TRIM32-mediated ABI2 downregulation promotes anoikis resistance and metastasis by regulating EMT in lung adenocarcinoma

Abstract

Metastasis largely drives the high mortality of lung adenocarcinoma (LUAD), with anoikis resistance acting as a critical mediator in this process. This study aims to reveal the role and mechanism of ABI2 in anoikis resistance and LUAD metastasis. ABI2 expression levels and clinical value were analyzed using multiple databases. Functional studies were conducted in LUAD cells under attached or detached culture. The level of apoptosis was assessed by flow cytometry and Calcein-AM/EthD-1 staining. The metastasis ability was detected by transwell assay and lung metastasis model by tail vein injection. The ubiquitination effect of TRIM32 on ABI2 was analyzed by coimmunoprecipitation. This study revealed that ABI2 was significantly downregulated in LUAD tissues and anoikis-resistant LUAD cells. Functionally, ABI2 overexpression suppressed the anoikis resistance and LUAD metastasis, while ABI2 knockdown exerted the opposite effects. Mechanistically, ABI2 inhibited anoikis resistance by regulating epithelial-mesenchymal transition (EMT), thereby inhibiting tumor metastasis. EMT activation could rescue the effects of ABI2 on anoikis and metastasis. In addition, anoikis resistance modulated ABI2 ubiquitination by altering TRIM32 localization. Clinically, patients with low ABI2 and high TRIM32 experienced the worst prognosis in TCGA-LUAD. And a negative correlation between ABI2 and TRIM32 was further observed in both the public cohorts and the real-world cohort. In conclusion, ABI2 exerts tumor-suppressive effects in LUAD by inhibiting anoikis resistance and metastasis. The TRIM32‑ABI2‑EMT axis shows encouraging preclinical prospects, yet sufficient follow-up studies are indispensable to support its clinical therapeutic potential.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00018-026-06300-3.

Background

Lung cancer remains the leading contributor to cancer-related incidence and mortality worldwide. Lung adenocarcinoma (LUAD), its most common histological subtype, constitutes roughly 40% of all lung cancer cases [1]. Radical operation offers a potential cure for patients diagnosed in the early stages; however, the therapeutic landscape for those with unresectable or metastatic disease is severely constrained [2, 3]. Since metastasis is the primary driver of mortality in LUAD, elucidating the molecular mechanisms underlying LUAD metastasis is critical for developing targeted strategies to improve patient survival and prognosis.

Emerging evidence suggests that only primary cancer cells undergo a highly complex and multi-step process to erode into the surroundings and evade apoptosis in the circulation could form metastatic lesions in distant sites. A critical barrier to this process is anoikis, a form of detachment-induced apoptosis related to insufficient cell-extracellular matrix interplay [4, 5]. Numerous studies have demonstrated that anoikis resistance is a prerequisite for metastasis [6]. Tumor cells can resist anoikis through multiple pathways, such as enhanced PIK3-AKT and NF-κB signaling [7, 8], increased intracellular oxidative stress levels [9], and activation of epithelial-mesenchymal transition (EMT) [10, 11], ultimately leading to anoikis resistance and promoting metastasis. Studies have revealed that anoikis resistance also exists in lung cancer, but the specific molecular mechanisms remain to be elucidated.

Abelson interactors (ABIs), including ABI1 and ABI2, act as the substrate of the eponymous Abelson non-receptor tyrosine kinase (c‐ABL) [12, 13]. Previous studies have proved that ABI1 is mainly distributed in the central nervous system and plays an important role in synaptic maturation and actin cytoskeleton remodeling [14, 15]. ABI2 is almost ubiquitously expressed and has been shown to function in a variety of cancers. Li et al. proved that ABI2-SLC17A9 axis maintained the cancer stem cells population in liver cancer [16]. Lv et al. also found that ABI2 could regulate breast cancer cell via PI3K/AKT and EMT signaling [17]. CRISPR screening uncovered ABI2 as a key determinant of multiple drug resistance in pancreatic cancer [18]. However, few studies have focused on the roles of ABI2 in LUAD during anchorage-independent anoikis.

The aim of this study was to determine the role of ABI2 in anoikis and LUAD metastasis, as well as to investigate the potential molecular mechanisms underlying the upstream and downstream regulation of ABI2. The results highlight the novel role of ABI2 in anoikis resistance, which may provide clues for developing therapeutic strategies against advanced LUAD after further systematic validation.

Materials and methods

Cell culture

Normal lung epithelial cells (BEAS-2B) and lung cancer cells (A549, H1975, PC9, H358, H1299) were purchased from the cell bank of the Chinese Academy of Science. All cell lines were maintained in RPMI 1640 or DMEM medium (Gibco, USA) containing 10% FBS (Gibco, USA) and 1% penicillin/streptomycin (Yeasen, China) at 37 °C in 5% CO₂ atmosphere. To establish anoikis-resistant cell lines, cells were seeded in ultra-low attachment culture plates to prevent cell adhesion. After incubation for 48 h, suspended viable cells were collected, reseeded in routine adherent culture flasks for recovery, and then re-exposed to anchorage-independent conditions. This suspension-adhesion cycle was repeated for 3 rounds to obtain stable anoikis-resistant sublines [19, 20]. In order to induce epithelial-mesenchymal transition, cells were incubated with 0.2 ng/mL TGF-β1 (MCE, China) for 24 h.

Specimen collection

A total of 16 LUAD tumor specimens from patients who underwent surgical resection at the Xiangya Hospital were obtained and then stored at -80℃. Randomization and blinding are not applicable in this study. The studies were approved by the Ethics Committees of Xiangya Hospital (ID: 202402128). All participants have given informed consent. The clinicopathological characteristics are summarized in Table S1.

RNA interference and plasmid transfection

The small interfering RNAs (siRNAs) and plasmids overexpressing ABI2 were obtained from GeneChem (China). Lipofectamine 8000 (Beyotime, China) was used as the transfection reagent. Specific procedures: Cells at the logarithmic growth stage were inoculated into a 6-well plate to make the cell density 60–70%. 2.5 µg of DNA or 100 pmol siRNA were diluted in Opti-MEM medium (Gibco, USA) and gently mixed. Subsequently, 4 µL Lipo8000 reagent was added and gently mixed. After 20 min of incubation at room temperature, the resulting complexes were added to the cells and incubated at 37 °C for 2 days. The efficiency of siRNA and plasmid was confirmed at protein levels. A non-targeting vector was used as a control. The siRNA sequences: siABI2 (GGAUGGAAUCUUCAAUCAAUC), siTRIM32 (GCACUAUUAUGUAAAGAUAAU).

Invasion and migration assay

The invasion and migration ability was assessed using the transwell chamber. 5 × 10⁴ cells (200 µL per well) were seeded into the upper chamber, which was either uncoated for migration or pre-coated with matrigel (Corning, USA) for invasion. The upper chamber was filled with serum-free medium, while the lower chamber contained 600 µL of medium with 10% FBS. Following 48 h of incubation at 37 °C, cells on the upper membrane were wiped away with cotton swab. Cells on the lower membrane were fixed in 4% paraformaldehyde (Servicebio, China), stained with 0.1% crystal violet (Servicebio, China), and observed under a microscope. The invasion and migration cells were counted using ImageJ.

RNA extraction and quantitative real-time PCR

Total RNA from tumor tissues of LUAD patients was extracted with Trizol reagent (Invitrogen, USA). RNA concentration was quantified via NanoDrop™ One, and 1.5 µg RNA was reverse-transcribed into cDNA using the Evo M-MLV RT Kit (AG, China). Quantitative real‑time PCR (qPCR) was performed on a QuantStudio™ Real-Time PCR System using the SYBR Green PCR Master Mix (AG, China). The qPCR thermal cycling conditions were set as follows: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and 60 °C for 30 s. Relative mRNA expression was analyzed using the 2^−ΔΔCt method with GAPDH as an internal reference. Accession numbers and primer sequences of the corresponding genes are listed in Table S2.

Western blot and immunoprecipitation

For Western blotting (WB), denatured proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using 10% resolving gels, then transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, USA) via wet transfer (300 mA, 90 min). After sequential incubation with primary and secondary antibodies, target proteins were visualized using chemiluminescence substrate. For immunoprecipitation (IP), protein lysates were pre-incubated with target antibodies, then mixed with protein G agarose beads (CST, USA) and rotated at 4 °C for 6 h. Immunoprecipitated proteins were eluted, analyzed by WB, and visualized using an AMERSHAM ImageQuant 800 system. The grayscale values were quantitatively analyzed using ImageJ. Antibody details are listed in Table S3.

Immunohistochemical and immunofluorescence

For IHC, tissues were fixed with 4% paraformaldehyde, dehydrated, paraffin-embedded, and prepared into tissue chips. After dewaxing and rehydration, antigen retrieval was performed with 10 mM sodium citrate buffer (95 °C, 15 min), followed by 3% H₂O₂ blocking (30 min, room temperature) and 5% BSA (Beyotime, China) blocking (30 min, room temperature). The tissue chips were incubated with diluted antibodies overnight at 4 °C, then with biotin-labeled secondary IgG. DAB (Solarbio, China) and hematoxylin were then used for staining.

For IF, cells cultured on coverslips were fixed with 4% paraformaldehyde for 15 min, permeabilized in 0.2% Triton X-100 (Solarbio, China) for 10 min, blocked with 5% BSA for 1 h at room temperature, and then incubated with primary antibodies overnight at 4 °C. On the second day, the cells were incubated with fluorescent secondary antibodies for 1 h at room temperature (protected from light). Nuclei were counterstained with DAPI (MCE, China) for 5 min. The mean fluorescence intensity was quantitatively analyzed using ImageJ. Antibody details are listed in Table S3.

Apoptosis analysis

Annexin V-FITC Apoptosis Detection Kit was purchased from Beyotime (China). Cells were cultured in standard adherent dishes or ultra-low attachment dishes for 24 h. All cells, including the supernatant, were harvested and resuspended in 100 µL binding buffer. Amounts of 2.5 µL Annexin V-FITC and 5 µL PI were added to the cell suspension and incubated at room temperature for 15–20 min in the dark. Flow cytometry was used to detect apoptotic cells, with Annexin V-FITC⁺/PI⁻ (early apoptosis) and Annexin V-FITC⁺/PI⁺ (late apoptosis/necrosis) populations analyzed via FlowJo.

Animal model

A tail vein metastasis model was used to study anoikis in vivo. BALB/c nude mice were purchased from Hunan SJA Laboratory Animal Co., Ltd. (China). BALB/c nude mice were housed in a specific pathogen-free (SPF) environment under controlled conditions: ambient temperature of 22 ± 2 °C, relative humidity of 50 ± 5%, and a 12 h light/12 h dark cycle. All mice were given free access to standard sterilized chow and autoclaved water. All animal experiments were approved by the Animal Ethics Committee of Xiangya Hospital, Central South University (Approval No. XY20251213003). Mice were randomly divided into control group (n = 5) and ABI2 overexpression group (n = 5). 1 × 10⁶ cells with GFP tag were injected into the tail vein of 4-week-old nude mice. Radiance (Photons), acquired by LivingImage 4.4 software, was used to semi-quantitatively reflect lung metastatic potential. The mice were euthanized on day 28, and the lungs were harvested and fixed in 4% paraformaldehyde. Tumor nodules on the lung surface were counted under a dissection microscope.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 7.0. Each experiment was performed independently three times and the quantitative data were shown as mean ± SD unless specified otherwise. Two-group comparisons were analyzed by Student’s t-test, while multiple group comparisons used one-way ANOVA with Tukey’s test. Categorical variables were evaluated using the χ² test or Fisher’s exact test where appropriate. Overall survival (OS) was estimated using the Kaplan–Meier method and compared by Cox regression. Correlations between variables were assessed with Spearman’s correlation coefficient. P value of less than 0.05 was deemed statistical significance.

Results

Down-regulation of ABI2 correlates with adverse clinical outcomes in LUAD

Pan-cancer analysis suggested that ABI2 was upregulated in multiple cancers, mainly including colon adenocarcinoma (COAD), esophageal carcinoma (ESCA), head and neck squamous cell carcinoma (HNSC), liver hepatocellular carcinoma (LIHC) and stomach adenocarcinoma (STAD). However, downregulation was observed in bladder urothelial carcinoma (BLCA), breast invasive carcinoma (BRCA), glioblastoma multiforme (GBM), kidney renal clear cell carcinoma (KIRC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD) and uterine corpus endometrial carcinoma (UCEC). In additon, ABI2 was correlated with prognosis in KIRC, LIHC and LUAD (Fig. 1A-B).

Fig. 1.

Analysis of clinical relevance of ABI2 in LUAD. (A) The expression levels of ABI2 in pan-cancer. (B) Prognostic heatmap of ABI2 in pan-cancer. (C) The mRNA expression of ABI2 in the paired LUAD tissues of TCGA (n = 58). (D) The protein expression of ABI2 in CPTAC database (n = 111). (E) Typical images of normal lungs and LUAD tissues stained with ABI2 by IHC. Scale bar, 50 μm. (F) ABI2 had great diagnostic potential in TCGA-LUAD (AUC = 0.783). (G) Kaplan–Meier curves of OS of LUAD patients represented in the TCGA database. (H-I) Kaplan–Meier curves of overall survival (OS) of LUAD patients with advanced disease (Pathologic stage III&IV) or lymph node metastasis. (J) Multivariate Cox regression analyses of ABI2 for OS. OS, overall survival. Data are presented as the means ± SD of three independent experiments; ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

To further evaluate the role of ABI2 in LUAD, we analyzed three publicly available databases. The TCGA database indicated that ABI2 mRNA expression was significantly lower in LUAD tissues than that in paired normal lung tissues (Fig. 1C). The CPTAC and HPA databases reflected the downregulation of ABI2 protein levels in LUAD, which is consistent with mRNA expression (Fig. 1D-E). The ROC curve showed that ABI2 had great diagnostic potential in LUAD (AUC = 0.783) (Fig. 1F). Kaplan-Meier analyses revealed that patients with lower ABI2 levels exhibited worse overall survival (OS) compared to those with higher levels (Fig. 1G). Further, ABI2 was associated with prognosis in patients with advanced disease (Fig. 1H) or lymph node metastasis (Fig. 1I). The Cox regression analyse established that the down-regulation of ABI2 served as an independent prognostic risk factor for OS in LUAD (Fig. 1J).

ABI2 is involved in the anoikis process of LUAD cells

The expressions of ABI2 were assessed across several lung cancer cell lines. The results suggested that ABI2 expression exhibited a downregulated trend in lung cancer cells compared with normal lung epithelial cells (BEAS-2B) (Fig. 2A). The A549 and H358 cell lines were subsequently selected for ABI2 knockdown and overexpression experiments (Fig. 2B-C). In addition, the gene set enrichment analysis (GSEA) indicates that ABI2 is involved in multiple biological processes (Figure S1). This study focuses on anoikis (Fig. 2D). To investigate the role of ABI2 in the process of anoikis, we cultured A549 and H358 cells in suspension and assessed the expression of ABI2. The western blotting results showed that the expression of ABI2 was elevated in a time-dependent manner upon detachment (Fig. 2E). Next, we plated both control cells and ABI2-knockdown/overexpressed cells in ultra-low attachment plates for 48 h and subsequently analyzed apoptosis using flow cytometry (Figure S2). Our data showed that ABI2 had no obvious effect on apoptosis under normal growth conditions. However, the apoptosis level was decreased in ABI2-knockdown cells compared to control cells under detached conditions. Conversely, the overexpression of ABI2 significantly induced apoptosis under detached conditions (Fig. 2F-G). CCK8 assay also suggested that ABI2 inhibited cell survival under detached conditions (Fig. 2H). These results collectively support the conclusion that ABI2 is involved in anoikis in LUAD.

Fig. 2.

ABI2 is involved in the anoikis process of LUAD cells. (A) The expressions of ABI2 were assessed across several lung cancer cell lines. (B-C) The A549 and H358 cell lines were selected for ABI2 knockdown and overexpression experiments. (D) Gene set enrichment analysis suggests that ABI2 may be involved in anoikis. (E) The level of ABI2 in A549 and H358 cells with detached culture for the indicated time. (F-G) Apoptosis levels of control and ABI2 knockdown/ overexpressed cells cultured in normal dishes (Attached) or ultra-low attachment dishes (Detached) for 48 h. Cell apoptosis was monitored by Annexin V-FITC/PI staining. (H) The survival rate of control and ABI2 knockdown/ overexpressed cells with detached culture. Data are presented as the means ± SD of three independent experiments; ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

ABI2 inhibits LUAD metastasis by regulating anoikis

Anoikis-resistant (AR) cell models were established to explore the effect of ABI2 on metastatic behaviors of LUAD cells. The results suggested that ABI2 was downregulated in anoikis-resistant (AR) A549 and H358 cells as compared to the parental normal cells (Fig. 3A-B). Next, anoikis was quantified by Calcein-AM/EthD-1 staining (Fig. 3C). The live cells are detected with Calcein AM, while cell death is detected with Ethidium Homodimer (EthD-1). We found that the red fluorescence of AR cells was notably decreased as compared to the normal cells, indicating the less apoptosis in AR cell lines. In line with this observation, the AR cells acquired more invasion and migration ability (Fig. 3D-E). Next, we simultaneously investigated the role of ABI2 via downregulating ABI2 in A549 AR cells, and upregulating ABI2 in H358 AR cells (Fig. 3A). We found that knockdown of ABI2 inhibited anoikis propelled cell death and led to increase of invasion and migration ability in AR cells (Fig. 3C-E). The opposite results were obtained by overexpressing ABI2.

Fig. 3.

ABI2 inhibits LUAD metastasis by regulating anoikis. (A-B) The expression of ABI2 in anoikis-resistant cell (AR) and its parental cell (NC). The transfection of siABI2 could significantly knockdown its expression in A549-AR cells. The transfection of ABI2 plasmid could significantly overexpress its expression in H358-AR cells. (C) The anoikis was quantified by Calcein-AM/EthD-1 staining. The live cells are detected with Calcein AM, while cell death is detected with Ethidium Homodimer (EthD-1). Scale bar, 100 μm (D-E) The effect of AR and ABI2 knockdown/overexpression on invasion and migration was detected by transwell assay. Scale bar, 100 μm. (F) Fluorescence imaging and statistical analysis of lung metastasis ability. (G) The representative image of the lung metastatic lesion on the surface of the lungs. Data are presented as the means ± SD of three independent experiments; ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

To further validate the role of ABI2 in anoikis in vivo, the tail vein metastasis models were established with BALB/c-nude mice. The result revealed a significant reduction in metastatic outgrowth in the lungs of mice injected with ABI2-overexpressing cells (Fig. 3F). The lung metastases on the lung surface were also decreased in the ABI2-overexpressing group (Fig. 3G). Collectively, these findings suggest that ABI2 inhibits LUAD metastasis by regulating anoikis.

ABI2 affects tumor metastasis by regulating the hybrid E/M phenotype of anoikis resistant cells

It is generally accepted that the anoikis and epithelial-mesenchymal transition (EMT) are initial and critical steps during the tumor distant metastasis process. We observed that anoikis-resistant (AR) cells co-express mesenchymal markers (TWIST1, SNAIL1, Vimentin) and epithelial markers (cytokeratin 8 and 14: KRT8, KRT14), suggesting that AR cells exhibit a hybrid E/M phenotype. Furthermore, AR cells exhibited reduced adhesion (lower E-Cadherin, EpCAM and Claudin-1), accompanied by enhanced metastasis ability (Fig. 4A-B). In addition, knockdown of ABI2 resulted in increased Vimentin and decreased E-cadherin in AR cells. However, overexpression of ABI2 inhibited the transition of AR cells to mesenchymal phenotype (Fig. 4C-D). Immunofluorescence staining also obtained consistent conclusions (Fig. 4E-F). To further demonstrate that ABI2 affects anoikis by regulating EMT, H358 AR cells were treated with TGF-β, the most important and common inducer of EMT. Flow cytometry analysis suggested that ABI2 promoted anoikis and could be rescued by EMT activation (Fig. 4G-H). Similarly, the attenuated migration (Fig. 4I-J) and invasion (Fig. 4K-L) ability caused by ABI2 overexpression could be rescued by EMT activation in AR cells.

Fig. 4.

ABI2 affects tumor metastasis by regulating the hybrid E/M phenotype of anoikis resistant cells. (A-B) Western blot analysis of epithelial and mesenchymal markers in NC and AR cells. NC, normal control; AR, anoikis resistant. (C-D) The expression levels of Vimentin and E-Cadherin in AR cells with ABI2 knockdown/overexpression. (E-F) The expression levels of Vimentin and E-Cadherin in AR cells were certified by immunofluorescent staining. Scale bar, 10 μm. (G-H) Apoptosis levels of control and ABI2 overexpressed cells cultured with/without TGF-β (EMT activitor) in H358-AR cells. Cell apoptosis was monitored by Annexin V-FITC/PI staining. (I-J) The effect of ABI2 overexpression combined with/without TGF-β treatment on migration ability in H358-AR cells. Scale bar, 100 μm. (K-L) The effect of ABI2 overexpression combined with/without TGF-β treatment on invasion ability in H358-AR cells. Scale bar, 100 μm. Data are presented as the means ± SD of three independent experiments; ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

Anoikis resistance modulated ABI2 ubiquitination by altering TRIM32 localization

More investigations are necessary to explore the mechanism of ABI2 downregulation in anoikis-resistant (AR) cells. We found that MG132 (a proteasomal inhibitor) inhibited the downregulation of ABI2 in AR cells; however, CHX (an inhibitor of protein synthesis) had no significant effect on ABI2 (Fig. 5A). The results indicate that ABI2 downregulation may result from increased degradation. Previous study has reported the interaction between ABI2 and TRIM32 (E3 ubiquitin-protein ligase tripartite motif-containing protein 32) in HEK293T cells and head and neck cancer cells [21]. Our endogenous CO-IP experiments confirmed this interaction in A549 lung cancer cells (Fig. 5B). To investigate the ubiquitination in ABI2, both normal A549 cells and A549-AR cells were transfected with HA-ABI2. Figure 5C suggested that the ubiquitination level of ABI2 was increased in AR cells. Therefore, we investigated whether AR regulates ABI2 expression through TRIM32. Compared with normal cells, AR did not result in significant changes in TRIM32 total protein levels (Fig. 5D). Subsequently, TRIM32 knockdown was performed in both normal and AR cells (Fig. 5E). The results suggested that TRIM32 knockdown caused the upregulation of ABI2 in AR cells, but not in NC cells (Fig. 5F). It is well known that protein localization affects its functional execution. We further isolated cytosolic and nuclear proteins. Results showed that AR promoted TRIM32 translocation to the cytoplasm (Fig. 5G). This finding was also confirmed by immunofluorescence (Fig. 5H). In addition, TRIM32 knockdown reduced the ubiquitination level of ABI2 (Fig. 5I). AR enhanced the interaction between TRIM32 and ABI2, resulting in increased ubiquitination of ABI2 (Fig. 5J).

Fig. 5.

TRIM32 mediates ubiquitination of ABI2 in anoikis-resistant cells. (A) MG132 treatment but not CHX inhibited the downregulation of ABI2 in AR cells. AR, anoikis-resistant. (B) A549 cells were immunoprecipitated with ABI2 antibody or normal IgG, then subjected to immunoblotting with ABI2 or TRIM32 antibody. (C) The ubiquitination of ABI2 is increased in A549-AR cells. (D) Anoikis resistance did not affect the total protein level of TRIM32. (E) The efficiency of TRIM32 knockdown was verified in NC and AR cells. NC, normal control; AR, anoikis-resistant. (F) TRIM32 knockdown increased ABI2 protein level in AR cells, but not in NC cells. (G) Western blot analysis of TRIM32 protein in the nucleus and cytoplasm. Loading control: LaminB1 for nucleus, GAPDH for cytoplasm. (H) The immunofluorescence showed that anoikis resistance promoted TRIM32 translocation to the cytoplasm. Scale bar, 100 μm. (I) TRIM32 knockdown reduced the ubiquitination of ABI2. (J) AR enhanced the interaction between TRIM32 and ABI2, resulting in increased ubiquitination of ABI2. ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

ABI2 correlates negatively with TRIM32 in multiple LUAD cohorts

To confirm the clinical correlations between ABI2 and TRIM32, we examined the expression of TRIM32 in LUAD samples derived from the TCGA and CPTAC database.

The results showed that TRIM32 is up-regulated in LUAD (Fig. 6A-B). The correlation analysis indicated that ABI2 was negatively correlated with TRIM32 in TCGA (R = -0.39) and CPTAC (R = -0.31) database (Fig. 6C-D). Furthermore, the survival analysis showed that the combined expression of low ABI2 and high TRIM32 was associated with poor prognosis (Fig. 6E). To further demonstrate the above observation, we performed the qPCR and immunohistochemistry to detect the expression of ABI2 and TRIM32 in LUAD tissues from Xiangya hospital. The results illustrated that the expression of TRIM32 was higher in the ABI2 low group (Fig. 6F), and the expression of ABI2 and TRIM32 was negatively correlated in Xiangya LUAD (Fig. 6G-H).

Fig. 6.

ABI2 is negatively correlated with TRIM32 in multiple LUAD cohorts. (A-B) The expression of TRIM32 in LUAD patients based on TCGA and CPTAC database. (C-D) The correlation between ABI2 and TRIM32 in LUAD patients based on TCGA and CPTAC database. (E) Relationship between combined ABI2 and TRIM32 expression and prognosis in TCGA-LUAD patients using KM plotter. (F) The mRNA expression of TRIM32 in ABI2-high (n = 8) and ABI2-low (n = 8) groups from Xiangya-LUAD cohort. (G) The correlation between ABI2 and TRIM32 in Xiangya-LUAD cohort. (H) Immunohistochemical representative plots of ABI2 and TRIM32 in Xiangya-LUAD cohort. Scale bar, 50 μm. (I) Diagrammatic representation of the TRIM32/ABI2/EMT axis in anoikis resistance and metastasis. ***P < 0.001, **P < 0.01, *P < 0.05, ns-not significant

Discussion

Tumor metastasis involves detachment from the primary tumor, invasion into surrounding tissues, migration through the circulatory or lymphatic system, and ultimately adhesion and colonization at distant organs [22]. Anoikis is a type of apoptosis triggered by cell detachment from their native environment. It is an extrinsic-to-intrinsic process initiated by the disruption of cytoplasmic junctions, such asextracellular matrix (ECM) and cadherin adhesions. The acquisition of an anoikis-resistant phenotype represents a critical initial step in tumor metastasis [23, 24]. In this study, LUAD cells with anoikis-resistance (AR) were efficiently established and the pivotal role of ABI2 in controlling the anoikis and metastasis in AR-LUAD cells was demonstrated for the first time. This provides novel evidence of the mechanism by which LUAD cells adapt to the external environment for anoikis resistance and metastasis promotion.

Anoikis resistance in tumor cells is regulated by numerous genes and signal transduction pathways. For instance, GDF15 enhances anoikis resistance of gastric cancer through protective autophagy [25]. Tian et al. found that fatty acid metabolic enzyme CPT1A promotes anoikis resistance through redox homeostasis in esophageal squamous cell carcinoma [26]. In addition, EWSR1/Ezrin axis is involved in ER stress-mediated anoikis in bladder cancer [27]. However, research on the mechanisms of anoikis in LUAD is relatively limited. A recent study based on multicenter cohorts demonstrated that the Anoikis and EMT-Related Optimal Model exhibits excellent prognostic predictive performance in LUAD [28]. Furthermore, the synergistic prediction of anoikis and EMT has also been explored in gastric cancer and liver cancer [29, 30].

Our study found that anoikis resistant cells exhibit a hybrid E/M phenotype, which is a plastic state characterized by co-expression of E and M markers and high metastatic properties [31, 32]. It is generally accepted that E-Cadherin and Vimentin are typical markers of epithelial cells and mesenchymal cells. Previous studies have shown that ABI2 regulates EMT process in breast cancer [33] and nasopharyngeal carcinoma [34]; however, the effect of ABI2 on EMT in LUAD cells, especially in anoikis resistant LUAD cells, remains unclear. In this report, ABI2 was shown to be involved in the metastatic ability of anoikis resistant cells by inhibiting EMT. More importantly, activation of EMT reversed the promoting effects of ABI2 on anoikis and metastasis. Nevertheless, the detailed mechanism of ABI2–EMT interaction has not been fully clarified in our current work. In addition, the tail vein injection metastasis model preliminarily verifies the inhibitory effect of ABI2 on LUAD metastasis. However, this model fails to fully recapitulate the process of spontaneous tumor dissemination, which constitutes an unavoidable limitation of our in vivo animal experiments.

The tripartite motif (TRIM) protein family play important roles in a variety of cancers mainly as versatile E3 ubiquitin ligases [35–37]. For example, TRIM69 suppresses the anoikis resistance of gastric cancer through ubiquitin‒proteasome-mediated degradation of PRKCD [38]. TRIM31 has been elucidated to promote anoikis resistance of hepatocarcinoma cells through regulation of p53-AMPK axis [39]. Although the exact molecular mechanism by which ABI2 regulates EMT remains unclear in this study, we demonstrated that ubiquitination of ABI2 by TRIM32 plays an important role in anoikis-resistant LUAD cells. Previous studies have confirmed that TRIM32 is upregulated in lung cancer tissues and regulates tumor malignant characteristics and chemotherapy resistance [40, 41]. In addition, the recent finding by Xu et al. suggests that TRIM32 promotes anoikis resistant by increasing IL-6 secretion [42]. This opposite effect of TRIM32 and ABI2 on anoikis supports the results of the present study and also suggests that TRIM32-ABI2 axis may be a potential target in LUAD. Future studies will focus on elucidating the molecular mechanism underlying TRIM32 regulation by anoikis signaling. Given the crucial role of the TRIM32-ABI2-EMT axis in LUAD, future challenges include extending these findings to other lung cancer subtypes or other tumor types.

Combining clinical data with in vitro and in vivo experiments is a necessary step to establish a causal relationship. Although preliminary correlation conclusions have been drawn through the of TCGA and CPTAC analysis, a critical and unavoidable limitation of the present study is the small clinical sample size (n = 16), which leads to low statistical power and greatly impairs the reliability of causal inference. Moreover, small samples are prone to the influence of selection bias or individual patient characteristics (such as age, gender, tumor stage, and treatment regimen). These confounding factors may distort the true association between TRIM32/ABI2 and clinical prognosis. The present study primarily relies on in vitro and in vivo experimental evidence to support causal inference, while the clinical data serve solely as preliminary observational validation to suggest potential clinical relevance.

Given the aforementioned limitations, more future perspectives need to be developed to enhance this study. First, a large-sample and multicenter independent cohorts with complete clinicopathological information is needed to systematically validate the prognostic and metastatic value of the TRIM32/ABI2 axis. Second, we will employ more physiologically relevant animal models, such as orthotopic xenograft models and spontaneous metastasis models, to faithfully recapitulate the full cascade of tumor metastasis. Last but not least, future studies will further elucidate the precise molecular mechanism of ABI2 in regulating EMT, and explore novel intervention strategies based on the TRIM32-ABI2-EMT axis, such as screening small-molecule compounds or polypeptides capable of downregulating TRIM32 or stabilizing ABI2 protein. Moreover, combinatorial regimens with conventional chemotherapy or immunotherapy will be explored to provide preclinical evidence for novel anti-metastatic therapies.

Conclusions

Our study presents an investigation into the novel role of ABI2 in anoikis resistance and metastasis of LUAD (Fig. 6I). In short, our findings suggest the TRIM32‑ABI2‑EMT axis bears promising preclinical potential; nevertheless, extensive further studies are needed to validate its feasibility as a therapeutic target for metastatic LUAD prior to clinical translation.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Yongting Liu designed and drafted the manuscript. Yongting Liu and Zhaohui Jiang performed experiments. Yulong Zhao collected and analyzed the data. Qiong Chen supervised the study and provided financial support. All authors read and approved the final manuscript.

Funding

This work was supported by the grants from the National Natural Science Foundation of China (No. 82370055), the National Key R&D Program of China (No. 2020YFC2008600), the National Multidisciplinary Cooprative Diagnosis and Treatment Capacity Building Project for Major Diseases (Lung Cancer, ID: z027002), the Fundamental Research Funds for the Central Universities of Central South University (No. 2025ZZTS0238, 2022ZZTS0840) and the Hunan Provincial Innovation Foundation for postgraduate (No. CX20250358).

Data availability

The data that support the findings of this study are included in the article or Supplementary material. Further access is available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committees of Xiangya Hospital and carried out in accordance with the principles of the Declaration of Helsinki. All participants have given informed consent.

Consent for publication

This article has been authorized by all authors and agreed to be published.

Conflict of interest

The authors declare that they have no competing interests.