Nuclear Beclin 1 Destabilizes Retinoblastoma Protein to Promote Cell Cycle Progression and Colorectal Cancer Growth

Simple Summary The role of autophagy core-protein Beclin 1 in colorectal cancer (CRC) development remains controversial. Here, we show that nuclear Beclin 1 is up-regulated in CRC with a negative correlation to RB protein expression. Silencing of BECN1 up-regulates RB expression resulting in cell cycle G1 arrest and inhibition of xenograft tumor growth independent of p53. Ablation of BECN1 facilitates MDM2–MDMX complex formation to promote MDMX polyubiquitination and degradation, consequently leading to RB protein stabilization. These results reveal that nuclear Beclin 1 can promote CRC growth through modulation of RB protein stability and imply that nuclear Beclin 1 may be a prognostic indicator in human colorectal cancer. Abstract Autophagy is elevated in colorectal cancer (CRC) and is generally associated with poor prognosis. However, the role of autophagy core-protein Beclin 1 remains controversial in CRC development. Here, we show that the expression of nuclear Beclin 1 protein is upregulated in CRC with a negative correlation to retinoblastoma (RB) protein expression. Silencing of BECN1 upregulates RB resulting in cell cycle G1 arrest and growth inhibition of CRC cells independent of p53. Furthermore, ablation of BECN1 inhibits xenograft tumor growth through elevated RB expression and reduced autophagy, while simultaneous silencing of RB1 restores tumor growth but has little effect on autophagy. Mechanistically, knockdown of BECN1 promotes the complex formation of MDM2 and MDMX, resulting in MDM2-dependent MDMX instability and RB stabilization. Our results demonstrate that nuclear Beclin 1 can promote cell cycle progression through modulation of the MDM2/X-RB pathway and suggest that Beclin 1 promotes CRC development by facilitating both cell cycle progression and autophagy.


Introduction
Colorectal cancer (CRC) is the third most commonly diagnosed malignancy and the second leading cause of cancer death. The CRC is attributable to environmental risk factors including obesity, alcohol drinking, smoking, and inflammatory bowel disease. Inactivation of the tumor suppressor genes such as APC, TP53, and RB1 (the retinoblastoma gene) are critically important for CRC development [1]. The RB1 inactivation yields a mouse model of malignant colorectal cancer [2]. RB1 encodes prototypical tumor suppressor retinoblastoma protein (RB), which plays pivotal roles in the regulation of the cell cycle; DNA replication; DNA damage response; and repair, senescence, apoptosis, and differentiation [3]. RB binds to the E2F family of transcription factors and forms an active transcription repressor complex that blocks the expression of genes involved in DNA replication and cell cycle progression. The cell-cycle-dependent kinases (CDKs)-mediated hyperphosphorylation

Flow Cytometry Analysis
Cells were trypsinized, washed with cold PBS, and fixed in 70% ethanol at 4 • C overnight. Cells were stained with 50 µg/mL propidium iodide supplemented with 100 µg/mL RNase A at 37 • C in the dark for 1 h. Cells were then subjected to flow cytometry analysis by FACScan Flow Cytometer [6] (Becton Dickson, San Jose, CA, USA), and the data were analyzed using Cell Quest software (Becton Dickson).

Colony Formation Assay
Aliquots of 500 HCT116 or HCT116 p53 −/− cells were seeded in 6-well plates and cultured for 20 days. The colonies were fixed for 15 min with cold methyl alcohol, stained with 0.1% crystal violet for 15 min at room temperature, and photographed. Colonies of >50 cells were used for quantitative counts.

Real-Time Cell Analysis (RTCA)
For cell index (CI) analysis, experiments were performed using an xCELLigence Real-Time Cell Analysis (RTCA) [18] DP instrument (Roche Diagnostics GmbH, Mannheim, Germany) at 37 • C with 5% CO 2 . To measure the cell index of HCT116 or HCT116 p53 −/− cells in real-time, cells were seeded on gold microelectrodes embedded at the bottom of 16-well microplates (E-plates; Roche Diagnostics, Basel, Switzerland) at a density of 8.0 × 10 3 cells/well for HCT116 or HCT116 p53 −/− cells. The impedance was recorded at 15 min intervals. Cell index values were evaluated by the RTCA-DP software (Roche Diagnostics GmbH, Metroplaza, Germany).

Mouse Xenograft Studies
All the mouse strains were kept in standard, infection-free housing conditions, with 12 h light:12 h dark cycles and 3-5 mice per cage. Animals were housed in a pathogen-free barrier environment throughout the study. All animal experiments in this study were approved by the Institutional Animal Care and Use Committee of Sichuan University (IACUC), and the operating procedures were carried out in accordance with the guidelines formulated by China's Council on Animal Care.
For in vivo tumor progression, Antares2 [15]-expressing HCT116 p53 −/− (5.0 × 10 6 ) or their derivatives were subcutaneously transplanted into the right front flanks of 6-week-old female nude mice (n = 5/group) (Model Animal Research Center of Nanjing University, China). Tumor size was measured with a caliper every other day, and tumor volume was calculated by width 2 × length × 1 2 . A total 0.1 mM Diphenylterazine (DTZ, MCE) (the substrates of Antares2) in 100 µL PBS was injected into tumor regions in anesthetized mice using tribromoethanol (250 mg/kg intraperitoneally (i.p.)). Mice were recovered on heat pads for 5 min and bioluminescence was subsequently imaged using a Caliper IVIS Lumina (Perkin Elmer, Waltham, MA, USA). The images were processed using the Living Image 4.5.5 software.

Statistical Analyses
Data from at least three independent experiments were presented as mean ± SD, and data from animal experiments were presented as mean ± SEM. Two-tailed unpaired Student's t-test was used for comparing two groups of data. One-/two-way ANOVA with Tukey's test was used to compare multiple groups of data. p values of less than 0.05 were considered significant.

Elevated Expression and Nuclear Location of Beclin 1 Is Negatively Correlated with RB Expression in Malignant Human Colorectal Cancer Specimens
The tumor suppressor RB protein plays a pivotal role in the negative control of the cell cycle and tumor progression [23]. However, the clinic correlation between RB expression and human colorectal cancer remains controversial [24,25]. We therefore analyzed the current TCGA database for RB1 expression in human colorectal cancer. As shown in Supplementary Figure S1a, low expression of RB1 was associated with poor overall survival (OS) in human colorectal cancer patients. Furthermore, we found that high expression of Beclin 1 protein, but not mRNA, was associated with poor OS and relapse-free survival (RFS) in CRC patients (Figure 1a). To investigate the correlation between RB and Beclin 1, we then analyzed the correlation between Beclin 1 and RB protein expression of colorectal cancer patients in the TCGA-COADREAD dataset obtained from LinkedOmics [22]. Our analyses reveal a moderate negative correlation (r = −0.3245, Pearson's correlation coefficient) between RB and Beclin 1 proteins (revised Figure 1b) in contrast to mRNA levels that showed a weak positive correlation (data from cBioPortal [19,20], GDC TCGA COAD [26], and GEPIA [21]) (Figure 1b and Supplementary Figure S1b-d). We further examined the correlation of RB and Beclin 1 protein expression using the human colorectal cancer tissue microarrays (TMA) consisting of colorectal cancer specimens at different clinical stages. As shown in Figure 1c-e, reduced RB protein expression and increased Beclin 1 expression were observed in human colorectal cancer specimens at stages II and III compared with the specimens at stage I. Notably, the expression of Beclin 1 protein was dramatically increased in the nucleus and was negatively correlated with RB protein expression.

Ablation of BECN1 Leads to Cell Cycle G1 Arrest and Cell Growth Retardation Independent of p53
We then examined the effects of BECN1 ablation on the growth of HCT116 or HCT116 p53 −/− cells. As shown in Figure  Interestingly, while the expression of Beclin 1 L184A+L187A -a mutant defective in promoting autophagy due to its exclusive nuclear localization [13]-failed to restore autophagy as expected, it could effectively reverse the knockdown of BECN1-induced altered expression of MDMX and RB, resulting in significantly reduced G1/S arrest and restored colony formation (Figure 2f-h). Together, these results indicate that silencing of BECN1 leads to upregulation of RB protein expression, concomitant with the cell cycle G1 arrest and growth inhibition independent of p53, and importantly, nuclear Beclin 1 can impact cell growth through modulation of RB protein expression.   (c,d) Human colorectal cancer tissue microarrays consisting of cancer specimens from three cancer stages (stage I, n = 6; II, n = 14; III, n = 7) were subjected to IHC staining for RB and Beclin 1, with quantitative analyses using average optical density (AOD). (e) Pearson's correlation between RB and Beclin 1 in colorectal cancer patients was analyzed. Scale bar = 50 µm.

Ablation of BECN1 Downregulates MDMX Expression to Induce RB-Dependent Cell Cycle Arrest and Inhibition of Cell Growth
To investigate whether RB plays a causal role in cell cycle G1/S arrest induced by knockdown of BECN1, a set of rescuing experiments was performed. As shown in Figure 3a-e, knockdown of RB1 could significantly, but not completely, restore the colony formation, rate of cell growth, and cell cycle progression, all of which were inhibited by BECN1 ablation. Furthermore, restoration of MDMX, but not MDMX C437A -a mutant defective in the RB degradation [6]-significantly reversed the expression of RB, which was upregulated by knockdown of BECN1, accompanied by markedly reduced G1/S arrest and largely restored colony formation or rate of cell growth independent of p53 (Figure 3f-i).   Figure 2. Ablation of BECN1 leads to cell cycle G1/S arrest and growth inhibition through upregulation of RB protein expression. HCT116 or HCT116 p53 −/− cells stably expressing shRNA specific for BECN1 (#1 or #2) or control shRNA (shC) were subjected to western blot (a), colony formation assays (b), real-time cell analyses (RTCA) (c), and FACS analyses (d). Quantifications of cell cycle derived from FACS analyses (e). HCT116 p53 −/− cells stably expressing shBeclin 1 were infected with lentivirus expressing Beclin 1 or Beclin 1 L184A+L187A followed by western blot (f), colony formation assays (g), and FACS analyses (h). Data derived from three independent experiments were presented as mean ± SD. * p < 0.05, ** p < 0.01, *** p < 0.001. The uncropped blots and molecular weight markers are shown in Supplementary File S1. h.

Ablation of BECN1 Facilitates MDM2-MDMX Interaction to Promote MDMX Degradation Leading to RB Protein Stabilization
Our aforementioned results show that BECN1 ablation leads to the downregulation of MDMX and upregulation of RB. Notably, MDMX can rescue BECN1 ablation-induced RB upregulation. Since MDMX has been shown to promote RB protein degradation [6], we postulated that MDMX might mediate Beclin 1-induced regulation of RB protein instability. To this end, we first investigated the effect of BECN1 knockdown on the mRNA expression of RB1 or MDMX and on respective protein stability. As shown in Supplementary Figure S1h, knockdown of BECN1 had little effect on the steady-state RB1 or MDMX mRNA levels. By contrast, knockdown of BECN1 significantly extended the protein half-life of RB (Figure 4a,b). On the other hand, inhibition of proteasome by MG132 significantly rescued the shBeclin 1-induced downregulation of MDMX, suggesting that BECN1 depletion leads to accelerated MDMX proteasome degradation (Figure 4c). Indeed, silencing of BECN1 significantly facilitated MDMX polyubiquitylation (Figure 4d). We next investigated the molecular basis on which Beclin 1 impacts MDMX polyubiquitylation and RB protein stability. We have previously shown that MDM2 binds to and promotes proteasomemediated degradation of RB and that MDMX binds to and promotes RB degradation in an MDM2-dependent manner [5,6]. Notably, it has been shown that MDM2 binds to and promotes MDMX polyubiquitylation and degradation [27,28]. We thus hypothesized that Beclin 1 might affect the interaction of the MDM2-MDMX-RB complex. As shown in Figure 4e, endogenous Beclin 1-MDM2-MDMX-RB quadruple protein complexes were readily detected in HCT116 or HCT116 p53 −/− cells. Furthermore, ablation of BECN1 led to a significant increase in the interaction of MDM2 and MDMX (Figure 4f). We next examined the domains involved in the interaction between Beclin 1 and MDM2, MDMX, or RB. As shown in Figure 4g-j, Beclin 1 lacking the BH3 domain significantly reduced its interaction with MDM2 or RB, while it was fully capable of interaction with MDMX. Beclin 1 lacking the BH3 and CCD domains was unable to interact with either MDMX, MDM2, or RB. Furthermore, while Beclin 1 could easily form stable complexes with RB, silencing of MDM2, but not silencing of MDMX, led to significantly reduced Beclin 1 interaction with RB ( Figure 4k). Silencing of both MDM2 and MDMX led to even weaker Beclin 1-RB interaction ( Figure 4k). Together, these results suggest that the CCD domain of Beclin 1 is involved in MDMX interaction while both the BH3 and CCD domains of Beclin 1 are involved in interaction with MDM2. RB is likely piggy-backed to form the RB-MDM2-MDMX-Beclin 1 complex, in which MDM2-MDMX promotes RB protein degradation, keeping with our previous observation [6]. Ablation of BECN1 facilitates MDM2-MDMX complex formation and promotes MDMX polyubiquitylation and degradation, consequently leading to RB protein stabilization (Figure 4l).

Knockdown of BECN1 Suppresses Xenograft Tumor Growth through Activation of RB
We next investigated the role of the Beclin 1-RB pathway in colorectal cancer growth in xenograft mouse models. HCT116 p53 −/− -Antares2 cells [15] bearing BECN1 shRNA were stably expressed in RB1 shRNA and were transplanted subcutaneously into flanks of nude mice. As shown in Figure 5a-f, depletion of BECN1 led to significantly reduced xenograft tumor growth, concomitant with elevated RB and markedly reduced Ki67 (Figure 5d,e). Notably, depletion of BECN1 also led to reduced autophagy, as evidenced by increased expression of p62 and reduced expression of LC3 aggregates (Figure 5d, insets). Importantly, the knockdown of RB1 remarkably rescued the tumor growth inhibited by the BECN1 depletion. Collectively, these results demonstrate that depletion of BECN1 suppresses colorectal cancer growth, at least in part, via elevation of RB expression.

Discussion
In the present study, several new findings were observed regarding the role of nuclear Beclin 1 in colorectal cancer. First, the expression of Beclin 1 protein was negatively correlated with tumor suppressor RB protein expression in later stages of colorectal cancers. The second important finding was that nuclear Beclin 1 is critically involved in cell cycle progression through modulation of RB protein stability. Ablation of BECN1 leads to elevation of RB protein expression, resulting in inhibition of cell cycle progression and retardation of colorectal cancer cell growth in vitro and in vivo. Mechanistically, knockdown of nuclear Beclin 1 promoted MDM2-MDMX interaction, resulting in reduced MDMX expression and upregulated RB expression.
Beclin 1 has an important role in the growth and metastasis of human colorectal cancer. However, the role of Beclin 1 linked to autophagy in CRC development has been extensively studied with conflicting results. Loss of Beclin 1 expression defines poor prognosis by promoting anti-apoptotic pathways, whereas overexpression of Beclin 1, being linked with tumor hypoxia and acidity, also defines subgroups of tumors with aggressive clinical behavior [29]. Meta-analysis indicated that the elevated Beclin 1 expression is associated with tumor metastasis and a poor prognosis in CRC patients [30]. Importantly, in KRAS-mutated CRC, nuclear Beclin 1 expression was associated with a significantly decreased OS [31]. Here, we showed that the protein expression of Beclin 1 negatively correlates with OS and RFS of CRC patients, while there is no significant correlation between the BECN1 mRNA expression and the OS and RFS of CRC patients, which suggests that Beclin 1 protein abundancy is likely influenced by post-translational modifications such as ubiquitylation/deubiquitylation in CRC progression [32]. Since cytoplasmic Beclin 1 is fundamentally important in autophagy, it is conceivable that nuclear Beclin 1 plays an autophagy-independent role in colorectal tumorigenesis. At present, the only known study of nuclear Beclin 1 reported that Beclin 1 directly interacts with DNA topoisomerase IIβ and can be recruited to the DNA double-strand break sites to facilitate DNA repair progression [33]. In this study, we show that nuclear Beclin 1, unable to induce autophagy, can modulate RB protein expression, thereby regulating cell cycle and colorectal cancer cell growth. Beclin 1 functions as an adapter for recruiting multiple proteins in different biological processes. The interaction between Beclin 1 and MDMX has been preliminarily revealed in a cancer-focused protein-protein interactions network research [34], without validation. In this study, we show that Beclin 1 deficiency leads to an elevated RB protein stability in an MDMX-dependent manner. Restoration of nuclear Beclin 1 (Beclin 1 L184A+L187A ) significantly rescues MDMX expression, which facilitates the degradation of RB.
MDM2 and MDMX are deregulated in many human cancers and exert their oncogenic activity predominantly by inhibiting p53 and RB [35]. MDM2 directly binds to the RBC-pocket, thus preventing RB-E2F interaction and also promoting RB proteasomal degradation [4,5]. MDMX is a close structural analog of MDM2; however, MDMX lacks intrinsic E3 ligase activity. MDM2 binds to MDMX and promotes MDMX proteasomal degradation [27,28]. In our previous study, we showed that MDMX binds to MDM2 and enhances MDM2 ubiquitin E3 ligase activity towards RB, thereby enhancing RB proteasomal degradation [6]. In this work, we showed that ablation of BECN1 facilitates the interaction between MDM2 and MDMX, resulting in the degradation of MDM2-mediated MDMX degradation.
Autophagy has been reported to exert both tumor-suppressive and tumor-promoting effects in colorectal cancer [9]. Autophagy can prevent the transformation from normal to malignant during the early stages of tumor formation [9]. In this study, we found that elevated Beclin 1 expression correlates with reduced RB protein in advanced stages of colorectal cancer. Silencing of BECN1 leads to RB-induced G1/S cell cycle arrest, which is independent of autophagy. However, restoration of nuclear Beclin 1 can not fully rescue BECN1 silencing-mediated reduction of colon cancer cell colony formation. In addition, in the xenograft tumor mouse model, BECN1 ablation leads to robust inhibition of tumor growth, attributable to both upregulated RB expression and reduced autophagy. Knockdown of RB1 cannot completely restore tumor growth nor restore autophagy. These results suggest that ablation of BECN1 inhibits colorectal cancer growth through inhibition of both cell growth and autophagy.

Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.