br BAZ A bromodomain adjacent to zinc finger
BAZ1A (bromodomain adjacent to zinc finger domain, 1A), also known as ACF1, is an accessory, non-catalytic subunit of ACF (the ATP-dependent chromatin assembly factor) . BAZ1A is thought to en-hance nucleosome sliding eﬃciency and regulate template specificity of the ATPase subunit SNF2H [33,34]. BAZ1A has been reported to modulate the ATPase activity of ACF complex and participate in gene transcription, DNA damage checkpoint and double-strand break repair [35–37]. BAZ1A can also function in neurodevelopment  and spermatogenesis through regulating expression of associated genes . Intriguingly, knockdown of SMARCA5 (SNF2H), a catalytic sub-unit of ACF which always functions incorporation with BAZ1A, leads to slower cell proliferation and migration rate in breast cancer PFK-158 . Nevertheless, whether BAZ1A functions in cell proliferation and even cellular senescence remains unclear.
Based on the findings that dynamic changes of histone modifications and DNA methylation occur during cellular senescence in mammals [40,41], we wonder whether other epigenetic factors, especially chro-matin remodeling factors could play a functional role in mammal cel-lular senescence. In this study, we are excited to discover the chromatin remodeling factor BAZ1A as a novel factor regulating senescence-as-sociated phenotypes in cancer and normal cells. It adds another layer of gene expression regulation for cellular senescence and may also have implication in cancer prevention.
2.1. BAZ1A is downregulated during cellular senescence in human and mouse
Dynamic expression variance of BAZ1A was first examined at mRNA level in multiple cellular senescence models. We adopted the replicative senescence system of mouse embryo fibroblasts (MEFs), which ex-hibited typical senescence markers including increased expression of cyclin-dependent kinase inhibitor p21 (coded by CDKN1A), p16 (coded by CDKN2A) and decreased expression of cell proliferation marker MKI67 (Supplemental Fig. 1A). We further validated the senescent state of passage 6 (P6) of MEFs compared to passage 3 (P3) via SA-β-Gal staining and calculating the percentage of positive stained cells (Supplemental Fig. 1B–C). Decreased expression of Baz1a in senescent MEFs was observed at both mRNA (Fig. 1A) and protein levels (Sup-plemental Fig. 2). Notably, the decreased expression of Baz1a at mRNA level was not specific to mouse cells, similar expression trend was Life Sciences 229 (2019) 225–232
observed in replicative senescent human umbilical vein endothelial cells (HUVECs) (Fig. 1B). Moreover, decreased BAZ1A mRNA was also found in multiple human replicative cellular senescence systems based on analysis of public RNA-seq datasets , including human foreskin fibroblasts (HFF), WI-38 and MRC-5 (Fig. 1C), whose senescent states were also featured by the expression of various senescence-associated markers (Supplemental Figs. 3–5). The above observations combined to indicate that BAZ1A exhibits down-regulated expression in multiple cellular senescence models.
2.2. Knockdown of BAZ1A induces cellular senescence
To examine whether down-regulation of BAZ1A contributes to cel-lular senescence, we stably knocked down (KD) BAZ1A by lentivirus mediated short hairpin RNA (shRNA) in three human cell lines (A549, a lung adenocarcinoma cell line; U2OS, Human Bone Osteosarcoma Epithelial Cells; HUVECs) and two mouse cell lines (NIH3T3, a murine embryonic fibroblast cell line; MEFs). The success of stable knockdown was validated at both mRNA and protein levels (Fig. 2A and Supple-mental Figs. 6A, 7A, 8A and 9A). CCK-8 assay showed BAZ1A-KD cells had decreased proliferation rate compared to control cells (Fig. 2B and Supplemental Figs. 6B, 7B and 8B). In addition, cell cycle was arrested at G1 phase in BAZ1A-KD cells (Fig. 2C and Supplemental Figs. 6C, 7C and 8C). Noteworthy, all of the five BAZ1A-KD cell lines showed in-creased percentage of positive SA-β-Gal stained cells (Fig. 2D–E and Supplemental Figs. 6D–E, 7D–E, 8D–E and 9B–C). Moreover, reduced EdU incorporation rate was also observed in BAZ1A-KD cells (Fig. 2F and Supplemental Fig. 6F), reflecting decreased level of DNA synthesis, which is also a well-known molecular phenotype of senescent cells . Furthermore, cellular senescence induced by BAZ1A knockdown was further proved by other molecular senescence-associated markers including increased expression of cyclin-dependent kinase inhibitor p21 (CDKN1A), decreased expression of cell cycle related Cyclin B2 (CCNB2) or cell cycle related Cyclin D1 (CCND1) and cellular pro-liferation marker MKI67 (Fig. 2G and Supplemental Fig. 6G). Besides, increased expression of CDKN1A in BAZ1A-KD cells was also validated by western blot (Fig. 2H and Supplemental Fig. 6H). These results de-monstrated that BAZ1A deficiency can induce senescence-associated phenotypes in both normal and cancer cells.