ACTL6A suppresses p21Cip1 tumor suppressor expression to maintain an aggressive mesothelioma cancer cell phenotype

Mesothelioma is a poor prognosis cancer of the mesothelial lining that develops in response to exposure to various agents including asbestos. Actin-Like Protein 6A (ACTL6A, BAF53a) is a SWI/SNF regulatory complex protein that is elevated in cancer cells and has been implicated as a driver of cancer cell survival and tumor formation. In the present study, we show that ACTL6A drives mesothelioma cancer cell proliferation, spheroid formation, invasion, and migration, and that these activities are markedly attenuated by ACTL6A knockdown. ACTL6A expression reduces the levels of the p21Cip1 cyclin-dependent kinase inhibitor and tumor suppressor protein. DNA binding studies show that ACTL6A interacts with Sp1 and p53 binding DNA response elements in the p21Cip1 gene promoter and that this is associated with reduced p21Cip1 promoter activity and p21Cip1 mRNA and protein levels. Moreover, ACTL6A suppression of p21Cip1 expression is required for maintenance of the aggressive mesothelioma cancer cell phenotype suggesting that p21Cip1 is a mediator of ACTL6A action. p53, a known inducer of p21Cip1 expression, is involved ACTL6A in regulation of p21Cip1 in some but not all mesothelioma cells. In addition, ACTL6A knockout markedly reduces tumor formation and this is associated with elevated tumor levels of p21Cip1. These findings suggest that ACTL6A suppresses p21Cip1 promoter activity to reduce p21Cip1 protein as a mechanism to maintain the aggressive mesothelioma cell phenotype.


INTRODUCTION
Mesothelioma is an aggressive asbestos-triggered cancer arising from the mesothelial layer of the pleura and peritoneum that is considered incurable and is associated with poor life expectancy [1,2]. Surgical resection and chemotherapy are standard of care and drug resistance and cancer recurrence is a common problem [2,3]. Our goal is to identify proteins that promote and maintain the aggressive mesothelioma cancer cell phenotype as potential therapy targets [4,5].
In addition, ACTL6A stabilizes the YAP1/TAZ pro-cancer transcriptional regulators, which are part of the Hippo signaling cascade [18,19], and reduces expression of the p21 Cip1 tumor suppressor [20][21][22][23]. We are interested in the role of ACTL6A in mesothelioma, as YAP1/TAZ function is required for optimal mesothelioma cancer cell survival [24]. Moreover, expression of the p21 Cip1 cyclin-dependent kinase inhibitor is induced by various anti-cancer agents and p21 Cip1 expression is associated with reduced mesothelioma cancer cell survival [25][26][27]. Our present study suggests that ACTL6A maintains the aggressive mesothelioma cancer phenotype by interacting with the p21 Cip1 gene promoter to inhibit transcription and reduce p21 Cip1 expression. Moreover, elevated p21 Cip1 is required to suppress the cancer phenotype, as the phenotype persists when the p21 Cip1 increase is inhibited in ACTL6A deficient cells. Our findings support a model where ACTL6A interacts with the Sp1 and p53 response elements in the p21 Cip1 promoter to reduce p21 Cip1 expression, and that loss of p21 Cip1 tumor suppressor enhances the mesothelioma cancer phenotype.

RESULTS
ACTL6A maintains the aggressive mesothelioma cancer phenotype ACTL6A has been reported to maintain an aggressive phenotype in cancer cells [9][10][11][12]. Enhanced spheroid formation, invasion through Matrigel and migration on plastic are properties associated with the aggressive mesothelioma cancer phenotype [4] and so we measured the impact of ACTL6A on these properties. As shown in Fig. 1A, ACTL6A knockdown is associated with increased levels of p21 Cip1 and this is associated with reduced spheroid size and number, invasion and migration ( Fig. 1B-D). To further characterize the role of ACTL6A, we characterized ACTL6A knockout cells. The ACTL6A-KOc1-1 and ACTL6A-KOc1-5 clonal cell lines display a marked increase in p21 Cip1 expression (Fig. 1E) and this is associated with reduced cell proliferation, spheroid number and diameter, and cell invasion (Fig. 1F-H). We next performed the inverse experiment and restored ACTL6A expression in ACTL6A knockout cells. We show that restoring ACTL6A level is Fig. 1 ACTL6A maintains the aggressive cancer phenotype. A ACTL6A was successfully knocked down using ACTL6A-siRNA which leads to an increase in p21 Cip1 protein level. B-D ACTL6A knockdown decreases spheroid number and size, invasion and migration. For spheroid assay, 40,000 cells were seeded on ultra-low attachment plates and counted at Day 4. Spheroid diameters were measured using ImageJ. E-H ACTL6A knockout cell clones, which lack ACTL6A, display elevated p21 Cip1 expression and this is associated with reduced cell proliferation, spheroid formation and invasion. I, J ACTL6A overexpression in ACTL6A knockout cell clones reduces p21 Cip1 protein and mRNA levels. Asterisks indicate a significant decrease (n = 3, p ≤ 0.001). associated with reduced p21 Cip1 mRNA and protein expression (Fig. 1I, J) suggesting that ACTL6A controls p21 Cip1 level by regulating p21 Cip1 RNA turnover or gene transcription. p21 Cip1 suppresses the peritoneal and pleural mesothelioma cancer cell phenotype The studies in Fig. 1 show that the p21 Cip1 cyclin-dependent kinase inhibitor level is markedly increased in ACTL6A knockdown and knockout cells, suggesting that p21 Cip1 may have a role in suppressing the cancer phenotype. To test this, we treated Meso-1 cells with control-and p21 Cip1 -siRNA and monitored the impact of p21 Cip1 loss on the cancer phenotype. Figure 2A confirms p21 Cip1 knockdown and shows that loss of p21 Cip1 minimally impacts ACTL6A level. Figure 2B, C show that p21 Cip1 knockdown enhances cancer cell aggression as evidenced by increased spheroid formation and invasion. To understand the relationship between ACTL6A and p21 Cip1 level, and the cancer cell phenotype, we treated cells with combinations of ACTL6Aand p21 Cip1 -siRNA and monitored the impact on cell invasion. Figure 2D shows that ACTL6A knockdown increases p21 Cip1 and Fig. 2E shows that this is associated with reduced invasion. Moreover, preventing the increase in p21 Cip1 in ACTL6A Fig. 2 p21 Cip1 suppresses the cancer phenotype. A-C p21 Cip1 knockdown enhances spheroid formation and invasion. D, E ACTL6A knockdown increases p21 Cip1 protein level and reduces invasion, and these actions are partially reversed following simultaneous p21 Cip1 -siRNA and ACTL6A knockdown. F-I ACTL6A knockout lines display enhanced p21 Cip1 levels and reduced spheroid formation. p21 Cip1 knockdown in these lines increases spheroid formation, invasion, and migration. This suggests that p21 Cip1 is a mediator of ACTL6A suppression of the cancer phenotype. Single asterisks indicate a significant decrease (n = 3, p ≤ 0.001) and double asterisks indicate a significant increase (n = 3, p ≤ 0.001).
knockdown cells (using p21 Cip1 -siRNA) (Fig. 2D) partially restores cell invasion (Fig. 2E), suggesting that the increase in p21 Cip1 is responsible for suppression of the cancer phenotype in ACTL6A knockdown cells. We also studied the role of p21 Cip1 using ACTL6A knockout cells. Figure 2F shows, as expected, that p21 Cip1 level is elevated in ACTL6A knockout cells and that p21 Cip1 is reduced by treatment with p21 Cip1 -siRNA. Figure 2G-I show that p21 Cip1 knockdown in the ACTL6A knockout environment partially restores spheroid formation, invasion, and migration. These findings suggest that ACTL6A suppresses p21 Cip1 to enhance the cancer phenotype.
We next examined the role of ACTL6A and p21 Cip1 in regulating the cancer phenotype in another mesothelioma cell line, NCI-Meso-17 cells. Figure 3A-D shows that ACTL6A knockdown increases p21 Cip1 level and that this is associated with reduced NCI-Meso-17 cell spheroid formation, invasion, and migration. Moreover, as shown in Fig. 3E-H suppression of ACTL6A increases p21 Cip1 and reduces the cancer phenotype, p21 Cip1 knockdown enhances the cancer phenotype, and suppression of both ACTL6A and p21 Cip1 produces an intermediate phenotype for spheroid formation, invasion, and migration.

ACTL6A control of p21 Cip1 expression
To understand the mechanism whereby ACTL6A regulates p21 Cip1 expression, we compared p21 Cip1 mRNA level in wild-type, ACTL6A knockdown and ACTL6A knockout Meso-1 cells. ACTL6A knockdown leads to an increase in p21 Cip1 protein and mRNA (Fig.  4A, B), a finding that is consistent with evidence that restoring ACTL6A in ACTL6A knockout cells reduces p21 Cip1 protein and mRNA (Fig. 1J). In addition, Fig. 4C shows that p21 Cip1 mRNA level Fig. 3 ACTL6A maintains the NCI-Meso-17 cancer cell phenotype. A-D ACTL6A knockdown increases p21 Cip1 protein level and is associated with reduced spheroid formation, invasion, and migration. E-H ACTL6A knockdown increases p21 Cip1 levels and reduces spheroid formation, invasion, and migration and p21 Cip1 knockdown partially reverses the ACTL6A-knockdown dependent reduction in spheroid formation, invasion, and migration. These findings suggest that ACTL6A suppresses p21 Cip1 to maintain the aggressive NCI-Meso-17 cell cancer phenotype. Single asterisks indicate a significant decrease (n = 3, p ≤ 0.001) and double asterisks indicate a significant increase (n = 3, p ≤ 0.005).
is markedly increased in ACTL6A knockout cells which is likely due to altered mRNA turnover or increased gene expression [28,29].
We next examined the effect of ACTL6A on p21 Cip1 transcription. The p21 Cip1 promoter map (Fig. 4D) shows the six Sp1 sites located proximal to the transcription start site and the more distally located p53-1 and p53-2 sites [28,29]. Activity of the p21-2326 (full-length) and p21-124 (proximal) promoter constructs are markedly increased in ACTL6A knockout cells suggesting activation of transcription (Fig. 4E). In contrast, the p21-2326 Sp1(Δ1-6) construct, in which the six Sp1 sites are mutated, is inactive, Fig. 4 ACTL6A suppresses p21 Cip gene transcription to maintain the mesothelioma cancer phenotype. A, B ACTL6A knockdown increases Meso-1 cell p21 Cip1 mRNA expression and protein level. C ACTL6A knockout Meso-1 cells express increased p21 Cip1 mRNA. D Map of the fulllength p21 Cip1 promoter showing the p53-1 and p53-2 sites in the distal promoter and six Sp1 site in the proximal promoter. The top sequence indicates the authentic p53 site sequences. Our previous report indicates the mutations that inactivate these sites [28]. E ACTL6A knockout Meso-1 cells display enhanced p21 Cip1 promoter activity for the full-length (p21-2326) and proximal (p21-124) promoter, and mutation of the six Sp1 sites eliminates the response. F Chromatin IP, using Meso-1 extracts, shows ACTL6A binding to the p21 Cip1 promoter p53 binding sites and the proximal promoter Sp1 sites. G-I ACTL6A knockdown in NCI-Meso-17 cells increases p21 Cip1 mRNA level and p21 Cip1 promoter activity and this is associated with ATL6A binding to the p21 Cip1 promoter p53 and Sp1 sites. As observed for Meso-1 cells, mutation of the proximal Sp1 sites eliminates transcriptional activity. Double asterisks indicate a significant increase (n = 3, p ≤ 0.001).
suggesting that the Sp1 elements are required for basal promoter activity and ACTL6A regulation of transcription. We next measured ACTL6A interaction at the proximal Sp1 sites, and the p53-1 and p53-2 sites because they constitute major p21 Cip1 promoter response elements [22,28,[30][31][32]. ChIP analysis shows that ACTL6A is enriched on these regulatory elements (Fig. 4F) suggesting that ACTL6A interacts at these sites to suppress p21 Cip1 transcription.
We next examined ACTL6A regulation of p21 Cip1 level in NCI-Meso-17 cells [33]. We show that ACTL6A knockdown increases p21 Cip1 mRNA level (Fig. 4G) and promoter activity (Fig. 4H). In addition, the lack of activity of the p21-2326 Sp1(Δ1-6) construct ( Fig. 4H), where all six Sp1 sites are inactivated, confirms that the promoter proximal Sp1 sites are required for basal and ACTL6A regulated promoter activity. ChIP analysis (Fig. 4I) shows that ACTL6A interacts at the p21 Cip1 proximal Sp1 elements and the p53-1 and p53-2 sites. These findings strongly suggest that ACTL6A regulates p21 Cip1 gene expression via a transcriptional mechanism in both Meso-1 and NCI-Meso-17 cells.
Role of p53 p53 is a well characterized regulator of p21 Cip1 expression that binds at the p53 sites on the p21 Cip1 promoter to increase gene expression [28,[30][31][32]. It is thus possible that ACTL6A loss leads to an increase in p53 which stimulates p21 Cip1 transcription. We therefore examined the role of p53. Figure 5A shows that ACTL6A-siRNA reduces Meso-1 cell ACTL6A mRNA and that this is associated with an increase in p53 and p21 Cip1 mRNA. Moreover, this is also associated with an increase in p21 Cip1 protein (Fig. 5B). However, a surprising finding is that p53 protein is not detected in ACTL6A expressing or knockdown cells (Fig. 5B) which is consistent with the lack of p53 binding to the p21 Cip1 promoter (Fig. 5C) and the fact that treatment with p53-siRNA does not alter Meso-1 spheroid formation or invasion (Fig. 5D). To provide additional evidence for the lack of p53 protein in Meso-1 cells, we confirmed an absence of p53 in ACTL6A knockout cells (Fig.  5E). It is interesting that p53 protein is absent even though p53 mRNA levels are increased in ACTL6A knockout cells (Fig. 5F). Taken together, these findings indicate that p53 is absent in Meso-1 cells and therefore has no role in regulating p21 Cip1 gene expression.
We next examined the role of p53 as a regulator of p21 Cip1 expression in NCI-Meso-17 cells. In contrast to Meso-1 cells, NCI-Meso-17 cells express p53, and p53 protein level is increased following ACTL6A knockdown (Fig. 5G). Moreover, treatment with p53-siRNA attenuates the p53 increase in ACTL6A knockdown cells (Fig. 5G). This suggests that p53 may have a role in mediating the increase in p21 Cip1 following ACTL6A knockdown. If p53 has a role in this induction, we would anticipate that it should interact with the p21 Cip1 promoter. In Fig. 5H we confirm that both ACTL6A and p53 interact with the p21 Cip1 promoter p53 binding sites, but that only ACTL6A interacts at the proximal promoter Sp1 sites. We next examine the impact of ACTL6A, p53 and combined knockdown on NCI-Meso-17 cell spheroid formation and invasion. Figure 5I, J show that ACTL6A knockdown reduces and p53 knockdown increases spheroid formation and invasion, and that combined ACTL6A and p53 knockdown reduces spheroid formation and invasion. We also examined the impact of ACTL6A knockdown on p53 mRNA level in NCI-Meso-17 cells and found no impact (Fig.  5K). Taken together, these findings are consistent with a role for p53 in mediating ACTL6A suppression of p21 Cip1 levels in NCI-Meso-17. This suggests two potential mechanisms of ACTL6A action. First, the loss of ACTL6A increases p53 protein level (via stabilization or enhanced translation) which stimulates p21 Cip1 gene expression. Second, reduced ACTL6A binding to the p53 sites on the p21 Cip1 promoter, following ACTL6A knockdown, likely permits p53 to drive p21 Cip1 promoter activity in an unimpeded manner.
ACTL6A is required for optimal tumor growth To assess the role of ACTL6A in tumor formation, we examined the ability of monolayer cultured wild-type and ACTL6A null cells to form tumors in immune-compromised mice. Figure 6A, B show that ACTL6A knockout cells (Meso1-ACTL6A-KOc1-5-1) display reduced tumor growth. Figure 6C shows that loss of ACTL6A in tumors is associated with increased p21 Cip1 expression, a result that is consistent with the findings from cultured cells. However, we were surprised to detect residual low level ACTL6A expression in Meso1-ACTL6A-KOc1-5-1 cells. To confirm that ACTL6A knockout status of these cells, we harvested cells from Meso-1 and Meso1-ACTL6A-KOc1-5-1 tumors and grew them in cell culture for 2 weeks prior to extract preparation for ACTL6A immunoblot. As shown in Fig. 6D, no ACTL6A was detected in the cultured Meso1-ACTL6A-KOc1-5-1 cells. This confirms that Meso1-ACTL6A-KOc1-5-1 cells retained the ACTL6A-null phenotype and that the ACTL6A detected in Fig. 6C is derived from connective tissue cells in the tumor.
ACTL6A and p21 Cip1 in patient tumor samples We next checked ACTL6A and p21 Cip1 levels in patient tumors. Figure 6E compares ACTL6A and p21 Cip1 levels in NCI-Meso-17 cells with tumor samples derived from three pleural mesothelioma patients. The highest level of ACTL6A is observed in NCI-Meso-17 cells which are negative for p21 Cip1 expression. In contrast, ACTL6A levels are lower in the patient samples and this is associated with basal levels of p21 Cip1 expression. Based on our model, where high ACTL6A expression reduces p21 Cip1 , we interpret these results as suggesting that the tumor samples express enough ACTL6A to partially, but not completely suppress, p21 Cip1 expression.

DISCUSSION
ACTL6A enhances the mesothelioma cancer phenotype Mesothelioma is an aggressive and incurable disease that is associated with poor life expectancy [1,2]. This disease profile has stimulated efforts to identify new treatment targets. ACTL6A has been reported to drive cancer progression by stabilizing the YAP1/ TAZ pro-cancer transcriptional regulators [18,19], and by suppressing expression of the p21 Cip1 cyclin-dependent kinase inhibitor [20][21][22][23]. Moreover, ACTL6A maintains the stem-like phenotype in keratinocytes [13], stimulates neural progenitor cell proliferation [34,35], enhances the osteosarcoma and hepatocellular carcinoma cell cancer phenotype [16,17], and drives an aggressive epidermal squamous cell carcinoma phenotype [22]. To our knowledge ACTL6A function has not been studied in mesothelioma; however, it is important to examine the role of ACTL6A in mesothelioma, since it regulates key pathways, including YAP1/TAZ/TEAD and p21 Cip1 signaling, that have been implicated as important in mesothelioma cancer progression [24][25][26][27].
We show that transient ACTL6A knockdown or stable knockout reduces mesothelioma cell proliferation, spheroid formation, invasion, and migration. This suggests that ACTL6A maintains an aggressive mesothelioma cancer phenotype. To understand the mechanism of action, we show that p21 Cip1 level is suppressed in ACTL6A competent cells and is increased in ACTL6A knockout cells, suggesting that ACTL6A may maintain the cancer phenotype by suppressing p21 Cip1 expression. A functional role for p21 Cip1 was confirmed by showing that suppressing p21 Cip1 level in ACTL6A knockdown or knockout cells, attenuates the anti-cancer response associated with ACTL6A loss. A key role of p21 Cip1 was confirmed by showing that p21 Cip1 knockdown results in increased cancer cell invasion and migration thereby confirming that the increase in p21 Cip1 in ACTL6A negative cells suppresses the cancer phenotype. These findings strongly suggest that p21 Cip1 is a biologically important downstream mediator of ACTL6A action.  5 Role of p53 in mediating ACTL6A regulation of p21 Cip1 expression. A, B ACTL6A-knockdown increases p53 and p21 Cip1 mRNA levels and p21 Cip1 protein level, but p53 protein is not detected in ACTL6A competent or knockdown Meso-1 cells. C Chromatin IP assay shows ACTL6A binding to the distal p53 binding sites and the proximal promoter Sp1 binding sites in Meso-1 cell extracts, but no binding of p53 is detected. D p53 knockdown does not alter Meso-1 cell spheroid number or invasion. E, F p53 protein is absent in ACTL6A knockout cells even though p53 mRNA levels are increased. G ACTL6A knockdown increases p53 in NCI-Meso-17 cells. H ACTL6A binds to the p21Cip1 promoter Sp1 and p53 sites. p53 binds to the p53 binding sites (p21-p53 sites 1 and 2) but not the Sp1 binding sites (p21-proximal). I, J ACTL6A knockdown reduces spheroid formation and invasion, p53 knockdown increases spheroid formation and invasion, and combination treatment with ACTL6A-and p53-siRNA does not alter ACTL6A suppression of spheroid formation and invasion. This suggests that ACTL6A suppresses the ability of p53 to activate p21 Cip1 promoter transcription. K ACTL6A knockdown does not increase p53 mRNA level. Single asterisks indicate a significant decrease (n = 3, p ≤ 0.001) and double asterisks indicate a significant increase (n = 3, p ≤ 0.001). All statistical comparisons are to the control group. Single asterisks indicate a significant decrease (n = 3, p ≤ 0.001) and double asterisks indicate a significant increase (n = 3, p ≤ 0.001).

ACTL6A regulation of p21 Cip1 level
Sp1 and p53 transcription factors are key regulators of p21 Cip1 gene expression that bind to key DNA elements on the p21 Cip1 promoter to increase expression [30,36]. We therefore examined ACTL6A interaction at these sites. Mechanistic studies show that in Meso-1 cells ACTL6A binds to the p21 Cip1 gene Sp1 site-rich proximal promoter and the distal p53-1 and 53-2 sites, and that this is associated with reduced p21 Cip1 mRNA and protein levels. This is consistent with a role for ACTL6A as a dominant suppressor of p21 Cip1 expression. Since, p53 is an important factor that binds to the p21 Cip1 promoter to activate transcription, we assessed its role. Our findings show that ACTL6A loss in Meso-1 cells is associated with increased p53 mRNA, but that p53 protein is not produced and is not found bound to the p21 Cip1 promoter p53 binding sites. Thus, p53 does not have a p21 Cip1 regulatory role in Meso-1 cells. This finding in Meso-1 cells is consistent with a previous report showing p53 is not required for ACTL6A regulation of p21 Cip1 in epidermal squamous cell carcinoma cells [22]. Taken together, these finding suggest that in Meso-1 cells ACTL6A interaction at these sites suppresses ability of other, presently unknown, transcription factors to activate p21 Cip1 expression.
ACTL6A was also found to suppress p21 Cip1 expression in NCI-Meso-17 cells where it is bound to the p21 Cip1 gene promoter at Sp1 and the p53 binding sites where it likely acts to suppress p21 Cip1 expression. Moreover, ACTL6A knockdown resulted in an increase in p21 Cip1 and an attenuated cancer phenotype. These findings suggest that ACTL6A binds to the Sp1 and p53 binding sites to reduce the ability of other transcription factors to activate expression. However, in contrast to Meso-1 cells, we found that NCI-Meso-17 cells do express p53 protein. We show that although p53 mRNA levels are not increased in NCI-Meso-17 ACTL6A knockdown cells, p53 protein level is increased. p53 and ACTL6A were both found to be able to bind at the p21 Cip1 promoter p53 binding sites. Taken together, these findings suggest that ACTL6A Fig. 6 ACTL6A suppresses p21 Cip1 level to maintain tumor growth. A, B Wild-type (Meso-1) and ACTL6A knockout (Meso1-ACTL6A-KOc1-5-1) cells (3 million/site) were injected into each front flank in five NSG mice per treatment group (10 tumor/group) and tumor formation was monitored for 0-6 week. The images are two paired sets of tumors harvested at 6 weeks. Differences in tumor formation were analyzed using the Student's t test and values are presented as mean ± SEM and p values are indicated. Single asterisks indicate a significant decrease (n = 10 tumor, p ≤ 0.001). C Meso-1 cell tumors display low levels of p21 Cip1 while p21 Cip1 levels are elevated in Meso1-ACTL6A-KOc1-5-1 cell tumors. D Meso-1 and Meso1-ACTL6A-KOc1-5-1 tumors cells were dissociated and grown in culture for detection of ACTL6A and p21 Cip1 . These blots show that the Meso1-ACTL6A-KOc1-5-1 cells retained their ACTL6A-null status during tumor formation. E ACTL6A and p21 Cip1 levels were compared between NCI-Meso-17 cells and tumors isolated from three patients. F Schematic describing the role of ACTL6A in regulating cancer cell function. ACTL6A binds to the p21 Cip1 promoter p53-1 and 2 sites and the Sp1 sites to suppress p21 Cip1 transcription. The resulting reduction in p21 Cip1 level leads to enhanced cell proliferation, spheroid formation, Matrigel invasion, migration, and tumor formation.
suppresses p53 protein level. In addition, ACTL6A binds to the p21 Cip1 promoter p53 binding site where it likely suppresses p53dependent activation of p21 Cip1 transcription. This is consistent with the previous evidence that ACTL6A can interfere with p53 function [23]. Additional studies will be required to fully understand the relationship between ACTL6A, p53 and p21 Cip1 in NCI-Meso-17 cells.
The role of the p21 Cip1 proximal promoter Sp1 sites We have previously identified an important role for the cluster of six Sp1 sites located in the proximal p21 Cip1 promoter [28]. The present studies indicate that these sites must be intact for p21 Cip1 basal transcription since mutation of the Sp1 sites eliminates promoter activity. Thus, these sites are required for basal promoter activity and are also likely important for ACTL6A suppression of transcription, as chromatin IP analysis reveals ACTL6A binding at these sites in both Meso-1 and NCI-Meso-17 cells. Although it appears that ACTL6A binding to these sites may inhibit Sp1 activation of p21 Cip1 transcription, additional studies will be necessary for a detailed description of the mechanism of action. It is important to note that we have focused our studies on major regulatory sites (p53-1, p53-2, and the Sp1 cluster) that are known to drive p21 Cip1 transcription. However, additional studies will ultimately be necessary, as it is very likely that ACTL6A can interact with other transcriptional regulators that bind to other regions of the p21 Cip1 promoter and that these interactions will influence p21 Cip1 levels.

ACTL6A regulation of tumor formation
To assess the biological relevance of these findings, we tested the impact of ACTL6A knockout on tumor formation. Wild-type Meso-1 cells and ACTL6A knockout cells (Meso1-ACtL6A-KOc1-5-1) cells were injected into immunocompromised mice to monitor the impact of ACTL6A loss on tumor formation. We observe a dramatic reduction in tumor size for ACTL6A knockout tumors. Biochemical analysis of tumor extracts confirms ACTL6A knockout and shows that this is associated with increased levels of p21 Cip1 . Thus, ACTL6A loss results in increased p21 Cip1 levels and this is associated with an attenuated cancer phenotype. The fact that ACTL6A knockout leads to increased p21 Cip1 levels and reduced tumor growth argues that ACTL6A suppresses p21 Cip1 to maintain aggressive tumor growth. These findings are completely consistent with the experiments in cultured cells and show that ACTL6A operates in the tumor context to suppress p21 Cip1 level to drive an aggressive cancer phenotype.
Based on these findings we propose the model shown in Fig. 6F where ACTL6A interacts at the p53 and Sp1 transcription factor binding sites to interfere with positive transcriptional regulators (such as p53) to reduce p21 Cip1 promoter expression and enhance the cancer phenotype.

Statistics
The two-tailed Student's t test was used for statistical assessment. The values in plots are mean ± SEM. Single and double asterisks indicate, respectively, a significant reduction and increase in response using the Student's t test based on triplicates for cell cultures studies and 5 mice/ group for animal tumor growth assays. No data was excluded from analysis.

Cell culture and bioassays
Meso-1 and NCI-Meso-17 cells are derived, respectively, from peritoneal and pleural mesothelioma [5,33] and mycoplasma testing is performed when new cell stocks are thawed for use. RPMI1640 medium (27519003) containing L-glutamine, penicillin-streptomycin solution, and 0.25% trypsin-EDTA were purchased from Gibco (Gaithersburg, MD). Growth medium is RPMI1640 supplemented with penicillin-streptomycin and 5% fetal bovine serum (FBS, 19B416) which was purchased from Sigma (St. Louis, MO). To study cell growth, 20,000 cells were plated into 35 mm plates in triplicate in growth medium and cell number was monitored from 0-2 days. For spheroid formation assay, cells were plated at 40,000 cells per 35 mm well in ultra-low attachment dishes [37,38]. To study invasion, 25,000 cells were plated into Millicell (1 cm diameter, 8 mm pore-size) chambers (353097) atop a 100 μl layer of 250 μg/ml matrigel (354234) [39]. Millicell chambers and matrigel were purchased from BD Biosciences (San Diego, CA). Growth medium containing 1% FCS (top chamber) or 10% FCS (bottom chamber) was added and the ability of cells to pass through the membrane was monitored over 0-18 h. The membrane was fixed with 4% paraformaldehyde and stained with DAPI (D9542, Sigma Aldrich, Milwaukee, WI) and fluorescent nuclei were detected by fluorescence microscopy [39]. For migration assay, scratch wounds were created in confluent monolayer cultures of cells using a pipette tip and closure of the wound was monitored from 0 to 48 h.

Electroporation
Cells (1.2 million), suspended in 100 μl of nucleofection reagent (VPD-1002, Lonza, Williamsport, PA) containing 3 μg of control-or target specific-siRNA, were electroporated using the AMAXA Electroporator on the T-018 setting [40]. The cells were permitted to recover for 48 h and then electroporation was repeated. After a 12 h recovery, the cells were plated to assay spheroid formation, invasion, and migration [37,38]. For ACTL6A-FLAG expression studies, cells were single electroporated with 1 µg of pCMV3-ACTL6A-FLAG expression vector and ACTL6A-FLAG expression and biological responses were monitored after 48 h.

Immunology related methods
Immunoblot was performed as previously described [28]. Chromatin immunoprecipitation (ChIP) was performed as described in the Diagenode LowCell# ChIP kit (kch-maglow-G48, Diagenode, Inc., Denville, NJ) protocol. Cells (1 million) were harvested and washed with phosphate buffered saline. Sodium butyrate was omitted from the protocol. The cells were crosslinked with 1% formaldehyde for 10 min, quenching with 125 nM glycine and washed with phosphate-buffered saline before lysis in Diagenode lysis buffer supplemented with 1x protease inhibitor. The samples were chilled and sonicated using a 550 Sonic Dismembrator (five 30 s pulses at 30% amplitude with 30 s between pulses) to obtain 200-1000 bp DNA fragments of sheered DNA. The samples were diluted in Diagenode buffer containing 1x protease inhibitor. ChIP grade antibodies (1 μg) were incubated with protein G coated magnetic beads for 3 h at 4°C, and sheared chromatin (100,000 cells equivalent) was added for overnight 4°C incubation. After 24 h, the DNA was extracted using Diagenode DNA isolation buffer containing proteinase K. p21 Cip1 promoter DNA were detected by qPCR using LightCycler 480 SYBR green I master mix and sequence specific primers. p21 Cip1 promoter primers targeted the proximal Sp1 elements (forward 5′-GCTGGGCAGCCAGGAGCCTG and reverse 5′-CTGCTCACACCTCAGCTGGC), the p53-1 site (forward 5′-GTGGCTCTGATTGGCTTTCTG and reverse 5′-CTGAAAACAGGCAGCCCAAG) and the p53-2 site (forward 5′-CCGAGGTCAGCTGCGTTAGAGG and reverse 5′-AGAACCCAGGCTTGGAGCAGC).

Tumor xenografts
Cancer cells (3 million), derived from monolayer cultures, were resuspended in 200 µl of phosphate-buffered saline containing 30% matrigel and 100 µl was injected subcutaneously into each front flank of five 8week-old NOD/scid/IL2 receptor γ chain knockout mice (NSG) per treatment group. Tumor growth was monitored by measuring tumor diameter and calculating tumor volume = 4/3π × (diameter/2)3. Tumor samples were harvested to prepare extracts for immunoblot [4]. The twotailed Student's t test was used for statistical analysis. Power analysis, assuming a power of 0.8, standard deviation of 12%, and a minimum difference in means of 30%, predicts that five animals per group (2 tumors/ mouse = 10 tumors) are sufficient to provide statistically relevant outcomes. Human pleural mesothelioma tumor samples were obtained from patients as discarded tissue samples.