SPIB suppresses protective autophagy via the IFIT2/PINK1/Parkin axis to promote anoikis in colorectal cancer.

Background
Colorectal cancer (CRC) ranks as the third most common malignant tumor worldwide and represents one of the leading causes of cancer-related mortality [1]. Despite advancements in treatment modalities including surgical resection, chemotherapy, and targeted therapies, the 5-year survival rate for advanced CRC patients remains unsatisfactory, primarily due to tumor metastasis and drug resistance [1–3]. Metastasis constitutes a critical determinant of CRC patient mortality, with cancer cell acquisition of anoikis resistance emerging as a pivotal step in this process [4, 5].
Anoikis, a specialized form of programmed cell death triggered by detachment from the extracellular matrix (ECM) or loss of cell-cell adhesion, serves as a physiological safeguard against ectopic cell survival [6]. However, CRC cells develop anoikis resistance (AR) through multiple mechanisms, enabling their survival in circulation and subsequent formation of distant metastases [5, 7]. For instance, STAT3 activation upregulates V-ATPase to promote anoikis resistance and tumor metastasis [8], while CPT1A-mediated fatty acid oxidation facilitates CRC cell metastasis by suppressing anoikis [9]. Consequently, investigating strategies to restore anoikis sensitivity in CRC cells has become a crucial research direction for metastasis inhibition.
Autophagy, a highly conserved cellular self-degradation process, provides energy and maintains cell survival under nutrient deprivation or stress conditions by degrading damaged organelles and proteins [10]. In cancer, autophagy exhibits dual roles: while suppressing tumorigenesis in early stages, it promotes cancer cell survival through protective autophagy activation [11, 12]. Numerous studies demonstrate that detached cancer cells upregulate key autophagy molecules to maintain energy supply and suppress apoptotic signals, thereby resisting anoikis. For example, PTK6 promotes CRC metastasis by activating autophagy via HNRNPH1 dissociation to inhibit apoptosis, while CircCEMIP enhances protective autophagy in prostate cancer cells to foster anoikis resistance [13, 14]. Additionally, CPX induces ROS-dependent cell death in CRC while simultaneously triggering protective autophagy, with autophagy inhibition further potentiating its anticancer effects [15]. These findings suggest that targeting protective autophagy may represent an effective strategy to restore anoikis sensitivity in CRC cells.
SPIB (Spi-b transcription factor), belonging to the ETS family, is a DNA-binding protein involved in gene expression regulation. Expressed in immune cells (e.g., B cells, plasmacytoid dendritic cells) and certain epithelial cells, SPIB primarily modulates cell proliferation, differentiation, and survival [16]. Recent studies reveal aberrant SPIB expression in various cancers, associating it with tumor progression, metastasis, and drug resistance. SPIB inhibits Claudin-2 transcription to suppress early dissemination of primary lung cancer cells and mediates apoptosis in diffuse large B-cell lymphoma via the PI3K/AKT pathway [17, 18]. Furthermore, SPIB promotes anoikis resistance in lung cancer cells by enhancing lysosomal processes [19]. In CRC, SPIB generally functions as a tumor suppressor. Zhao et al. demonstrated that SPIB activates NF-κB and JNK pathways through MAP4K1, while Xu et al. reported that tsRNA-GlyGCC promotes CRC progression and 5-FU resistance by modulating SPIB, both mechanisms inhibiting CRC advancement [20, 21]. However, SPIB’s mechanistic role in regulating anoikis resistance in CRC remains unclear.
This study identifies SPIB as downregulated in CRC tissues, where it suppresses CRC cell proliferation, invasion, and metastasis. Mechanistically, SPIB restores mitochondrial membrane potential and inhibits autophagy via the IFIT2/PINK1/Parkin axis, thereby reducing anoikis resistance in CRC cells. Consequently, SPIB sensitizes CRC cells to anoikis, functioning as a metastasis suppressor gene that may provide novel insights for CRC diagnosis and treatment.

Methods

Bioinformatics analysis
Expression data and clinical metadata were obtained from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). DEGs (differentially expressed genes) between tumor and normal tissues were identified using limma, followed by visualization with pheatmap (heatmap) and “ggplot2/ggrepel” (volcano plot). WGCNA was performed to identify co-expression modules. Lasso regression was used to develop diagnostic models. Anoikis-related genes were obtained from public databases (https://www.gsea-msigdb.org/gsea/msigdb/).

Clinical samples collection
A total of 46 paired colorectal cancer and adjacent normal tissue specimens were collected at the Third Affiliated Hospital of Chongqing Medical University between 2023 and 2024. The protocols used in this research were evaluated and approved by the Ethics Committee of the Third Affiliated Hospital of Chongqing Medical University. All methods of this study were carried out in accordance with the Declaration of Helsinki.

Cell culture
Human CRC cell lines (HCT116, SW480, LoVo, HCT8, HT29) and normal human colonic epithelial cells (NCM460) were cultured in high-glucose DMEM medium (supplemented with 10% FBS [fetal bovine serum; Gibco] and 1% penicillin/streptomycin [Thermo Fisher Scientific]) at 37 °C in a 5% CO₂ humidified atmosphere. All cell lines were authenticated by STR (short tandem repeat) profiling.
To induce anoikis-resistant conditions, parental cells were continuously cultured for 48 h in ultra-low attachment 6-well plates (Corning, NY, USA) to simulate detachment from the ECM, followed by subsequent experiments.

RNA extraction and qRT-PCR
Total RNA was extracted from tissues and cell lines using a column-based RNA extraction kit according to the manufacturer’s protocol. cDNA was synthesized using the RT Master Mix for qPCR kit (MCE, China). Quantitative real-time PCR (qPCR) was performed with 2X Universal SYBR Green Fast qPCR Mix (ABclonal, China), and all data were analyzed using the Bio-Rad CFX Maestro PCR system (Bio-Rad, USA). The primer sequences are listed as follows:
SPIB-F: 5’-CACATCTAGGGCTCCTCCA-3’.
SPIB-R: 5’-TGACTCCCCATCCTCCAC-3’.
IFIT2-F: 5’-ATCTGCGGTATGGCAACT-3’.
IFIT2-R: 5’-GGTGGATGGCCTTGTCT- 3’.
β-actin-F: 5’-CTACCTCATGAAGATCCTGACC-3’.
β-actin-R: 5’-CACAGCTTCTCTTTGATGTCAC-3’.

Western blot
Cells were lysed using RIPA buffer (Beyotime, China) supplemented with PMSF (Beyotime) and phosphatase inhibitors (Beyotime). After complete lysis on ice for 30 min, the lysates were centrifuge at 12,000 rpm and 4 °C for 15 min. The supernatant was collected, and protein concentrations were determined using the BCA assay kit (Beyotime). Subsequently, 5X loading buffer was added at the appropriate ratio. Store at −20 °C for future use.
Total protein extracts were separated by SDS-PAGE gels of appropriate concentrations based on target protein molecular weights, followed by transfer onto PVDF membranes. The membranes were blocked with 5% skim milk in TBST for 1 h at room temperature, then incubated with primary antibodies at 4 °C overnight. Subsequently, the membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit or anti-mouse IgG secondary antibodies (Proteintech, China) for 1 h at room temperature. Protein bands were visualized using an ECL kit (Biosharp, China) and the Bio-Rad ChemiDoc imaging system (Bio-Rad, USA). Band intensity was quantified using ImageJ software. Primary antibodies used are listed as follows:
SPIB (ab309346 1:1000) Caspase3 (TA6311 1:1000) Cleaved-Caspase-3 (TA7022 1: 1000) E-Cadherin (20874-1-AP 1:1000) N-Cadherin (22018-1-AP 1:1000) Vimentin (10366-1-AP 1:1000) p62 (T55546 1:5000) LC3B (ab192890 1:2000) β-actin (20536-1-AP 1:5000) GAPDH (10494-1-AP 1:5000) COX Ⅳ(ET1701 −63 1:1000) IFIT2 (12604-1-AP 1:1000) PINK1 (A7131 1:1000) Parkin (A0968 1:1000).

Mitochondrial protein extraction by differential centrifugation
Cell samples were homogenized in pre-chilled HB homogenization buffer (0.25 M sucrose, 10 mM Hepes pH 7.4) containing protease and phosphatase inhibitors. The homogenate was subjected to initial low-speed centrifugation at 1,100 × g for 10 min at 4 °C. The supernatant was collected and centrifuged at 11,000 × g for 15 min at 4 °C to obtain the cytosolic protein fraction in the resulting supernatant. The pellet was resuspended in 1mL HB buffer and centrifuged again at 11,000 × g for 15 min at 4 °C. After discarding the supernatant, the mitochondrial pellet was gently resuspended in 100µL HB buffer to obtain the mitochondrial protein fraction. All procedures were performed at 4 °C. Protein concentration was determined by BCA assay, and extraction quality was verified by Western blot analysis of the mitochondrial marker protein COX IV.

Establishment of stable transfected cell lines
SPIB-targeting siRNA was designed and synthesized by Sangon Biotech (Shanghai). Transfection was performed using Lipofectamine® 2000 (Thermo Fisher Scientific). Sh-SPIB and sh-IFIT2 lentiviral vectors were constructed by Hanbio Biotechnology based on siRNA sequences. The SPIB overexpression plasmid was synthesized by Tsingke Biotechnology. For lentiviral packaging, 293 T cells were co-transfected with the SPIB plasmid, packaging plasmids, and envelope plasmid at optimized ratios using Lipofectamine 2000. Viral supernatants collected 48–72 h post-transfection and filtered through 0.45 μm membranes to remove cellular debris, yielding SPIB-overexpressing lentiviral particles. Lentiviral transduction was performed according to the manufacturer’s protocol. All siRNA sequences are listed as follows:
si-SPIB-1: sense (5’−3’): CCAGCUACCCUGAUUCAGATT.
si-SPIB-1: antisense (5’−3’): UCUGAAUCAGGGUAGCUGGTT.
si-SPIB-2: sense (5’−3’): UCUUCCAGUUCUCCUCCAATT.
si-SPIB-2: antisense (5’−3’): UUGGAGGAGAACUGGAAGATT.
si-IFIT2: sense (5’−3’): AAGGACUUGGGAAAUGUCAUUGATA.
si-IFIT2: antisense (5’−3’): UAUCAAUGACAUUUCCCAAGUCCUUGC.

CCK-8 assay
Cells were seeded in 96-well plates at a density of 5 × 10³ cells per well (triplicate wells for each condition). At the indicated time points (0,24,48, and 72 h),10µL of CCK-8 reagent (APEBIO) was added to each well. After 2 h of incubation at 37 °C, the absorbance at 450 nm was measured using a microplate reader (BioTek, USA).

EdU assay
Cells were plated in 96-well plates and incubated with 10µM EdU (Beyotime Cell Proliferation Detection Kit) for 2 h at 37 °C. Following incubation, cells were fixed with 4% paraformaldehyde (PFA) for 15 min and permeabilized with PBS of 0.3% Triton X-100 for 10 min at room temperature. The Click reaction was then performed by incubating Click Additive Solution for 30 min in the dark. Nuclei were counterstained with Hoechst 33,342 for 5 min. EdU-positive cells were imaged and documented using fluorescence microscopy (ZEISS, Ger).

Wound healing assay
CRC cells were cultured in 6-well plates until reaching 90%–100% confluence. To eliminate the potential confounding effect of cell proliferation on migration, cells were pretreated with mitomycin C (10 µg/mL) for 2 h prior to wounding. A uniform linear wound was then created in the cell monolayer using a sterile 10 µL pipette tip. After washing twice with PBS to remove detached cells and residual drug, cells were maintained in low-serum medium (2% FBS) to further suppress proliferation while permitting migration. And wound closure was photographed at 0 and 48 h. The migration distance was quantified using ImageJ software.

Transwell invasion assay
Transwell chambers (8 μm pore, Corning) were pre-coated with 50µL Matrigel (1:8 dilution) for 4 h at 37 °C. Serum-starved cells (2 × 10⁴) were seeded in the upper chamber with 600µL complete medium in the lower chamber. After 48 h, invaded cells were fixed, stained, and counted in three random fields (10× magnification).

Subcutaneous xenograft model
HCT116 cells (1 × 10⁷ cells/mL) in logarithmic growth phase were subcutaneously injected into 4-week-old BALB/c nude mice (100µL/mouse). Tumor dimensions were measured twice weekly (Volume = L×W²×0.5). Endpoint criteria: tumor diameter ≥ 2 cm or impaired mobility. Tumors were excised, photographed, weighed, and stored in liquid nitrogen. Animal experiments were approved by the Chongqing Medical University IACUC (IACUC-CQMU-2024–10152) following a comprehensive review.
The Chongqing Medical University IACUC stipulates that the maximum longitudinal diameter of subcutaneous tumors shall not exceed 2 cm. All tumor longitudinal diameters in this experiment were maintained within this prescribed limit.

Spleen-liver metastasis model
Under anesthesia, 1 × 10⁶ HCT116 cells in 50µL PBS were slowly injected into the spleen of 4–6-week-old BALB/c nude mice. After 2–4 weeks, liver tissues were harvested for metastatic nodule counting (gross examination) and histopathological analysis (H&E staining).

Anoikis assay
Cells (2 × 10⁵) were cultured in ultra-low attachment 6-well plates (Corning). After 48 h, floating cells were collected for Annexin V-FITC/PI staining and analyzed by flow cytometry (CytoFLEX) using Kaluza Analysis software.

Transmission electron microscopy (TEM)
Suspended cells were collected and enriched by low-speed centrifugation, followed by supernatant removal. The pellet was resuspended in room temperature EM fixative(2.5% glutaraldehyde), transferred to a 1.5mL conical tube with fixative, and immediately centrifuged at 12,000 g for 10–12 min. After careful supernatant aspiration, fresh EM fixative was slowly added along the tube wall to avoid disturbing the cell pellet. The sample was incubated for 30 min and stored at 4 °C for subsequent processing.
Collected cells were pelleted by centrifugation and pre-fixed with 2% paraformaldehyde − 2.5% glutaraldehyde, followed by post-fixation with 1% osmium tetroxide in buffer. Dehydration, infiltration, and embedding were performed sequentially as follows: 30% ethanol (10 min), 50% ethanol (10 min), 70% ethanol (15 min), 90% ethanol (15 min), absolute ethanol (15 min), acetone (15 min, twice), resin: acetone (1:3 mixture, 30 min), resin: acetone (1:1 mixture, 30 min), resin: acetone (3:1 mixture, 1 h), pure resin (overnight at 37 °C), fresh resin (8 h at 40 °C), and polymerization (48 h at 60 °C). Embedded blocks were sectioned at 0.8 μm for semi-thin slices, stained with toluidine blue, and examined by light microscopy to select regions with optimal cellular morphology. Ultrathin Sect. (90 nm) were mounted on 200-mesh formvar-coated copper grids. Conventional electron staining was performed with uranyl acetate (15 min) and lead citrate (5 min), followed by air drying. Samples were imaged using a transmission electron microscope (TEM, HITACHI, Japan).
Autophagic structures and autolysosomes were identified and imaged as previously described [22].

CUT&Tag sequencing
Per the manufacturer’s instructions, DNA libraries were prepared using NovoNGS® CUT&Tag 4.0 High-Sensitivity Kit (Novoprotein) and sequenced by ApexBio Co., Ltd.

RNA sequencing
Cells were prepared and washed with PBS. Subsequently, 1 mL of TRIzol reagent was added, and the cells were repeatedly pipetted until complete lysis was achieved. The criterion for complete lysis was that the lysate exhibited low viscosity, no visible cell clumps, and good fluidity. The entire TRIzol lysate was transferred to a conical-bottom centrifuge tube and rapidly frozen in liquid nitrogen. Sequencing analysis was performed by ApexBio Co., Ltd.

Dual-luciferase reporter assay
Potential SPIB binding sites in the IFIT2 promoter were predicted using JASPAR. The IFIT2 promoter region was cloned into pGL3-Basic vector (with mutant construct as negative control). Cells were co-transfected with reporters, effectors (pcDNA3.1-SPIB or empty vector), and normalization (pRL-TK) plasmids. Luciferase activity was measured using the Dual-Luciferase Reporter Assay Kit (Beyotime) 24–48 h.

JC-1 mitochondrial membrane potential assay
Cells at 60–70% confluence were stained with JC-1(Solarbio) for 20 min at 37 °C. Mitochondrial depolarization was assessed by fluorescence microscopy (Zeiss,200×༉。.

Statistical analysis
Data are presented as mean ± SEM. Comparisons were made using two-tailed t-tests or one-way ANOVA (GraphPad Prism 9.0)p < 0.05 was considered statistically significant.

Results

SPIB is downregulated in colorectal cancer and serves as a key regulator of anoikis
First, the read count expression matrix and corresponding clinical information were downloaded from the TCGA database. The R (limma) was used to identify DEGs between tumors and adjacent normal tissue groups (Table S1). The R (pheatmap, dplyr, ggplot2, and ggrepel) were employed to generate heatmaps and volcano plots, respectively (Fig. 1A). WGCNA was performed using the R WGCNA package. Gene co-expression matrices were constructed by calculating pairwise gene correlation coefficients through Pearson correlation tests. Subsequently, following the scale-free network principle, we selected an appropriate soft-thresholding power value to construct the scale-free network and transformed the adjacency matrix into a topological overlap matrix (TOM) (Fig. 1B). Cluster analysis was then conducted to identify gene modules. The module(green-yellow) with the smallest p-value was selected as the disease phenotype-related module, from which the gene set was extracted and defined as the WGCNA gene set (Fig. 1C, Table S2). Anoikis-related gene sets were obtained from public databases (www.gsea-msigdb.org/gsea/msigdb/), and a three-way intersection analysis was performed, yielding 15 overlapping genes (Fig. 1D, Table S3). Subsequently, LASSO (Least Absolute Shrinkage and Selection Operator) regression was employed to construct a binomial logistic regression model for further refinement, ultimately identifying four candidate genes: SPIB, CLDN1, TNFRSF12A, and VEGFA (Fig. 1E). Notably, SPIB’s function in colorectal cancer cell anoikis has not been previously reported, prompting us to focus on it for subsequent studies.

We investigated SPIB expressions in colorectal cancer tissues from the TCGA database, which revealed significantly lower SPIB levels in tumor tissues compared to normal controls (Fig. 1F). To further validate the differential expression of SPIB between CRC tissues and adjacent normal tissues, we collected 46 paired CRC and adjacent normal tissue samples. Quantitative real-time PCR (qPCR) results from all 46 paired samples showed that SPIB expression was significantly downregulated in CRC tissues compared with their matched normal counterparts (Fig. 1G). Combined analysis of SPIB mRNA expression and clinical data revealed that colorectal cancer patients in the SPIB-low expression group exhibited larger tumor size (p < 0.05), higher TNM stage (p < 0.05), and increased lymph node metastasis rate (p < 0.05) (Table S4). Additionally, we examined the protein expression levels of SPIB in 18 out of the 46 paired samples. The results demonstrated a significant decrease in SPIB protein levels in CRC tissues relative to the corresponding adjacent normal tissues (Figs. 1H, S1A, and S3A). Notably, compared with normal human colonic epithelial cells (NCM460), both the mRNA and protein levels of SPIB were significantly downregulated in colorectal cancer cell lines (HCT8 and HCT116) (Figs. 1G and I, and S1A).
In summary, our findings demonstrate that SPIB expression is downregulated in CRC patients and is associated with the TNM stage of their tumors. Consequently, we propose that SPIB influences the malignant behavior of CRC and may serve as a potential therapeutic target for CRC patients.

SPIB suppresses colorectal cancer progression by inhibiting cell proliferation, migration, invasion, and EMT process
To investigate the functional role of SPIB in CRC progression, we established stable SPIB-knockdown(sh-SPIB)and SPIB-overexpressing(OE-SPIB) cell lines in HCT8 and HCT116 cells using lentiviral transduction. The efficiency of SPIB knockdown and overexpression was confirmed by qPCR and Western blot (Figs. 2A, B and 3A, B, S1B). Subsequently, we assessed cellular proliferation capacity through CCK-8 and EdU assays, which demonstrated that SPIB overexpression significantly inhibited the proliferation of colorectal cancer cells HCT8 and HCT116 compared to control groups (Fig. 2C, D). Wound healing and Transwell assays revealed that SPIB overexpression significantly inhibited migration and invasion in HCT8 and HCT116 cells versus controls (Fig. 2E, F). In contrast, SPIB depletion significantly enhanced the proliferative, migratory, and invasive capacities of HCT8 and HCT116 cell lines (Fig. 3C-F).

Mechanistically, WB analysis of epithelial-mesenchymal transition (EMT) markers demonstrated that SPIB knockdown reduced E-cadherin expression while upregulating N-cadherin and vimentin, indicative of enhanced metastatic potential. In contrast, SPIB overexpression reversed this EMT phenotype, attenuating the cells’ metastatic capabilities (Figs. 3G, S1C). These findings collectively establish SPIB as a critical suppressor of CRC progression through its multifaceted inhibition of proliferation, motility, invasion, and EMT.

SPIB suppresses colorectal cancer proliferation and metastasis in vivo
To evaluate the tumor-suppressive function of SPIB in vivo, we established xenograft models by subcutaneously injecting BALB/c nude mice with HCT116 cells stably expressing SPIB-knockdown, SPIB-overexpressing cells, and their respective controls. Quantitative analysis demonstrated that SPIB knockdown significantly increased tumor volume and weight compared to control groups, whereas SPIB overexpression produced opposite effects (Fig. 4A, B). We further validated SPIB expression levels in a subset of subcutaneous tumors using immunohistochemistry (IHC) (Fig. 4C). Subsequently, we established a nude mouse splenic liver metastasis model to validate the impact of SPIB overexpression on the in vivo metastatic potential of HCT116 cells. Results demonstrated that the SPIB-overexpressing group exhibited fewer hepatic metastatic nodules compared to controls. Histopathological features of liver nodules were confirmed by H&E staining (Fig. 4D, E).

SPIB promotes anoikis and suppresses protective autophagy in colorectal cancer cells
To investigate SPIB’s role in anoikis regulation, we cultured SPIB-knockdown and SPIB-overexpressing CRC stable cell lines (HCT8 and HCT116) in ultra-low attachment plates to simulate detachment conditions. Flow cytometry analysis after 48 h revealed that SPIB knockdown significantly inhibited anoikis, while SPIB overexpression enhanced it (Fig. 5A). Western blot analysis of apoptotic markers showed decreased cleaved caspase-3 levels in SPIB-knockdown cells and increased levels in SPIB-overexpressing cells compared to controls (Figs. 5B, S1D). To study the role of SPIB in anoikis regulation of CRC cells, Using transmission electron microscopy, we observed that normal CRC cells (HCT8/HCT116) under suspension culture exhibited more autophagosomes than adherent controls (Fig. 5C). Corresponding Western blot analysis demonstrated: Reduced SPIB expression in suspended cells, decreased p62 and increased LC3-II/LC3-I ratio (indicating autophagy activation) (Figs. 5D, S1E). Notably, SPIB-knockdown cells showed similar autophagy activation, while SPIB-overexpressing cells displayed autophagy suppression (Figs. 5E, S1F). These findings demonstrate that SPIB promotes anoikis by inhibiting autophagy.

SPIB regulates downstream target genes through integrated CUT&TAG-Seq and RNA-Seq analysis
We performed CUT&Tag assays using a commercial kit (NovoNGS® CUT&Tag 4.0 High-Sensitivity Kit), followed by DNA library preparation and high-throughput sequencing. In CUT&Tag data analysis, peak calling refers to the computational identification of characteristic enrichment peaks across the genome, which represent potential binding sites of target proteins (transcription factors or histone modifications). These candidate peaks serve as genomic anchors for subsequent functional validation. The functional regions of the genome are categorized into promoters, downstream regions, coding exons, introns, and distal intergenic regions.
Genomic annotation of binding sites was performed by assigning each site to its nearest gene, thereby characterizing the genome-wide distribution patterns of these protein-DNA interactions. Subsequently, we performed genome-wide functional annotation of the sequencing results (Fig. 6A). Transcription factors ༈TFs༉ are typically enriched near transcription start sites (TSS), a region commonly referred (TFs) the TSS-proximal region. We calculated the average signal intensity of peaks within ± 3 kb windows centered on all gene TSS or transcription end sites (TES), and visualized the read density distribution using deepTools (Fig. 6B). This analysis employed MEME (Multiple EM for Motif Elicitation) to predict conserved binding motifs associated with the target protein, providing mechanistic insights into how transcription factors or protein modifications regulate gene expression (Fig. 6C).

Volcano sualized the distribution of DEGs between SPIB-knockdown and control groups. In the plot, the x-axis represents the log2-transformed fold change (log2FC) of gene expression between treatment and control groups, while the y-axis indicates the statistical significance of differential expression (-log10 adjusted p-value or -log10 p-value). Upregulated genes are denoted by red dots, and downregulated genes are marked with blue dots (Fig. 6D). A heatmap was generated using log2(TPM + 1) transformed values of DEGs. The plot displays TPM values of the top 50 DEGs with the smallest adjusted p-values (or p-values) from DESeq2 or edgeR analysis, where red and blue colors represent high and low TPM expression levels, respectively (Fig. 6E). To refine our screening, we defined candidate genes within ± 3 kb of TSSs from CUT&Tag data as the ‘CUT&Tag gene set’. Concurrently, we selected the top 50 genes from transcriptomic sequencing to generate the ‘transcriptome gene set’. Intersection of these sets yielded six candidate genes (IFIT2, E2F3, PDE4B, SNAP23 and SYTL2). Notably, IFIT2 has been previously reported to associate with mitochondrial membrane potential [23], whose alteration serves as a key trigger for mitophagy activation. Potential SPIB binding sites in IFIT2 were predicted using the JASPAR database (https://jaspar.elixir.no/) (Fig. 6). Dual-luciferase reporter assays demonstrated direct regulatory binding of SPIB to IFIT2 (Fig. 6H). We further performed Western blot analysis on the same 18 pairs of CRC tissues and their matched adjacent normal tissues. The results revealed that, consistent with the expression pattern of SPIB, the protein level of IFIT2 in colorectal cancer tissues was significantly lower than that in the corresponding adjacent normal tissues (Figs. 6I, S2A, S3A). Consistent with this, SPIB knockdown reduced IFIT2 levels versus controls, while SPIB overexpression elevated IFIT2 expression (Figs. 6J, S2B). We therefore identified IFIT2 as a downstream target gene of SPIB that participates in regulating anoikis in colorectal cancer cells.

IFIT2 mediates the tumor-suppressive effects of SPIB in colorectal cancer progression
This study aims to investigate whether the SPIB/IFIT2 axis coordinately regulates colorectal cancer cell proliferation, migration, and invasion. Western blot analysis validated the knockdown efficiency of IFIT2 in HCT8 and HCT116 cells transduced with sh-IFIT2 lentivirus (Figs. 7A, S2C). For functional rescue experiments, HCT8 and HCT116 cells were co-transduced with OE-SPIB, sh-IFIT2, corresponding control lentiviruse. Similarly, Western blot analysis was performed to detect IFIT2 expression levels in co-transduced cells (Figs. 7F, S2D).

Functional rescue assays demonstr, HCT8 IFIT2 knockdown significantly enhanced colorectal cancer cell proliferation, migration, and invasion compared to controls. Notably, in SPIB-overexpressing cells, IFIT2 depletion reversed SPIB-mediated suppression of these malignant phenotypes (Fig. 7B-E). Western blot analysis of epithelial-mesenchymal transition ༈EMT༉ markers corroborated these functional findings, showing that IFIT2 silencing reversed OE-SPIB-induced upregulation of E-cadherin downregulation of N-cadherin/vimentin (Figs. 7G, S2E). These results establish IFIT2 as an essential downstream effector through which SPIB exerts its tumor-suppressive functions in CRC progression.

SPIB promotes anoikis in colorectal cancer cells by suppressing protective autophagy via the IFIT2/PINK1/Parkin axis
Using suspension culture to mimic anoikis conditions, flow cytometric analysis of apoptosis revealed that IFIT2 knockdown significantly reduced the apoptotic rate compared to controls. Notably, in SPIB-overexpressing cells, IFIT2 knockdown downregulated the pro-anoikis effect of SPIB in colorectal cancer cells (Fig. 8A). Consistently, Western blot analysis of apoptosis-related proteins yielded concordant result (Figs. 8B, S2F). Previous studies have reported that IFIT2, as a mitochondria-associated gene, contributes to sepsis progression by inducing mitochondrial dysfunction through mechanisms including ΔΨm dissipation [23]. To investigate the role of IFIT2 in ΔΨm regulation in colorectal cancer cells, JC-1 staining revealed that IFIT2 knockdown significantly reduced ΔΨm compared to controls. Notably, in SPIB-overexpressing cells, IFIT2 depletion reversed the SPIB-mediated ΔΨm elevation (Fig. 8D).

Additional studies have demonstrated that ΔΨm depolarization activates mitophagy via the PINK1/Parkin pathway. For instance, TUBB4A suppresses glioma genesis by modulating the ROS-PINK1/Parkin-mitophagy axis [24]. Mechanistically, ΔΨm depolarization triggers PINK1 stabilization on the outer mitochondrial membrane (OMM), which relieves Parkin autoinhibition and promotes its translocation from the cytosol to mitochondria. We therefore isolated cytosolic and mitochondrial protein fractions by differential centrifugation to assess the subcellular localization of PINK1 and Parkin. The results demonstrated significantly higher PINK1 and Parkin expression levels in mitochondrial fractions of IFIT2-knockdown cells compared to controls. Consistently, IFIT2 depletion in SPIB-overexpressing cells reversed SPIB-mediated suppression of PINK1/Parkin protein accumulation in mitochondria (Figs. 8E, S2G). Western blot analysis of autophagy markers (LC3-II and p62) further demonstrated that IFIT2 knockdown activated autophagic flux, while concurrently reversing SPIB-mediated autophagy suppression (Figs. 8C, S2H). We conclude that in colorectal cancer, like the schematic diagrams, low SPIB expression activates the ΔΨm depolarization-dependent PINK1/Parkin pathway by downregulating IFIT2, thereby promoting mitophagy and ultimately enhancing anoikis resistance in colorectal cancer cells, which facilitates cancer metastasis (Fig. 9).

Discussion
Colorectal cancer (CRC) remains a leading cause of cancer-related mortality worldwide. Despite advances in surgical techniques, chemotherapy, targeted therapies, and immunotherapy, the prognosis for advanced and metastatic cases remains poor [1, 25, 26]. Metastasis in CRC involves complex mechanisms including signaling pathway activation, epigenetic alterations, metabolic reprogramming, and immune evasion [2, 27]. AR represents a critical capability for disseminated tumor cells to survive in circulation and establish metastases [5, 7, 28]. Therefore, identifying reliable biomarkers of anoikis resistance is crucial for molecular diagnosis and prognostic evaluation in CRC.
In this study, we identified SPIB as a novel anoikis-related gene through bioinformatics analysis. We demonstrated that SPIB, downregulated in CRC compared to adjacent normal tissues, functions as a tumor suppressor by inhibiting CRC cell proliferation, migration, and invasion. Importantly, we established for the first time that the transcription factor SPIB promotes anoikis by directly regulating IFIT2 expression to suppress protective autophagy. This finding aligns with existing evidence that tumor cells often activate autophagy upon ECM detachment to maintain energy homeostasis and evade anoikis, suggesting a potential therapeutic target.
While IFIT2, as an interferon-stimulated gene (ISG) family member, has been primarily studied for its antiviral and pro-apoptotic functions [29–31], our mechanistic investigation revealed its novel role in mitochondrial regulation. Building on reports identifying IFIT2 as a mitochondria-associated differentially expressed gene involved in ΔΨm depression during sepsis [23], and its depletion promoting cancer stem-like properties in oral cancer [32], we demonstrated that the SPIB/IFIT2 axis maintains ΔΨm to prevent PINK1 accumulation on mitochondrial outer membranes and subsequent Parkin recruitment, thereby inhibiting mitophagy and reducing anoikis resistance. This mechanism is consistent with recent findings that mitophagy activation through PINK1/Parkin signaling promotes tumor cell survival [33]. This study demonstrates that the SPIB/IFIT2 axis suppresses mitochondrial membrane potential depolarization in colorectal cancer cells, thereby inhibiting PINK1 accumulation on mitochondrial outer membranes and Parkin recruitment, which consequently blocks mitophagy initiation and reduces anoikis resistance.
An important consideration in interpreting these findings is the specific cellular context. The characterization of SPIB expression levels in our cellular models is based on consistent data validated internally through repeated qPCR and Western blot analyses. We observed discrepancies between our internal SPIB expression profiles and the RNA-Seq data from public databases such as CCLE (e.g., in HCT116 cells) (Fig. S3C). This suggests that genetic drift may occur in subclones of cell lines maintained in different laboratories, leading to variations in baseline gene expression profiles. Consequently, the molecular mechanisms elucidated in this study primarily apply to CRC models defined as ‘SPIB-low’ (such as the HCT116 and HCT8 cells used here). This finding not only underscores the importance of internally validating the expression of key molecules in functional studies but also highlights the complexities associated with cellular models in biological research.
While our study has established IFIT2’s regulatory role in maintaining ΔΨm, several mechanistic questions remain unresolved:1 the precise molecular mechanism through which IFIT2 modulates ΔΨm, 2 whether IFIT2 directly interacts with the PINK1/Parkin pathway or indirectly regulates mitophagy via metabolic intermediates. These unresolved aspects represent critical directions for future investigation to fully elucidate the SPIB/IFIT2-mediated regulatory network in colorectal cancer metastasis.

Conclusion
In summary, this study elucidates that the transcription factor SPIB suppresses protective autophagy and promotes anoikis in CRC cells by regulating IFIT2 expression to inhibit the PINK1/Parkin signaling pathway, thereby identifying a novel therapeutic target for preventing CRC metastasis. Although mechanistic challenges remain, these findings establish a theoretical foundation for developing precision therapies targeting the metabolic-immune-apoptotic regulatory network. Future research should focus on clinical validation using patient-derived samples and exploring combinatorial targeting strategies to facilitate the clinical translation of this approach.

Supplementary Information