. a Overview of the Adamtsl3 locus showing coding and non-coding exons. CRISPR-Cas9 mediated editing resulted in a 5 bp deletion in exon 2. b Sequence around the 5 bp deletion target site. The underlined sequences indicate the primers used for genotyping. c Schematic timeline of mouse study. Wild-type (WT) and Adamtsl3 knock-out (L3-KO) mice underwent aortic banding (AB) or sham surgery, and were followed with echocardiography and magnetic resonance imaging (MRI) for 6 weeks. d Kaplan-Meier survival curves post-AB. e Body weight (BW) change from baseline (%), in WT and L3-KO post-AB or sham surgery (n = 8 WT and n = 12 L3-KO). f Absolute heart weight (HW) post-AB. g Left ventricle mass (LVM) post-AB. h Interventricular septal thickness in diastole (IVSd), i LV posterior wall thickness in diastole (LVPWd), j LV inner diameter in diastole (LVIDd), at baseline, and post-AB compared to sham. k LV end-diastolic volumes (LVEDV) post-AB. l Fractional shortening (FS) at baseline and post-AB. m LV ejection…

• 

  Fig. 3. RNA sequencing highlights cardiomyopathy, TGFβ, and cardiomyocyte dysfunction pathways in Adamtsl3 knock-out hearts in response to pressure overload. Pressure overload of the left ventricle (LV) was induced in Adamtsl3 knock-out (L3-KO, n = 10) and wild-type (WT, n = 6) mice by aortic banding (AB) for 1 week, and RNA sequencing was performed on LV tissue. 233 differentially expressed genes (DEGs, FDR < 0.05) were analysed for enrichment of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO) terms using Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.8 analysis wizard49. The 10 most enriched a KEGG pathways, b GO biological pathways, c GO molecular functions, and d GO cellular components, identified among the 233 DEGs are shown. e Scatterplot comparing log reads per kilobase million (RPKM) of 233 DEGs between L3-KO and WT. Red dots correspond to annotated DEGs in the graph. The complete list of DEGs (gray and red dots) can be found in Supplementary Data 1. f-g Log2-fold change of mean RPKM of DEGs related to cardiomyocyte (CM) structure and function (f), and extracellular matrix (ECM) and cardiac fibrosis (g). h-i Predicted…

• 

  Fig. 4. ADAMTSL3 modulates cardiac TGFβ signalling in response to increased LV afterload in vivo and in human cardiac fibroblasts in vitro. a–c Wild-type (WT) and Adamtsl3 knock-out (L3-KO) mice were subjected to aortic banding (AB) or sham surgery for a total of 6 weeks. a,b Representative immunoblots and quantifications of active TGFβ1/total protein (a) and phosphorylated SMAD2 (pSMAD2)/total SMAD2 (b) in LV lysates, 1 week (sham: n = 4 WT and n = 4 KO, AB: n = 6-8 WT and n = 12 KO) and 6 weeks (sham: n = 4 WT and n = 4 KO, AB: n = 12 WT and n = 8 KO) post-AB. c mRNA/Rpl32 levels of TGFβ (Tgfb1), connective tissue growth factor (Ctgf), periostin (Postn) and latent TGFβ binding protein 1 (Ltbp1) in AB mice vs sham 6 weeks post-AB (Sham: n = 8 WT and n = 8 KO, AB: n = 10 WT and n = 7 KO). Statistical analyses waere performed using the one-way ANOVA with Tukey’s multiple comparisons test, with *p < 0.05, **p < 0.01, and ***p < 0.001 for AB vs. sham, and $p < 0.05, $$p < 0.01, and $$$p < 0.001 for L3-KO AB vs. WT AB. d-k Human foetal cardiac fibroblasts cultured for 7 days, producing a rich extracellular matrix (ECM) network, and transduced with ADAMTSL3 (L3) or control (vehicle, Veh)…

• 

  Fig. 5. ADAMTSL3 regulates deposition of insoluble collagen in the heart, and production of ECM molecules by cardiac fibroblasts in vitro. a–f Wild-type (WT) and Adamtsl3 knock-out (L3-KO) mice were subjected to aortic banding (AB) or sham surgery for 1 and 6 weeks. a Left ventricle (LV) % total collagen, calculated from Picrosirius Red, Fast Green, and Alcian Blue (RGB)-stained histology mid-ventricular sections at 1 week post-AB (Sham: n = 5 WT and n = 5 KO, AB: n = 5 WT and n = 5 KO). b LV type-I collagen content, as measured by peak hydroxyproline HPLC, 1-week post-AB (Sham: n = 6 WT and n = 5 KO, AB: n = 6 WT and n = 7 KO). c LV mRNA/Rpl32 levels of collagens type I (Col1a1, Col1a2) and type III (Col3a1) and lysyl oxidase (Lox) (Sham: n = 8 WT and n = 8 KO, AB: n = 10 WT and n = 7 KO). d LV % total collagen in mid-ventricular sections at 6 weeks post-AB (Sham: n = 3 WT and n = 3 KO, AB: n = 4 WT and n = 4 KO). e Levels of total collagen, soluble collagen, insoluble collagen and collagen crosslinking in LVs 6 weeks post-AB, using colorimetric and enzymatic procedures on mid-ventricular sections (n = 7 WT and n = 7 KO). (f) mRNA/Rpl32 levels of fibrillin-1/2 (Fbn1, Fbn2), tropoelastin…

• 

  Fig. 6. ADAMTSL3 regulates cardiac myofibroblast differentiation in vivo and in vitro. a–d Wild-type (WT) and Adamtsl3 knock-out (L3-KO) mice were subjected to aortic banding (AB) or sham surgery for 1 and 6 weeks. a mRNA/Rpl32 levels of α-smooth muscle actin (α-SMA, Acta2). b Representative immunoblots and quantification of α-SMA in LV protein extracts of WT and L3-KO 1 and 6 weeks post-AB or sham surgery. c mRNA/Rpl32 levels of osteopontin (OPN, Spp1). d Representative immunoblots and quantification of OPN in LV protein extracts of WT and L3-KO 1 and 6 weeks post-AB or sham surgery. At 1-week post-AB, n = 4-5 WT and n = 4 KO sham, and n = 6–7 WT and n = 8–9 KO AB. At 6 weeks post-AB, n = 4-8 WT and n = 4-8 KO sham, and n = 10-12 WT and n = 7-8 KO AB. Statistical analyses were performed using the one-way ANOVA with Tukey’s multiple comparisons test (a-d). P-values are reported as *p < 0.05, **p < 0.01, and ***p < 0.001 for AB vs. sham, and $p < 0.05, $$p < 0.01, and $$$p < 0.001 for L3-KO AB vs. WT AB. e–k Human foetal cardiac fibroblasts (hfCFBs) were cultured for 7 days, producing a rich extracellular matrix (ECM) network, and transduced with ADAMTSL3 (L3) or control (vehicle, Veh)…