BEX2 influences the MCL1-Hedgehog signaling axis to regulate the potential of stemness characterization in colorectal cancer.

Introduction
Colorectal cancer (CRC) is a malignant tumor that is prevalent globally and poses a substantial health challenge1,2. Despite advances in treatment, relapse remains a major issue that is largely due to cancer stem cells (CSCs)3–5. CSCs exhibit strong self-renewal, resistance to therapy, and can drive tumor heterogeneity and recurrence through high expression of markers, like OCT4, Nanog, EpCAM, and SOX26–8.
Dysregulation of the Hedgehog (Hh) pathway is crucial for maintaining stemness, cell fate determination, and self-renewal in CSCs9–11. Recent studies have shown that myelogenous leukemia-1 (MCL1) interacts with the Gli inhibitor, SUFU, to block inhibition of Gli1 and thus activating the Hh pathway12,13. MCL1 is often aberrantly expressed in malignant tumors14–16 but the role of MCL1 in the Hh pathway and CSC regulation has not been established.
BEX2, an intrinsically disordered protein, is epigenetically silenced in primary gliomas and re-expression of BEX2 inhibits tumor proliferation17,18. Our previous research showed that silencing BEX2 enhances migration and invasion in CRC cells, suggesting a role in regulating tumor malignancy under some conditions19. However, the comprehensive role of BEX2 in regulating CRC stemness is unknown.
This study determined the impact of BEX2 knockdown or overexpression on cancer stemness, Hh signaling, and therapeutic resistance in CRC in vitro and in vivo. BEX2 was shown to reduce stemness in CRC cell lines by promoting MCL1 degradation and inhibiting the Hh pathway. Conversely, MCL1 enhanced Hh activity and CRC stemness. Inhibiting MCL1 expression counteracted the enhanced stemness in BEX2 knockdown cells. The findings herein highlight BEX2 as a promising target for CRC therapy.

Materials and methods

Cell culture, transfection, and transduction
Human CRC cell lines (DLD1 and HCT116) were obtained from the American Type Culture Collection [ATCC] (Manassas, VA, USA). These cells were cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum [FBS] (Gibco), 100 U/mL of penicillin, and 100 μg/mL of streptomycin. BEX2 stable knockout cells (BEX2-KO-DLD1) were established previously19. HEK 293 cells, also procured from the ATCC, were cultured in Dulbecco’s modified Eagle’s medium (Gibco, Carlsbad, CA) with the same supplements.
Transfection was carried out using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). Transduction was performed using lentivirus at a multiplicity of infection of 10. Stable clones were selected with 4 μg/mL of puromycin (Sigma-Aldrich, Beyotime, Hangzhou, China). Constructs were transformed into HEK 293T cells with lentivirus packaging reagents (Addgene, Invitrogen, Carlsbad, CA) following the manufacturer’s instructions. The shRNA sequences utilized are listed in Table S1.

Establishment of stable BEX2 overexpression cell lines
BEX2 overexpression in HCT116 cells was achieved using the lentiviral-based vector, pLent-RFP-Blasticidin-CMV (Addgene, Invitrogen, Carlsbad, CA). The lentiviral-based vector, pLKO.1-Puro (Addgene), was utilized for shRNA expression. BEX2 was amplified and cloned into the pCMV-HA vector (Addgene, Invitrogen, Carlsbad, CA) using a BEX2 DNA fragment as a template in the transfection of HEK 293 cells. Similarly, MCL1 was amplified using an MCL1 DNA fragment as a template and cloned into the pCMV-3xflag vector (Addgene, Invitrogen, Carlsbad, CA).

Bioinformatic analyses
Three GEO datasets (GSE14333, GSE24549, and GSE24551) containing patient clinical characteristics and RNA-seq data from colon adenocarcinoma (COAD) tissues were utilized. Pearson’s correlation coefficients were used to describe the relationship between BEX2 and disease progression. Normalized RNA-seq data for BEX2, OCT4, NANOG, CD44, and CD133 from the TCGA database (https://portal.gdc.cancer.gov/) were analyzed to assess the correlations in COAD tissues.

Flow cytometry analysis
Cells were harvested and resuspended in PBS containing 2% FBS, then stained with fluorescein Isothiocyanate (FITC)-conjugated CD44 antibody (Biolegend, Shanghai, China) or allophycocyanin (APC)-conjugated CD133 antibody (Biolegend) for 30 min at 4°C in the dark. DAPI (Sigma-Aldrich) was used for dead cell exclusion. Flow cytometry analysis was performed using a Beckman Coulter DxFLEX and data were analyzed using FlowJo software.

Tumor sphere formation assay
Cells (5 × 103 cells per well) were seeded in 6-well ultra-low attachment plates (Corning, NY, USA) in serum-free DMEM/F12 (1:1) medium (Gibco) containing 2% B27 supplement (Gibco), 20 ng/mL human fibroblast growth factor [FGF] (PeproTech, MA, USA), and 20 ng/mL epidermal growth factor [EGF] (PeproTech, MA, USA). Cells were fixed with 4% formaldehyde after 12–14 d and stained with 0.1% crystal violet, followed by photography. Spheres with diameters > 75 μm were counted. Each experiment was performed at least three times.

In vivo xenograft assay
Six-week-old female NOD/SCID mice (SLAC Laboratory Animal Co., Ltd., Shanghai, China) were subcutaneously injected with a total of 1 × 102–1 × 105 DLD1 or HCT116 cells embedded in Matrigel (BD Biosciences, Sigma, MO, USA). Tumor size and weight were evaluated 7 weeks post-injection upon sacrificing the mice. All experiments were performed following the Institutional Guidelines for Animal Care. The animal experiment protocol was approved by the Ethics Committee of the Second Affiliated Hospital of Zhejiang University School of Medicine [Approval No.: 006 (2021)].

Ubiquitylation assay
The HEK 293 cells were transfected with the indicated constructs, including Flag-MCL1, BEX2-MYC, K48R, FBXW7, and BTRC, for 24 h, preceded by a 15 μM MG132 (MCE) pre-treatment for 8 h. The BEX2-NC-HCT116 and BEX2-OE-HCT116 cells were transfected with the indicated constructs and also pre-treated with 15 μM MG132 (MCE) for 8 h. Proteins from whole-cell lysates were subsequently extracted, sonicated, and diluted 10 times with NP-40 lysis buffer, then supplemented with protease inhibitor cocktail (MCE) and immunoprecipitated with Flag beads. The ubiquitinated form of MCL1 was detected by western blot using anti-Ub antibody (Cell Signaling Technology, Beverly, MA, USA).

Co-immunoprecipitation (Co-IP) and immunoblotting (IB)
Cells were lysed using RIPA buffer [50 mM Tris-HCl (pH 7.5), 1% NP40, 0.35% DOC, 150 mM NaCl, 1 mM EDTA, and 1 mM EGTA supplemented with protease and phosphatase inhibitor cocktail; (MCE, Shanghai, China)] or IP lysis buffer [10 mM Tris-HCl (pH 7.5), 2 mM EDTA, 0.5% NP40, and 150 mM NaCl supplemented with protease and phosphatase inhibitor cocktail; (MCE, Shanghai, China)]. The cell lysates were centrifuged at 20,000 × g and 4°C for 10 min. The supernatants were then incubated with prepared bead-antibody complexes [10 μg of antibody and 50 μL of protein A/G magnetic beads; (Selleck, Shanghai, China)] or 5 μL of antibodies against Flag M2 magnetic beads (Sigma-Aldrich) at 4°C overnight. The proteins were electrophoresed in 12% SDS-PAGE, transferred to a PVDF membrane (Bio-Rad, Hercules, CA, USA), then incubated with a primary antibody at 4°C overnight and an appropriate secondary antibody for 1 h at room temperature (RT). The protein signal was visualized using an enhanced chemiluminescence substrate (YEASEN, Shanghai, China) and scanned using the Tanon 5200 chemiluminescence imaging system. The primary antibodies included rabbit anti-BEX2 (Sigma-Aldrich), rabbit anti-Flag (Cell Signaling Technology, Beverly, MA, USA), rabbit anti-MCL1 (Cell Signaling Technology), mouse anti-MYC (Cell Signaling Technology), mouse anti-Ub (Cell Signaling Technology), rabbit anti-NANOG (Proteintech, Rosemont, USA), rabbit anti-OCT4 (Proteintech), mouse anti-β-actin (Huabio), mouse anti-Lamin B1 (Proteintech), and rabbit anti-GAPDH (Proteintech).

Immunofluorescence analysis
Cells on coverslips were washed twice in PBS and permeabilized with 4% (w/v) paraformaldehyde (PFA) for 10 min at RT. The cells were permeabilized with 0.5% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) in PBS for 15 min. Next, the cells were blocked with 1% bovine serum albumin [BSA] (Sigma-Aldrich, St. Louis, MO, USA) for 1 h, followed by incubation with anti-HA (Cell Signaling Technology) and anti-MCL1 antibodies (Cell Signaling Technology) at RT for 2 h. The cells were then incubated with Alexa Fluor 488- or Alexa Fluor 555-conjugated secondary antibodies (1:500; Abcam, Shanghai, China) for 1 h at RT. Nuclei were stained with DAPI for 5 min at RT, then observed under a Zeiss LSM 710 laser-scanning confocal imaging system (Carl Zeiss, Germany).

Quantitative real-time PCR (qRT-PCR)
Total RNA was extracted using TRIzol® reagent (Invitrogen) and cDNA was reversed-transcribed using the PrimeScript™ RT Master Mix (Takara, Beijing, China). Quantitative PCR was performed using the SYBR® Premix Ex Taq™ GC (Takara). GAPDH mRNA was used to normalize RNA inputs. The primers used for real-time PCR in the current study are listed in Table S2.

Transwell migration and Matrigel invasion assays
Cell migration was determined using the Transwell migration assay (8.0 μm pore size; Corning, NY, USA). DLD1 (2 × 104 and 4 × 104) or HCT116 cells (3 × 104 and 6 × 104) were seeded in serum-free medium in the upper chamber with or without Matrigel and the lower chamber was filled with RPMI-1640 medium containing 20% FBS. Non-migrating cells were removed after 36 h with cotton buds and the migrated cells were fixed with methanol, stained with crystal violet dye, and counted in nine different fields.

Wound healing assay
Cells were seeded in 6-well plates (Corning, NY, USA) for 24 h before being scratched with a 200-μL pipette tip. Wound healing within the same scraped line was examined and photographed at the indicated time points (0, 24, and 48 h) after rinsing the cells with PBS to remove cellular debris. Each experiment was repeated three times.

Cell viability assay
Cell viability was assessed using the Cell Counting Kit-8 [CCK-8] (Dojindo, Kumamoto, Jappan) following the manufacturer’s protocol. Briefly, 5,000–7,000 cells per well were seeded in 96-well plates (Corning, NY, USA) and cultured for the specified duration. Subsequently, CCK-8 solution was added and the absorbance at 450 nm was determined using a SpectraMax microtiter plate reader (Molecular Devices, USA).

Gene set enrichment analysis (GSEA)
GSEA was performed on two groups (BEX2-WT and BEX2 KO) using the GSE112590 dataset to identify the most enriched and depleted KEGG pathways from the differentially expressed genes20. Differential genes between the high and low expression groups were subjected to enrichment analysis (FDR < 0.25 indicates statistical significance). The top 10 significantly enriched pathways were selected for further visual analysis.

Flow cytometric sorting
Cells were fully digested using trypsin (Gibco) and resuspended using 2% cell staining buffer (Biolegend, Shanghai, China), then labeled with APC-conjugated CD133 antibody (Biolegend). Flow cytometric sorting (BD FACSAria™ III, BD, SJ, CA, USA) was utilized using the BD FACSAria™ III to isolate specific DLD1 cells based on CD133 expression.

Extreme limiting dilution assay (ELDA)
Cells with reduced cell density were cultured in ultra-low adherence 96-well plates with fresh medium added every 3 d. The number of positive wells with spheroids was counted after 2 weeks of cell culture. Limiting dilution analysis was performed (http://bioinf.wehi.edu.au/software/elda/).

Nuclear and cytoplasmic protein extraction
Nuclear and cytoplasmic proteins were obtained using the Nuclear and Cytoplasmic Protein Extraction Kit (P0028; Beyotime, Shanghai, China) following the manufacturer’s instructions.

Statistical analysis
All data were statistically analyzed using GraphPad Prism 6.0 (GraphPad Software, Inc., San Diego, CA, USA) and SPSS 20.0 (SPSS Inc, Chicago, IL, USA). A two-tailed t-test was utilized to compare the differences between the two groups. Pearson’s test was used to determine correlations between clinicopathologic parameters and protein expression. Data are presented as the mean ± standard deviation (SD) or standard error of the mean (SEM). Statistical significance was set at a P < 0.05.

Results

Correlation of BEX2 expression with stemness markers and clinical characteristics of CRC
TCGA analysis revealed significantly reduced BEX2 expression in CRC tissues (Figure 1A). Kaplan-Meier analysis in GSE24551 indicated shortened disease-free survival (DFS) in the BEX2 low-expression group (P = 0.02211; Figure 1B). Meta-analysis across GEO datasets (GSE14333, GSE24549, and GSE24551) consistently demonstrated a direct correlation between higher BEX2 expression and improved disease-free survival (DFS) in CRC patients (Figure S1). Notably, BEX2 expression had a negative association with stemness markers (CD133 and CD44; Figure 1C). Molecular analysis using TCGA and GTEx databases indicated that BEX2 overexpression was associated with reduced CD44 and PROM1 (CD133) expression (Figure 1D, E). Therefore, there was a negative correlation between BEX2 and the stemness phenotype in CRC.

BEX2 knockout enhances CSC phenotypes in CRC in vitro and in vivo
BEX2-knockout cell lines were established in DLD1 cells (DLD1 KO and BEX2-KO-DLD1) and BEX2-overexpressing cell lines in HCT116 cells (HCT116 OE and BEX2-OE-HCT116) to further verify the relationship between BEX2 expression and stemness potential in CRC cells. Sphere-forming assays showed increased spheroid tumor proliferation in BEX2-KO-DLD1 cells and reduced proliferation in BEX2-OE-HCT116 cells (Figure 2A, B). Flow cytometry analysis revealed a higher proportion of CD133+ cells in the BEX2-knockout group compared to the control group (Figure 2C, D). CD133-positive cell populations in BEX2-OE-HCT116 cells showed significantly smaller and fewer spheres (Figure S2A). Pluripotency transcription factors (Nanog and OCT4) were more strongly expressed in CRC cells with low BEX2 levels (Figures 2E, F and S2B). BEX2-KO-DLD1 cells exhibited enhanced chemotherapy resistance to oxaliplatin (Figure 2G) and increased invasion and migration abilities (Figure 2I, K). In contrast, BEX2-OE-HCT116 cells exhibited a marked reduction in chemoresistance to oxaliplatin (Figure 2H) with significantly impaired invasive and migratory capabilities (Figure 2J, L). Rescue experiments confirmed that the observed effects were specifically due to BEX2 itself (Figure S3). ELDAs demonstrated that BEX2 knockout cells had a stronger tumor presence compared to wild-type cells (Figure S4A). BEX2-OE-HCT116 cells were less tumorigenic than the control group (Figure S4B). In vivo studies showed that BEX2 knockout significantly increased tumorigenicity with larger tumors formed from as few as 1 × 104 cells in the BEX2-KO-DLD1 group (Figures 3A–D and S5A). Conversely, the BEX2-OE-HCT116 group exhibited reduced tumorigenesis with the same cell count in NOD/SCID mice (Figures 3E and S5C). Tumorigenesis was noted in DLD1 cells with 1 × 102–1 × 105 cells and in HCT116 cells with 1 × 102–1 × 105 cells (Figure S5A–D). Collectively, these findings underscored that BEX2 knockout enhances stem cell marker expression and spheroid growth under stem cell-permissive conditions in CRC cells.

BEX2 inhibits the potential of cancer stemness characterization by up-regulating the Hh signaling
Gene Set Enrichment Analysis [GSEA] (Figure 4A) was performed to explore the mechanistic intricacies of the BEX2 impact on CRC cell stemness characterization. The analysis was performed using CRC tissue data from GSE112590 with BEX2 expression subgroups defined by the median level of BEX2 expression as the cut-off. Pathways related to African trypanosomiasis, fat digestion and absorption, nitrogen metabolism, primary bile acid biosynthesis, and vitamin digestion and absorption were most significantly enriched, as shown in Figure 4A. The analysis highlighted a significant enrichment of the Hh signaling pathway in the BEX2 low-expression group (Figure 4B). Notably, Gli1, PTCH1, and HHIP (Figure 4C) were substantially upregulated in BEX2-KO-DLD1 cells compared to the control group, whereas BEX2-OE-HCT116 cells showed a significant decrease in these markers (Figure 4D). Western blot analysis corroborated the elevated expression of Gli1 and PTCH1 in BEX2-KO-DLD1 cells and the reduction in BEX2-OE-HCT116 cells (Figure 4E, F). Importantly, the Hh pathway inhibitor, GDC-0449, effectively reversed the effects induced by low BEX2 expression, substantiating the involvement of the Hh signaling pathway (Figure 4G). Treatment with GDC-0449 not only inhibited sphere formation in BEX2-KO-DLD1 and HCT116 cells (Figures 4H, I and S6) but also confirmed that BEX2 regulates the potential of CRC cell stemness characterization through an Hh signaling-dependent mechanism. These findings underscore the intricate interplay between BEX2 and the Hh signaling pathway in modulating the stemness properties of CRC cells.

BEX2 attenuates MCL1 stability via the ubiquitin-mediated proteasomal system
Previous studies have suggested interactions between BCL-2 family members and BEX221. The crosstalk between BEX2 and MCL1 was explored given the established role of MCL1 in maintaining cancer stemness15,16. BEX2-KO-DLD1 cells showed elevated MCL1 expression compared to controls, while BEX2 overexpression in CRC cells led to reduced MCL1 levels that was confirmed by western blot (Figure 5A, B). Immunoblot analysis further substantiated a negative correlation between BEX2 and MCL1 expression (Figure 5C). Co-IP analysis confirmed the interaction between BEX2-MYC and FLAG-MCL1 co-expressed in HEK 293 cells (Figure 5D, E). Semi-endogenous BEX2 and endogenous MCL1 were co-precipitated in HCT116 cells under physiologic conditions (Figure 5F, G). In addition, co-localization of MCL1 and BEX2 in the cytoplasm was demonstrated through immunofluorescence in HEK 293 cells (Figure 5H). Co-IP analysis was used to confirm the interaction between BEX2 and MCL1 in protein extracts from the cytoplasm and nucleus of HCT116 cells. The results demonstrated increased binding between BEX2 and MCL1 in the cytoplasm (Figure 5I). The results showed that BEX2 binds to MCL1 in the cytoplasm and nucleus.
The molecular mechanism by which BEX2 interacts with MCL1 expression was subsequently probed. Surprisingly, the MCL1 mRNA level showed no significant difference between BEX2-OE-HCT116 and control cells (Figure S7A). Notably, chloroquine (CQ) treatment did not alter MCL1 expression in HCT116 cells overexpressing BEX2 (Figure S7B), suggesting that BEX2-mediated MCL1 regulation operates independent of autophagy inhibition. Conversely, treatment with cycloheximide (CHX) resulted in a time-dependent decrease in MCL1 protein levels (Figure 6A), suggesting that BEX2 influences the stability of MCL1. A previous study showed that the ubiquitin-proteasome system may regulate the stability of MCL122, so whether BEX2 promotes MCL1 degradation in a proteasome-dependent manner was tested. Notably, the proteasome inhibitor, MG132, inhibited the degradation of endogenous MCL1 protein under conditions of BEX2 overexpression (Figure 6B), indicating that BEX2 may promote ubiquitin-proteasome-dependent degradation of MCL1. Further analysis revealed that BEX2 overexpression led to a typical ubiquitination smear in the presence of MG132, indicating enhanced polyubiquitination of MCL1 (Figures 6C, D and S7C). Considering that FBXW7 and BTRC are known E3 ligases implicated in the ubiquitination-mediated degradation of MCL123–25, FBXW7 plasmids were introduced and increased ubiquitination of MCL1 was noted (Figure 6E, F). Co-IP assays were performed in HEK 293 cells co-transfected with BEX2-MYC and Flag-BTRC or Flag-FBXW7. The results demonstrated the interaction between BEX2 and BTRC but not with FBXW7 (Figure S7D, E). These results collectively suggested that BEX2 promotes the ubiquitination of MCL1 through BTRC, leading to MCL1 degradation.

BEX2 inhibits the potential of CRC stemness characterization by facilitating ubiquitination of MCL1
An MCL1-specific inhibitor (MIM1) was used in BEX2-KO-DLD1 cells to discern the impact on stemness-related genes and further elucidate the regulatory role of MCL1 in the potential of CRC cell stemness characterization orchestrated by BEX2. Treatment with MIM1 significantly suppressed the expression of key markers associated with CSCs in CRC cells (Figures 7A and S8A). By leveraging MCL1 shRNA to silence MCL1 in BEX2-KO-DLD1 cells, knockdown efficiency was subsequently confirmed through qRT-PCR and western blot analyses (Figure S8B, C). Suppression of MCL1 expression in BEX2-KO-DLD1 cells partially reversed the increase in the CD44+CD133+ cell population (Figure 7B). The formation of spheres in shMCL1-BEX2-KO-DLD1 cells was significantly diminished compared to the control group (Figure 7C) and accompanied by a substantial reduction in NANOG and OCT4 expression (Figure 7D). Flow cytometric sorting by CD133 revealed significantly reduced levels of OCT4 and NANOG in shMCL1-BEX2-KO-DLD1 cells compared to shCtrl-BEX2-KO-DLD1 cells (Figure S8D). Notably, the migratory and invasive capabilities of shMCL1-BEX2-KO-DLD1 cells were attenuated compared to the control group (Figures 7E and S8E). Moreover, shMCL1-BEX2-KO-DLD1 cells exhibited heightened sensitivity to chemotherapy (Figure 7F). In addition, knockdown of MCL1 expression in BEX2-KO-DLD1 cells partially mitigated the enhanced sphere formation capability observed in ELDAs, as shown in Figure 7G. Mice were injected with DLD1 cells and BEX2 depletion in the presence or absence of MCL1 shRNA expression to further examine the role of BEX2 impacting the MCL1-Hh signaling axis in regulating the potential of CRC cell stemness. Depletion of BEX2 enhanced tumor growth, which was counteracted by the expression of MCL1 shRNA (Figures 7H and S8F). These findings suggested that BEX2 reduces CRC cell stemness by promoting MCL1 ubiquitination and degradation.

Discussion
CRC remains a major health threat due to high rates of metastasis, recurrence, and resistance to treatment, all of which contribute to poor patient outcomes26. The CSCs within CRC are known for heightened resistance to treatments and greater metastatic potential27–29. These CSCs are critically involved in the recurrence of CRC following conventional therapies30,31. Notably, chemotherapy can increase the number of CSCs in CRC tissues, highlighting the clinical importance32. Understanding the molecular mechanisms that regulate and maintain CSCs in CRC is crucial for developing effective therapeutic strategies.
Recent studies have underscored the heterogeneity of CRC CSCs with respect to phenotype, function, and transcriptomics33. Markers, such as CD4434 and CD13335, are widely used to identify and characterize these CSCs, which helps in understanding the diverse nature of these cells. Exploring the genes and regulatory mechanisms affecting CSCs in CRC could significantly advance clinical treatment approaches.
The current study revealed that BEX2, a gene associated with malignant characteristics in CRC19,36, has a critical role in regulating the stemness potential of CRC. Reduced BEX2 expression was shown to be correlated with heightened stemness features and poorer prognosis in CRC patients. Specifically, BEX2 was shown to facilitate degradation of MCL1 and inhibit Hh signaling, thereby diminishing the stemness potential of CRC cells. MCL1, a member of the BCL-2 anti-apoptotic family, is crucial for cancer cell survival by resisting apoptosis. MCL1 is frequently amplified and overexpressed in various cancers, particularly in patients with chemotherapy-resistant tumors. For example, MCL1 gene amplification is prevalent in drug-resistant triple-negative breast cancer (TNBC) after neoadjuvant chemotherapy37 and elevated MCL1 levels have been reported in chemotherapy-resistant ovarian cancer patients14. The current study highlighted the importance of MCL1 in the stemness regulation mechanism mediated by the BEX2-MCL1-Hh signaling pathway. MCL1 modulates the expression of stemness-related proteins, chemoresistance, invasive potential, and tumorigenicity in CRC cells.
Interestingly, MCL1 interacts with the Hh inhibitory factor, SUFU, disrupting the inhibitory effect on Gli1 and activating the Hh pathway12. Co-IP confirmed direct binding between MCL1 and Gli1, as evidenced by immunoprecipitation of MCL1 followed by Gli1 immunoblotting (Figure S9). The Hh, Wnt/β-catenin, and Notch signaling pathways are known to significantly impact tumor cell stemness characteristics38–40. The current study adds to this knowledge by demonstrating that BEX2 inhibits Hh signaling by facilitating MCL1 degradation. This finding suggests a mechanism in which BEX2 negatively regulates Hh signaling to attenuate CRC stemness.
Furthermore, alterations in the CD133+CD44+ and CD44+CD133− cell subsets, were noted, suggesting a complex interplay between MCL1 and CRC CSC markers. Inhibition of MCL1 may selectively affect CD133+CD44+ CSCs, promoting quiescence or differentiation while enhancing the proliferation of the CD44+CD133− subpopulation. This finding highlights the multifaceted role of MCL1 in regulating CRC stemness and the need for further research to fully elucidate its therapeutic implications.
The suppression of BEX2 in specific CRC subtypes, particularly those with mesenchymal or stem-like phenotypes, likely reflects convergent regulatory mechanisms at multiple levels, including epigenetic silencing, transcriptional repression, and post-transcriptional control41. Emerging evidence indicates that DNA hypermethylation of promoter CpG islands is a hallmark of CRCs exhibiting the CpG island methylator phenotype (CIMP), leading to silencing of tumor-suppressive genes early in tumorigenesis. For example, promoter methylation of PRMT542 and PTGER443 has been proposed as diagnostic biomarkers in CIMP-positive CRCs, underscoring the role of DNA methylation in subtype-specific gene downregulation44. By analogy, hypermethylation of the BEX2 promoter could account for the diminished expression in stemness-prone CRC subgroups, thereby stabilizing MCL1 and promoting Hh pathway activation. In addition, transcriptional repression by oncogenic factors, such as FOXA1, which modulates chromatin accessibility in metastasis-associated gene networks45, may further curtail BEX2 expression in aggressive CRC subtypes. Finally, post-transcriptional regulation via miRNA networks could contribute to BEX2 mRNA degradation, as has been documented for analogous stemness-associated genes in CRC46,47.
Abnormal activation of the Hh signaling pathway in CSCs is a critical driver of CRC treatment resistance48,49. Targeting molecules involved in this pathway holds promise for reversing resistance in adenocarcinoma therapy4. In the initial study, BEX2 knockout enhanced CRC cell migration and metastasis via nuclear translocation of Zic2, which activated the Hh signaling pathway19. The findings herein indicated that BEX2, by inhibiting the Hh signaling pathway, reduces the stemness potential of CRC cells (Figure 8). In additon, targeting MCL1 could provide a novel approach to overcome drug resistance in CRC. The apparent paradox of BEX2 dual roles is reconciled by partitioned interactions, as follows: cytoplasmic sequestration of Zic2 to inhibit metastasis versus scaffolding MCL1 for degradation to suppress stemness, reflecting stage- and context-dependent substrate availability and signaling priorities. Consequently, therapeutic strategies aimed at restoring BEX2 function or mimicking its interactions could simultaneously block metastatic spread and eradicate cancer stem cells, offering a multifaceted approach to improve CRC patient outcomes. This effect is attributable not only to the MCL1’ interaction with the Hh signaling pathway but also to the critical role in altering the metabolic state of tumor cells and regulating stemness through mitochondrial respiration15.
In summary, the current study elucidates the regulatory network involving BEX2, MCL1, and the Hh signaling pathway in modulating CRC stemness characteristics. BEX2 was shown to regulate Hh signaling and MCL1, contributing to a reduction in CRC cell stemness potential. This regulatory mechanism appears to function independently of the Wnt and Notch signaling pathways. In addition, inhibiting MCL1 genetically or pharmacologically diminishes stemness potential and enhances chemotherapy efficacy. These findings highlight the potential of BEX2 and MCL1 as novel therapeutic targets, particularly for drug-resistant CRC cases characterized by MCL1 amplification. While the current findings underscore the therapeutic potential of targeting the BEX2-MCL1-Hh axis in CRC, several limitations warrant consideration. First, the mechanisms underlying BEX2 suppression in specific CRC subtypes have not been completely elucidated. Further validation is required to confirm whether BEX2 downregulation in stemness-prone subtypes arises from epigenetic modifications or transcriptional repression by oncogenic regulators. Second, clinical translation of MCL1 inhibitors faces challenges, including on-target toxicity in normal tissues and compensatory activation of alternative anti-apoptotic pathways. Third, while BEX2 exhibit stemness-suppressive properties, BEX2 overexpression promotes invasion and migration. Finally, the interplay between BEX2 and other stemness-related pathways, such as Wnt/β-catenin or Hippo signaling, remains unexplored, necessitating multi-omics approaches to delineate context-dependent crosstalk. Addressing these limitations will refine subtype-specific therapeutic strategies and enhance the clinical feasibility of targeting this axis.

Conclusions
Overall, the current study demonstrated that BEX2 functions as a key negative regulator of stemness characteristics in CRC cells. The findings revealed that BEX2 interacts with MCL1 and promotes ubiquitin-mediated degradation via the proteasomal pathway. This degradation of MCL1 subsequently leads to inhibition of the Hh signaling pathway, ultimately suppressing cancer stemness properties, such as tumor sphere formation, expression of stemness markers, chemoresistance, and invasive capabilities. Use of the Hh pathway inhibitor, GDC-0449, or knockdown of MCL1 effectively reversed the stemness-enhancing effects induced by BEX2 deficiency. These findings not only elucidate a novel molecular mechanism by which BEX2 modulates CRC stemness but also suggest that BEX2 could serve as a potential prognostic biomarker. Targeting the BEX2-MCL1 axis may offer a promising strategy for developing new therapeutic interventions against CRC.

Supporting Information