Deacetylation via SIRT2 prevents keratin-mutation-associated injury and keratin aggregation
Jingyuan Sun, Pei Li, Honglian Gui, Laure Rittié, David B. Lombard, Katrin Rietscher, Thomas M. Magin, Qing Xie, Li Liu, M. Bishr Omary
Jingyuan Sun, Pei Li, Honglian Gui, Laure Rittié, David B. Lombard, Katrin Rietscher, Thomas M. Magin, Qing Xie, Li Liu, M. Bishr Omary
View: Text | PDF
Research Article Dermatology Hepatology

Deacetylation via SIRT2 prevents keratin-mutation-associated injury and keratin aggregation

Abstract

Keratin (K) and other intermediate filament (IF) protein mutations at conserved arginines disrupt keratin filaments into aggregates and cause human epidermolysis bullosa simplex (EBS; K14-R125C) or predispose to mouse liver injury (K18-R90C). The challenge for more than 70 IF-associated diseases is the lack of clinically utilized IF-targeted therapies. We used high-throughput drug screening to identify compounds that normalized mutation-triggered keratin filament disruption. Parthenolide, a plant sesquiterpene lactone, dramatically reversed keratin filament disruption and protected cells and mice expressing K18-R90C from apoptosis. K18-R90C became hyperacetylated compared with K18-WT and treatment with parthenolide normalized K18 acetylation. Parthenolide upregulated the NAD-dependent SIRT2, and increased SIRT2-keratin association. SIRT2 knockdown or pharmacologic inhibition blocked the parthenolide effect, while site-specific Lys-to-Arg mutation of keratin acetylation sites normalized K18-R90C filaments. Treatment of K18-R90C–expressing cells and mice with nicotinamide mononucleotide had a parthenolide-like protective effect. In 2 human K18 variants that associate with human fatal drug-induced liver injury, parthenolide protected K18-D89H– but not K8-K393R–induced filament disruption and cell death. Importantly, parthenolide normalized K14-R125C–mediated filament disruption in keratinocytes and inhibited dispase-triggered keratinocyte sheet fragmentation and Fas-mediated apoptosis. Therefore, keratin acetylation may provide a novel therapeutic target for some keratin-associated diseases.

Authors

Jingyuan Sun, Pei Li, Honglian Gui, Laure Rittié, David B. Lombard, Katrin Rietscher, Thomas M. Magin, Qing Xie, Li Liu, M. Bishr Omary

×

Figure 1

Parthenolide (PN) normalizes K18-R90C filament disruption and protects against Fas-induced apoptosis in cultured cells and mice.

Options: View larger image (or click on image) Download as PowerPoint
Parthenolide (PN) normalizes K18-R90C filament disruption and protects a...
(A) A549 cells were transduced with lentivirus expressing GFP-tagged K18-R90C and K8-WT (cotransfection with type I and II keratins is needed for physiologic heteropolymer filament formation). After 2 days, cells were treated with DMSO or PN (5 μM, 48 hours) and then fixed and stained with DAPI. Average percentage of cells with dots/group for DMSO vs. PN is shown. P < 0.001, from 3 independent experiments). Scale bar: 50 μm. (B) GFP-K18-WT and GFP-K18-R90C lentiviruses were transduced into A549 cells followed by treatment with DMSO or PN (5 μM, 48 hours) and then apoptosis was induced using IFN-γ and Fas ligand (Fas-L). Cell lysates were analyzed by blotting using antibodies against the indicated antigens (Coomassie stain shows equal loading). The average relative intensity of the indicated bands from 3 individual experiments is included below each blot. *P < 0.05, ***P < 0.001 using unpaired, 2-tailed Student’s t test (comparing DMSO with PN). (C) Representative TUNEL staining of GFP-K18-R90C–transduced A549 cells treated with DMSO or PN and then challenged with IFN-γ plus Fas-L. The percentage of TUNEL+ cells in DMSO vs. PN was 82% vs. 33%, respectively (P < 0.001). Scale bar: 50 μm. For panels A–C, all experiments were repeated at least 3 times. *P < 0.05, **P < 0.005 using unpaired, 2-tailed Student’s t test. (D) Transgenic mice that express K18-R90C were treated daily with DMSO or PN (1 mg/kg mouse weight; i.p.) for 4 days. Liver sections were double stained with anti-K18 antibody and DAPI. The percentage of cells with dots/group is shown in Supplemental Figure 1F (P < 0.001). Scale bar: 50 μm. (E) Mice were treated with DMSO or PN as in panel D and then challenged with Fas-L. Livers were collected after 5 hours and analyzed by immunoblotting (each lane represents a different liver). (F) Alanine transaminase (ALT) was measured using sera collected from the indicated number of mice. **P = 0.002 using 1-way ANOVA followed by Tukey’s post hoc test. (G) Representative images of hematoxylin and eosin–stained liver sections. Arrows highlight areas of hemorrhage. Scale bar: 100 μm. (H) Quantification of the hemorrhage after Fas-L treatment of the mice is included. *P = 0.013 using unpaired, 2-tailed Student’s t test. For each group (DMSO or PN), 10 sections for each liver and 7 independent livers were used.