Plectin elevation in tumor hepatocytes is associated with HCC progression and poor prognosis.

(A) Meta-analysis of differential plectin (PLEC) mRNA expression in non-tumor (NT) liver and hepatocellular carcinoma (HCC) patients. Blue squares indicate the standardized mean difference (SMD) and 95% confidence interval of individual datasets. The black diamond shows the mean and 95% confidence interval for the combined SMD, while the whiskers indicate the 95% prediction interval. (B) Kaplan-Meier curve of recurrence-free survival of HCC patients with low PLEC (lower 2 tertiles, n = 219) and high PLEC (top tertile, n = 108) mRNA expression. Log-rank test; P < 0.05. (C) Representative images of human HCC tissue sections immunolabeled for plectin (green). Nuclei, DAPI (blue). Dashed line, the borderline between non-tumor (NT) and tumor (T) area. Boxed areas, x4 images. Scale bars, 200 and 100 µm (boxed areas). Boxplot shows quantification of plectin fluorescence intensities in NT and T areas. The box represents the median, 25th, and 75th percentile; whiskers reach the last data point; dots, individual patients; N = 19. Paired two-tailed t-test; †P < 0.001. (D) Immunoblot analysis of indicated HCC cell lines with antibodies to plectin, E-cadherin, and vimentin. GAPDH, loading control. (E) Quantification of the speed of indicated HCC cell lines migrating in the scratch-wound assay. Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, fields of view; n = 15 (Huh7), 13 (HepG2), 15 (Hep3B), 15 (PLL), 15 (SNU-398), 15 (SNU-423), 15 (SNU-475) fields of view; N = 3. (F) Hepatocarcinogenesis was induced in two-week-old C57Bl/6J mice by intraperitoneal injection of DEN. Representative image of the livers with multifocal HCC at 46 weeks post-induction. Scale bar, 1 cm. (G) Representative image of DEN- induced HCC section immunolabeled for plectin (green). Nuclei, DAPI (blue). Dashed line, the borderline between non-tumor (NT) and tumor (T) area. Boxed areas, x2 images. Scale bars, 200 and 100 µm (boxed areas). Quantification of plectin fluorescence intensities in NT and T areas. Boxplot shows the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, fields of view; n = 16 fields of view; N = 4. Paired two-tailed t-test; **P < 0.01. (H) Immunoblot analysis of NT and T liver lysates. The boxplot shows relative plectin band intensities normalized to GAPDH. The box represents the median, 25th, and 75th percentile; whiskers reach the last data point; dots, individual mice; N = 8.

Plectin promotes HCC growth.

(A) Representative MRI images of Plefl/fl and PleΔAlblivers at 32 and 44 weeks post-DEN injection. Dashed circles, tumors. Scale bar, 500 µm. (B, C) Quantification of tumor number (B) and volume (C) in Plefl/fl and PleΔAlb livers shown in (A). Boxplot shows the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual mice; N = 6 (WT), 5 (KO). Two-tailed t-test; *P < 0.05. (D) Representative images of Plefl/fl and PleΔAlblivers at 44 weeks post-induction. Dashed circles, tumors. Scale bar, 1 cm. (E, F) Quantification of the number (E) and size (F) of Plefl/fl and PleΔAlb tumors shown in (D). Boxplot shows the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual mice; N = 8 (WT), 3 (KO). Two-tailed t-test; *P < 0.05; †P < 0.001. (G) Representative images of colonies from WT, KO, ΔIFBD, and PST-treated WT (WT+PST) Huh7 and SNU-475 cells grown in soft agar. Scale bar, 500 µm. Boxplots show the number of Huh7 (upper graph) and SNU-475 (lower graph) cell colonies. The box represents the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, agar wells; n = 9 agar wells; N = 3. Two-tailed t-test; **P < 0.01; †P < 0.001. (H) Six-week-old NSG mice were subcutaneously injected with indicated Huh7 cells and were kept either untreated (WT, KO, and ΔIFBD) or daily treated by orogastric gavage of plecstatin (WT+PST) as indicated in upper bar. Mice were sacrificed 4 weeks post-injection and xenografts were dissected. The graph shows the time course of xenograft growth. Data are shown as mean ± SEM; N = 4 (WT), 4 (KO), 5 (ΔIFBD) and 6 (WT+PST). Two-way ANOVA. The table shows the number (N), percentage, and representative images of formed xenografts. Scale bar, 2 cm.

CRISPR/Cas9- or PST-mediated plectin inactivation attenuates HCC oncogenic potential through FAK, Erk1/2, and PI3K/Akt axis.

(A) Schematic of MS-based proteomic analysis of WT, KO, and PST-treated WT (WT+PST) SNU-475 cells. (B,C) Ingenuity Pathway Analysis (IPA) canonical signaling pathways predicted from differentially expressed proteins identified by proteomics (left) and phosphoproteomics (right) in WT vs. KO (B) and WT vs. WT+PST (C) proteomes. Venn diagrams show relative proportions of differentially expressed proteins. Two-sided Student’s t-test with multiple testing correction: FDR < 0.05; s0 = 0.1; triplicates. (D) Quantification of FAK, phospho-Tyr397-FAK (pFAK), Akt, phospho-Ser473-Akt (pAkt), Erk1/2, and phospho-Thr202/Tyr204-Erk (pErk) in indicated Huh7 and SNU-475 cell lines by immunoblotting. GAPDH, loading control. The numbers below lines indicate relative band intensities normalized to average WT values. Boxplots show relative band intensities normalized to GAPDH or non-phosphorylated protein. The box represents the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual experiments; N = 9. Two-tailed t-test; *P < 0.05; **P < 0.01; †P < 0.001. (E) Schematic representation of immunoblot analyses of adhesome-associated signaling shown in (F) and (Extended Data Fig. 2h). Proteins with significantly reduced expression levels and/or phosphorylation status (P) upon plectin inactivation in both HCC cell lines are highlighted in red, proteins with significantly reduced expression levels upon plectin inactivation in either Huh7 or SNU-475 cells are highlighted in pink.

Disruption of cytoskeletal networks upon plectin inactivation accounts for reduced contractility and aberrant adhesions in HCC cells.

(A) Representative confocal images of crossbow-shaped fibronectin micropattern-seeded WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells stained for F-actin (red), vinculin (green), and vimentin (grey). Nuclei, DAPI (blue). Arrows, dorsal stress fibers; arrowheads, ventral stress fibers. Scale bar, 10 μm. (B) Quantification of the percentage of cells (shown in (A)) with well-formed, bundled, and clump-containing vimentin networks. Data are shown as mean ± SEM; n = 60 (WT), 68 (KO), 55 (ΔIFBD), 50 (WT+PST) cells; N = 4 (WT, KO, IFBD), 3 (WT+PST). (C,D) Quantification of the number of dorsal (C) and ventral (D) actin stress fibers in cells shown in (A). Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual cells; n = 60 (WT), 68 (KO), 55 (ΔIFBD), 50 (WT+PST); N = 4 (WT, KO, IFBD), 3 (WT+PST). Two-tailed t-test; †P < 0.001. (E) Quantification of FAs located within the interior of cells (central) shown in (A). Boxplot shows the median, 25th and 75th percentile with whiskers reaching the last data point; dots, individual cells; n = 25 (WT), 26 (KO), 23 (ΔIFBD), 28 (WT+PST); N = 3. **P < 0.01; †P < 0.001. (F) Representative confocal images of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells immunolabeled for vinculin (green). Nuclei, DAPI (blue). Boxed areas, representative FA clusters shown as segmented binary maps in x2 enlarged insets. Scale bar, 30 µm. (G) Quantification of FA number in cells shown in (F). Boxplot shows the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual cells; n = 15 (WT), 18 (KO), 20 (ΔIFBD), 19 (WT+PST); N = 3. Two-tailed t-test; †P < 0.001. (H) Pseudocolor spatial maps of contractile energy determined by TFM in WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells. Scale bar, 50 µm. (I) Quantification of contractile energy in cells shown in (H). Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual cells; n = 54 (WT), 53 (KO), 41 (ΔIFBD), 24 (WT+PST) cells; N = 4. Two-tailed t test; *P < 0.05; **P < 0.01; †P < 0.001.

Plectin links migration potential of HCC cells to cell shape dynamics.

(A) Representative phase contrast images of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells migrating in the scratch-wound assay for 14 hours. Note individual, highly polarized WT cells frequently migrating into scratch areas. Scale bar, 200 µm. (B) Quantification of migration speed of indicated Huh7 (upper graph) and SNU-475 (lower graph) cells. Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, fields of view; n (Huh7) = 59 (WT), 51 (KO), 58 (ΔIFBD), 43 (WT+PST); n (SNU-475) = 47 (WT), 47 (KO), 50 (ΔIFBD), 24 (WT+PST); N (Huh7) = 3; N (SNU-475) = 5 (WT, KO, ΔIFBD), 3 (WT+PST). Two-tailed t-test; † P < 0.001. (C) Representative confocal images of F-actin stained WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells grown on fibronectin-coated coverslips (2D) or in collagen (3D) and classified as round (#1), intermediate (#2), and polarized (#3) shape. Quantification of the percentage of cell shape categories in indicated 2D and 3D SNU-475 cell cultures. Data are shown as mean ± SEM; N (2D) = 3; N (3D) = 5 (WT), 3 (KO, ΔIFBD), 2 (WT+PST). (D) Spider plots with migration trajectories of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells tracked during 16 hours of EGF-guided migration; dots, the final position of each single tracked cell. (E) Quantification of processivity indices of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells. Boxplot shows the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual cells; n = 15 (WT), 15 (KO), 19 (ΔIFBD), 14 (WT+PST); N = 3. Two-tailed t-test; **P < 0.01;† P < 0.001. (F) Representative time sequences of the WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cell contours during EGF-guided migration. Color coding indicates the time of cell position acquired in 10-minute intervals. Scale bar, 20 µm. (G) Representative phase contrast image of SNU-475 cell with protrusions (green) segmented from superimposed contours used in morphodynamic analysis. Extension vectors (purple arrows) were drawn from the center of the cell nucleus towards individual protrusions and related to the direction of cell motion (black arrow). Scale bar, 20 µm. (H) Rose graphs show the percentage of extension vector directions in 30° cones, normalized to the directions of random and EGF-guided (directed) motions (0°; arrows) of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells. n = 9752 extensions in 22 cells (WT random), 4167 extensions in 15 cells (WT directed), 8394 extensions in 19 cells (KO random), 5107 extensions in 15 cells (KO directed), 8362 extensions in 21 cells (ΔIFBD random), 5809 extensions in 19 cells (ΔIFBD directed), 9450 extensions in 20 cells (WT+PST random), 4350 extensions in 14 cells (WT+PST directed); N = 3. Bar graphs show the percentage of cell extensions formed either in the direction of motion (frontal, 30° to −30° cones) or along the rest of the cell perimeter (rest). Data are shown as mean ± SEM; dots, biological replicates; N = 3. Two-tailed t-test; *P < 0.05; **P < 0.01;† P < 0.001.

Plectin inactivation inhibits HCC invasion and metastasis.

(A) Representative images of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 spheroids grown for 3 days in collagen mixture. Insets, superimposed binary masks of initial (red) and final (white) spheroid area. Scale bar, 200 µm. (B) Quantification of the invaded area calculated as the percentage of the initial spheroid area from day 0. Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, individual spheroids; n = 47 (WT), 44 (KO), 34 (ΔIFBD), 25 (WT+PST) spheroids; N = 5 (WT, KO), 4 (ΔIFBD), 3 (WT+PST). Two-tailed t-test; **P < 0.01; †P < 0.001. (C) Quantification of the number of indicated cells invaded in Matrigel transwell assay. Boxplots show the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, fields of view; n = 51 (WT), 45 (KO), 38 (ΔIFBD), 31 (WT+PST) fields of view; N = 4 (WT, KO), 3 (ΔIFBD, WT+PST). Two-tailed t-test; †P < 0.001. (D) Representative confocal micrographs of WT, KO, ΔIFBD, and PST-treated WT (WT+PST) SNU-475 cells grown on FITC-labeled gelatin (grey) for 24 hours and stained for F-actin (red). Nuclei, Hoechst (blue). Insets, segmented binary masks of FITC-gelatin signal. Black regions correspond to gelatin areas degraded by individual cells. Scale bar, 30 µm. (E) Representative confocal images of WT and KO SNU-475 cells during the Matrigel invasion assay, stained for plectin (magenta) and F-actin (yellow). Nuclei, Hoechst (blue). See Video 1. Boxed areas, x3 images. Scale bars, 100 and 30 µm (boxed areas). Boxplot shows the invaded area calculated as the percentage of the initial area covered by WT and KO cells. The box represents the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, fields of view; n = 29 fields of view; N = 2. Rose graphs show the percentage of extension vector directions in 30° cones, normalized to the directions of cell motions (0°; arrow) during matrigel invasion. n = 857 extensions in 18 cells (WT), 623 extensions in 12 cells (KO); N = 2. Two-tailed t-test; †P < 0.001. (F) Relative plectin (PLEC) mRNA expression in samples collected from HCC patient meta-cohort clustered based on TNM classification (stage I-IV). The meta-cohort includes 6 different datasets from 5 platforms (for details, see Materials and methods section). The numbers of participants per stage are indicated in the graph. Scattered boxplots show individual data points, median, 25th, and 75th percentile; N = 978. Wilcoxon rank-sum test; *P < 0.05; **P < 0.01. (G) The 5-week-old NSG mice were injected (t.v.i.) with WT and KO RedFLuc-GFP-expressing Huh7 cells generated for lung colonization assay. Kaplan-Meier curves show the overall survival of mice injected with the cells indicated. N = 14 (WT), 13 (KO). Long-rank test, P < 0.05. (H) The 5-week-old NSG mice were injected (t.v.i.) with indicated RedFLuc-GFP-expressing Huh7 cells. WT cell-bearing mice were kept either untreated or every second day provided with orogastric gavage of plecstatin (WT+PST) as indicated. Five weeks post-injection mice were screened by whole-body bioluminescence imaging (BLI). Representative BLI images of WT, KO, and PST-treated WT (WT+PST) SNU-475 cells-bearing mice are shown. Scale bar, 2 cm. (I) Representative images of lungs dissected from mice shown in (H). Scale bar, 1 cm. Representative lattice light sheet fluorescence image of clear, unobstructed brain imaging cocktails (CUBIC)-cleared lung lobe immunolabeled with antibodies against GFP (green). Autofluorescence visualizing the lobe structures is shown in red. Scale bar, 2 mm. Representative magnified images from lung lobes with GFP-positive WT, KO, and WT+PST Huh7 nodules. Insets, segmented binary masks of GFP-positive metastatic nodules. Scale bar, 400 µm. Boxplots show metastatic load in the lungs expressed as the number (upper graph) and relative volume (lower graph) of indicated GFP-positive (GFP+) nodules. The box represents the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, lung lobes; n = 8 lung lobes; N = 4. Two-tailed t-test; *P < 0.05; **P < 0.01; †P < 0.001.

Genetic and pharmacological plectin targeting prevents hepatocarcinogenesis through signatures shared by animal models and patients.

(A) HCC-predisposing lesions were introduced by hydrodynamic gene delivery via tail vein injection (HDTVi) of transposon vector encoding cMyc in conjunction with CRISPR/Cas9 construct targeting Tp53 (Myc;sgTp53) in Plefl/fl and PleΔAlb cohorts of 7-week-old male mice. Plefl/fl mice were kept either untreated or every second day provided with orogastric gavage of plecstatin (Plefl/fl+PST) and the development of HCC was monitored by MRI at 11, 13, and 16 week, as indicated. Representative MRI images of Plefl/fl and PleΔAlb and Plefl/fl+PST tumors acquired at indicated time points. Dashed circles, tumors. Scale bar, 2 cm. Graphs show the average number of tumors (upper graph) and percentual tumor incidence (lower graph) inferred from MRI images. N = 5 (Plefl/fl), 7 (PleΔAlb), 4 (Plefl/fl+PST). (B) Myc;sgTp53 HCC was induced in Plefl/fl, PleΔAlb, and PST-treated Plefl/fl(Plefl/fl+PST) male mice as in (A). Shown are representative images of Plefl/fl, PleΔAlb, and Plefl/fl+PST livers from mice with fully developed multifocal HCC sacrificed 6 weeks post-induction. Scale bar, 1 cm. Boxplots show tumor burden in the livers expressed as the liver/body weight ratio (left graph) and number of tumors per mouse (right graph). The box represents the median, 25th, and 75th percentile with whiskers reaching the last data point; dots, mice; N = 12 (Plefl/fl), 9 (PleΔAlb), 10 (Plefl/fl+PST). Two-tailed t-test; *P < 0.05. (C) Representative images of H&E-stained Plefl/fl, PleΔalb, and Plefl/fl+PST liver sections. Note darker areas corresponding to HCC lesions. Boxed areas, x12 images. Scale bars, 5 and 1 mm (boxed areas). (D) Gene set enrichment analysis of differentially regulated proteins in Plefl/fl vs PleΔAlb and Plefl/flvs Plefl/fl + PST livers from the cohort shown in (A). Prediction of canonical signaling pathways in Plefl/fl vs PleΔAlb (left) and Plefl/fl vs Plefl/fl + PST (right) proteomes. (E) Association of plectin-dependent signatures compiled from human HCC-derived cells (see Fig. 3B-E) and mouse models (see Fig. 7D) with plectin (PLEC) mRNA expression in HCC patients. Right panel shows the levels of selected signatures in patients grouped into quartiles of PLEC expression level. N = 1268. P values were generated from an analysis of variance (ANOVA).