TBC1D25 Regulates Cardiac Remodeling Through TAK1 Signaling Pathway

TBC1 Domain Family Member 25 (TBC1D25) expression is upregulated in response to pressure overload

To investigate the potential role of TBC1D25 in cardiac hypertrophy, we first analyzed the altered expression levels of TBC1D25 in pressure overload mouse heart. In Figure 1A, the Western blot analysis demonstrated that the protein level of TBC1D25 was upregulated in hearts of mice subjected to transverse aortic constriction (TAC). Atrial natriuretic peptide (ANP), a hypertrophic marker protein was also increased dramatically in these mice (Figure 1A). Furthermore, in vitro study, TBC1D25 and ANP protein levels were significantly increased in H9C2 cells or neonatal rat cardiomyocytes (NRCMs) stimulated by Ang II (Figure 1B and 1C). Taken together, these findings indicated that TBC1D25 may participate in pathological cardiac remodeling.

TBC1D25 deficiency accelerates cardiac hypertrophy induced by TAC

TBC1D25 expression was significantly increased in experimental cardiac hypertrophy, which indicated TBC1D25 may regulate pathological cardiac remodeling in this process. To further evaluate the function of TBC1D25 in cardiac remodeling, TBC1D25 -KO mice were utilized in our research. Compared with wild-type (WT) control mice, the loss of TBC1D25 protein in TBC1D25-KO mice was confirmed by Western blotting (Figure 2A). Subsequently, TBC1D25-KO mice and WT mice were subjected to TAC or sham operation. The heart structure and function were examined after 4 weeks of TAC or sham operation. No comparable differences were noticed in the sham-operated TBC1D25-KO mice and WT controls. Compared with WT mice, TBC1D25-KO mice manifested increased ratios of heart weight to body weight (HW/BW; Figure 2B), lung weight to body weight (LW/BW; Figure 2C) and heart weight to tibia length (HW/TL; Figure 2D). Furthermore, the echocardiography assay confirmed that left ventricular end-diastolic diameter (LVEDd), left ventricular end-systolic diameter (LVESd), EF% and FS% were significantly decreased in TBC1D25-KO mice compared with WT control mice (Figure 2E-H).

Moreover, the heart cross-sections of WT and TBC1D25-KO mice were stained with haematoxylin and eosin (H&E) or picrosirius red (PSR). After 4 weeks of TAC, TBC1D25-KO mice exhibited a significantly larger heart and cardiomyocytes cross-sectional area (Figure 3A) and more prominent interstitial and perivascular fibrosis(Figure 3B). While, there were no comparable differences in the sham-operated TBC1D25-KO mice and WT mice (Figure 3A-B). RNA expression of hypertrophy markers (ANP, BNP and MYH7) and fibrosis markers (Collagen Iα, Collagen III and CTGF). The TBC1D25-KO mice showed greatly increased mRNA expression levels of these markers compared with WT mice (Figure 3C and 3D). These data demonstrated that loss of TBC1D25 exacerbated cardiac dilation and dysfunction greatly.

TBC1D25 suppresses angiotensin II-induced cardiomyocytes hypertrophy in vitro

Cardiomyocyte enlargement is the mainly characteristic changes of cardiac remodeling⁸. We next examined the contribution of TBC1D25 to cardiac hypertrophy in cardiomyocytes. To address this issue, H9C2 cells and neonatal rat cardiomyocytes (NRCMs) were investigated. H9C2 cells have similar biological characteristics (including morphology, protein expression, signaling and hypertrophy related characters) with adult cardiomyocytes⁹. Many studies use H9C2 cells to investigate cardiac hypertrophy^{10, 11}. In this study, we infected viruses containing TBC1D25 and control particles with H9C2 cardiomyocytes to overexpress TBC1D25 (Figure 4A). The infected cells were selected and then treated with Ang II or PBS for 48 h. Then, the cell surface area and expression of ANP, BNP were detected. The data demonstrate that Ang II had pro-hypertrophic effect in vitro. However, TBC1D25 overexpression dramatically decreased the hypertrophic phenotypes induced by Ang II, compared with control cells (Figure 4B and 4C). To further confirm the function of TBC1D25 in cardiomyocyte hypertrophy, NRCMs were isolated and infected with AdTBC1D25 or AdGFP, respectively (Figure 4D). Similarly, AdTBC1D25-infected NRCMs exhibited decreased individual cell size and mRNA levels of hypertrophic markers (Figure 4E-F). These findings confirmed that TBC1D25 plays an important antihypertrophic role at the cellular level.

TBC1D25 regulates TAK1 signaling during cardiac hypertrophy both in vivo and vitro

The protective role of TBC1D25 in cardiac hypertrophy had been demonstrated. Furthermore, we explored the molecular mechanisms by which TBC1D25 regulated cardiac hypertrophy. Many researches proved that MAPK signaling pathways played an important role in cardiac remodeling^{12, 13}. So we evaluated the potential involvement of MAPK pathways in this process. Therefore, we monitored the expression of TAK1, ERK, JNK, and p38 in heart of TAC-operated WT and TBC1D25-KO mice. Our data indicated that the phosphorylation levels of TAK1, JNK, and p38 were significantly increased in TBC1D25-KO mice, compared with TAC-operated WT mice. But phosphorylated ERK was not affected by TBC1D25 deficiency (Figure 5A). At the same time, we detected phosphorylation and total level of TBK1 and ASK1 in vivo study. And the results demonstrated that TBC1D25 knockout did not affect phosphorylation of TBK1 and ASK1(Figure 5B). In vitro experiments, TBC1D25-overexpressing H9C2 cells and NRCMs were treated with Ang II at specified points in time. Our data indicated that TBC1D25 overexpression remarkably suppressed the level of phosphorylated TAK1, JNK, and p38 (Figure 5C-D).

Inhibiting TAK1 is the key step of cardiac remodeling regulated by TBC1D25

Furthermore, to verify whether inhibition of TAK1 is required in TBC1D25 mediated anti-hypertrophic response, we infected AdcaTAK1 (adenovirus expressing constitutively active TAK1) into AdGFP or AdTBC1D25 infected NRCMs, respectively (Figure 6A). AdcaTAK1 infection could partially reverse the protective effects of AdTBC1D25 in Ang II-induced cardiomyocyte hypertrophy (Figure 6B-C). These findings robustly verified that TBC1D25 regulate pathological cardiac remodeling through TAK1 signaling pathway.

To further investigate the molecular mechanism of TBC1D25-regulated cardiac remodeling. We examined whether TBC1D25 could directly regulates TAK1 activation. The result showed that TBC1D25 and TAK1 were completely co-localized in the cytoplasm of 293T cells by immunofluorescent staining (Figure 7A). And then, we transfected 293T cells with Flag-tagged TBC1D25 and HA-tagged TAK1. The direct interaction of TBC1D25 and TAK1 was demonstrated by Immunoprecipitation assay (Figure 7B-C). Furthermore, Glutathione S-transferase (GST) pull-down assays were performed to confirm this interaction between TBC1D25 and TAK1(Figure 7D-E). Subsequently, we constructed a series of TBC1D25 and TAK1 mutants to further identify the binding domain for TBC1D25 and TAK1 direct interaction. The results showed that the amino acids from at least 138 to 226 in the C-terminal region of TBC1D25 and from 1 to 300 in the C-terminal region of TAK1 was required for TBC1D25 and TAK1 direct interaction(Figure 7F and 7G). All these data indicated that TBC1D25 directly regulate TAK1-JNK/p38 signaling pathway, which is important in the process of cardiac remodeling (Figure 8).

Animals use for study

Experimental procedures were performed according to the NIH Guide for the Care and Use of Laboratory Animals published by the US National Institute of Health (NIH publication, 8th edition, 2011). All animal usage protocols were approved by the Animal Care and Use Committee of The First Affiliated Hospital of Zhengzhou University. The related procedures were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Constitutive TBC1D25 knock-out (TBC1D25-KO) mice were purchased from the Texas A&M Institute for Genomic Medicine.

Animal surgery

Cardiac hypertrophy was induced in mice through partial transverse aortic constriction (TAC) of aortic arch, as previously described with some adaptations³⁵. Briefly, in both groups, 9-to 11-week-old male mice were fixed in a supine position after anesthetized with sodium pentobarbital via an intraperitoneal injection, and the skin in the middle of the chest was opened to expose the aortic arch through the right side of clavicle after toe pinch reflex disappeared. Body temperature was maintained as close as possible to 37.0 °C throughout the experiment using a self-regulating heating pad. Subsequently, a specific needle (26-G for body weights of 25-27 g) was placed on the aortic arch and ligating with 7-0 silk suture at same degree, then needle was removed rapidly before the closure of the skin. Mice were observed until recovery in a 37.0 °C heated cage.

Echocardiographic assessment

Mice were anesthetized with isoflurane (1.5-2%), and echocardiography was performed using a MyLab 30CV ultrasound system (Biosound Esaote Inc.) using a 15-MHz transducer. The size of the left ventricular (LV) cavity and LV wall thickness were acquired from at least 3 consecutive cardiac cycles. The end-systole and end-diastole were defined as the phases in which the smallest or largest LV area was obtained, respectively. The LV end-diastolic dimension (LVEDd) and LV end-systolic dimension (LVESd) were measured from the LV M-mode tracing, with a sweep of 50 mm/s at the mid-papillary muscle level. Fractional shortening (FS%) was calculated using the formula: FS%= (LVEDd-LVESd)/LVEDd×100%.

Histological analysis

Hearts were obtained from experimental animals 4 weeks after TAC surgery that had been perfused with a 10% potassium chloride solution to induce cardiac arrest at the end of diastole and then harvested and fixed with a 10% formalin solution. After being embedded in paraffin, the hearts were cut into 5-μm transverse sections. Sections were stained with hematoxylin and eosin (H&E) to measure the myocyte cross-sectional area, and the abundance of collagen was assessed after Picrosirius red (PSR) staining. Fibrosis was expressed as a percentage of the average positively stained area relative to the total area. More than 40 fields per group were examined. A quantitative digital image analysis system (Image-Pro Plus 6.0) was used for the image measurements.

Plasmids

TBC1 Domain Family Member 25 (TBC1D25, NM_002536.4) and TAK1 (NM_003188.3) gene were amplified by nested PCR and then cloned into pHAGE-3×flag vector, or pcDNA5-flag and pcDNA5- HA vector, or GST-HA vector. Primer sequences used for cloning the coding sequences of TBC1D25 and TAK1 were as following: TBC1D25 forward, '-TCGGGTTTAAACGGATCCGGGATGGCAACAGCCTCC-3'; TBC1D25 reverse, 5'-GGGCCCTCTAGACTCGAGTCAAGATGCGGCTGTGGCCT-3'. TAK1 forward, 5'-TCGGGTTTAAACGGATCCATGTCTACAGCCTCTGCCGCCT-3'; TAK1 reverse, 5'-GGGCCCTCTAGACTCGAGTCATGAAGTGCCTTGTCGTTTCT-3'.

Cell cultures and lentivirual infections

Primary neonatal rat cardiomyocytes (NRCMs) were prepared as described previously³⁶. NRCMs isolated from the hearts of 1- to 2-day old Sprague-Dawley rats were enriched, and seeded into 6-well plates at a density of 3×10⁵ cells/well. Cultured them in DMEM/F12 containing 20 % FCS, BrdU (0.1 mM, inhibits the proliferation of fibroblasts), and penicillin/streptomycin at 37℃ for 48 h. Then, replaced the culture medium with serum-free DMEM/F12 for 12 hours. For further experiments, AdGFP (used as the non-targeting control) or AdTBC1D25 adenoviruses were generated and infected into NRCMs at a multiplicity of infection (MOI) of 10 for 24h. Similarity, AdcaTAK1 (adenovirus expressing constitutively active TAK1) were co-infected into AdGFP or AdTBC1D25 infected NRCMs respectively.

H9C2 cardiomyocytes and HEK-293T cell lines were maintained in DMEM medium containing 10% fetal bovine serum and 1% pencillin/streptomycin. For lentiviral infection, recombinant constructs were firstly co-transfected into HEK-293T cells with lentiviral helper plasmids (pMD2.G and psPAX2) and PEI transfection reagent. After 6 to 16 hours, fresh media were replaced and lentiviral particles were harvested 48 hours later and then used for target cell infection. Briefly, viruses containing TBC1D25 and control particles were added to H9C2 cardiomyocytes in the presence of polybrene (8μg/mL) respectively. Infected cells were selected by puromycin (2μg/mL) for 48 hours and then used for indicated assays.

Immunofluorescence staining

Immunofluorescence staining was performed to determine the surface area of the cell. H9C2s and NRCMs were stimulated with PBS or Ang II (1 μmol/L) for 48 h under a condition of 37.0 °C, 5% CO₂ and fixed with 3.7% formaldehyde. After permeabilization with 0.1% Triton X-100 in PBS and blocking with a 10% BSA solution at room temperature, cells were immunostained with an α-actinin antibody (05-384, Merck Millipore, 1:100 dilution), followed by staining with a fluorescent secondary antibody (donkey anti-mouse IgG [H+L] secondary antibody, A21202, Invitrogen, 1:200). Image-Pro Plus 6.0 software was used to measure the surface area of the cell.

Quantitative real-time (RT) PCR and Western blotting

For the RT-PCR assay, total mRNA was extracted from ventricular tissues or cells with TRIzol reagent (15596-026, Invitrogen). Then, cDNAs were reverse transcribed from RNAs using the Transcriptor First Strand cDNA Synthesis Kit (04896866001, Roche). Quantitative real-time PCR was used to detect the expression of selected genes with SYBR Green PCR Master Mix (04887352001, Roche). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the reference gene. The primer pairs used in this study are listed in Supplementary table 1.

For the Western blot analyses, total proteins were extracted from ventricular tissues or cell samples using RIPA lysis buffer (720 μL of RIPA buffer, 20 μL of phenylmethylsulfonyl fluoride, 100 μL of complete protease inhibitor cocktail, 100 μL of Phos-stop, 50 μL of NaF and 10 μL of Na3 VO4 in a final volume of 1 mL), and the protein concentration was determined with a BCA Protein Assay Kit (23225, Thermo Scientific). After fractionation by using SDS-PAGE, proteins were transferred to PVDF membranes and blocked with 5% non-fat milk at room temperature for 1h. After incubated with multiple antibodies overnight at 4 °C, the secondary antibodies were added next day, and bands were visualized using a Bio-Rad ChemiDoc XRS+ system (Bio-Rad). The levels of specific proteins were normalized to the levels of GAPDH on the same PVDF membrane. The antibodies used in this study are listed in Supplementary table 2.

Immunoprecipitation (IP) assay

IP was performed as previously described²⁶. Briefly, after transient transfection with plasmid, HEK-293T cells were lysed in ice-cold IP buffer (20 mM Tris-HCl, pH7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100) containing protease inhibitor cocktail tablets (04693132001, Roche). The cell lysates were incubated with the indicated antibody-conjugated beads (Anti-Flag M2 Magnetic Beads, M8823, Sigma; Pierce Anti-HA Magnetic Beads, 88836, Thermo Fisher Scientific) at 4°C overnight. The immunoprecipitates were subjected to immunoblotting using the indicated primary and corresponding secondary antibodies.

GST pull-down assay

In glutathione S-transferase (GST) pull-down assay³⁷, immunopurified Flag- TBC1D25 or Flag- TAK1 was incubated in elution buffer for 2 h at 4ºC. The elution buffer was composed of 50mM HEPES (pH 7.4), 500mM NaCl, 1% Triton X-100 and 3 μg μl-1 FLAG peptide (F4799, Sigma).Transfect vectors pGEX-4T-1- GST- HA- TAK1 or GST- HA- TBC1D25 into Rosetta (DE3) Escherichia coli, and then induced them with 1mM isopropyl-β-D-thiogalactopyranoside (IPTG) at an optical density of 600nm (OD600) of 0.6. Escherichia coli extracts were dealt with PBS containing protease inhibitor cocktail tablets (04693132001, Roche). Subsequently, incubated them with glutathione-sepharose 4B beads (17075601, GE Healthcare Biosciences AB) for 1h at 4 ℃. The proteins binding to the beads were washed five times with 1ml of PBS, and then incubated them for an additional 4h at 4℃ with immunopurified Flag- TBC1D25 or Flag-TAK1. Then washed the glutathione-sepharose 4B beads with 1ml of the NaCl wash buffer for three times. And the eluted proteins were analyzed by Western blotting using indicated antibodies. In the same conditions, the GST tag was used as the negative control.

Immunofluorence and confocal microscopy assays were detected as previously described ¹⁸, HEK-293T cells were cultured on gelatin-coated coverslips in 24-well plates and then co-transfected with Flag-TBC1D25 and HA-TAK1. 24h later, cells on coverslips were fixed with 4% fresh paraformaldehyde for 15 min and permeabilized subsequently with 0.2% Triton X-100 in PBS for additional 5min. Then coverslips were incubated with indicated primary and fluprescent secondary antiboies sequently. The coverslips were washed three times with PBS and stained with DAPI (10mg/ml) before mounted by mounting solution (D2522, Sigma). Fluorescence photographs were acquired and analyzed by Laser Scanning Confocal Microscopy (LSM880, ZEISS).

Statistical analysis

Data are presented as means ± S.D. Comparisons between two groups were performed using a two-tailed Student's t-test. Differences among more than 2 groups were assessed using one-way analysis of variance (ANOVA), followed by the Bonferroni test (equal variances assumed) or Tamhane's T2 (equal variances not assumed) test. A value of P<0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using SPSS (Statistical Package for the Social Sciences) software, version 21.0.

Author contributions

Sen Guo conceived, designed and performed the experiments, and wrote the manuscript. Yuan Liu performed the animal experiments. Lu Gao performed the cell experiments. Fankai Xiao, Jihong Shen and Shiying Xing checked experimental data. Fan Yang, Wencai Zhang and Qiangwei Shi analyzed the experimental data. Yan Li and Luosha Zhao conceived and designed the experiments, and revised the manuscript.