Ufm1-Specific Ligase Ufl1 Regulates Endoplasmic Reticulum Homeostasis and Protects Against Heart Failure

Abstract

Background: Defects in protein homeostasis are sufficient to provoke cardiac remodeling and dysfunction. Although posttranslational modifications by ubiquitin and ubiquitin-like proteins are emerging as an important regulatory mechanism of protein function, the role of Ufm1 (ubiquitin-fold modifier 1)-a novel ubiquitin-like protein-has not been explored in either the normal or stressed heart.

Methods and results: Western blotting revealed that Ufl1 (Ufm1-specific E3 ligase 1)-an enzyme essential for Ufm1 modification-was increased in hypertrophic mouse hearts but reduced in the failing hearts of patients with dilated cardiomyopathy. To determine the functional role of Ufl1 in the heart, we generated a cardiac-specific knockout mouse and showed that Ufl1-deficient mice developed age-dependent cardiomyopathy and heart failure, as indicated by elevated cardiac fetal gene expression, increased fibrosis, and impaired cardiac contractility. When challenged with pressure overload, Ufl1-deficient hearts exhibited remarkably greater hypertrophy, exacerbated fibrosis, and worsened cardiac contractility compared with control counterparts. Transcriptome analysis identified that genes associated with the endoplasmic reticulum (ER) function were dysregulated in Ufl1-deficient hearts. Biochemical analysis revealed that excessive ER stress preceded and deteriorated along with the development of cardiomyopathy in Ufl1-deficient hearts. Mechanistically, Ufl1 depletion impaired (PKR-like ER-resident kinase) signaling and aggravated cardiomyocyte cell death after ER stress. Administration of the chemical ER chaperone tauroursodeoxycholic acid to Ufl1-deficient mice alleviated ER stress and attenuated pressure overload-induced cardiac dysfunction.

Conclusions: Our results advance a novel concept that the Ufm1 system is essential for cardiac homeostasis through regulation of ER function and that upregulation of myocardial Ufl1 could be protective against heart failure.

Keywords: endoplasmic reticulum stress; heart failure; hypertrophy; ubiquitin-like proteins.

Figure 1.. Expression of Ufm1 and Ufl1 in developing and diseased hearts.

A, Western blots of Ufm1 and Ufl1 in the hearts of C57bl/6 mice at 1 day (1d), 3 weeks (3w), 2 months (2m) and 1 year (1yr) of age. Total protein (Pro.) stain serves as loading control. B, Representative Western blot (left) and densitometric quantification (right) of ufmylated and Ufl1 proteins in mouse hearts at 2 weeks post sham (n=4)- or transverse aortic constriction (TAC, n=6)-operation. C, Western blots (top) and quantification (bottom) of Ufl1 in mouse hearts subject to sham or ischemia-reperfusion (I/R) surgery (30min ischemia followed by 24hr reperfusion. n=3 per group). D, Immunostaining of Ufl1 (green) in I/R mouse hearts. Evans blue dye (EBD, red) was injected into the mice 24 hrs prior to the surgery to identify ischemic CMs that lost membrane integrity following I/R. Bar, 50 μm. E, Representative Western blots of Ufl1 in the hearts of patients with dilated cardiomyopathy (DCM, n=4) or without cardiovascular disorders (non-HF, n=4). * P<0.05.

Figure 2.. Generation of Ufl1^CKO mice.

A , Targeting strategy of genomic region of Ufl1 gene. Cardiac-specific knockout of Ufl1 was achieved by intercrossing of Ufl1^flox mice with αMHC-Cre transgenic (αMHC^Cre) mice. B-C, Detection of Ufl1 mRNA (B) and protein levels (C) in the hearts of 2-month-old CTL (Ufl1^flox/+ or Ufl1^f/f, n=4), heterozygous (αMHC^Cre:Ufl1^f/+, Het, n=4) and Ufl1^CKO (αMHC^Cre:Ufl1^f/f, KO, n=6) mice by qRT-PCR and Western blot, respectively. The quantification of (C) is shown under the blots. D, Representative Western blot (top) and the quantification (bottom) of ufmylated proteins in CTL (n=4) and KO (n=4) hearts. Arrow, a significantly reduced ufmylated band. * P<0.05, *** P<0.001 versus CTL.

Figure 3.. Ufl1^CKO causes dilated cardiomyopathy and heart failure.

A, Heart weight (Hw) to tibial length (TL) ratio. B, qRT-PCR detection of the indicated fetal genes. C, Representative M-mode echocardiographs of 6-month-old CTL and KO mice. Left ventricle (LV) chamber dimensions at systole and diastole (double head arrows) are marked, respectively. D, Temporal echocardiography analysis of CTL and KO mice at the indicated ages. LV systolic or diastolic internal diameter (LVIDs or LVIDd), LV diastolic posterior wall thickness (LVPWd), and LV ejection fractions (EF) are shown. E-G, Gross morphology (E) of the hearts from 6-month-old CTL and Ufl1^CKO (KO) mice and hematoxylin and eosin (H&E) (F) and Masson’s trichrome (G) staining of the myocardium sections. Bars in (E-G) represent 1 mm, 1 mm and 100 μm, respectively. H, Quantification of myocardial fibrotic area in 6-month-old mice. *P<0.05, **P<0.01, ***P<0.001 vs CTL at the respective age.

Figure 4.. Ufl1^CKO hearts were more susceptible to TAC-induced pathological cardiac remodeling and heart failure.

Two-month-old CTL and Ufl1^CKO mice were subjected to either sham (n=5 for CTL and 5 for KO) or TAC (n=8 for CTL and 11 for KO) operation for 2 weeks. A-B, Temporal echocardiography analysis of LV functions. Representative M-mode images at 2 weeks post operation (A), the pressure gradients across constricted aorta and the temporal changes of LV morphometry and function (B) are shown. *P<0.05, **P<0.01, ***P<0.001 vs CTL+TAC at indicated times. C-E, Representative images of H&E (C) and Masson’s trichrome (D) stained myocardium sections. Bars in C-D are 1 mm, and 100 μm, respectively. E, Ratios of heart weight (Hw) to tibial length (TL) and lung weight to TL. F-G, The quantification of fibrotic area (F) and Mhy7 mRNA expression (G). *P<0.05.

Figure 5.. Global transcriptome analysis of Ufl1^CKO hearts.

RNAs isolated from 2-month CTL (n=4) and Ufl1^CKO (n=3) hearts were used for RNA sequencing analysis. A, Illustration of RNA sequencing analysis. B, Gene clustering by Z-score showing consistency within groups. C, KEGG (Kyoto Encyclopedia Genes and Genomes) analysis shows significantly enriched pathways in Ufl1^CKO hearts. Gene set related to ER homeostasis is highlighted. D, Volcano plot of downregulated (green) and upregulated (red) transcripts in Ufl1^CKO hearts. ER-related, differentially expressed transcripts, as well as Ufl1 transcript, are marked by dark dots and blue triangle, respectively. Transcripts that were not differentially expressed are in gray. E, qPCR confirmation of mRNA expression of ER-related, differentially expressed genes (n=4). *P <0.05 vs CTL. n.s., not significant. u.d., not detected.

Figure 6.. Elevated ER stress in the Ufl1^CKO hearts.

A-D, Western blots (A, C) and the quantification (B, D) of the indicated ER stress-responsive proteins in heart extracts from 2- and 6-month-old mice (A, B. n=3) or 2-month-old mice subjected to sham or TAC for 2 weeks (C, D. n=4). In (A), an anti-KDEL antibody (Santa Cruz, #58774) was used to simultaneously detect Grp94 and Bip. In (C), an anti-Bip antibody (Cell Signal, #3183S) was used to specifically detect Bip.*, a non-specific band. E, Electron micrographs of ventricular sections from 4-month-old mice. Abnormal ER cisternae (arrows) are shown. Three mice per group were analyzed. F-G, Representative images (F) and the quantification (G) of TUNEL+ (red, arrows) cardiomyocytes in myocardium sections (n=3 for sham group, n=4 for TAC group). Cardiomyocytes and nuclei were counterstained by phalloidin (green) and DAPI, respectively. Bar, 50 μm.

Figure 7.. Ufl1 modulates unfolded protein response and protects against ER stress-induced apoptosis.

Neonatal cardiomyocytes were transfected with siRNAs against luciferase or Ufl1 for 48 hr and then treated with indicated ER stress inducers for another 6 or 12 (A-B) and 48 hr (C-D). TM, tunicamycin (1 μg/mL). TG, thapsigargin (1 μM). Shown are representative results of three repeats. A-B, Western blots of indicated UPR transducers and downstream targets (A) and the quantification (B). N=3 per group for each repeat. C, Western blots of indicated apoptosis markers. Densitometric quantifications are denoted under the blots. N=2 per group for each repeat. D, Cell viability measured by CellTiter-Glo assay. N=4 per group for each repeat. *P<0.05, **P<0.01 vs siLuci at respective time points.

Figure 8.. Inhibition of ER stress attenuates cardiac dysfunction in Ufl1^CKO hearts after pressure overload.

Mice at 2 months of age were subjected to TAC surgery and immediately received TUDCA (250 mg/kg/day, n=5 for CTL and 7 for KO) or vehicle (n=5 for CTL and 8 for KO) through daily intraperitoneal injections for 2 weeks. A-C, Gross morphology (A), heart weight to body weight (Hw/Bw) ratio (B) and lung weight to body weight (Lung/Bw) ratio (C) of indicated mice. D, Echocardiography analysis of cardiac function. E, Quantification of TUNEL+ cardiomyocytes. n=3 mice per group. F-G, Western blot analysis of PDI in heart extracts (F) and the quantification (G). n=3 mice per group. *P<0.05, **P<0.01, ***P<0.001.

Figure 9.. A model illustrating the role of ufmylation in the heart.

Protein ufmylation is mediated by the Ufm1-specific enzymes UBA5, UFC1 and Ufl1, and is dysregulated under pathological conditions. Myocardial Ufl1 is indispensable for the maintenance of normal heart function and protect against pressure overload-induced cardiac remodeling and heart failure, which is in part through ensuring ER homeostasis by sustaining adaptive PERK signaling.