Human molecular genetic and functional studies identify TRIM63, encoding Muscle RING Finger Protein 1, as a novel gene for human hypertrophic cardiomyopathy

Abstract

Rationale: A delicate balance between protein synthesis and degradation maintains cardiac size and function. TRIM63 encoding Muscle RING Finger 1 (MuRF1) maintains muscle protein homeostasis by tagging the sarcomere proteins with ubiquitin for subsequent degradation by the ubiquitin-proteasome system (UPS).

Objective: To determine the pathogenic role of TRIM63 in human hypertrophic cardiomyopathy (HCM).

Methods and results: Sequencing of TRIM63 gene in 302 HCM probands (250 white individuals) and 339 control subjects (262 white individuals) led to identification of 2 missense (p.A48V and p.I130M) and a deletion (p.Q247*) variants exclusively in the HCM probands. These 3 variants were absent in 751 additional control subjects screened by TaqMan assays. Likewise, rare variants were enriched in the white HCM population (11/250, 4.4% versus 3/262, 1.1%, respectively, P=0.024). Expression of the mutant TRIM63 was associated with mislocalization of TRIM63 to sarcomere Z disks, impaired auto-ubiquitination, reduced ubiquitination and UPS-mediated degradation of myosin heavy chain 6, cardiac myosin binding protein C, calcineurin (PPP3CB), and p-MTOR in adult cardiac myocytes. Induced expression of the mutant TRIM63 in the mouse heart was associated with cardiac hypertrophy, activation of the MTOR-S6K and calcineurin pathways, and expression of the hypertrophic markers, which were normalized on turning off expression of the mutant protein.

Conclusions: TRIM63 mutations, identified in patients with HCM, impart loss-of-function effects on E3 ligase activity and are probably causal mutations in HCM. The findings implicate impaired protein degradation in the pathogenesis of HCM.

Figure 1. TRIM63 mutations

A. Pedigrees of families with TRIM63 mutations. The full circles (female) and squares (male) indicate the affected individuals. Those with the mutations are identified with the + sign. A broken connecting line indicates an adopted person. The slash through sign indicates a deceased person. B. Electrophoregram of three TRIM63 mutations identified in HCM probands. C and D. Topographic location of the TRIM63 mutations on exons and protein, respectively. E. Multiple sequence alignment showing evolutionary conservation of the affected amino acids.

Figure 2. Effects of the mutations on auto-ubiquitination of TRIM63

A. Immunoblot showing expression of the wild type (WT) and mutant TRIM63 proteins in HeLa cells transduced with the recombinant lentiviruses. B. Auto-ubiquitination of the WT and mutant TRIM63 in MG132 treated HeLa-His/Bio-Ub cells detected by Co-immunoprecipitation (Co-IP). Biotinylated ubiquitin was precipitated with streptavidin-conjugate agarose beads and probed with an anti Flag antibody. C. Immunoblot of the input protein probed with an anti Flag (TRIM63) antibody. D. Immunofluorescence staining of auto-ubiquitinated TRIM63 in HeLa-His/Bio-Ub cells treated with MG132. TRIM63 is detected using an anti Flag antibody and biotinylated ubiquitin with a streptavidin-conjugated anti ubiquitin antibody. Percent auto-ubiquitinated TRIM63 aggregates, as determined by the percentage of co-localized ubiquitin and TRIM63 proteins, were 76.7 ± 13%, 31.5 ± 8%, 21.0 ± 5% and 0% in the TRIM63^WT, TRIM63^A48V, TRIM63^I130M and TRIM63^Q247*, respectively (N= 20 cells per group, ANOVA p<0.001 and p<0.05 any of the mutant TRIM63 vs TRIM63^WT).

Figure 3. Effects of the TRIM63 mutations on ubiquitination of MYH6, MYBPC3 and PPP3CB in adult cardiac myocytes

A. IP was performed using an anti ubiquitin antibody and immunoblotting with anti MYH6, anti MYBPC3 and anti PPP3CB antibodies. B. Blots representing input proteins and α-TUBULIN (loading control). C and D. Immunoblots of MYH6 and MYBPC3 in the absence or presence of MG132, respectively. E. Quantitative data for MYH6 and MYBPC3 levels in the absence (black columns) or presence (open columns) of MG132 (N=3 per group, # p<0.05 compared to cells alone and *<0.05 compared to TRIM63^WT).

Figure 4. Reduced localization of mutant TRIM63 proteins to Z disks in cardiac myocytes

A. Low-magnification immunofluorescence images of transduced adult cardiac myocytes expressing either a WT or a mutant TRIM63 protein. B. Deconvolution images of myocytes at Z disk regions after co-staining with anti α-ACTININ (green) and anti Flag (red) antibodies and the corresponding overlay images. C. Quantitative spectral display of Z disks stained for α-ACTININ and TRIM63 in the overlay images. Correlation coefficient (r value) between the two colors in each image is shown (A value of 0 indicates no correlation and 1 a perfect co-localization).

Figure 5. Levels of selected signal regulators of cardiac hypertrophy in adult cardiac myocytes

Immunoblots showing levels of p-MTOR and total MTOR (A), p-S6K1 and total S6K1 (B), and p-AKT1 and total AKT1 (C) in the presence and absence of MG132. D. Represents expression of WT and mutant TRIM63 in the experimental groups. E, F and G. show quantitative levels of p-MTOR, p-S6K1 and p-AKT1, respectively, in the absence (black columns) and presence (open columns) of MG132 (N=3 per group, # p<0.05 TRIM63^WT compared to control cells; * p<0.05 Mutant TRIM63 groups compared to TRIM63^WT).

Figure 6. Cardiac and myocyte hypertrophy in Doxycycline-inducible WT and mutant TRIM63 transgenic mice

A. Immunoblots showing expression of the transgene protein detected using an anti Flag antibody in double transgenic lines (without Doxycycline). Bands other than those corresponding to TRIM63 are thought to represent nonspecific antibody reactivity. There were no discernible differences in the expression levels of the transgene among different lines of the same gene, with the exception of the line 5033 in the TRIM63^A48V group, which had a lower level of transgene protein. Therefore, TRIM63^WT line 5067, TRIM^A48V line 5016, TRIM63^I130M line 5048 and for TRIM63^Q247* line 5057 were used in the analysis. B. Gross cardiac morphology. C. Left ventricular weight/body weight ratio (mean ±SD, N=6 to 12 mice per group). D. Representative WGA stained thin myocardial sections reflective of myocyte cross-sectional area. E. Myocyte cross sectional area (mean ± SD), was measured in over 2,000 myocytes per mouse and three to four mice per group. # p<0.05 compared to non-transgenic mice, * p<0.05 compared to TRIM63^WT. Molecular size markers on the right side of the panels relate only to the TRIM63^Q247* group.

Figure 7. Reduced ubiquitination of cardiac proteins in mutant TRIM63 transgenic mice

A. Immunoblot showing levels of ubiquitinated cardiac proteins detected probed with an anti ubiquitin antibody. The lower panel represents ubiquitinated TRIM63 enriched for ubiquitination and probed with an anti Flag antibody. B. Immunoblots showing levels of TRIM63, MYH6, MYBPC3 and α-TUBULIN proteins, the latter as a control for loading conditions. C. Quantitative data on MYH6 and MYBPC3 levels. D. Levels of p-MTOR, total MTOR, p-S6K1 and total S6K1 along with panels showing expression of the transgene proteins and α-TUBULIN. E and D. Quantitative data on protein levels of p-MTOR and p-S6K1. N=3-4 mice per group, # p<0.05 compared to non-transgenic mice, * p<0.05 compared to TRIM63^WT.

Figure 8. Induction of cardiac hypertrophic markers upon turning off and on expression of the mutant TRIM63 protein

A. Immunoblots representing levels of the transgene TRIM63, p-MTOR, total MTOR, p-S6K1, total S6K1, RCAN1.4 and α-TUBULIN upon withdrawal (turning on) and administration of Doxycycline (turning on) expression of the TRIM63^Q247*. B. Quantitative data for the p-MTOR, p-S6K1 and RCAN1.4. The comparisons were made between off and on states of the transgene and between transgenic on mode and non-transgenic groups. TRIM63^WT mice were not included in these experiments. C. qPCR data on relative mRNA levels of selected hypertrophic markers Myh7, Nppa and Nppb. (N= 3 per group, # p<0.05 compared to non-transgenic mice, * p<0.05 compared to TRIM63^WT).