Chromatin Remodeler Smarca5 Is Required for Cancer-Related Processes of Primary Cell Fitness and Immortalization

Abstract

Smarca5, an ATPase of the ISWI class of chromatin remodelers, is a key regulator of chromatin structure, cell cycle and DNA repair. Smarca5 is deregulated in leukemia and breast, lung and gastric cancers. However, its role in oncogenesis is not well understood. Chromatin remodelers often play dosage-dependent roles in cancer. We therefore investigated the epigenomic and phenotypic impact of controlled stepwise attenuation of Smarca5 function in the context of primary cell transformation, a process relevant to tumor formation. Upon conditional single- or double-allele Smarca5 deletion, the cells underwent both accelerated growth arrest and senescence entry and displayed gradually increased sensitivity to genotoxic insults. These phenotypic characteristics were explained by specific remodeling of the chromatin structure and the transcriptome in primary cells prior to the immortalization onset. These molecular programs implicated Smarca5 requirement in DNA damage repair, telomere maintenance, cell cycle progression and in restricting apoptosis and cellular senescence. Consistent with the molecular programs, we demonstrate for the first time that Smarca5-deficient primary cells exhibit dramatically decreased capacity to bypass senescence and immortalize, an indispensable step during cell transformation and cancer development. Thus, Smarca5 plays a crucial role in key homeostatic processes and sustains cancer-promoting molecular programs and cellular phenotypes.

Keywords: ATAC-seq; MEF; RNA-seq; Smarca5; Snf2h; cell cycle; cell immortalization; homologous recombination; non-homologous end-joining; senescence.

Figure 1

Characterization of the Smarca5 deletion system. (A) Schematics of the Smarca5 conditional knockout cassette. The cS5^fl/fl cells have the exon 5 of both Smarca5 alleles floxed, while cS5^fl/wt MEFs have a single floxed Smarca5 allele. The floxed allele(s) is excised upon 4-hydroxytamoxifen (4-OHT) addition. Some elements of the figure were created using BioRender.com. (B) Genotyping of the MEFs to confirm Smarca5 deletion status using PCR. The wildtype Smarca5 allele, the floxed allele and the allele with the exon 5 deletion differ in size and can thus be differentiated from each other, revealing the cell identity. DNA was isolated from cells 48 h after adding 4-OHT. Genotyping was done in biological triplicates, each line represents MEF cell line isolated from different embryos with CreER cS5^fl/fl or CreER cS5^fl/wt genotype. The low-representation bands below 300 bp are non-specific background amplicons. (C) Immunoblots (in the same triplicates used for genotyping) for anti-Smarca5 antibody reveal protein level depletion over time (48 h and 96 h) in the 4-OHT induced, Smarca5 knockout MEFs compared to control without 4-OHT. A gradual protein level depletion is observed over time. (D) Confirmation of Cre-activation upon 4-OHT administration as assessed by expression of the EYFP, 96h after induction. Experiments were done in triplicates, MEF isolated from 3 embryos with CreER^+/− cS5^fl/fl EYFP^+/+ genotype. Percent of cells expressing EYFP indicate percent of cells that are also positive for Smarca5 knockout. Dark curve represents cells without 4-OHT and light curve represents same MEFs 96 h after 4-OHT induction, WT/4-OHT = untreated/treated with 4-OHT. ns = p > 0.05; * = p ≤ 0.05; ** = p ≤ 0.01; **** = p ≤ 0.0001.

Figure 2

Epigenomic and transcriptomic remodeling upon the gradual loss of Smarca5. (A) Heatmap showing differentially accessible ATAC peaks displaying higher accessibility in the knockout cells, p-value–0.025. Notice the gradual increase in accessibility from the single allele knockout (cS5^fl/wt, 4-OHT) to the double allele knockout (cS5^fl/fl, 4-OHT) compared to their respective wildtypes (WT), HASs = higher accessibility sites. (B) Heatmap showing gradually upregulated genes from single to double-allele knockout MEFs, p-value–0.05. DEGs = differentially expressed genes. Notice the gradual increase in gene expression from cS5^fl/wt, 4-OHT to the cS5^fl/fl, 4-OHT compared to their respective wildtypes. (C) Heatmap showing differentially accessible ATAC peaks displaying lower accessibility in the knockout cells, p-value–0.025. Notice the gradual decrease in accessibility from the single allele knockout to the double allele knockout. LASs = lower accessibility regions. (D) Heatmap showing the gradually downregulated genes in the knockout cells, p-value–0.05. (E) Gene Ontology pathways (GO) in upregulated genes as interrogated by DAVID. Text in red shows events (pathways/genes) common in both chromatin accessibility and gene expression dataset as inspected by ATAC-seq and RNA-seq. (F) GO pathways in downregulated genes as interrogated by the DAVID tool (human orthologue gene symbols are shown as a default output of DAVID). Text in red font shows events (pathways/genes) common to both chromatin accessibility and gene expression datasets. WT = cells untreated with 4-OHT; log2 FC–log2-transformed fold-change from the WT baseline condition; R1, R2 = experimental replicates 1 and 2.

Figure 3

Effects of Smarca5 loss on senescence markers and phenotypes. (A) Induction of 50 direct p53 target genes upon treatment with 4-hydroxy-tamoxifen (4-OHT). (B) Immunoblotting and RNA-seq analyses of p21/Cdkn1a, a p53-dependent inducer of senescence. (C) RNA-seq analysis of senescence and cell-cycle exit markers. (D) Staining for senescence–associated SA-β-gal activity observed in the 4-OHT-treated cultures (already at 96 h post-treatment for Smarca5 double-knockout (cS5^fl/fl)). Scale bar = 100 μm.

Figure 4

Impact of the stepwise Smarca5 deletion on cell growth and cell viability upon genotoxic insults. (A) Growth curve for the Smarca5 knockout MEFs when grown in culture for prolonged periods of time, (left): single allele knockout, (right): double allele knockout. Red dotted lines indicate the point of 4-OHT induction, arrowheads indicate the onset of senescence-like state (SLS). (B) Cell viability of single and double allele knockout Smarca5 MEFs as measured by MTS assay upon exposure to different doses of AA in triplicates (left) and MNNG (right). NOTE: p-values in (B) are based on paired t-test for each genotype group.

Figure 5

Large-scale analysis of the Smarca5 requirement for primary cell immortalization. (A) A schema showing senescence bypass and clonal immortalization in MEFs (top panel) and the representative cell culture images from the Smarca5 single and double allele knockout (bottom panel). Black arrowheads in the middle panels indicate the SA-β-gal staining of cells with typical senescent morphology. Scale bar = 50 μm. (B) Immortalization setup for Smarca5 knockout MEFs, twenty-five flasks for each condition were cultured over periods of several weeks until they bypassed senescence or died, to check the ability of the MEFs to overcome senescence and immortalize; created with BioRender.com. (C) Bar graphs show the total number of cultures out of twenty-five starting cultures that managed to immortalize when cells were grown for prolonged periods of time. NOTE: the P-value in (C) reflects the significance of difference between the conditions and it was calculated based on Χ² test with the degree of freedom equal to 3.