PR55α-controlled protein phosphatase 2A inhibits p16 expression and blocks cellular senescence induction by γ-irradiation

PR55α inhibits the expression of p16 in human pancreatic normal and cancer cells

The p16 tumor suppressor is a major CDK4/6 inhibitor that plays an essential role in the induction of cellular senescence [30]. The loss of p16 occurs in >90% of pancreatic cancer and has been identified as a driver mutation in pancreatic cancer tumorigenesis [31]. We have previously reported an essential role for PR55α in maintaining the tumorigenicity and metastatic potential of human pancreatic cancer cells [23, 27]. Here, we assessed the impacts of PR55α on the expression and function of p16 in both HPNE cells and CD18/HPAF cells. HPNE cells are human normal pancreatic ductal cells that we previously immortalized with telomerase [32, 33]. CD18/HPAF cells are human pancreatic cancer cells driven by the KRAS oncogene and mutant p53^P151S [34].

Through retroviral transductions, stable lines of normal HPNE cells were engineered to express a Dox-inducible PR55α, while stable lines of CD18/HPAF cells were engineered to express a Dox-inducible PR55α-shRNA. Since p16 plays an essential role in most forms of cellular senescence, from replicative senescence to stress-induced premature senescence, we assessed the effects of PR55α on p16 expression and premature senescence-induced by IR in the context of both normal cells and cancer cells of the pancreas. To induce senescence in these experiments, we have used a high dose of radiation (10 Gy), similar to doses used in extreme hypofractionation radiation therapy, now made possible by advances in stereotactic radiotherapy [35, 36]. In radioresistant forms of cancer, such as pancreatic cancer, this approach may be producing better outcomes.

In the normal HPNE cells, PR55α induction abolished the expression of p16 (Figure 2A). After exposure to γ-irradiation, p16 expression was slightly increased in both the control and PR55α-overexpressing HPNE cells. However, the magnitude of the minor increase of p16 induced by IR was the same in the control cells and PR55α-overexpressing cells (Figure 2A). For validation, we tested the effect of PR55α on p16 expression in CD18/HPAF pancreatic cancer cells, which harbor a p53^P152S mutant thus lacking a functional p53/p21 pathway [34]. Although CD18/HPAF cells express a p16 protein that contains an in-frame deletion from amino acids (aa) 29 to 34 [37], this mutation does not appear to affect p16 mRNA stability, protein expression, or even p16 function since it does not interfere with the ability of p16 to bind and inhibit CDK4/6, a function mediated by its Ankyrin repeats AR2 (aa 36–69) and AR3 (aa 70–100) [38]. As predicted, the knockdown of PR55α by shRNA in CD18/HPAF cells resulted in the opposite effect compared to the overexpression of PR55α in the normal HPNE cells. Again, the high levels of PR55α expression were associated with low levels of p16 (Figure 2B). IR exposure also resulted in a subtle increase in p16 levels in CD18/HPAF cells over time, but the magnitude of this induction was the same in the control and PR55α-knockdown cells (Figure 2B). It was also noticed that there was a transient increase of the p16 level in control cells at 2 hours following IR. The cause of the effect is unknown, which could be due to a temporary increase in the protein stability of p16 prompted by IR-activated stress signaling.

Figure 2. PR55α suppresses p16 protein expression in normal and cancer cells originating from the human exocrine pancreas. (A) Human pancreatic normal ductal (HPNE) cells were stably transduced with a retroviral vector expressing a Dox-inducible PR55α cDNA (PR55α). As a control, HPNE cells stably transduced with a relevant empty retroviral vector were included in the analysis (Control). Following the induction of ectopic PR55α expression by Dox (1 μg/ml) for 2 days, cells were exposed to 10 Gy of ionizing radiation (IR), incubated for the indicated hours, and analyzed by immunoblotting for the levels of p16, p14, p53, and p21. GAPDH in the lysates was measured as an internal control. (B) Human pancreatic ductal adenocarcinoma cells (CD18/HPAF) were stably transduced with a lentiviral vector expressing a Dox-inducible shRNA against either PR55α (PR55α-shRNA) or an irrelevant negative control (Control-shRNA). Following induction of the shRNA with Dox (2 μg/ml) for 5 days, the cells were exposed to 10 Gy IR, incubated for the indicated hours, and analyzed by immunoblotting for the levels of the indicated proteins. GAPDH in the lysates was again used as an internal control. The difference in the p16 levels between HPNE-Control and HPNE-PR55α cells, as well as between the CD18/HPAF-Control-shRNA and CD18/HPAF-PR55α-shRNA cells were determined to be statistically significant (HPNE-Control vs. HPNE-PR55α, p < 0.001; CD18/HPAF-Control-shRNA vs. CD18/HPAF-PR55α-shRNA, p = 0.004).

Because p14 and p16 are essentially transcribed from the CDKN2A locus, albeit from distinct promoters (Figure 1), we assessed the effects of PR55α on p14 expression after IR. In both the HPNE and CD18/HPAF cells, the manipulation of PR55α levels did not result in a similar effect on p14 as it did on p16. In the HPNE cells, for example, PR55α overexpression abolished p16 expression but it slightly increased p14 levels. These results show that PR55α specifically inhibits p16 expression but not p14 expression.

PR55α enhances the response of the p53/p21 pathway to IR exposure

The p53/p21 pathway plays a key role in DNA damage-induced cell cycle checkpoint response and senescence induction [39]. Loss of p53 function via mutations is the most frequent event occurring across all cancer types and is a main driver of pancreatic cancer development and progression [40]. We previously reported on the role of p53 as a negative regulator of the protein stability of PR55α, whose function supports several oncogenic pathways, including ERK, β-catenin, c-Myc, and YAP [23–28]. Since p53′s role as a tumor suppressor counteracts that of PR55α in the promotion of oncogenic pathways, we examined the impacts of PR55α on the response of the p53/p21 pathway to IR.

As shown in Figure 2A, ectopic PR55α expression in HPNE cells caused a 3.2-fold elevation in the steady-state level of p53, as well as markedly enhanced p53 induction by IR in HPNE cells. Within 2 h post-IR, p53 was induced in both the HPNE/control and HPNE/PR55α cells, but the induction was 6–8 fold higher in the latter (Figure 2A). To validate these results, we assessed the effects of PR55α on p21 expression, which is known to be induced by the transcriptional activity of p53 [41]. Consistently, ectopic PR55α expression in HPNE cells led to higher steady-state levels of p21, which is in line with its effect on p53 (Figure 2A). Like p53 levels, p21 levels were also induced by IR, albeit with a kinetic that was delayed by 2 hours compared to the induction of p53 (Figure 2A). Similar to p53, the magnitude of the induction of p21 by IR was higher in HPNE/PR55α cells compared to the HPNE/control cells (Figure 2A).

For comparison, we assessed the effect of PR55α level on mutant p53 expression in CD18/HPAF pancreatic cancer cells. The mutant p53 expressed in CD18/HPAF cells harbors a P151S mutation in the DNA binding domain of p53, thus abolishing its activity as a transcription factor [34]. As shown in Figure 2B, the knockdown of PR55α by shRNA did not affect the level of the p53^P151S mutant protein, either in the presence or absence of IR (Figure 2B), which was in contrast to the situation observed in HPNE cells known to express wild-type (wt) p53 protein (Figure 2A). In agreement with the lack of functional p53 in CD18/HPAF cells, levels of p21 were extremely low in these cells and were not particularly affected by either the knockdown of PR55α or exposure to IR (Figure 2B). These results indicate that PR55α only increases the expression of wt-p53, but not mutant p53. Taken together with our analyses of CDKN2A gene products, these results show that PR55α inhibits p16 expression while simultaneously enhancing the expression of wt p53. However, this enhancement of p53 expression is not associated with changes in p14 levels and is thus independent of p14.

PR55α suppresses p16 expression by inhibiting its mRNA transcription

To define the mechanism by which PR55α inhibits p16 expression, we first examined the impacts of PR55α on the protein stability of p16 using the cycloheximide (CHX)-chase assay, as described in our previous work [28]. CD18/HPAF cells expressing the control- or PR55α-shRNA were incubated with 15 μg/ml cycloheximide to block protein synthesis to allow measurements of p16 protein decay. As depicted in Figure 3A, PR55α knockdown by shRNA did not result in a significant change in the p16 protein half-life in CD18/HPAF cells, indicating a lack of effect of PR55α on the p16 protein decay process.

Figure 3. PR55α inhibits p16 expression by suppressing its mRNA transcription. (A) PR55α does not affect p16 protein stability. CD18/HPAF cells stably transduced with Dox-inducible PR55α-shRNA or Control-shRNA were cultivated in media containing 2 μg/ml Dox for 5 days to induce the shRNA, after which cells were exposed to cycloheximide (CHX, 15 μg/ml) to halt protein synthesis. Cell lysates collected at the indicated time points after CHX addition were analyzed for changes in p16 protein levels. The α-tubulin protein has a long half-life and was used as an internal control. Relative p16 protein levels were determined after normalization with the α-tubulin levels and these normalized values were used to calculate the protein half-life of p16. Half-lives were estimated by linear regression analysis of p16 normalized levels against time using SigmaPlot. (B) PR55α inhibits p16 mRNA expression. Left panel: HPNE expressing the Dox-inducible PR55α (HPNE/PR55α), or empty vector (HPNE/Control) was treated with Dox (1 μg/ml) for 3 days. Right panel: CD18/HPAF cells expressing the Dox-inducible PR55α shRNA (B4-2, B4-11) or Control shRNA (Control) were treated with Dox (2 μg/ml) for 5 days. At the end of each treatment, RNA was isolated and analyzed by quantitative RT-PCR for differences in p16 mRNA. The relative abundance of the p16 mRNA was calculated by normalizing the p16 mRNA levels with those of the GAPDH mRNA, with the data represented as mean ± S.D. (bar graphs). Statistical significance was calculated by a Student’s t-test (HPNE cells) or one-way ANOVA (CD18/HPAF). The difference with the Control group (n = 6/group) was determined to be statistically significant at ^*p < 0.001 or ^**p < 0.005. (C) PR55α suppresses p16 promoter activity. HPNE and CD18/HPAF cells in the presence/absence of ectopic PR55α and PR55α-shRNA expression, respectively, were co-transfected with a Firefly luciferase reporter under the control of the p16 promoter and a control Renilla luciferase reporter driven by the SV40 promoter. Two days after transfection, Firefly, and Renilla luciferase activities were measured separately in each lysate, as described in the Materials and Methods. p16 promoter activity was calculated by normalizing the activity of Firefly luciferase over that of Renilla luciferase. The graphs show relative p16 promoter activities in the indicated cell samples and are expressed as the mean ± S.D. of two independent experiments done in duplicates. ^*Statistically significant in a Student’s t-test with p < 0.001. (D) HPNE/Control and HPNE/PR55α cells were incubated in the presence of 1 μg/ml Dox for 48 h to induce PR55α expression, after which cells were transfected with either a non-targeting siRNA (Control) or AUF1 siRNA. Two days later, cells were analyzed by immunoblotting for differences in levels of AUF1 and p16. GAPDH was used as an internal standard. The levels of p16 and GAPDH were quantified using Fiji-ImageJ software and relative p16 levels in the samples were determined after normalizing it with GAPDH levels.

We next assessed the effects of PR55α on p16 mRNA expression using real-time quantitative Reverse Transcription and PCR (qRT-PCR). As shown in Figure 3B, the p16 mRNA level was decreased by ~70% in the HPNE cells expressing ectopic PR55α compared to the control HPNE cells. Conversely, the PR55α knockdown by shRNA resulted in a 3–5 fold induction in p16 mRNA level in CD18/HPAF cells compared to the control cells (Figure 3B). Statistically, these changes were all determined to be highly significant. To further define the mechanism by which PR55α reduces p16 mRNA expression, we analyzed the impact of PR55α on p16 promoter activity using a luciferase reporter assay. HPNE and CD18/HPAF cells were transiently co-transfected with two reporter vectors: a vector expressing firefly luciferase under the control of the p16 promoter (nucleotides −3243 to −165) and a control vector expressing Renilla luciferase from the SV40 promoter (Figure 3C). As shown in Figure 3C, ectopic PR55α expression in HPNE normal cells resulted in >90% inhibition of the p16 promoter activity compared to the control cells. Conversely, shRNA knockdown of PR55α in CD18/HPAF cancer cells resulted in approximately 85-fold induction in the p16 promoter activity. Together, these results indicate that the inhibition of p16 expression by PR55α is at least in part mediated by the repression of its promoter activity.

Knocking down AUF1 does not affect the inhibition of p16 expression by PR55α

AUF1 is an RNA-binding protein that has been reported to promote p16 mRNA decay during replicative senescence [13]. Thus, we examined the effect of AUF1 on the suppression of p16 expression by PR55α. AUF1 was knocked down by siRNA in both HPNE control and PR55α overexpressing cells and the resulting cells were analyzed for differences in the levels of AUF1 and p16 after 48 h incubation. In HPNE control cells, the knockdown of AUF1 resulted in a 6-fold induction of p16 protein expression (Figure 3D), which is consistent with previous findings [13]. However, in the PR55α overexpressing cells, the knockdown of AUF1 did not suffice to reallow the expression of the p16 protein (Figure 3D). Thus, the AUF1-mediated mRNA decay mechanism is not responsible for the inhibition of p16 expression by PR55α.

PR55α promotes RB phosphorylation

RB plays a key role in the induction of both replicative senescence and stress-induced premature senescence, both of which are known to exhibit a permanent cell cycle arrest at the G1 phase caused by the sequestration of the E2F transcription factor by RB [42–44]. During cell proliferation, RB is phosphorylated by CDK4/6 and then CDK2 kinases, which free E2F to induce the transcription of genes needed for the S phase [45]. In senescent cells, CDK4/6 and CDK2 are permanently inhibited by the induction of the p16 and p21 proteins, respectively [46] (Figure 1). Since PR55α affected the expression of both p16 and p21 but in different ways (Figure 2), we investigated the net effect of PR55α on RB phosphorylation.

Consistent with the preponderant role played by p16 in cellular senescence [47], the effects of PR55α on RB phosphorylation mirrored those observed with the p16 protein in both the normal and cancer cells. In HPNE cells, where PR55α overexpression suppresses p16 expression (Figure 2A, lane 6 vs. 1), the phosphorylation of RB-Ser780 (Figure 4A, lane 6 vs. 1) was elevated, indicative of increased CDK4/6 kinase activity [48, 49]. Similar to the situation observed with p16 in the HPNE/PR55α cells (Figure 2A), IR only had minor impacts on the Ser780 phosphorylation of RB, which remained high across all time points post-IR (Figure 4A).

Figure 4. PR55α expression increases RB phosphorylation in normal and malignant pancreatic cells. (A) A schematic depiction of the regulation of the p16/RB cascade by PR55α-controlled PP2A enzymes. PR55α suppresses the expression of p16, which acts as the primary inhibitor of the CDK4/6 kinases that phosphorylate and inactivate RB, thereby resulting in its dissociation from the E2F transcriptional factor. The net effect of PR55α is promoting the G1/S transition by allowing the phosphorylation of RB and the release of E2F. (B) HPNE cells transduced with Dox-inducible PR55α, or control vector were incubated with 1 μg/ml Dox for 3 days to induce ectopic PR55α expression. The cells were exposed to 10 Gy IR and then incubated for the times indicated. Harvested cells were analyzed for RB-Ser708 phosphorylation, total RB level, and GAPDH (as an internal control). (C) CD18/HPAF-transduced with the Control or PR55α shRNA were incubated with 2 μg/ml Dox for 5 days to knockdown PR55α expression, after which cells were analyzed by immunoblotting for differences in RB-Ser708 phosphorylation, total RB, and GAPDH.

For validation, we examined the effect of PR55α on RB phosphorylation in CD18/HPAF pancreatic cancer cells. The knockdown of PR55α resulted in a ~50% decrease in the steady-state phosphorylation of RB-Ser780 in the log-phase CD18/HPAF cells (Figure 4B, lane 6 vs. 1), which is consistent with the elevated p16 levels in these cells relative to the control cells (Figure 2B, lane 6 vs. 1). Furthermore, IR exposure of control CD18/HPAF cells resulted in the induction in RB-Ser780 phosphorylation, whereas IR exposure of PR55α-knockdown CD18/HPAF cells rather resulted in a time-dependent decrease in RB-Ser780 phosphorylation (Figure 4B). Together, these results support a role for PR55α in the suppression of p16/RB signaling (Figure 4C).

PR55α impedes the cell cycle response of pancreatic normal and cancer cells to IR exposure

In response to DNA damage, cells elicit a cell cycle checkpoint response, leading to either cell cycle arrest for DNA repair, or replicative cell death (apoptosis or else senescence) if the damage is irreversible [50, 51]. Normal cells possess wild-type p53, which can be rapidly activated by DNA damage to induce p21 expression and arrest the cell cycle at the G1 phase while promoting DNA repair [52]. However, most cancer cells harbor p53 mutations and are thus defective in the G1 checkpoint that requires the p53/p21 pathway. In these cancer cells, DNA damage only activates the G2 checkpoint, thereby blocking the G2/M transition to allow DNA repair [53]. Yet, both the G1 and G2 checkpoints are dependent on the activation of RB to block cell cycle progression and this activation can be triggered by either the p53/p21 pathway and/or the p16/RB pathway (Figure 1) [8]. Because PR55α inhibits p16 expression in both the normal HPNE cells expressing wild-type p53 and the CD18/HPAF cancer cells expressing mutant p53, we decided to examine the impacts of PR55α on the IR-induced cell cycle response. For this purpose, exponentially proliferating HPNE and CD18/HPAF cells, with/without PR55α manipulations, were exposed to increased doses of IR and analyzed 24 hours later for changes in DNA content by Fluorescence-activated cell sorting (FACS).

As shown in Figure 5, IR exposure of HPNE/Control cells resulted in a small but statistically significant increase in the percent cells in the G1 phase (p = 0.016), which was accompanied by a concurrent decrease in cells in the S phase (p = 0.021) and little change in the cells at the G2/M phases, all indicative of a G1 cell cycle arrest. In contrast, in PR55α-overexpressed HPNE cells (HPNE/PR55α), IR exposure produced little, if any, effects on the cell cycle profile compared to the unirradiated cells (Figure 5). Because HPNE/Control cells express both wild-type p53 and wild-type p16, their cell cycle response to IR was controlled by the combined activities of the p53/p21 and p16/RB pathways. In contrast, in HPNE/PR55α cells, since the expression of p16 was inhibited to near totality while the p53/p21 was further enhanced (Figure 2A), the effect of IR on the cell cycle response of these cells could only be attributed to the p53/p21 pathway.

Figure 5. Effect of PR55α on the cell cycle response of normal and cancer cells to IR. Log-phase growing HPNE (Control- or PR55α-transduced) and CD18/HPAF cells (Control-shRNA or PR55α-shRNA transduced) were incubated in media containing 1 μg/ml Dox for 3 days and 2 μg/ml Dox for 5 days, respectively, to manipulate PR55α expression. The cells were exposed to IR at indicated doses, incubated for 24 hours, and then stained with propidium iodide (PI) and analyzed for DNA content using FACS. (A) Representative FACS analyses are shown. The location and percent of cells in the G1, S, and G2/M phases of the cell cycle are indicated by arrows. (B) Bar graphs depicting the percent cells in the G1, S, and G2/M phases of the cell cycle. Each bar represents the mean ± S.D. of two separate experiments, each done in duplicates.

Next, we examined the effect of PR55α on the cell cycle response of CD18/HPAF cells, which express the p53^P151S mutant along with very low levels of p16 [37, 54], making these cells deficient in both p53 and p16 functions and lacking a functional G1 cell cycle checkpoint as part of their DNA damage response [53, 55]. However, CD18/HPAF cells still possess an intact G2 checkpoint as part of their DNA damage response, which can function independently of p53 and p16 [56, 57]. As shown in Figure 5, IR exposure of CD18/HPAF control cells (CD18/Control) resulted in dose-dependent increases in cells at the G2/M phases that were accompanied by concurrent diminutions in percent cells at the G1 and S phases, which is indicative of a G2/M cell cycle arrest. In the PR55α knockdown cells (CD18/PR55α-shRNA), while IR still led to a G2/M cell cycle arrest (Figure 5), the percent cells arrested at the G2/M phases was much reduced compared to the irradiated control cells (CD18/Control), by more than a factor of 3 (22.6% versus 78.6%). In the PR55α knockdown cells, IR also did not substantially deplete the percent cells in the G1 or S phases of the cell cycle. Compared to only 8.4% of CD18/Control cells that were retained in the G1 phase after IR, 54.9% of CD18/PR55α-shRNA were still detected in the G1 phase after IR, which corresponds to a 6.5-fold difference in the percent cells in the G1 phase between the two isogenic cell populations (Figure 5). Collectively, these results indicate that, although the parental CD18/HPAF cancer cells lack a functional G1 checkpoint and thus instead respond to IR with the induction of a G2/M cell cycle arrest, the knockdown of PR55α by shRNA was sufficient to induce high levels of p16 expression, which may have then allowed the bulk of the cells to remain in G1 after IR. These results from CD18/HPAF cells are again consistent with the negative regulation of p16 expression by PR55α and subsequent impacts on RB phosphorylation and cell cycle checkpoints.

PR55α inhibits p16-dependent cellular senescence induction by IR

Our findings in Figures 2–5 reveal PR55α-controlled PP2A as a potent inhibitor of p16 expression and G1 checkpoint activation. In normal human cells, activation of the G1 checkpoint by genotoxic assaults is required for the induction of cellular senescence, both replicative and premature [7, 13, 58], to prevent the damaged cells from dividing and subsequently becoming malignant [59–61]. The induction of senescence commonly involves two key regulators: the p16/RB and/or p53/p21 pathways, depending on the insult and the cell type [4–6, 58]. To examine the effects of PR55α on senescence induction by IR, we measured the activity of senescence-associated β-galactosidase (SA-β-gal), a hallmark of senescent cells [62]. In control HPNE cells, SA-β-gal activity was undetectable before IR exposure (0 Gy) but detected in ~60% and ~80% of cells after exposure to a single (7 Gy) or double (2 × 7 Gy) doses of IR, respectively. In contrast, PR55α overexpression in HPNE cells markedly diminished senescence induction by IR, with less than 10% of irradiated HPNE/PR55α cells staining positive for SA-β-gal activity after either the 1 × 7 Gy or 2 × 7 Gy treatment (Figure 6A). Likewise, while control CD18/HPAF cells exposed to 7 Gy IR only exhibited <1% SA-β-gal positive cells at 7 days post-IR, < 90% of irradiated PR55α-knockdown CD18/HPAF cells were positive for SA-β-gal activity post-IR (Figure 6B). Senescent HPNE or CD18/HPAF cells were also typically larger and flatter in shape compared to their respective unirradiated controls (Figures 6A, 6C). Collectively, in either p53-wt HPNE normal cells or p53-mutant CD18/HPAF cancer cells, senescence induction in response to IR requires a low level of PR55α expression, thereby revealing PR55α as a potent inhibitor of IR-induced cellular senescence. Also, this regulation of p16 expression by PR55α was independent of the p53/p21 pathway, since the IR-activated p53/p21 cascade was further enhanced by the overexpression of PR55α in HPNE/PR55α cells, in which IR still failed to induce senescence. Conversely, despite CD18/HPAF cells harboring a mutant p53, the knockdown of PR55α still resulted in increased p16 expression and induction of senescence in response to IR exposure (Figures 2 and 6).

Figure 6. PR55α inhibits IR-induced cellular senescence. (A) Ectopic PR55α overexpression prevents senescence induction by IR in normal HPNE cells. HPNE/PR55α and HPNE/control cells were incubated in the presence of 1 μg/ml Dox for 2 days (to induce ectopic PR55α in the HPNE/PR55α cells) and then exposed to 7 Gy IR or left unirradiated as control (0 Gy). When the second radiation dose was applied, the interval was 24 hours between the two doses. Seven days post-IR, cells were assessed for SA-β-gal activity and photographed. Scale bar = 1 μm. The bar graph expresses the percent of senescent cells in the indicated samples and represents the mean ± S.D. of two separate experiments with each done in duplicate samples. (B) PR55α-knockdown sensitizes CD18/HPAF pancreatic cancer cells to senescence induction by IR. CD18/HPAF cells expressing Dox-inducible PR55α-shRNA or Control-shRNA were cultivated in the presence of 2 μg/ml Dox for 5 days, to allow time to silence PR55α expression, and then exposed to 7 Gy IR, or left unirradiated as a control (0 Gy). After 7 days, the cells were assessed for SA-β-gal activity and photographed. Scale bar = 1 μm. The graphs express the percent of senescent cells in the indicated samples and represent the mean ± S.D. of two separate experiments with each one in duplicate samples. (C, D) Normal HPNE and CD18/HPAF pancreatic cancer cells, with/without PR55α manipulation, were exposed to 7 Gy IR, or left unirradiated as a control (0 day). When the second radiation dose was applied, the interval was 24 hours between the two doses. The irradiated cells were incubated for the times indicated and analyzed by immunoblotting for the differences in levels of p16, p14, p53, and p21. GAPDH was used as an internal control.

To confirm that p16 expression is required for senescence induction by IR in PR55α knockdown cells, we utilized the AsPC-1 pancreatic cancer cells, which carry a homologous deletion of the CDKN2A locus [37] and thus lack both p16 and p14 expression (Figure 7A). AsPC-1 cells express p53 from an allele that carries a single nucleotide deletion (TGC→GC) at Cys135 that produces a frameshift [37]. Hence, these cells express a truncated and dysfunctional p53 protein, which is also unstable, and are thus devoid of p21 expression (Figure 7B). Subsequently, we tested the effect of PR55α-knockdown on senescence induction by IR in AsPC-1 cells using the SA-β-galactosidase assay. As shown in Figures 7B, 7C, IR exposure of AsPC-1 cells did not lead to senescence induction either in the presence or absence of the PR55α knockdown by shRNA. These results, together with the data obtained from CD18/HPAF cells, which also express mutant p53 but possess a functional p16 (Figures 2B and 6B), suggest a role for PR55α in the inhibition of the p16/RB pathway-dependent senescence induction by IR.

Figure 7. Knockdown of PR55α does not result in senescence induction by IR in pancreatic cancer cells that lack p16 expression. (A) The CDKN2A locus, which encodes the p16 and p14 genes, is deleted from AsPC-1 pancreatic cancer cells [37]. Log-phase growing AsPC-1 cells were analyzed for the presence of p16 and p14 proteins with CD18/HPAF pancreatic cancer cells as a positive control. (B) AsPC-1 cells stably transduced with the Dox-inducible PR55α-shRNA or Control-shRNA were induced with 2 μg/ml Dox for 3 days to knock down PR55α. The cells were then exposed to 7 Gy IR, or left unirradiated (0 time point), and incubated for an additional 1, 4 and 7 days. The cells were analyzed by immunoblotting for the levels of PR55α, p53, and p21. GAPDH level was measured as an internal loading control. (C) The irradiated cells incubated for 7 days were analyzed for senescence by SA-β-gal activity assay and photographed. Scale bar = 1 μm.

PR55α expression inversely correlates with the expression of p16 in normal human tissues

We analyzed the co-expression of PR55α and p16 in serial sections of human normal tissue specimens derived from various organs (bladder, colon, fallopian tube, heart, lung, liver, kidney, pancreas, and skin). Since p16 expression increases as a function of age, we sub-divided specimens into two cohorts based on the donor’s age: a “young” cohort consisting of tissues derived from individuals younger than 43 years old (y/o) and an “old” cohort consisting of the tissues derived from individuals older than 68 y/o. As shown in Figure 8, there was a statistically significant inverse correlation between PR55α and p16 expression in both young and old cohorts of tissues, with PR55α levels detected at higher levels in tissues of young donors compared to old donors and p16 levels exhibiting the opposite relationship. Indeed, PR55α levels were significantly lower in the old cohort compared to the young cohort. Conversely, the p16 levels were significantly higher in the old cohort compared to the young cohort.

Figure 8. PR55α level is much lower in human normal tissue specimens of older individuals compared to younger individuals and inversely correlates with p16 levels. Human normal tissue specimens derived from various organs/sites were analyzed for differences in PR55α and p16 expression by IHC. (A) Representative images of adjacent tissue sections stained with anti-PR55α and anti-p16 antibodies. OSTSE–tonsil; GMST-seminiferous tubules; ET-thyroid; GIE-esophagus. Young, ≤43 y/o; Old, ≥68 y/o; (B) Box plot shows the H-Score quantification of PR55α and p16 expression from adjacent sections.

Cell culture and treatment

The HPNE cells are primary human pancreatic normal ductal cells immortalized with human telomerase hTERT [32, 33]. The Doxycycline (Dox)-inducible HPNE-PR55α cell line was established by transducing HPNE cells with the pRevTet-On retroviral vector (Clontech Laboratories, Mountain View, CA, USA) modified to express the PR55α cDNA and selecting the cells with hygromycin 200 μg/ml. The ectopic expression of PR55α was induced by 1 μg/ml Dox for 48–72 h. The HPNE isogenic cell lines were maintained in Medium D growth medium (3 parts high glucose Dulbecco’s modified Eagle’s medium (DMEM) (GE Healthcare Life Sciences, Pittsburgh, PA, USA) to one part of M3 medium (INCELL, San Antonio, TX, USA) supplemented with 10% Tet-free FBS and 10 ng/mL human recombinant EGF (Invitrogen, Waltham, MA, USA)) [32, 33]. Human pancreatic cancer cell lines CD18/HPAF and AsPC-1 were obtained from ATCC (Manassas, VA, USA) and maintained in DMEM (Life Technologies, Carlsbad, CA, USA) with 10% fetal bovine serum (FBS) in an atmosphere containing 5% CO₂. IR exposure was performed by treating exponentially growing cells with a Mark I 68A Cesium-137 Irradiator (JL Shepherd and Associates, San Fernando, CA, USA) at the indicated doses and incubated for the specified times at 37°C.

Antibodies, Western blot analysis, and protein stability assessment

The primary antibodies used in this study were mouse IgG against p16 (#55079, BD Bioscience, USA), GAPDH (#32233, Santa Cruz, USA), and AUF1 (#12382, Cell Signaling Technologies, USA). Western blot analyses were performed as described previously [23, 28]. In brief, protein samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was then probed with the primary antibody against p16, AUF1, or GAPDH, followed by detection using an anti-mouse IgG secondary antibody conjugated to horseradish peroxidase. Chemiluminescent signals were captured and analyzed with a BioRad Chemidoc System (Hercules, CA, USA). Alternatively, they were revealed after exposure to an x-ray film and scanning of the film with an EPSON Perfection 4490 PHOTO scanner. Signal quantitation was done using the Fiji-ImageJ analytical program (NIH, Bethesda, MD, USA).

To analyze the protein half-life of p16, we exposed cells to protein synthesis inhibitor cycloheximide (CHX, dissolved in water) (Sigma-Aldrich, St. Louis, MO, USA) at 15 μg/ml and we began monitoring the levels of p16 protein by immunoblotting, as described previously [23, 28]. α-Tubulin served as a loading control on the Western blots. The levels of p16 on the blot were normalized by control protein levels, and calculated for half-life times (t(1/2)) by linear regression analysis of log (protein level) against time using the SigmaPlot graphing and data analysis software, as described in our studies [23, 28].

Senescence-associated β-galactosidase (SA-β-gal) assay

The SA-β-gal activity was measured per the manufacturer’s instructions (Catalog #200488, Agilent Technologies, La Jolla, CA, USA). The SA-β-gal activity was determined by using X-gal (5-bromo-4-chloro-3-indolyl β-D-galactosidase) staining at pH 6.0. Ten randomly selected fields per sample were photographed and quantified for the cells positive in SA-β-gal staining under a light microscope.

Cell cycle analysis

Cell cycle analysis was carried out as described previously [80, 81]. Briefly, 24 h after IR exposure, the cells were harvested, washed with phosphate-buffered saline (PBS), and fixed in ice-cold ethanol. Cells were stained with propidium iodide (PI) and analyzed for DNA content by fluorescence-activated cell sorting (FACS) analysis using the FACS Calibur instrument (Becton Dickinson, Mansfield, MA, USA). At least 20,000 cells from each sample were evaluated.

Luciferase reporter constructs and luciferase assay

pGL2-p16-Luc is a firefly luciferase reporter for p16 promoter activity. It carries a firefly luciferase gene under the control of the p16 promoter (from nucleotides −3243 to −165 upstream of the p16 initiation codon). p16 promoter sequences were initially amplified by PCR using primers p16A (5′-cggatatcacgcgtccacccaaggatgccataat-3′; MluI site underlined) and p16B (5′-ctgaagatctcccgccgccggctccat-3′; BglII site underlined). The PCR product was digested with MluI and BglII and inserted into the same sites of the pGL2-Basic vector (Promega, Madison, WI, USA). Next, this plasmid was digested with SacII and BglII and then self-ligated to eliminate the segment −164 to +17 containing the p16 initiation codon. pRL-SV40-Luc (Promega) is a control Renilla luciferase reporter that contains the SV40 promoter. pGL2-p16-Luc and pRL-SV40-Luc were co-transfected into the cells using Lipofectamine 2000 (Invitrogen, Waltham, MA, USA) according to the manufacturer’s suggestions. Firefly luciferase and Renilla luciferase activities were measured at 48 h following transfections using a Luciferase Assay System, as instructed by the manufacturer (Promega). Firefly luciferase activity was normalized for transfection efficiency using the activity of Renilla luciferase.

RNA isolation and real-time quantitative reverse transcription (qRT)-PCR

RNA was isolated from cells using TRIzol and reverse transcribed using the iScript Reverse Transcription kit. Further, the mRNA was quantified by Real-Time qRT-PCR using SsoAdvanced TM SYBR Green Supermix (Bio-Rad) following the manufacturer’s protocols. GAPDH was used as the internal control. The following primers were used for the PCR: p16 forward primer: 5′-GACCTGGCTGAGGAGCTG-3′ and p16 reverse primer: 5′-CAATCGGGGATGTCTGAGGG-3′; GAPDH forward primer: 5′-TTCCACCCATGGCAAATTCC-3′ and reverse primer: 5′-TGGCAGGTTTTTCTAGACGG-3′.

Small hairpin RNAs, short interfering RNAs, and transfection

The small hairpin RNAs (shRNA) sequences targeting human PR55α (PPP2R2A) mRNA are 5′-ATGGCTAGCAGACATGGAG-3′, and 5′-CAACTATCTCAACTAAGCA-3′. The non-targeting control shRNA was designed to target no known genes in humans, mice, or rats. The control shRNA sequence is 5′-ATCTCGCTTGGGCGAGAGTAAG-3′ (Dharmacon, Lafayette, CO, USA). The siGENOME/AUF1 short interfering RNAs (siRNAs) SMARTpool (Dharmacon) consists of four siRNAs targeting multiple sites on AUF1 mRNA: 5′-CAAAUUUGGUGAAGUUGUA-3′, 5′-GGAAGGUGAUUGAUCCUAA-3′, 5′-AGACUGCACUCUGAAGUUA-3′, and 5′-CGUGGGUUCUGCUUUAUUA-3′. Control siGENOME nontargeting siRNA (Dharmacon) was designed to target no known genes in humans, mice, or rats. The sequence of the control siRNA is 5′-UAGCGACUAAACACAUCAA-3′. siRNA transfection of cells was performed with DharmaFECT-1 (Thermo Fisher Scientific, Waltham, MA, USA) as instructed by the manufacturer. Transfected cells were incubated for 48–72 hours before analysis.

Human normal tissue specimens and immunohistochemistry

Human normal tissue samples (CHTN_Norm3 - Normal Tissue Survey) for immunohistochemistry (IHC) analyses were provided by the Cooperative Human Tissue Network (Mid-Atlantic Division) funded by the National Cancer Institute. The normal tissue samples were derived from various human organs/sites including the skin, bladder, breast, digestive tracts, heart, kidney, liver, ovary, pancreas, spleen, and tonsil of both young (≤43 years old) and old (≥68 years old) individuals (Supplementary Table 1). The IHC analysis of PR55α used the anti-PR55α antibody 100C1 (Cell Signaling Technology, Danvers, MA, USA) at the dilution of 1:400. The IHC analysis of p16 used the anti-p16 JC2 mouse monoclonal IgG (ready use, MM156-10, Statlab, USA). The IHC analyses were performed by the Tissue Science Core of the University of Nebraska Medical Center (UNMC). PR55α and p16 IHC staining were evaluated by a UNMC pathologist who was blinded to the clinical information. Immunohistochemistry (IHC) staining was quantified by a Histoscore (H-score) based on both signal intensity (0 = no staining, 1 = weak, 2 = moderate, 3 = strong immunoreactivity) and percentage of positive cells (1 = 0–25%, 2 = 26–50%, 3 = 51–75%, 4 = 76–100%), as we have done in previous studies [27, 28, 82].

Statistical analysis

Student’s t-test and one-way ANOVA methods were used to compare experimental groups using SigmaPlot software (Palo Alto, CA, USA), and the results were expressed as mean ± SD. The statistical analyses were described in the respective figure legends and P-values ≤ 0.05 were considered significant.