BMAL1 knockdown triggers different colon carcinoma cell fates by altering the delicate equilibrium between AKT/mTOR and P53/P21 pathways

Decreased BMAL1 altered expression of some circadian genes in primary CRC cell lines

Three CRC cell lines, two primary cell lines (HCT116 and SW480) and a metastatic cell line SW620, were transduced with lentiviruses encoding a scrambled shRNA (shScr) or a shRNA targeting BMAL1 (shBMAL1). After transduction, cells were selected by one-week puromycin treatment to remove non-transduced cells. Successful transduction was confirmed by flow cytometry of GFP expressing cells. The GFP positive cell population was used immediately for analysis as BMAL1-KD or control (Ctr) cells.

BMAL1 expression was significantly decreased compared to control at mRNA (Figure 1A, qRT-PCR) and protein levels (Figure 1B, Western blot) in all three BMAL1-KD cell lines, despite the fact that the two primary CRC cell lines exhibited much higher BMAL1 expression than the metastatic CRC cell line SW620.

Expression of BMAL1 target circadian genes (NR1D1, PER2, CRY1 and CRY2) and its functional partner CLOCK, were also analyzed following BMAL1-KD. In the two primary BMAL1-KD CRC cell lines (HCT116 and SW480), but not in BMAL1-KD SW620 cells, NR1D1 expression was reduced and CLOCK expression was increased (Figure 1C, 1D). No significant modification of PER2, CRY1 or CRY2 expression was identified in any of the three cell lines (Data not shown).

BMAL1-KD thus modified circadian gene expression profile in two primary CRC cell lines (HCT116 and SW480) but not in the metastatic CRC cell line SW620.

BMAL1-KD increased AKT/mTOR activation in primary CRC cell lines

BMAL1 knockout mice manifest increased mTORC1 (mammalian target of rapamycin complex 1) activity and associated premature aging [27]. We thus investigated whether BMAL1-KD could modify AKT/mTOR signaling in CRC cell lines.

BMAL1-KD increased AKT activation in colon cancer cell lines. Compared to their controls, the percentage of AKT phosphorylation, measured by the ratio between phosphorylated AKT (pAKT) and total AKT, was greatly increased in the primary CRC lines, HCT116 BMAL1-KD (p=0.0207) and SW480 BMAL1-KD (p=0.0302), and slightly increased in the metastatic SW620 BMAL1-KD (p=0.0141) cells (Figure 2A).

Moreover, mTOR activation, measured by the ratio between phosphorylated mTOR (pmTOR) and total mTOR, was significantly increased in HCT116 BMAL1-KD cells (p=0.0298) and in SW480 BMAL1-KD cells (p=0.0291) in comparison to their respective controls. However, SW620 BMAL1-KD cells only showed trend to increased mTOR activation (p=0.0897) (Figure 2B).

To further evaluate mTOR activity, we measured the phosphorylation of 40S Ribosomal protein S6, a major mTOR effector. Western blot analysis revealed a significant increase of S6 phosphorylation (pS6/S6 total) in HCT116 BMAL1-KD cells (p=0.0324) and in SW480 BMAL1-KD cells (p=0.0052) compared to controls. However, only a trend to increased phosphorylation was found in SW620 BMAL1-KD (p=0.0634; Figure 2C). These results were further confirmed by flow cytometry analysis, which revealed that the mean fluorescence intensity of pS6-APC staining increased significantly in HCT116 BMAL1-KD cells (p=0.0316) and in SW480 BMAL1-KD cells (p<0.0001), but not in SW620 BMAL1-KD cells (p=0.4027) compared to their own controls (Figure 2D).

BMAL1-KD induced different cell proliferation patterns in CRC cell lines

The activation of mTOR pathway by AKT kinase is implicated in many fundamental cell functions, such as survival, proliferation and growth [28]. We thus examined whether BMAL1-KD influenced CRC cell proliferation by using an MTT Cell Proliferation and Viability Assay (Figure 3A) and cytometry cell counts (Figure 3B).

In primary CRC cells, corresponding to their higher AKT/mTOR activity, MTT analysis showed that HCT116 BMAL1-KD cells exhibited greater proliferation from 72h in culture, compared to its control (72h, p=0.0078; 96h, p<0.0001). This result was confirmed by cell counts, which showed more HCT116 BMAL1-KD cells after 48h compared to control (48h, p=0.0090; 72h, p=0.0317; 96h, p<0.0001). In contrast, despite increased AKT/mTOR activity, the SW480 BMAL1-KD cell line did not show greater cell proliferation compared to its control. At 96h, cytometry cell counts even revealed fewer BMAL1-KD cells compared to control (p=0.0259).

In the metastatic SW620 BMAL1-KD cell line, there was slight increase of AKT activity but without evident increase of mTOR activity. However, this cell line did proliferate faster at 96h when compared to control (MTT: p=0.0460; cell count: p=0.0090).

BMAL1-KD increased senescence only in SW480 cells

Increased mTOR activity also leads to accelerated aging under certain conditions, as shown in BMAL1 knockout mice [25–27]. SW480 BMAL1-KD cells demonstrated increased mTOR activity but no increased cellular proliferation, which led us to check whether BMAL1-KD induced senescence in these SW480 cells.

Of the three BMAL1-KD CRC cell lines, only SW480 BMAL1-KD cells showed an obvious increase of cell senescence, as identified by senescence-associated β-galactosidase activity (SA-β-gal) staining. These SA-β-gal-positive cells also demonstrated other senescence related alterations: enlarged cell size and flattened shape (Figure 4A). In contrast, compared to their controls, there was no evident increase in cellular senescence in HCT116 BMAL1-KD and SW620 BMAL1-KD lines.

Another common indicator of senescent cells is the marker of DNA double-strand breaks (DSB), phosphorylated H2AX (pH2AX) [29, 30]. Immunofluorescence (Figure 4B) and western blot (Figure 4C) analyses showed that phosphorylated H2AX was increased strongly in SW480 BMAL1-KD cells (p=0.0209) and only slightly in the SW620 BMAL1-KD cell line (p=0.0269).

Cell senescence is also mediated by, and can induce, P53/P21 activation [31]. Although P53 mRNA (Figure 5A, qRT-PCR) and protein (Figure 5B, western blot) levels did not change in the three BMAL1-KD CRC cell lines, there was a change of P53 localization in SW480 BMAL1-KD cells. Compared to control, cytoplasmic P53 decreased (p=0.0235; Figure 5C) and P53 nuclear expression increased (p=0.0183; Figure 5D) in SW480 BMAL1-KD cells, suggesting P53 activation. In contrast, P53 cytoplasmic expression did not change in HCT116 BMAL1-KD and SW620 BMAL1-KD cells. Moreover, P53 nuclear localization in HCT116 BMAL1-KD cell line even decreased compared to its own control (p=0.0249; Figure 5D).

To confirm the link between cellular senescence and P53 expression in our cell lines, we analyzed expression of P53 targets, P21 and MDM2 (murine double minute homolog 2). In agreement with the different P53 activation, qRT-PCR (Figure 6A, p=0.0409) and western-blot results revealed increased P21 (Figure 6B, p=0.0291) and MDM2 (p=0.0160; Figure 6C) expression only in SW480 BMAL1-KD cells compared to control. Moreover, MDM2 expression was even decreased in SW620 BMAL1-KD cells (p=0.0122).

In summary, after BMAL1-KD cellular senescence was only induced in one primary CRC line (SW480) in association with nuclear translocation of P53 and increased expression of P21 and MDM2.

BMAL1-KD increased apoptosis and P53 activation only in HCT116 cells

The previous results were performed on BMAL1-KD cell lines which were obtained after puromycin selection. However, phenomena such as apoptosis, will appear transiently after shBMAL1 begins to be expressed. GFP gene is incorporated in the lentivirus construction as a reporter of shRNA expression, and it becomes visible 24h post-transfection, indicating the onset of shBMAL1 expression. We thus tested for the onset of apoptosis with Annexin V labeling 48h after lentivirus transduction, i.e. around 24h after shBMAL1 expression begins.

Without puromycin selection, only GFP positive (GFP+) cells were considered as shRNA transduced. In the GFP+ population, the cells undergoing early apoptosis, i.e., Annexin V-positive and PI-negative cells, as well as the total apoptosis cells (Annexin V-positive cells) are all increased in shBMAL1 vs. shScr transduced HCT116 cells (p=0.0054 and p= 0.0393). However, shBMAL1 transduced SW480 (p=0.1858and p=0.1149) or SW620 (p=0.1705 and p=0.2601) cells only presented a trend to increase apoptosis (Figure 7A).

For HCT116 cells, we also measured P53 nuclear expression by flow cytometry, 48h after lentivirus transduction. For the GFP+ population, nuclei of shBMAL1 transduced cells exhibited increased P53 expression compared to shScr control cells (p=0.0070). However, this difference was not observed in the GFP negative (shRNA non-transduced) population (Figure 7B).

BMAL1-KD appropriately alters expression of other core circadian genes

BMAL1–KD induced similar, and expected, gene expression changes in the two primary CRC cell lines: decreased NR1D1 and increased CLOCK expression, but without modification of PER2, CRY1 or CRY2. Reduced NR1D1 expression is consistent with the exclusive control of its promoter by CLOCK–BMAL1 binding to E-box regulatory elements [32] and the presence of two BMAL1 binding sites in the gene (Supplementary Figure 1) [33]. This contrasts with only one binding site for BMAL1 in PER2, CRY1 or CRY2 genes [33], and explains why NR1D1 expression is more sensitive to the diminution of BMAL1. Similarly, NR1D1 (REV-ERBα) rapidly represses transcription of BMAL1 and CLOCK genes via REV-ERBΑ response elements (RREs) [34, 35]. So, decreased NR1D1 and increased CLOCK expression in the two primary BMAL1 knockdown CRC cells represent correct feedback regulation of core-clock gene expression [20–22].

BMAL1-KD increases mTOR activity in CRC cells

In the two primary CRC cell lines, BMAL1-KD increased activity of mTOR, a central regulator of cellular metabolism that links cellular energy and nutrients to cell division, growth, quiescence, senescence and death; and which is critically involved in cellular live, for example aging, diabetes and cancer [25, 26, 36]. Our result is coherent with observations in BMAL1 KO mice and the circadian rhythmicity of mTOR signaling [27, 37–39]. Moreover, we also demonstrate that BMAL1-KD increased phosphorylation of the mTOR effector, ribosomal S6 (Figure 2), whose protein kinase, S6K1, rhythmically phosphorylates BMAL1 so that it associates with cellular translational machinery for protein synthesis [39]. Thus our data support the links between BMAL1 and mTOR pathway, which suggests BMAL1 regulation of protein synthesis and the role of circadian timing in cancer development.

Different CRC cell fates triggered by BMAL1-KD depend not only on increased mTOR activity but also on P53 status of each cell line

BMAL1 knockout mice show increased mTOR activity, associated with age-related pathology and reduced lifespan, i.e. premature ageing [27, 40, 41]. Therefore, it is not surprising that cell senescence increased after BMAL1 knockdown and increased mTOR activity. However, why did this occur only in SW480 BMAL1-KD, but not HCT116 BMAL1-KD cells?

A possible explanation for this variance in cellular senescence between SW480 and HCT116 cells is their different P53 status. The HCT116 cell line expresses wild type (WT) P53, whereas the SW480 cell line carries a mutant P53 (R273H and P309S; mP53) which is only partially functional, such as inducing P21 expression [42]. Knockdown of this mP53 decreases colorectal cancer malignancy, indicating its important role in CRC development [43]. Blagosklonny et al. hypothesized that maximal P53 activity blocks mTOR, causing cellular quiescence or death, whereas partial P53 activity sustains mTOR activity and causes senescence [23–26]. In keeping with this hypothesis, we propose that the different P53 status (WT vs mutant) of HCT116 and SW480 cells underlies the different cell fates induced by mTOR activity after BMAL1-KD.

This proposal is supported by our findings. The reduction of BMAL1 in lentiviral transfected cells depends on the number of shRNA copies that are integrated, and will vary for each cell [44]. Thus just after lentiviral transfection, HCT116 BMAL1-KD and SW480 BMAL1-KD cultures will contain many cell populations each with distinct BMAL1, mTOR and P53 activation according to the number of integrated shBMAL1 copies. Thus, in HCT116 cells which integrate a high number of shBMAL1 copies, there will be high WT P53 activation, which can induce rapid apoptosis, as indeed we observed 24h after shBMAL1 expression (Figure 7). Consequently, during the one-week puromycin selection for stable BMAL1-KD, the HCT116 BMAL1-KD cells with strong P53 activation would be removed by apoptosis, leaving those with less P53. Such P53 expression will in turn induce expression of MDM2, which directs P53 for proteasome degradation to limit its expression level and activity [45]. As our results showed, stable HCT116 BMAL1-KD cells had high MDM2 and low P53 expression, consistent with functional MDM2-P53 negative feedback. Finally, in the absence of increased P53 activity, increased AKT and mTOR activity in the stable HCT116 BMAL1-KD cells can increase cellular proliferation (Figure 3).

Alternatively, for SW480 cells, those which integrate many shBMAL1 copies will stimulate mP53 activity, as demonstrated by nuclear translocation and expression of P21 and MDM2 (Figure 6). But mP53 has only partial function [42], as confirmed in our results by low expression of MDM2, which would limit MDM2-P53 negative feedback and explain the higher P53 expression in SW480 cells than HCT116 cells. Consequently, mTOR activity is sustained and provokes cell senescence. SW480 BMAL1-KD cells with fewer integrated shBMAL1 will only moderately activate mP53, not block mTOR and promote cell survival or proliferation. Our results, showing only a relatively small decrease in SW480 BMAL1-KD cell counts (Figure 3) despite profound cellular senescence (Figure 4), are consistent with cultures being transduced by a range of shBMAL1 copies.

Effect/influence of BMAL1 KD on CRC cell life requires functional circadian rhythm

In contrast to the two primary CRC cell lines, metastatic SW620 cells showed very little response to BMAL1-KD: a small increase in phosphorylated AKT without associated mTOR or S6 activation (Figures 2). A likely explanation for this muted response is an abnormal circadian timing system and low baseline expression of BMAL1 and other cell regulatory proteins (e.g. AKT, P21) in this cell line. In contrast to the two primary CRC cell lines (HCT116 and SW480), the metastatic cell line SW620 has severely diminished core-clock gene oscillation, indicating dysfunction of the transcription-translation feedback loops among different circadian proteins [20–22]. Although SW480 and SW620 cell lines were derived from primary and metastatic sites of the same patient, there was only 5.5% overlap of genes with oscillating expression profiles, indicating the loss of intact circadian clock during tumor progression [22]. Consistent with this, BMAL1 knockdown in metastatic SW620 cells did not alter NR1D1 and CLOCK expression, in contrast to the two primary CRC cells. Thus, a dysregulated core clock in SW620 cells, as indicated by poor oscillation of clock-controlled genes and low level of BMAL1 protein can explain why BMAL1-KD had less influence on SW620 cells than on SW480 cells in our experiments.

BMAL1 KD-associated oxidative stress reveals impaired P53 signaling in SW620 cells

Global deletion of BMAL1 induces an aging phenotype associated with oxidative stress [46], which damages DNA, including DNA strand breaks.

In stable HCT116 BMAL1-KD cells there was no evident oxidative stress (H2AX phosphorylation), consistent with vulnerable HCT116 BMAL1-KD cells with high WT P53 activation being eliminated by apoptosis just after shBMAL1 transduction, and P53 activation in surviving cells being blocked by MDM2-P53 negative feedback.

In contrast, our results show phosphorylated H2AX in SW480 and SW620 BMAL1–KD cells, indicating oxidative stress. In SW480 BMAL1-KO cells, there was a large increase in phosphorylated H2AX and mP53 activation. In contrast, SW620 BMAL1-KD cells only slightly increased phosphorylated H2AX but without associated mP53 activation: no P53 nuclear translocation or P21 expression, and even decreased MDM2 expression (Figures 4–6), which suggested impaired P53 signaling in SW620 cells. In the absence of appropriate P53 activation and its induction of apoptosis or senescence, the moderate increase of AKT activity permits faster growth (Figure 8).

Cell culture

HCT116, SW480 and SW620 cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with GlutaMAX (GIBCO, Life Technology, CA, USA) and 10% fetal bovine serum (FBS, Hyclone, UT, USA).

The shRNA sequence cloning in a lentiviral vector

A short hairpin RNA sequence (shRNA) against human BMAL1 (shBMAL1; Forward: 5'-CCGGAGAACCCAGGTTATCCATATTCTGCAGAATATGGATAACCTGGGTTCTTTTTT-3'; Reverse: 5'-CTAGAAAAAAAGAACCCAGGTTATCCATATTC TGCAGAATATGGATAACCTGGGTTCT-3') or a control scrambled sequence (shScr) were inserted separately into a lentiviral vector (pLKO-shBMAL1-GFP-puro or pLKO-Scr-GFP-puro), which also encode the reporter protein GFP and the puromycin resistance gene.

Lentivirus production and cell transduction

Lentivirus particles were produced as previously described [48] but without concentration. After lentiviral production, the lentivirus supernatant was filtered with 0.45 μm filters and stored at -80°C. For target cell transduction, 1 mL filtered viral supernatant mixed with 1 mL DMEM (10% Fetal Bovine Serum, FBS) was added to each well of 6-well plates containing 50% confluent cells in the presence of 8 μg/mL polybrene (Sigma, MO, USA) and incubated with for 24 hours.

Transduced cells were either used directly 48h after transfection, or after selection by adding 2 μg/mL puromycin (Thermo Fisher) to the medium 72h after transfection and culturing for 1 week. After puromycin selection, flow cytometry analysis was performed to confirm that the entire cell population was GFP positive. These GFP positive cells, named as BMAL1-KD or control (Ctr) cells, were used for subsequent analysis.

Analyses with or without puromycin selection are repeated with the cells from three independent lentivirus transductions.

MTT cell proliferation assay

The BMAL1-KD or control CRC cell lines were seeded in 96-well plates at an initial density of 2x10³ cells per well. Cell proliferation was measured daily during 4 days by Vybrant MTT Cell Proliferation Assay Kit (V13154, Molecular Probes, Invitrogen). Every time point was repeated 3-4 times in independent experiments. Two-way ANOVA was used for statistical analysis of 8 independent experiments.

Cell proliferation curve analysis

The BMAL1-KD or control CRC cell lines were seeded in 24-well plates at an initial density of 1x10⁴ cells per well. After trypsinization (Trypsin-EDTA, Thermo Fisher, MA, USA) and suspension in PBS containing 0.5% bovine serum albumin (BSA) and 2 mM EDTA, the number of living cells from each well was counted over 4 days by a Miltenyi Biotec AutoMACS cytometry. Three independent experiments were statistically analyzed by two-way ANOVA.

Quantitative RT-PCR (qRT-PCR)

Cells were collected and total RNA was extracted as previously described [49]. Reverse transcription was performed with Superscript II RT-kit (Invitrogen, CA, USA). Quantitative real time PCR was performed by using LightCycler 480 SYBR Green I master kit (Roche, Bâle, Switzerland). Primers of BMAL1 and 36B4 were used for gene amplification were previously described [49]. P21 primer: Forward-GACACCACTGGAGGGTGACT and Reverse-CAGGTCCACATGGTCTTCCT. P53 primer: Forward-GTTCCGAGAGCTGAATGAGG and Reverse-TCTGAGTCAGGCCCTTCTGT.

The relative quantification of target RNA by using 36B4 as a reference was computed with the Relquant software (Roche, Bâle, Switzerland) with the “delta delta Ct” (ΔΔCt) method. T-test was used for statistical analysis.

Flow cytometry analysis of S6 phosphorylation

Cells were detached with Trypsin-EDTA (Thermo Fisher, MA, USA), fixed in ice-cold 70% ethanol, washed twice in ice-cold PBS and then centrifuged at 300g for 10 min. Cells were incubated with anti-Phospho-S6-APC (#14733; Cell Signaling, MA, USA) in PBS containing 0.5% bovine serum albumin (BSA) and 2 mM EDTA for 30 min at 4°C and labelled with Alexa Fluor® 647 or Alexa Fluor® 546 conjugated secondary antibodies (Molecular Probes) and Hoechst 33342 (B2261, Sigma MO, USA). After labelling, cells were washed once time and analyzed in an LSR Fortessa™ cell analyzer (Becton Dickinson, NJ, USA).

Flow cytometry analysis of nuclear P53 expression

Cells were trypsinized and centrifuged at 1200 rpm for 5 min and suspended very gently in citrate solution (0.1% Trisodium citrate and 0.058% NaCl, pH=7.6) containing 0.001% NP40 and 10 mg/mL Hoechst 33342, incubated for at least 2h (maximum 24h) at 4°C to extract and stain cell nuclei. Cell nuclei were then incubated with anti-P53-PE (130-109-570, Miltenyi Biotec) in a PBS solution containing 0.5% BSA and 2 mM EDTA for 30 min at 4°C. After washing, nuclear P53 expression was analyzed by flow cytometry in an LSR Fortessa^TM cell analyzer (Becton Dickinson, NJ, USA). T-test was used for statistical analysis.

Apoptosis assay

48h after viral transduction, transfected cells were trypsinized and then 10⁶ cells were washed in 1 mL of 1x Binding Buffer (130-092-820, Miltenyi Biotec, Germany) and centrifuged at 300g for 10 min. After 15 min incubation with anti-Annexin V-Alexa Fluor® 647 (640943, BioLegend, CA, USA), cells were washed and suspended in 500 μL 1x binding buffer, 10 mg/mL propidium iodide (PI) solution was added immediately prior to analysis by flow cytometry. All the experiments or solution are realized at 4°C. T-test was used for statistical analysis.

Immunofluorescence and confocal microscopy

The cells were cultured on 17 mm coverslip until desired confluence (around 50%), fixed with 4% PFA, permeabilized with 0.1% Triton and then blocked with 0.3% BSA before being incubated with primary (Anti-pH2AX 9718, Cell Signaling) and secondary antibodies. The confocal images were captured by a confocal LEICA SP5-AOBS microscope with a 63X/1.4 NA oil-immersion objective. Hoechst 33342 was used for nuclear staining.

Cytoplasmic and nuclear extracts preparation

Cytoplasmic and nuclear extracts of different cell lines were separated and prepared with NE-PER™ Nuclear and Cytoplasmic Extraction Reagents (78833, Thermo Fisher) by following the kit instruction. The different extracts were stored at -80°C until western-blot analysis by anti-P53 (2527, Cell signaling). HSC70 expression was used as a control of cytoplasmic protein loading (SPA-815, Stressgen) and LAMIN B1 (ab16048, Abcam) was used as nuclear protein loading.

Western blot

Western blot analysis was performed with sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) as previously described [49] with different primary antibodies: Anti-HSC70 (SPA-815, Stressgen), or anti-BMAL1 (14020), anti-Phospho-S6 ribosomal protein (Ser240/244; 5456), anti-S6 ribosomal protein (2317), anti-phospho-AKT (Ser473; 4060), anti-AKT(4691), anti-pmTOR (Ser2448; 2971), anti-mTOR (2972), anti-P21 Waf1/Cip1 (2947), anti-MDM2 (86934), anti-P53 (2527) all from Cell Signaling Technology. The results were quantified by Image J. T-test was used for statistical analysis.