Dear Editor,

BAF155 is a subunit of the SWI/SNF chromatin remodeling complexes, which increase DNA accessibility by remodeling nucleosomes during gene transcription1. BAF155 plays an important role in development and disease. For instance, the BAF155-containing BAF complex is required to maintain self-renewal and pluripotency of embryonic stem cells through regulation of Oct4/Sox2-dependent transcription2. BAF155 has also been shown to promote breast cancer progression and metastasis by regulating the expression of c-Myc pathway genes3,4. However, the roles and mechanisms of BAF155 in cardiovascular disease remain unknown. Here, we found that BAF155 expression was notably upregulated in cardiac tissues of patients and mice with heart failure and in angiotensin II (Ang II)-treated cardiomyocytes (Fig. 1a; Supplementary Fig. S1a, b).

Fig. 1: BAF155 restrains WWP2-mediated PARP1 ubiquitination to promote cardiac hypertrophy and fibrosis.
figure 1

a Immunohistochemical staining of BAF155 and BNP proteins in heart tissues from healthy donors and patients with heart failure (n = 5). b Schematic of Ang II-induced mouse model of cardiac hypertrophy and fibrosis. BAF155-cWT and BAF155-cKO mice were administered with saline or Ang II (1.5 mg/kg/day) through a subcutaneously implanted osmotic minipump (0.5 µL/h) for 2 weeks. c EF% and FS% of BAF155-cWT and BAF155-cKO mice (n = 6). d H&E staining, TRITC-labeled WGA staining, and Masson’s trichrome staining of BAF155-cWT and BAF155-cKO hearts (n = 6). e Schematic of Ang II-induced mouse model of cardiac hypertrophy and fibrosis. BAF155-WT and BAF155-TG mice were administered with saline or Ang II (1.5 mg/kg/day) through a subcutaneously implanted osmotic minipump (0.5 µL/h) for 2 weeks. f EF% and FS% of BAF155-WT and BAF155-TG mice (n = 6). g H&E staining, TRITC-labeled WGA staining, and Masson’s trichrome staining of BAF155-WT and BAF155-TG hearts (n = 6). h Schematic showing workflow for quantitative proteome analysis. i Heatmaps of differentially expressed proteins among BAF155-cWT, BAF155-cKO, BAF155-WT, and BAF155-TG hearts (n = 3). j, k Volcano plots showing foldchanges of all detected proteins between BAF155-cWT and BAF155-cKO hearts (j), and BAF155-WT and BAF155-TG hearts (k). l Co-IP assay showing the interaction between WWP2 and PARP1 in BAF155-cWT and BAF155-cKO hearts. m Ubiquitination levels of PARP1 in BAF155-cWT and BAF155-cKO hearts. n Co-IP assay showing the interaction between WWP2 and PARP1 in BAF155-WT and BAF155-TG hearts. o Ubiquitination levels of PARP1 in BAF155-WT and BAF155-TG hearts. p, q Immunoblotting showing expression of PARP1 and levels of total PARylation in BAF155-cWT and BAF155-cKO hearts (p), and BAF155-WT and BAF155-TG hearts (q) (n = 6). r Working model showing the role of BAF155 in regulating cardiac homeostasis. Data represent means ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. Two-way ANOVA with Bonferroni multiple comparisons test (a, c, d, f, g, p, q).

Full size image

To explore the effect of BAF155 on heart disease, we generated conditional myocardium-specific Myh6Cre+;BAF155Fl/Fl (hereafter, BAF155-cKO) mice, and Myh6Cre–;BAF155Fl/Fl (hereafter, BAF155-cWT) mice were used as controls. We established a mouse model of cardiac hypertrophy and fibrosis by Ang II (1.5 mg/kg/day) infusion for 14 days (Fig. 1b; Supplementary Fig. S1c, d)5,6. Notably, compared with BAF155-cWT mice, BAF155-cKO mice displayed significantly alleviated Ang II-induced cardiac dysfunction, as reflected by increased ejection fraction (EF%) and fractional shortening (FS%) (Fig. 1c). Cardiac hypertrophy and fibrosis were also significantly mitigated in BAF155-cKO mice as revealed by hematoxylin and eosin (H&E), wheat germ agglutinin (WGA), scanning-source optical coherence tomography and Masson staining assays (Fig. 1d; Supplementary Figs. S1e, f, S2a). We found that compared to BAF155-cWT mice, expression levels of cardiac hypertrophy and heart failure markers (ANP and BNP), fibrosis markers (α-SMA and Col-1), cardiomyocyte death markers (cleaved-caspase-3, cleaved-caspase-9), and DNA damage response markers (p-ATM and p-ATR) were significantly decreased in BAF155-cKO mice (Supplementary Fig. S1g–j). Since Myh6-Cre is active in both cardiomyocytes and smooth muscle cells7, arteries were further collected for histology analyses. We found that knockout of BAF155 mitigated Ang II-induced hypertension, vascular thickening, and fibrosis (Supplementary Fig. S2b–e). These results suggested that Myh6-Cre-induced knockout of BAF155 in mice alleviated Ang II-induced cardiac hypertrophy and fibrosis, which improved heart function by modulating myocardial cells and reduced hypertension by modulating vascular smooth muscle cells.

To further examine whether BAF155 overexpression in cardiomyocytes can aggravate heart disease, we generated transgenic mice overexpressing BAF155 driven by the CAG promoter (hereafter, BAF155-TG) (Fig. 1e; Supplementary Fig. S3a). Notably, compared with BAF155-WT mice, BAF155-TG mice displayed a slight reduction in EF% and FS% under physiological conditions, and Ang II treatment further aggravated the impairment of EF% and FS%, as demonstrated by echocardiography imaging (Fig. 1f). Likewise, under physiological conditions, BAF155-TG mice displayed mild cardiac hypertrophy, which was significantly aggravated after Ang II treatment (Fig. 1g; Supplementary Figs. S3b, c, S4a). At the molecular level, under physiological conditions, BAF155-TG mice exhibited slightly increased expressions of α-SMA, cleaved-caspase-3, cleaved-caspase-9, p-ATM, and p-ATR compared with BAF155-WT mice. Moreover, compared with BAF155-WT mice, the expression levels of ANP and BNP were significantly increased in Ang II-treated BAF155-TG mice (Supplementary Fig. S3d–g). We also found that BAF155-TG mice displayed aggravated Ang II-induced hypertension, vascular thickening, and fibrosis (Supplementary Fig. S4b–e). These results confirmed that Ang II-induced cardiac hypertrophy and fibrosis were aggravated in BAF155-TG mice.

To identify the potential target of BAF155, cardiac tissues from Ang II-treated BAF155-cWT, BAF155-cKO, BAF155-WT, and BAF155-TG mice were collected for quantitative proteomic analysis (Fig. 1h). A total of 3982 proteins were identified (Supplementary Figs. S5, S6a–g), and differentially expressed proteins were shown in Fig. 1i–k. We further identified potential protein interactors of BAF155 using mass spectroscopy. Interestingly, we found that Poly(ADP-ribose) polymerase 1 (PARP1), a known player in cardiac hypertrophy and fibrosis, was a potential interacting protein of BAF155 (Supplementary Fig. S7 and Table S1). Co-IP assays confirmed the interaction between BAF155 and PARP1, which was increased with Ang II treatment (Supplementary Fig. S6h–j). Co-IP assays using a series of truncation variants suggested that BAF155 mainly interacted with amino acids 1–203 (Zinc finger domains 1/2) and 476–779 (PARP-A-helical domain) of PARP1 (Supplementary Fig. S6k).

We next explored the mechanism by which PARP1 is regulated by BAF155. As shown in Supplementary Fig. S8a, BAF155 overexpression resulted in increased expression of PARP1, whereas BAF155 silencing remarkably downregulated the expression of PARP1 (Supplementary Fig. S8b). In addition, treatment with cycloheximide (CHX), a protein translation inhibitor, resulted in decreased expression of PARP1 in a dose-dependent manner in control cells, whereas BAF155 overexpression could maintain PARP1 abundance in the presence of CHX (Supplementary Fig. S8c). Furthermore, treatment with MG132, a proteasome inhibitor, dose-dependently increased the expression of PARP1 in control cells; however, BAF155 overexpression maintained a high level of PARP1 expression (Supplementary Fig. S8d). Altogether, these findings indicated that BAF155 inhibited PARP1 degradation by blocking the proteasome pathway. Consistently, the ubiquitination level of PARP1 was decreased in BAF155-overexpressing cells (Supplementary Fig. S8e), whereas the ubiquitination level of PARP1 was increased by BAF155 knockdown (Supplementary Fig. S8f).

Our previous study found that K249 and K418 were key sites of PARP1 ubiquitination8. To further explore whether PARP1 ubiquitination was inhibited by BAF155, we overexpressed PARP1-WT, PARP1-K249R, or PARP1-K418R in control cells or BAF155 knockdown cells. As shown in Supplementary Fig. S8g, compared with the PARP1-WT, ubiquitination levels of the PARP1-K249R, PARP1-K418R were decreased in control cells, and BAF155 knockdown remarkably increased ubiquitination levels of PARP1-WT, but not PARP1-K249R or PARP1-K418R. Collectively, these results demonstrated that BAF155 might inhibit PARP1 ubiquitination at K249 and K418 sites.

Our previous work revealed that WWP2 is a specific E3 ubiquitination ligase of PARP1 and mediates the ubiquitination of PARP1 at K249 and K418 sites8. WWP2 is also an E3 ubiquitination ligase of BAF1559. Treatment with MG132 enhanced the interaction between BAF155 and WWP2, as well as that between PARP1 and WWP2 (Supplementary Fig. S8h). Furthermore, upon BAF155 knockdown, binding between PARP1 and WWP2 was increased (Supplementary Fig. S8i), whereas BAF155 overexpression decreased interaction between PARP1 and WWP2 (Supplementary Fig. S8j). When WWP2 was overexpressed, the ubiquitination level of PARP1 was decreased by overexpression of BAF155, compared with the control group (Supplementary Fig. S8k).

We further examined the regulation of PARP1 by BAF155 in the established mouse model of Ang II-induced cardiac hypertrophy and fibrosis. Our results showed that under Ang II treatment, the interaction between PARP1 and WWP2 was enhanced in cardiac tissues from BAF155-cKO mice compared with BAF155-cWT mice (Fig. 1l). In addition, under Ang II treatment, PARP1 ubiquitination was also increased in BAF155-cKO mice (Fig. 1m). In contrast, compared with BAF155-WT mice, the interaction between PARP1 and WWP2 was downregulated in BAF155-TG mice with Ang II treatment (Fig. 1n). The ubiquitination level of PARP1 was also decreased in cardiac tissues of BAF155-TG mice with Ang II treatment (Fig. 1o).

Interestingly, WWP2 may target the BAF155–PARP1 complex to regulate its ubiquitination and degradation, thereby protecting the heart against hypertrophy and fibrosis. To further understand the role of WWP2 in Ang II-induced cardiac hypertrophy and fibrosis, we generated WWP2-cKO and WWP2-cWT mice. Proteomic analysis for global ubiquitination was performed (Supplementary Figs. S9, S10a–h). Co-IP assays showed that the interaction between BAF155 and PARP1 was increased in WWP2-cKO mice with Ang II treatment, compared with WWP2-cWT mice (Supplementary Fig. S10i). In addition, PARP1 ubiquitination was downregulated in WWP2-cKO mice with Ang II treatment (Supplementary Fig. S10j). BAF155 ubiquitination was also decreased in WWP2-cKO mice with Ang II treatment (Supplementary Fig. S10k). These results suggested BAF155–PARP1 as a key physiological substrate of WWP2 in vivo.

PARP1 is an abundant nuclear protein involved in various DNA repair pathways. Previous studies have shown that PARP1 promotes cardiac hypertrophy by PARylation of downstream targets, including BRD4, HMGB1, CEBPβ, and FOXO3a10,11,12. We next examined the role of BAF155 in the regulation of PARP1 expression and total protein PARylation levels. We found that BAF155-cKO mice displayed decreased Ang II-induced expression of PARP1, total PARylation modification, and PARylation levels of BRD4, HMGB1, CEBPβ, and FOXO3a, compared with BAF155-cWT mice (Fig. 1p; Supplementary Fig. S11a–d). In contrast, compared with those in BAF155-WT mice, expression of PARP1, total PARylation modification, and PARylation levels of BRD4, HMGB1, CEBPβ, and FOXO3a were mildly increased in BAF155-TG mice under physiological conditions, which were significantly increased following Ang II stimulation (Fig. 1q; Supplementary Fig. S11e–h).

To determine whether PARP1 inhibition can restrain the detrimental effects of BAF155 overexpression on cardiac function, a PARP1 inhibitor (25 mg/kg/day for 14 days) was administered to Ang II-treated BAF155-TG mice (Supplementary Fig. S12a). Remarkably, PARP1 inhibition significantly improved cardiac EF% and FS% and alleviated cardiac hypertrophy and fibrosis (Supplementary Fig. S12b–d). Consistently, PARP1 inhibitor treatment downregulated expression levels of PARP1, cleaved-caspase-3, cleaved-caspase-9, p-ATM, p-ATR, and total PARylation levels in BAF155-TG mice with Ang II treatment (Supplementary Fig. S12e–g).

Collectively, our study shows that BAF155 promotes pathological cardiac hypertrophy and fibrosis by inhibiting WWP2-mediated PARP1 ubiquitination and degradation (Fig. 1r). Therefore, our findings suggest BAF155 as a potential target in the treatment of myocardial hypertrophy and fibrosis disease.