Histone deacetylase 6 controls cardiac fibrosis and remodelling through the modulation of TGF‐β1/Smad2/3 signalling in post‐infarction mice

The animal experiments were reviewed and approved by the Institutional Animal Care and Use Committee at Yanbian University (Protocol: YD20211128016). Eight‐week‐old HDAC6^+/+ male C57BL/6J (provided by Yanbian University Animal Center) and HDAC6^−/− mice provided by Shanghai Biomodel Organism Science & Technology Development Co. (Shanghai, China; protocol: 2022‐W5‐1109) that weighed 22–26 g were housed in an SPF‐level facility under a 12‐h light/dark cycle at 22 ± 3°C and a relative humid environment. All mice had free access to abundant food and water. The mice were monitored throughout the experimental period, and all experimental protocols involving the mice were performed by trained research staff.

The mouse MI model was created as described.17 In brief, HDAC6^+/+ and HDAC6^−/− mice as the MI group were each anaesthetised with 5% isoflurane inhalation with an air delivery system. After the mice became unconscious, the anaesthetic agent was switched to 2%–3% isoflurane to maintain the anaesthesia status. A small incision was made over the left chest, and the fourth intercostal space was exposed. With a clamp slightly open, a small hole was formed and the heart was manually and smoothly squeezed out of the thoracic cavity. Next, 7–0 silk suture was used to ligated the left main descending coronary artery (LAD) of the mice of both genotypes (named the HDAC6^+/+‐MI mice and HDAC6^−/−‐MI mice) at the same level at the lower edge of the left atrium from the LAD's origin. The heart was then immediately returned back into the thoracic cavity. Sham‐operated mice of both genotypes underwent the same procedure without LAD ligation (named the HDAC6^+/+‐sham and HDAC6^−/−‐sham mice).

Each mouse in the two MI groups was given a daily peritoneal injection of Tub A, 10 mg/kg body weight (Apexbio Technology, Houston, TX, USA) or vehicle, from 12 h before the surgery until 7 days post‐MI.

At the indicated time points before and after shame and MI operations, following hemodynamic analysis by an echocardiography (Figure 1A), all mice were anaesthetised with an intraperitoneal injection of chloral hydrate (0.1 mL/10 g), and blood samples were collected from the left ventricles. Following perfusion with 4% phosphate buffered saline (PBS) at the physical pressure, the whole hearts and left ventricles were successively isolated and weighted. For the biological analysis, the LV tissues was maintained in RNA later solution (for the gene assay) or stored at −80°C (for the protein assay). For the morphological analysis, after being immersed in fixative at 4°C, the LV tissues were embedded in optimal cutting temperature compound (Sakura Fine‐technical, Tokyo) and stored at −20°C. The blood was poured into a blood collection tube and centrifuged, and the plasma was collected for the evaluations of cardiac injury biomarkers (i.e. lactate dehydrogenase [LDH], creatine kinase‐MB [CK‐MB] and cardiac troponin I [cTnI]) and stored at −80°C. All surgical and sampling procedures followed were in accordance with institutional guidelines.

Transthoracic two‐dimensional (2D) parasternal short‐axis M‐mode echocardiography was performed at baseline and at day 14 after MI surgery, using a Vevo 2100 system (VisualSonics, Toronto, Canada) with a 30‐MHz transducer. In brief, the mouse was selected and placed into an anaesthesia induction kit full of 2% isoflurane until the mouse became unconscious. The mouse was then fixed in a supine position on a heating plate (37°–38°C) with its limbs dotted with pieces of adhesive tape. Hair removal cream was applied to remove the hair on the chest area. After the thorax of the mouse was cleaned, the thorax was coated with an ultrasonic coupling agent. When the long‐axis surface of the left ventricle was detected and rotated 90° clockwise, the image of the short axis of the left ventricle appeared. We observed the ventricular wall motion and saved the left ventricular (LV) long‐ and short‐axis views under the B‐mode and M‐mode. The left ventricular ejection fraction (LVEF), LV fraction shortening (LVFS), LV end‐diastolic diameter (LVDd), LV end‐systolic diameter (LVSd) and the mitral valve E peak/A peak ratio (E/A ratio) of the LV inflow were calculated with doppler measurements. The mouse heart rate (HR) was also calculated.

Total proteins were extracted from heart tissue and cardiac fibroblasts as samples for Western blotting.18 Samples were homogenized and lysed with RIPA buffer containing protease and phosphatase inhibitor cocktails for 30 min on ice. The lysates were then centrifuged at 13,000 g for 10 min, and the supernatants were collected and labelled. The total protein concentrations in the supernatants were measured by a bicinchoninic acid assay (Pierce BCA Protein Assay Kit; Thermo Scientific, Waltham, MA). Twenty micrograms of protein per lane were resolved by SDS‐PAGE, transferred to PVDF membranes (Bio‐Rad, Hercules, CA). Each membrane was blocked with 5% nonfat milk and incubated with primary antibodies overnight at 4°C. On the following day, the membranes were incubated with secondary antibodies (Cell Signaling Technology [CST], Danvers, MA).

Immunoreactive signals were detected using an Azure 500 analyzer (GE Healthcare, Milwaukee, WI). The following antibodies were used for Western blotting: rabbit anti‐phospho‐Smad2/3 (p‐Smad2/3) monoclonal antibody (mAb) (#8828S, 1:1000), rabbit anti‐Smad2/3 mAb (#8685S, 1:1000), rabbit anti‐TGF‐β1 mAb (#3711S, 1:500), anti‐HDAC6 mAb (#7558S, 1:1000), anti‐GAPDH mAb (#5174S, 1:6000), mouse anti‐extracellular signal regulated kinase1/2 (Erk1/2) mAb (#9107, 1;1000) and anti‐p‐Erk1/2^t202/t204 mAb (#4377, 1:1000) were purchased from Cell Signaling Technology (Danvers, MA). Mouse‐anti‐Nkx‐2.5 mAb (sc‐376565, 1:1000) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti‐TGF‐β1 receptor (TGF‐βR1) polyclonal antibody (pAb, #SAB4502958, 1:500) was purchased from Sigma‐Aldrich (St. Louis, MO). Rabbit anti‐Collagen I mAb (#EPR22894↗ ‐89, 1:1000) and rabbit anti‐Collagen III mAb (#EPR17673↗, 1:1000) were purchased from Abcam (Cambridge, MA). Anti‐α‐SMA antibody (#AF1032, 1:1000) was from Affinity Biosciences (Cincinnati, OH).

Total RNA was extracted from tissues using Trizol reagents (Invitrogen, Carlsbad, CA), and first‐strand cDNA was synthesized using a Thermo Script RT‐PCR synthesis kit (Fermentas, Burlington, ON, Canada).19 Quantitative PCR (qPCR) analyses of mRNA were performed using Thermo Script RT‐PCR kits (Fermentas). The qPCR was carried out under a standard protocol using the following primers. The PCR was performed in triplicate. Each PCR was performed at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 65°C for 1 min. The expression of GAPDH was measured in parallel with that of the genes of interest and was used as an internal standard for the quantitative comparison of mRNA levels. The generation of specific genes' expression changes was evaluated using the comparative Ct method, X = Z^−∆∆Ct. Fold‐alterations were evaluated using the ∆∆Ct method and compared with GAPDH. The sequences of mice targeted gene primers was listed in the Table 1.

On the 14th day post‐MI or sham operation, mice were injected 2% with Evans Blue dye (Sbjbio Life Sciences, Nanjing, China) via the caudal vein, and their hearts were excised and frozen at −80°C. Fifteen minutes later, the hearts were minced into approximately 1‐mm‐thick slices and submerged in a 1% triphenyl tetrazolium chloride (TTC) (Solarbio Science & Technology Co., Beijing, China) solution at 37°C for approximately 15 min. Healthy areas of a mouse heart are stained blue; the area at risk (AAR) turns red, and the infarct area (IA) remains unstained as white. To determine the sizes of the AAR and IA, we calculated the ratio between the IA and AAR with image analysis software (Image J).

Masson's trichrome staining was performed as described.20 After mice were sacrificed, cardiac tissue was sectioned (5–6 μm) from below the ligation plane of the heart toward the apex with the use of a cryostat (CM1950, Leica Biosystems, Wetzlar, Germany) and dehydrated in 20% sucrose solution. The slices were fixed in 4% polyformaldehyde for 24 h and then embedded in OCT. At first, transverse tissue sections were stained with haematoxylin and eosin (H&E) and the cross‐sectional area of myocytes was determined from cells that were cut transversely and exhibited both a nucleus and an intact cell membrane; at least 100 cells were assessed per specimen, and the average value was used for analysis (×200). And transverse tissue sections were stained with Masson's trichrome. Sections were imaged using an EVOS FL Auto 2 imaging system (Thermo Fisher Scientific). To determine the extents of interstitial fibrosis in the infarct area, we selected five fields at random and calculated the ratio of the area of Masson's trichrome‐stained fibrosis to the total area of the left ventricles with the Image J software.

The immunofluorescence assay was performed as described.21 Following blocking with bovine serum albumin for 30 min, the LV tissue sections and the cultured cells (cover‐glasses) were treated with the primary antibodies to HDAC6, TGF‐β1, TGF‐βR1, α‐actin, vimentin or CD206 (1:250 for each antibody), respectively, overnight. After being washed with PBS for three times, the tissue sections and the cells were treated with the fluorescent labelled secondary antibodies against anti‐rabbit IgG or goat anti‐mouse IgG (1:200 for each), respectively, for 1 h at room temperature, and then the nuclei were counterstained with an anti‐fluorescence quenching solution including DAPI. Sections were imaged using an EVOS FL Auto 2 imaging system (Thermo Fisher Scientific). The levels of HDAC6 and CD206 protein expressions in the LV tissues were evaluated in four random microscopic fields from three independent sections of each animal (n = 5) and were expressed as the percentages of the positive staining signal intensities per high‐power field (×400). And the levels of TGF‐β1 and its receptor TGF‐βR1 protein expressions of the targeted cells were evaluated in two‐three random microscopic fields from three independent sections of each cell seeded cover‐glass (n = 5) and were expressed as the numbers of the positive staining cells per high‐power field (×400).

The following commercially available antibodies were used for the immunofluorescence: Mouse anti‐α‐actinin mAb (#sc‐17829) and anti‐Vimentin mAb (#sc‐6260) were purchased from Santa Cruz Biotechnology. Rabbit anti‐TGF‐βR1 pAb (#SAB4502958) was purchased from Sigma‐Aldrich (Burlington, MA). Rabbit anti‐TGF‐β1 mAb (#ab315254) was purchased from Abcam (Cambridge, MA). Rabbit anti‐HDAC6 mAb (#7558S) was from Cell Signaling Technology. And the secondary goat anti‐rabbit IgG pAb (FITC 495) and anti‐mouse IgG pAb (Alexa Fluor 594) were purchased from APExBIO (Houston, TX).

Ethylenediaminetetraacetic acid (EDTA) (0.5%) was used to digest detached cardiac fibroblasts (CFs) (Cat no. CPR‐074; Procell Life Science & Technology Co., Wuhan, China), and the supernatant was collected in 10‐mL Falcon centrifuge tubes. Four of these tubes were then centrifuged to collect the supernatant. The dispersed cells were resuspended in six‐well plates containing 6 mL of Dulbecco's modified Eagle's medium (DMEM) with 15% fetal bovine serum (FBS), 1% penicillin and streptomycin. Two hours later, the plates were washed with warmed phosphate‐buffered saline (PBS) and replenished with cell culture medium.

The CFs were then cultured until they reached approx. 80%–90% confluence. After being cultured in serum‐free DMEM for 6 h, the CFs were treated with Tub A at the indicated concentrations (0, 5 and 10 μmol/L) under normoxic (5% CO₂ and 95% air) or hypoxic conditions (1% O₂, 5% CO₂ and 94% N₂), respectively, for 24 h and then subjected to biological analyses.

Cell transfection was performed as described.22 CFs were seeded on six‐well plates and cultured for until they were 50%–60% sub‐confluent. The cells were then transfected with short interfering (si)RNA against HDAC6 (siHDAC6, 100 nmol/L) or non‐targeting control siRNA (NC, both from Hanheng Biotechnology Co., Shanghai, China) using Lipofectamine 3000 (Thermo Fisher Scientific). After transfection for 48 h, the cells were cultured in normoxic or hypoxic conditions (1% O₂, 5% CO₂ and 94% N₂), respectively, for 24 h and then subjected to a Western blotting assay. The siRNA sequences are listed in Table 1.

All data are expressed as the mean ± standard error of the mean (SEM). We performed a one‐way analysis of variance (ANOVA) for comparisons of multiple groups, followed by Tukey's post hoc test or by Student's t‐test for comparisons of two independent sample groups with Prism 9.0 software. After we determined the status of the data distribution, the data were subjected to the statistical analysis. If the homogeneity of variance assumption was violated, the nonparametric Kruskal–Wallis test was used instead of Pearson's chi‐square or Fisher's exact test. Survival rate was analysed by the standard Kaplan–Meier method with a log‐rank test. Morphological and histological characteristics were evaluated by two observers in a blind manner, and the values they obtained were averaged. Probability (p)‐values <0.05 were considered significant.

To study the impact of the ischemic stress on HDAC6 expression, we randomly assigned HDAC6^+/+ mice to the sham operation and LAD ligation surgery and subjected the mice to morphological and biological analyses at the indicated time points (Figure 1A). The HR and LVEF were tested to evaluate the success of the model (Figure 1B–D). The results of the qPCR and Western blotting assays demonstrated that the expression level of HDAC6 was significantly increased in the LV infarct myocardium of MI group compared to the control group (Figure 1E–G), indicating that HDAC6 is likely to participate in the cardiac pathological remodelling in the post‐MI phase of the murine MI model.

Figure 2A provides representative heart photos from the four experimental groups at day 14 after the ligation or sham operation. We observed that the heart weight (HW)/body weight (BW) ratio was lower in the HDAC6^−/−‐MI mice compared to the HDAC6^+/+‐MI mice (Figure 2B). As shown in Figure 2C, there was no difference in HR between both genetic MI groups. The quantitative echocardiography data revealed that compared to the HDAC6^+/+‐MI mice, HDAC deficiency significantly ameliorated the reductions of the LVEF, LVFS, LVLd, LVDs and E/A ratio values in the post‐MI mice (Figure 2D,E). Consistently, HDAC6 deletion markedly lowered the infarct size and cardiomyocyte size and reduced the degree of fibrosis as well as the levels of serum LDH, cTnI and CK‐BM (Figure 2F–K), suggesting that HDAC6 deficiency was resistant to ischemic injury in mice. However, there were no significant differences in these nine parameters (LDH, cTnI, CK‐BM, HW/BW ratio, HR, LVEF, LVFS, LVLd, LVDs, E/A ratio, infarct size and fibrosis area) between the sham‐operated mice of both genotypes (Figure 2).

To explore the molecular mechanisms of HDAC6^−/−‐mediated cardioprotection in the post‐MI phase, we conducted a Western blotting assay to focus on the TGF‐β1/Smad2/3 signalling pathway. As anticipated, the HDAC6^−/−‐MI mice exhibited significant reductions in the levels of p‐Samd2/3, TGF‐β1, alpha‐smooth muscle action (α‐SMA) and collagen I and III proteins compared to the HDAC6^+/+‐MI mice (Figure 3A,B). The qPCR yielded the same conclusions for the collagen I and III gene expressions (Figure 4A). HDAC6 thus appeared to modulate cardiac remodelling and dysfunction through the modulation of TGF‐β1/Samd2/3 signalling pathway activation in the post‐MI phase.

Figure 5A provides representative heart photos from the three experimental groups at day 14 after the ligation or sham operation. As shown in Figure 5B, the TubA treatment also markedly reduced the ratio of HW to BW of the HDAC6^+/+‐MI mice. We observed that in the HDAC6^+/+‐MI mice, TubA loading resulted in reductions in the values of HR, LVEF, LVFS, LVDd, LVDs and the E/A ratio as well as the cardiac infarct size and cardiomyocyte size as well as collagen accumulation (Figure 5C–J). The quantitative data of the Western blotting images demonstrated that the protein levels of HDAC6, p‐Samd2/3, TGF‐β1, α‐SMA, collagen I and collagen III were markedly lower in the LV myocardium of the HDAC6^+/+‐TubA MI mice compared to the control non‐treated HDAC6^+/+‐MI mice (Figure 6). The qPCR yielded same conclusions regarding collagen I and III as well as HDAC6 gene expressions (Figure 4B). Immunofluorescent analysis revealed only a low positive staining signal of HDAC6 in the LV myocardium of control mice (Figure 7A,B). And the staining signal of HDAC6 was markedly increased in the infarcted myocardium of mice with AMI, with staining apparent in CFs and these changes were reduced by TubA treatment (Figure 7A,B). Similarly, the HDAC6^+/+‐Tub A MI mice had decreased levels of CD206 protein expression as compared to the control non‐treated HDAC6^+/+‐MI mice (Figure 7C,D). Collectively, these observations suggest that the up‐regulation of HDAC6 by ischemic injury could act a key modulator of cardiac response of post‐MI mice. However, we observed that pharmacological and genetic interventions targeted toward DAC6 had no effect on mice mortality and/or the levels of HADC1, SIRT3 and HADC4 genes in the infarcted myocardium (Figures 4B,7E,F).

To test our experimental results at the cellular level, we performed an in vitro experiment, and Western blotting was then performed in two different sets of the experiments. The hypoxic condition increased the expressions of HDAC6, p‐Smad2/3, p‐Erk1/2, TGF‐β1, α‐SMA, Nκx‐2.5 and collagen types I and III proteins, and these changes were rectified by HDAC6 silencing and TubA treatment in a dose‐dependent manor, providing evidence of HDAC6‐mediated CF collagen synthesis in response to ischemic stress (Figures 8 and 9). The data of immunofluorescence and Western blotting analysis revealed that hypoxic condition also increased the expressions of HADC6, TGF‐β1 and TGF‐βR1 proteins in CFs, whereas it exhibited a limited effect on these molecular expressions in H9C2 (Figures 10 and 11).

The data underlying this article will be shared on reasonable request to the corresponding author.