Cyclin-dependent kinase inhibitor 1 plays a more prominent role than activating transcription factor 4 or the p53 tumour suppressor in thapsigargin-induced G1 arrest

Cell Culture and drug treatments

HCT116 cells and sublines that carry deletions in the ATF4, TP53 or CDKN1A genes (referred to as ATF4-def, p53-null and p21-null, respectively) (van Zyl et al., 2021; Bunz et al., 1998) were grown in McCoys media (Hyclone, San Angelo, TX, USA) supplemented with a 12% mixture of Newborn Calf Serum and Fetal Bovine Serum (3:1) ratio (Gibco, ThermoFisher Scientific, Ottawa, Canada) and 90 units/ml penicillin, 90 µg/ml streptomycin (Hyclone). The p53-null, p21-null and their parental HCT116 strain were obtained from Dr. Bert Vogelstein (Johns Hopkins University) (van Zyl et al., 2021; Bunz et al., 1998). The ATF4-def cells were derived from HCT116 cell line obtained directly from the American Type Culture Collection (Cat #: CCL-247; Manassa, VA, USA) (van Zyl et al., 2021). Normal human neonatal foreskin fibroblasts expressing human telomerase (NFhTert) were obtained from Mats Ljungman (University of Michigan) (O’Hagan & Ljungman, 2004). These cells were maintained in DMEM supplemented with 10% fetal bovine serum and 90 units/ml penicillin, 90 µg/ml streptomycin (Hyclone). In all experiments, cells were seeded 24 h prior to treatment at 500,000 cells/well of a 6-well plate or 100,000 cells/12-well plate. Cells were left untreated, treated with up to 2 µM Tg (Millipore Sigma Canada, Oakville ON, Canada), or treated with a volume of dimethyl sulfoxide (DMSO) (Calbiochem) equal in volume to the highest Tg treatment as a vehicle control. Where indicated, 25 µM zVAD-fmk (Millipore Sigma Canada), was added alone or in combination with 2 µM Tg to block caspase activity.

Immunoblotting

Immunoblotting was completed as described previously (van Zyl et al., 2021). Briefly, media was removed, and cells were washed with phosphate buffered saline (PBS) pH 7.4. Cells monolayers were collected in RIPA buffer. Samples were sonicated and quantified using the Bio-Rad protein assay (Bio-Rad, Hercules, CA, USA). Samples were denatured in NuPAGE™ LDS Sample Buffer and NuPAGE™ Sample Reducing Agent (10X) (Thermo Fisher Scientific, Ottawa, Canada), heated for 10 min at 70 °C and equal amounts of protein were loaded onto NuPAGE 4–12% Bis-Tris gels (Thermo Fisher Scientific). The protein was transferred to a nitrocellulose membrane (Bio-Rad). 1mg/ml Ponceau S in 1% acetic acid was used to detect the transferred protein. The membrane was then blocked in 5% milk in TBST pH 7.6 for a minimum of 1 h at room temperature. The membrane was then incubated in primary antibody overnight at 4 °C. All primary antibodies were prepared in 0.5% milk in TBST at 1:200 for anti-p21WAF1 (Ab-1, Millipore Sigma) and anti-p53 (sc-6243, Santa Cruz) and 1:1,000 for anti-ATF4 (abcam) and anti-beta-Actin (A5316, Sigma Aldrich). The membrane was then washed 4 times for 5 min with TBST, and incubated in secondary antibody (1:10,000 for HRP-conjugated goat anti-mouse and 1:20,000 for HRP-conjugated goat anti-rabbit; Abcam, Cambridge, UK) for 2–3 h at room temperature. The membrane was washed 4 times with TBST for 5 min, followed by a 10-minute wash. Clarity™ Western ECL substrate (Bio-Rad) was added, and the membrane was imaged using the Fusion FX5 gel documentation system (Vilber Lourmat, Collégien, France). To visualize additional proteins, the membrane was subsequently washed 4 times in TBST, blocked in 5% milk in TBST again and incubated in primary antibody, as described above.

Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR)

Medium was removed and cells were rinsed twice with PBS (pH 7.4). RNA was extracted using the EZ-10 DNA away RNA Miniprep kit (Bio Basic Canada Inc, Markham, Canada). RNA concentration and quality was measured using a DeNovix DS-11 Spectrophotometer (DeNovix, Wilmington, DE, USA). Equal amounts of RNA were converted to cDNA using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Ottawa Canada). qPCR was performed using Bioline SensiFAST™ Probe HI-ROX Master Mix (FroggaBio Inc., Toronto, ON) with SYBR green. The primers used were obtained from integrated DNA technologies (IDT) and the sequences were CDKN1A (GGAGACTCTCAGGGTCGAAA, and GCTTCCTCTTGGAGAAGATCAG), and GAPDH (AGCCACATCGCTCAGACA, and GCCCAATACGACCAAATC).

Cell cycle analysis

Flow cytometric analysis of cell cycle distribution was completed as described previously (Vanzyl et al., 2018). Briefly, cells were treated with the indicated drugs for 24 h. For the final hour of drug treatment, the regular drug media was replaced with fresh media containing drug media supplemented with 30 µM 5′ Bromo- 2′ deoxyuridine (BrdU) (Sigma). After one hour, the media was removed, cells were washed with PBS and collected using trypsin and centrifugation. The pellet was washed with PBS and fixed in 70% ice cold ethanol for a minimum of 30 min at −20 °C. Following fixation, cells were pelleted and washed with PBS, then treated with 50 µg/ml RNAse A for 30 min at 37 °C. Samples were centrifuged, and cells were resuspended in 0.1M HCL 0.7% Triton X-100 and incubated on ice for 15 min. Cells were pelleted, resuspended in H₂O, and boiled for 15 min, then placed on ice for 15 min. 0.5% Tween 20 in PBS was added to the cells and then spun to pellet. The cell pellet was resuspended in Alexa Fluor 488-conjugated anti-BrdU antibody (BD Biosciences) in PBS, 5% FBS and 0.5% Tween 20 for 30 min in the dark at room temperature. The antibody was removed, and cells were resuspended in 30 µM PI with 50 µM RNAse A in PBS. Samples were stained at 4 °C for a minimum of 30 min. BrdU incorporation and DNA content were measured using a BD Accuri C6 flow cytometer using the FL1 and FL2 channels, respectively, with BD Accuri C6 software.

Sub G₁ assay

Flow cytometric analysis of sub G₁ DNA content was completed as described previously (Vanzyl et al., 2018; Vanzyl et al., 2020). Floating and adherent cells were collected and fixed in ice cold 70% ethanol for a minimum of 30 min at −20 °C. After fixation, cell pellets were washed twice with PBS and incubated in 30 µM PI with 50 µM RNAse A in PBS at 4 °C for 30 min. DNA content was measured using a BD Accuri C6 flow cytometer using the FL2 channel and analyzed using the BD Accuri C6 software.

The effect of ATF4 on Tg-induced changes in cell cycle distribution

ER stress and ATF4 overexpression led to G₁ arrest in a variety of cell types (Frank et al., 2010). However, the contribution of ATF4 to ER stress-induced alterations in cell cycle distribution had not been tested. To address this, we compared the effects of Tg on cell cycle distribution using HCT116 colon cancer cells and an ATF4-def subline that we generated recently (van Zyl et al., 2021) (Fig. 1A). We found that Tg exposure greatly increased ATF4 expression in parental HCT116 cells but that ATF4 was not detected before or after exposure to Tg in the isogenic ATF4-def subline (Fig. 1A). To assess the effects of Tg on cell cycle distribution, we used two-parameter flow cytometric analysis of BrdU incorporation versus DNA content. BrdU is an analog of thymidine, so it is incorporated into DNA during replication readily identifying cells in S phase during the BrdU labeling period. DNA content was estimated using PI. Together, this method is far more precise than estimates of cell cycle phase based on DNA content alone.

In the parental HCT116 cells, Tg led to the virtual depletion of S phase (Fig. 1B upper panels). The only BrdU positive cells remaining had close to 4 C DNA content and BrdU incorporation per cell was lower than the bulk of the S phase population in control samples, indicating that there was a small population of cells in late S phase near the S/G₂ boundary that incorporated some BrdU. Very few BrdU-positive cells were found in early S phase (Fig. 1B). There was also no significant accumulation of BrdU-negative cells with S phase DNA content (i.e., no population between G₁ and G₂/M). Quantitative analysis of several independent experiments indicates that the proportion of parental cells in G₁ increased substantially with a corresponding decrease in S phase (Fig. 1C). There was a slight increase in the proportion of cells in the G₂ phase following Tg treatment. These alterations indicate that Tg induced a strong G₁ arrest and a less prominent G₂ arrest.

Cell cycle analysis in the ATF4 deficient subline was performed in parallel. There were no differences in cell cycle distribution prior to Tg treatment (Fig. 1B). Like controls, we detected a large decrease in S phase and a substantial increase in the G₁ phase of the cell cycle that were virtually identical to the controls (Fig. 1B). When quantified over multiple experiments, there were small but significant differences in G₁ and S phases between cell lines following Tg treatment (Fig. 1C lower panels). Again, the few cells in S phase were in late S phase near S/G₂ boundary (Fig. 1C). The absence of BrdU positive cells with DNA content covering the spectrum from G₁ to G₂ indicates that Tg-induces G₁ arrest in the ATF4-deficient cells as well. The BrdU positive population of cells are likely those that entered S phase prior to the complete G₁ arrest and just reached the S/ G₂ boundary during the labeling period. There was a small but statistically significant increase in the proportion of cells incorporating BrdU in the ATF4-def cell line. However, ATF4-def cells accumulated in G₁ phase indicating that ATF4 contributed little to the Tg-induced G₁ arrest.

The effect of p53 on Tg-induced changes in cell cycle distribution

The p53 protein is a DNA damage-responsive transcription factor that positively regulates the cyclin-dependent kinase inhibitor p21^WAF1 leading to both G₁ and G₂ arrests (Bunz et al., 1998; el Deiry et al., 1993; Xiong et al., 1993; Polyak et al., 1996). In addition, exposure to Tg reportedly induces a PERK-dependent G₂ arrest through the upregulation of a specific isoform of p53 lacking the N-terminal transactivation domain  (Bourougaa et al., 2010). Therefore, we sought to determine whether p53 contributed to Tg-induced changes in cell cycle distribution. Immunoblot analysis confirmed that ATF4 could be induced by Tg in the p53-null cells (Fig. 2A). We again used two-parameter flow cytometric analysis to determine the effects of Tg on cell cycle distribution in HCT116 and isogenic p53-null cells. There was no significant difference in cell cycle distribution before treatment (Fig. 2B). Again, Tg exposure led to an increase in G₁ and decrease in S phase in both cell lines comparable to the ATF4 parental HCT116 cells in Fig. 1 (Fig. 2B). The Tg treatment led to a large increase in G₁ along with a concomitant decrease in S phase across all concentrations (0.1–2 µM). Therefore, loss of p53 had no significant effect on Tg-induced ATF4 expression or Tg-induced cell cycle alterations.

The effect of Tg on cell cycle distribution is largely p21^WAF1-dependent

The p21^WAF1 cyclin-dependent kinase inhibitor plays important roles in both G₁ and G₂ arrest induced by p53 (Bunz et al., 1998; el Deiry et al., 1993; Xiong et al., 1993; Polyak et al., 1996). In addition, it has been reported that ATF4 can regulate p21^WAF1 directly in response to ER stress (Inoue et al., 2017). Conversely, ER stress has also been reported to down regulate p21^WAF1 through the ATF4-regulated CHOP protein (Mihailidou et al., 2010). Therefore, we sought to determine if p21^WAF1 plays a role in Tg-induced alterations in cell cycle distribution. Immunoblot analysis HCT116 and an isogenic p21-null cell line indicated that p21^WAF1 was not detectable in p21-null cells even when p21^WAF1 was induced in the parental strain with the proteasome inhibitor MG132 (Fig. 3A). We went on to treat HCT116 and the p21^WAF1 null cells for 24 h with Tg. Again, the BrdU-positive S phase population of control cells was depleted in response to Tg (Fig. 3B). Remarkably, the Tg-treated p21^WAF1 null cells incorporated BrdU across S phase from the G₁/S to the S/G₂ boundary (Fig. 3B). Quantitative analysis of multiple experiments indicated that the Tg-induced G₁ arrest was abrogated in p21^WAF1 null cells and that the proportion of cells in S phase remained close to pre-treatment levels (Fig. 3C). The slight increase in G₂/M in the control cells was not detected in the p21^WAF1 null cells either. These results suggest that the G₁ and G₂ arrests resulting from Tg exposure were dependent on the p21^WAF1 protein.

As described above, ATF4 and p53 can both regulate p21^WAF1 expression, therefore it was important to assess the effect of Tg on CDKN1A mRNA and p21^WAF1 protein expression in these cell lines. We did detect a small but significant increase in CDKN1A mRNA at 24 h (Figs. 4A and 4B). However, immunoblot analysis indicates that p21^WAF1 levels were not significantly altered by Tg in control cells at 8 or 24 h (Figs. 4C–4F). The basal expression of p21^WAF1 was lower in the p53-null cells, as expected (Kaeser & Iggo, 2002; Tang et al., 1998), and wasn’t increased by Tg (Figs. 4C and 4E). The constitutive expression of p21^WAF1 in the ATF4-def line was a little higher than isogenic controls and again Tg had no additional effect on p21^WAF1 levels in the ATF4-def cells (Fig. 4D and 4F). So, there was discordance between changes in CDKN1A mRNA and p21^WAF1 protein levels. This is likely explained by the fact that Tg inhibits protein synthesis (van Zyl et al., 2021; Wong et al., 1993) through the PERK-dependent branch of the UPR (Hetz, 2012). The small increase in mRNA was unlikely to lead to a comparable increase in the encoded protein. Taken together, p21^WAF1 plays a key role in determining the effect of Tg on cell cycle progression but p21^WAF1 levels themselves were unaltered by Tg exposure.

p21^WAF1, p53 and ATF4-deficient cells are hypersensitive to Tg-induced apoptosis

Tg is a well-known inducer of apoptosis (Tsukamoto, Kaneko & Kurokawa, 1993; Jiang et al., 1994). This can occur through an ATF4-dependent manner through UPR activation and the upregulation of pro-apoptotic proteins including CHOP (Sehgal et al., 2017; Lindner et al., 2020; Teske et al., 2013; Matsumoto et al., 2013). Therefore, we sought to determine how ATF4, p53 and p21^WAF1 disruption affected sensitivity of HCT116 cells to Tg. We used one parameter flow cytometric analysis of PI-stained cells to estimate sub G₁ DNA content. Unexpectedly we found that the ATF4-def cells were more sensitive to Tg treatment than their parental controls across all concentrations of Tg (Fig. 5A). Similarly, we found that both the p21-null and p53-null cells were more sensitive to Tg treatment than their parental cells (Figs. 5B and 5C). The sub- G₁ apoptosis assay assesses the proportion of cells with fragmented DNA (Nicoletti et al., 1991). To ensure that the large increase in the proportion of cells with sub-G1 DNA was associated with apoptosis, we treated HCT116 cells and the p21-null cell line with Tg in the presence of the broad range caspase inhibitor (zVAD-fmk). This caspase inhibitor effectively blocked the Tg-induced increase in the sub- G₁ population of cells (Fig. 6). Therefore, the generation of the sub-G1 population was caspase dependent. These results indicate that ATF4, p21^WAF1 and p53 are acting in a protective manner to prevent Tg-induced apoptosis in HCT116 cells. This protective effect does not appear to be related to the differences in cell cycle regulation detected in this isogenic series of cell lines.

The effect of Tg on cell cycle distribution and apoptosis in human fibroblasts

The results above were all generated in HCT116 cells and genetically modified sublines. It was important to determine if the response of Tg-treated HCT116 cells were unique or common to other cell types. Exposure to Tg similarly caused non-transformed human fibroblasts (NFhTert cells) to arrest predominantly in the G₁ phase of the cell cycle with a concomitant decrease in S phase across all drug concentrations (Fig. 7A). Furthermore, the proportion of NFhTert cells undergoing Tg-induced apoptosis was similar to HCT116 cells (Fig. 7B) and could be blocked with zVAD-fmk (Fig. 7C). Therefore, the responses of HCT116 and NFhTert cells to Tg across all drug concentrations were similar.