Inhibition of the AT-Hook DNA-Binding Domain Attenuates HMGA2-Mediated Epithelial Mesenchyme Transition in Esophageal Cancer Cells

Abstract

Esophageal squamous cell carcinoma (ESCC) is a highly prevalent malignancy worldwide. Moreover, ESCC remains poorly characterized at the molecular level, which contributes to limited therapeutic options and an overall poor prognosis. In this context, HMGA family members, which are overexpressed in tumors but almost absent in healthy adult tissues, seem to represent promising therapeutic targets. These proteins act by binding to AT-hook DNA-binding motifs and may regulate the expression of several genes associated with tumor progression. Therefore, integrating in silico, translational, and in vitro approaches, we investigated the functional consequences of blocking HMGA2–DNA interaction in ESCC tumor progression by using netropsin, a site-specific ligand for AT-rich DNA regions. Our results demonstrate that netropsin treatment significantly reduced cell viability, migration, and cell cycle progression, thereby promoting apoptosis. Furthermore, netropsin treatment was capable of partially reverting Epithelial–Mesenchymal Transition (EMT) activation associated with HMGA2 expression, by downregulating EMT activators, such as Slug and Twist. Finally, the netropsin treatment sensitizes ESCC cells to chemotherapeutic treatment with 5-Fluorouracil. Taken together, our findings highlight that AT binding-specific blockade could be correlated with the inhibition of HMGA2 and may reveal a promising approach to better understand ESCC progression.

1. Introduction

Esophageal cancer (EC) is a highly aggressive malignancy characterized by elevated incidence and mortality, ranking as the 11th most common cancer and the 7th leading cause of cancer-related death worldwide [1]. It comprises two major histopathological subtypes, Esophageal Squamous Cell Carcinoma (ESCC) and Esophageal Adenocarcinoma (EAC), which differ in their geographic distribution, incidence, risk factors, and underlying pathogenesis [2]. ESCC accounts for approximately 90% of all esophageal cancer cases, with its principal risk factors including excessive alcohol consumption, tobacco smoking, the habitual intake of very hot beverages, and diets deficient in micronutrients [3,4]. Because ESCC is frequently diagnosed at advanced stages, it is associated with an unfavorable prognosis [5,6]. Despite recent advances in the understanding and management of this disease, further studies are required to elucidate the molecular mechanisms underlying esophageal carcinogenesis and to identify biomarkers for early detection, as well as novel therapeutic targets that may contribute to the development of more effective treatment strategies and ultimately improve patient outcomes. In this context, the High Mobility Group A (HMGA) family comprises three architectural chromatin-binding proteins—HMGA1, HMGA1b, and HMGA2—which are characterized by their ability to bind adenine- and thymine-rich DNA sequences (AT-rich regions) through their AT-hook domains within the minor groove of DNA [7]. Aberrant overexpression of these proteins has been implicated in the initiation and progression of multiple tumor types by altering chromatin architecture and indirectly regulating the transcription of numerous genes [8,9,10,11]. Specifically, in ESCC, HMGA2 overexpression has been reported in approximately 90% of tumor specimens, and its silencing markedly reduces cellular phenotypes associated with tumor progression [12], suggesting that functional inhibition of HMGA2 may represent a promising therapeutic strategy for this malignancy. In this context, netropsin is a naturally occurring cationic oligopeptide isolated from Streptomyces netropsis. Structurally, it consists of two N-methylpyrrole rings linked by amide bonds and exhibits high affinity for AT-rich DNA sequences within the minor groove, thereby preventing members of the HMGA protein family from binding to these regions [13], making it a promising antagonist of HMGA proteins. Notably, Miao and collaborators demonstrated that netropsin is capable of inhibiting HMGA2 activity with high specificity [14]. Given the established role of HMGA2 in the biology of several malignancies, including ESCC [12], the present study aimed to investigate the functional consequences of pharmacologically disrupting HMGA2 binding to DNA in an ESCC model.

2. Results

2.1. Netropsin Treatment Decreases the Malignant Progression of ESCC Cells

After determining the IC₅₀ values for the ESCC cell lines TE-1 (117.4 µM) and TE-13 (120.6 µM) (Figure 1A,B), functional assays were performed to investigate whether netropsin could antagonize malignant phenotypes classically associated with HMGA2 expression. Netropsin treatment significantly reduced both cell viability (Figure 1C) and migratory capacity (Figure 1D) in the two ESCC cell lines. Moreover, it increased the proportion of cells in the S and G2/M phases of the cell cycle (Figure 1F,G) in both cell lines and significantly elevated the percentage of Annexin V-positive cells (Figure 1H) compared with their respective control groups. Collectively, these findings indicate that netropsin impairs cell cycle progression and promotes apoptotic cell death in ESCC cells. These findings are consistent with the well-established effects of DNA-targeting agents, which activate the DNA damage response (DDR), induce S-phase and G2/M cell-cycle arrest, and, when DNA lesions cannot be efficiently repaired, trigger apoptotic cell death through sustained checkpoint activation, mitotic failure, and the inability to re-enter the G1 phase [15].

Figure 1.

Netropsin treatment impacts important cellular parameters associated with ESCC in vitro progression. (A,B) Dose–response curves evaluating cell viability (nor-malized absorbance) of TE-1 (A) and TE-13 (B) cell lines after treatment with different concentrations of netropsin (50, 100, 150, 200, and 250 µM) for 72 h to determine the IC₅₀. (C) Cell viability of ESCC cell lines TE-1 and TE-13 after treatment with netropsin determined IC₅₀ for 72 h and 96 h. (D,E) Migration assay (wound healing) in TE-1 (D) and TE-13 (E) cells at 0 h, 24 h, and 48 h after treatment with netropsin determined IC₅₀.Images are representative, and the quantification refers to the migration percentage over time. Scale bar = 300 µm for all images. (F,G) Cell cycle distribution analysis in TE-1 (F) and TE-13 (G) cell lines after 72 h of treatment with netropsin determined IC₅₀. (H,I) Assessment of apoptotic cell death (Annexin-V/7-AAD) in ESCC cell lines TE-1 (H) and TE-13 (I). Bar graphs quantify the ratio of Annexin-V-positive cells. All data are represented as mean ± standard deviation (SD) of three independent experiments. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns = non-significant.

2.2. HMGA2 Expression Is Associated with EMT Pathway Activation in ESCC

Since netropsin antagonized malignant phenotypes classically associated with HMGA2 expression, we next investigated, through in silico analyses, which HMGA2-regulated pathways and biological processes in ESCC could potentially be affected by this compound. To this end, we performed differential gene expression analysis followed by Gene Set Enrichment Analysis (GSEA) using the GEO dataset GSE53625, which identified the Epithelial–Mesenchymal Transition (EMT) pathway as one of the most significantly enriched biological programs (Figure 2A). To determine whether EMT pathway activation correlates with poor clinical outcomes, we used the MSigDB Hallmark EMT gene set to establish an EMT signature and applied single-sample gene set enrichment analysis (ssGSEA) to assess EMT-related gene expression in each sample. Based on these scores, survival analysis demonstrated that ESCC patients displaying high expression of EMT signature genes (Figure 2B) exhibited poorer overall survival. To further investigate the relationship between HMGA2 expression and the EMT pathway, a correlation analysis was performed using HMGA2 expression levels and the previously generated ssGSEA EMT scores. This analysis revealed an overall inverse correlation between HMGA2 expression and EMT pathway activity (Figure 2C). However, when EMT signature genes were analyzed individually, a distinct pattern emerged. Although HMGA2 expression was not globally associated with EMT signature genes, its overexpression was positively correlated with key EMT regulators, including CDH2, SNAI2, and VIM (Figure 2D). Since in silico analyses demonstrated a strong association between HMGA2 and the EMT pathway, the correlation between HMGA2 expression and major EMT transcriptional regulators was further evaluated in ESCC patient samples. A significant correlation was observed between HMGA2 and the genes encoding the transcription factors Snail and Twist (Figure 2E,F), but not Slug (Figure 2G).

Figure 2.

HMGA2 expression associates with the Epithelial–Mesenchymal Transition (EMT) pathway. (A) Gene Set Enrichment Analysis (GSEA) for the Hallmark EMT signature, comparing expression profiles of tumor samples and normal tissue. (B) Overall survival curve (Kaplan–Meier) of patients from the GSE53625 cohort, stratified by high (red) or low (blue) EMT pathway expression. (C) Scatter plot demonstrating the correlation between the single-sample enrichment score (ssGSEA) for the EMT pathway and HMGA2 expression levels (log2). (D) Overall ranking of the Spearman correlation (Rho) between HMGA2 expression and individual genes of the Hallmark EMT signature, highlighting structural and regulatory markers (CDH2, SNAI2 and VIM). (E–G) Validation of the correlation of relative mRNA expression (normalized by that of GAPDH) between HMGA2 and EMT-inducing transcription factors: Snail (E), Twist (F), and Slug (G) in tumor samples from patients with ESCC. Gene correlation analyses were evaluated with the respective coefficients (R or Rho) indicated in the panels. Differences in survival probability were calculated using the log-rank statistical test (p = 0.041).

2.3. Netropsin Treatment Partially Reverts the EMT Phenotype in ESCC Cells

Based on the association identified between HMGA2 expression and the EMT pathway in ESCC, we next investigated whether this molecular profile could be recapitulated in vitro through modulation of HMGA2 expression. Therefore, since netropsin binds with high affinity to the minor groove of AT-rich DNA sequences, which are the same regions preferentially recognized by the AT-hook domains of HMGA2 and sterically hinders HMGA2 from interacting with its target DNA sequences. As a result, HMGA2 loses its ability to modulate chromatin architecture and regulate the transcription of downstream target genes [14]. After confirming the efficiency of stable HMGA2 overexpression and knockdown (Figure 3A,B), we found that HMGA2 overexpression markedly increased the expression of SNAI2 and TWIST1, accompanied by a modest increase in SNAI1. Unexpectedly, HMGA2 overexpression also resulted in elevated CDH1 expression (Figure 3C). Conversely, HMGA2 silencing reversed the expression pattern of EMT-associated transcription factors while also markedly increasing CDH1 expression (Figure 3D). Collectively, these findings are consistent with our translational and in silico analyses and further support a close association between HMGA2 expression and EMT activation in ESCC. Next, we investigated whether netropsin treatment could modulate the expression of key transcription factors involved in EMT activation. Netropsin significantly reduced the expression of TWIST1 and SNAI2 while increasing SNAI1 expression in both ESCC cell lines (Figure 3E,F). Consistent with the elevated SNAI1 levels, CDH1 expression was reduced in TE-1 cells (Figure 3E). In contrast, TE-13 cells exhibited the opposite response, showing increased CDH1 expression following netropsin treatment (Figure 3F). Since netropsin modulated the expression of EMT-associated regulators, we next evaluated the mesenchymal marker vimentin [16] by immunofluorescence. In agreement with the gene expression findings, netropsin markedly reduced vimentin expression in both cell lines (Figure 4A,B). This effect was accompanied by reorganization and retraction of the actin cytoskeleton, particularly in TE-13 cells, as revealed by phalloidin staining (Figure 4A,B).

Figure 3.

HMGA2 modulation and netropsin treatment reshape the expression of EMT-associated transcription factors in ESCC cell lines. (A) qPCR validation of HMGA2 modulation in TE-1 cells stably overexpressing HMGA2 (OE-HMGA2) or expressing an HMGA2-targeting shRNA (shHMGA2), relative to their respective controls. (B) Representative Western blotting images confirming HMGA2 protein overexpression and knockdown in TE-1 cells; Lamin A/C was used as a loading control. (C,D) qPCR analysis of EMT-associated markers SNAI1, SNAI2, TWIST1, VIM and CDH1 in TE-1 cells upon HMGA2 overexpression (C) or silencing (D). (E,F) qPCR analysis of TWIST1, SNAIL2, SNAIL1 and CDH1 in TE-1 (E) and TE-13 (F) cells treated with IC₅₀ of netropsin for 72 h. * p < 0.05; ** p < 0.01; *** p < 0.001; ns = non-significant.

Figure 4.

Netropsin reduces vimentin expression and remodels the actin cytoskeleton in ESCC cells. (A) Immunofluorescence evaluating protein expression and distribution of vimentin (green) in TE-1 and TE-13 ESCC cell lines cultured in the absence (top) and presence (bottom) of netropsin IC₅₀ treatment for 72 h. Nuclei were contrasted with DAPI (blue), and the actin cytoskeleton was labeled with phalloidin (red). Images were acquired with a 20× objective using the EVOS M5000 Imaging System (Thermo Scientific). Scale bars: 150 µm. (B) Quantification of vimentin fluorescence intensity in TE-1 (left) or TE-13 (right) cells, expressed in arbitrary units (A.U.).

2.4. Netropsin Abrogates Malignant Behaviors Associated with EMT Pathway Activation

Subsequently, we investigated whether the molecular changes induced by netropsin were accompanied by functional alterations associated with EMT activation in ESCC cells. Given that netropsin negatively modulated the expression of key EMT-related regulators, we evaluated its effects on TGF-β-induced cell migration. Accordingly, migration assays were performed in the presence of the canonical EMT inducer TGF-β, either alone or in combination with netropsin. As expected, TGF-β treatment significantly enhanced the migratory capacity of TE-1 cells (Figure 5A). Notably, co-treatment with netropsin markedly attenuated the promigratory effects induced by TGF-β (Figure 5A). Furthermore, given that HMGA2 is frequently overexpressed in ESCC and that netropsin inhibits HMGA2 binding to DNA, thereby blocking its transcriptional activity [17], we next investigated the effects of netropsin in ESCC cells stably overexpressing HMGA2. Consistent with the findings obtained in the parental cell lines, netropsin significantly inhibited both cell migration and proliferation in HMGA2-overexpressing ESCC cells (Figure 5B,C). Finally, these results demonstrate that netropsin effectively counteracts HMGA2-driven malignant phenotypes, particularly those associated with EMT activation in ESCC. In addition, given previous evidence that aberrant HMGA2 expression influences the response of colorectal cancer cells to 5-fluorouracil (5-FU) [18], we investigated whether netropsin could modulate the sensitivity of ESCC cells to 5-FU. To this end, ESCC cell lines were exposed to netropsin and 5-FU, either individually or in combination. Co-treatment resulted in a greater reduction in cell growth than either agent alone (Figure 5D). To determine whether this enhanced effect reflected additive or synergistic interactions, drug combination analyses were performed using the Chou–Talalay method [19] across the IC₅₀ values of each compound and their respective fractional concentrations. Under these conditions, TE-1 cells exhibited synergistic interactions at the 1.0× and 0.75× IC₅₀ fractions, whereas TE-13 cells displayed synergism across all concentrations evaluated (Figure 5E).

Figure 5.

Netropsin reverses the TGF-β-induced migratory phenotype and exhibits synergism in 5-Fluorouracil (5-FU) sensitization. (A) Wound healing assay in TE-1 cells at 0 h, 24 h, and 48 h after treatment with the IC₅₀ (117.4µM) of netropsin, or TGF-β (10 ng/mL), or the IC₅₀ of netropsin + TGF-β (10 ng/mL). Images are representative, and quantification refers to the percentage of migration after 48 h. (B) Wound healing assay in TE-1 OE-Control and TE-1 OE-HMGA2 cells at 0 h, 24 h, and 48 h after treatment with the IC₅₀ of netropsin. Images are representative, and quantification refers to the percentage of migration after 48 h. Scale bar = 300 µm. (C) Evaluation of cell growth of TE-1 cells (OE-Control and OE-HMGA2) after 96 h of culture in the presence or absence of netropsin. (D) Evaluation of cell viability in TE-1 and TE-13 cell lines over 72 h of culture in the presence of netropsin, or 5-Fluorouracil (5-FU) or netropsin + 5-FU. (E) Analysis of the pharmacological interaction between 5-FU and netropsin in TE-1 and TE-13 cell lines, using Combination Index (CI) plots based on the Chou-Talalay method. The dashed line (CI = 1) delimits the additive effect; CI values < 1 indicate synergism and CI > 1 indicate antagonism in the different effect fractions (Fa). The data represent the mean ± standard deviation (SD) of three independent replicates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001; ns = not significant.

3. Discussion

Esophageal cancer remains one of the most lethal malignancies worldwide and is associated with a particularly poor prognosis [20]. HMGA2 is frequently overexpressed in ESCC [12] and represents an attractive therapeutic target due to its established role in tumor initiation and progression, as well as its association with aggressive disease and chemoresistance [21]. Indeed, several studies have shown that reducing HMGA2 expression attenuates malignant phenotypes across multiple tumor types [22,23]. Accordingly, pharmacological inhibition of HMGA2 has emerged as a promising therapeutic strategy, although its clinical potential remains to be fully explored [24,25,26,27]. In this study, we investigated the effects of netropsin, a DNA minor groove ligand that specifically binds adenine- and thymine-rich sequences, thereby preventing the interaction of HMGA proteins with their target DNA regions [24]. Using an in vitro model of ESCC, we demonstrated that netropsin significantly reduced key hallmarks of tumor progression, including cell viability, migration, and cell cycle progression, while promoting apoptotic cell death. Collectively, these findings are consistent with the well-established oncogenic functions of HMGA2, which promotes proliferation and cell cycle progression through multiple mechanisms [28,29], while also regulating signaling pathways involved in cell survival [30] and migration [31].

Furthermore, to gain mechanistic insights into the effects of HMGA2 inhibition following netropsin treatment, we performed in silico analyses that revealed a close association between HMGA2 expression and the EMT pathway in ESCC. When HMGA2 expression was correlated with the overall EMT signature, a weak inverse association was observed. However, gene-level analyses revealed a distinct pattern, with HMGA2 expression showing significant positive correlations with key EMT-associated genes, including CDH2, VIM, and SNAI2. To further validate these findings, we assessed the relationship between HMGA2 expression and the major EMT-inducing transcription factors in an independent cohort of ESCC patient samples. HMGA2 expression was positively correlated with SNAI1 and TWIST1, whereas no significant association was detected with SNAI2 (Slug). Together, these findings suggest that HMGA2 does not globally regulate the EMT transcriptional program in ESCC but rather promotes a selective EMT-related gene expression profile through the modulation of specific downstream targets. This interpretation is supported by previous studies demonstrating that HMGA2 promotes EMT through multiple signaling pathways, including TGF-β/SMAD and MAPK/ERK signaling [21]. In this context, it is important to emphasize that ESCC is a highly heterogeneous malignancy strongly influenced by distinct etiological factors, the prevalence of which varies considerably across geographic regions [2]. Accordingly, population-specific molecular characteristics should be taken into account when interpreting transcriptomic data. In the present study, the validation cohort consisted of Brazilian patients, whereas the ESCC cases included in TCGA are predominantly derived from Asian populations, particularly Chinese patients. Therefore, the apparent discrepancies observed in the correlation between HMGA2 expression and EMT-related regulators may, at least in part, reflect population-specific molecular differences. Supporting this hypothesis, previous studies have demonstrated that the TP53 mutational spectrum of Brazilian ESCC closely resembles that reported in French cohorts, while differing substantially from that observed in Chinese patients [32]. In this context, we investigated whether netropsin could modulate the expression of key transcription factors involved in EMT regulation in vitro. Our results demonstrated that netropsin significantly downregulated TWIST and SLUG, while also reducing the expression of the mesenchymal marker vimentin. Interestingly, however, SNAIL1 expression was significantly increased in both ESCC cell lines. Upregulation of SNAIL1 has been classically associated with the repression of E-cadherin expression during EMT [33], a pattern observed in TE-1 cells but not in TE-13 cells. Together, these findings suggest that pharmacological inhibition of HMGA2 by netropsin selectively modulates the EMT transcriptional program, attenuating several mesenchymal features while indicating that additional regulatory mechanisms may contribute to the differential control of E-cadherin expression in ESCC. In this regard, Tan et al. [34] demonstrated that HMGA2 promotes EMT through the coordinated regulation of SNAIL and TWIST. Importantly, the authors showed that these transcription factors exert partially redundant functions during EMT, such that activation of either factor alone is sufficient to sustain mesenchymal traits. These observations are partially consistent with our findings. In TE-13 cells, netropsin treatment induced a clear attenuation of the EMT phenotype, characterized by reduced TWIST and SLUG expression, decreased vimentin protein levels, and increased E-cadherin expression, despite the concomitant upregulation of SNAIL1. Interestingly, Tan et al. also demonstrated that TWIST suppression can influence SNAIL expression, suggesting that the increased SNAIL1 levels observed following netropsin treatment may represent a compensatory response to the marked downregulation of TWIST. Consistent with this hypothesis, HMGA2 overexpression in our model markedly increased TWIST expression, accompanied by a modest increase in SNAIL1, whereas HMGA2 silencing reversed the expression of both transcription factors. Moreover, Tan et al. demonstrated that HMGA2 represses E-cadherin expression through mechanisms that are, at least in part, independent of SNAIL and TWIST, indicating that HMGA2 can regulate epithelial identity through additional molecular pathways. This observation may also help explain our findings in TE-1 cells. Although netropsin treatment reduced TWIST and SLUG expression while increasing SNAIL1, E-cadherin expression was modestly decreased rather than restored. Therefore, the reduction in E-cadherin expression observed in TE-1 cells may reflect either compensatory SNAIL1 upregulation or the persistence of HMGA2-dependent regulatory mechanisms that operate independently of the canonical EMT transcription factors. Finally, these findings raise the possibility that the biological effects of netropsin are not exclusively mediated through HMGA2 inhibition. In addition to preventing HMGA2 binding to AT-rich DNA regions, netropsin has been reported to interact with other AT-rich DNA-binding proteins and regulatory elements [35]. Consequently, modulation of these additional targets may also contribute to the molecular and phenotypic changes observed in the present study. Additionally, these findings should be interpreted with caution, as the phenotype induced by netropsin appears to be compatible with a partial epithelial–mesenchymal transition (partial EMT) state [36]. Under these conditions, tumor cells acquire a hybrid epithelial–mesenchymal phenotype in which epithelial and mesenchymal transcriptional programs coexist, rather than being regulated in a strictly antagonistic manner [37]. Increasing evidence indicates that this intermediate cellular state is not merely a transitional phase but a stable and biologically relevant phenotype associated with enhanced cellular plasticity, collective migration, extracellular matrix remodeling, metastatic dissemination, and resistance to chemotherapy [38,39]. Therefore, the selective modulation of EMT markers observed following netropsin treatment may reflect a shift toward a partial EMT state rather than a complete reversal of the EMT program. Nevertheless, these findings reveal an apparently paradoxical pattern. Although SNAIL1 expression was consistently increased in both ESCC cell lines following netropsin treatment, the distinct effects observed on E-cadherin expression suggest that, in addition to the mechanisms discussed above, the intrinsic molecular heterogeneity of ESCC should also be considered. Given that TE-1 and TE-13 cells harbor distinct genetic and epigenetic backgrounds, different regulatory networks are likely to govern E-cadherin expression in each model. Consequently, HMGA2 inhibition may engage context-dependent molecular mechanisms that differentially influence the epithelial transcriptional program.

Indeed, although SNAIL1 directly binds to the CDH1 promoter, efficient repression of E-cadherin transcription requires the recruitment of multiple transcriptional co-repressors and chromatin-remodeling complexes, including histone deacetylases (HDACs) [40]. Consequently, increased SNAIL1 expression alone may not be sufficient to suppress CDH1 transcription, particularly in cellular contexts in which these co-repressors are absent, functionally impaired, or differentially regulated. This context-dependent regulatory mechanism may therefore explain the distinct patterns of E-cadherin expression observed in TE-1 and TE-13 cells following netropsin treatment. Since HMGA2-mediated EMT has been shown to involve activation of the TGF-β signaling pathway [41], and HMGA2 overexpression promotes cell migration and invasion through EMT induction [31], we next investigated whether netropsin could counteract these promalignant effects. To this end, migration assays were performed in ESCC cells stimulated with TGF-β or stably overexpressing HMGA2. In both experimental settings, netropsin effectively suppressed the enhanced migratory phenotype induced by either TGF-β stimulation or HMGA2 overexpression. Moreover, netropsin significantly reduced cell viability even in HMGA2-overexpressing cells, further supporting its ability to antagonize HMGA2-driven oncogenic signaling. Finally, given the reported association between HMGA2 overexpression and reduced sensitivity to 5-fluorouracil (5-FU) [42], we investigated whether netropsin could enhance the therapeutic response of ESCC cells to this chemotherapeutic agent. Combined treatment with netropsin and 5-FU produced a greater inhibitory effect on cell growth than either agent alone. Moreover, drug interaction analyses revealed a synergistic effect between the two compounds, suggesting that pharmacological inhibition of HMGA2 may enhance the antitumor efficacy of 5-FU in ESCC.

In conclusion, the present study demonstrates the antitumor potential of pharmacological inhibition of HMGA2 activity by netropsin in ESCC, particularly through the modulation of EMT-associated phenotypes. Nevertheless, several limitations should be considered when evaluating the translational potential of this approach. Netropsin has well-recognized drawbacks, including limited target specificity and dose-limiting toxicity [43]. Moreover, HMGA2 exerts biological functions that extend beyond its interaction with AT-rich DNA sequences, including protein–protein interactions and chromatin remodeling [29], which are unlikely to be completely abolished by preventing its DNA binding alone. These considerations should therefore be taken into account when interpreting the findings of the present study. Despite these limitations, our results demonstrate that pharmacological blockade of AT-rich DNA regions by netropsin attenuates multiple malignant phenotypes associated with HMGA2 overexpression in ESCC, particularly those related to EMT activation. These findings provide further evidence supporting HMGA2 as a therapeutically actionable target in ESCC and reinforce the rationale for the development of next-generation HMGA2 inhibitors with improved specificity and safety profiles. Given that aberrant HMGA2 expression is largely restricted to malignant tissues [44,45], targeting this protein represents a particularly attractive strategy for selective anticancer intervention.

4. Materials and Methods

4.1. Cell Lines

The TE-1 (CVCL_1759) and TE-13 (CVCL_4463) cell lines, derived from ESCC, were kindly provided by Dr. Pierre Hainaut (IARC, Lyon, France). The cells were cultured in RPMI-1640 medium (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10% fetal bovine serum (Gibco), 1% L-glutamine (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), and 1% penicillin/streptomycin (Invitrogen), and maintained at 37 °C in a humidified atmosphere containing 5% CO₂. The TE-1 OE-CTRL, TE-1 OE-HMGA2, TE-1 sh-Control and TE-1 sh-HMGA2 cell lines were established by our group transfecting vectors ORF-Control, ORF-HMGA2, sh-Control and sh-HMGA2 (GeneCopoeia, Rockville, MD, USA) using Lipofectamine 3000 (Invitrogen—L3000015) according to the manufacturer’s instructions and cultured under the same conditions described above, with the addition of 0.2% Geneticin (G418—Sigma-Aldrich, St. Louis, MO, USA) for stably HMGA2 overexpression cell lines and 0.5 µg/mL of puromycin (P8833—Sigma-Aldrich) for stably silenced HMGA2 cell lines.

4.2. Netropsin IC₅₀ Determination and Cell Viability

To determine the Inhibitory Concentration of 50% (IC₅₀) for netropsin (Sigma-Aldrich—N9653), TE-1 and TE-13 cell lines were seeded at a density of 5 × 10³ cells per well in 96-well plates and treated with increasing concentrations of netropsin (50, 100, 150, 200, and 250 µM). Seventy-two hours after treatment initiation, the culture medium was removed, and 200 µL of 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT—Sigma-Aldrich) solution at 125 µg/mL, previously diluted in culture medium, was added and incubated at 37 °C for 3 h. Subsequently, absorbance was measured at 595 nm. To evaluate cell viability following combined treatment with 5-fluorouracil (5-FU, Fauldflour, Libbs Farmacêutica, São Paulo, Brazil) and netropsin, TE-1 and TE-13 cell lines were seeded in 96-well plates at a density of 5 × 10³ cells per well. Twenty-four hours after seeding, cells were treated with the IC₅₀ concentration of netropsin and 5-FU for 72 h and then subjected to a viability assay using CCK-8 (TargetMol, Boston, MA, USA), according to the manufacturer’s instructions.

4.3. Cell Migration (Wound Healing Assay)

Cells were seeded in 24-well plates at a density of 1 × 10⁵ cells per well and allowed to grow until reaching 90–100% confluence. Subsequently, cells were treated with 10 µg/mL mitomycin C (Sigma Aldrich—M4287), diluted in culture medium, for 2 h. After treatment, the culture medium was removed, cells were washed with PBS, and a scratch was created across the cell monolayer using a 200 µL pipette tip. Cells were then incubated with either control culture medium or culture medium supplemented with the IC₅₀ concentration of netropsin or 10 ng/mL of TGF-β (PeproTech, Thermo Fisher Scientific, Cranbury, NJ, USA; 100-21C). Images were acquired at 0 h, 24 h, and 48 h, and wound closure was quantified using the EVOS M5000 Imaging System (Thermo Fisher Scientific).

4.4. Cell Cycle Analysis

In 6-well plates, 2 × 10⁵ cells were seeded for cell cycle analysis. Twenty-four hours after seeding, cells were treated with either control culture medium or culture medium supplemented with the IC₅₀ concentration of netropsin. After 72 h, cells were harvested and fixed for staining with propidium iodide (Invitrogen, F10797) according to the manufacturer’s protocol. Cell cycle distribution was analyzed using a flow cytometer (FACSCalibur, BD Biosciences, San Jose, CA, USA) and subsequently processed using Flowing Software 2.5.1 (Perttu Terho, Turku Bioscience Centre, University of Turku, Turku, Finland).

4.5. Apoptosis Assay

To evaluate apoptosis, 2.5 × 10⁵ cells per well were seeded in 6-well plates and allowed to adhere for 24 h. Cells were then treated with either control culture medium or culture medium supplemented with the IC₅₀ concentration of netropsin. Seventy-two hours after treatment initiation, cells and supernatant were collected and stained with Annexin V and 7-AAD according to the manufacturer’s instructions (eBioscience Annexin V Apoptosis Kit; Invitrogen, Thermo Fisher Scientific, San Diego, CA, USA). The percentage of Annexin V–positive cells was determined by using the web-based software Floreada.io (https://floreada.io, accessed on 24 February 2025).

4.6. Synergism Between 5-Fluorouracil and Netropsin

The interaction between netropsin and 5-FU was evaluated using the Chou–Talalay method [19]. Briefly, cells were treated for 72 h with concentrations corresponding to 0.25×, 0.50×, 0.75×, and 1.0× of the IC₅₀ of each drug, alone or in combination. Cell viability was assessed using the MTT assay, as previously described. The resulting data were analyzed using CompuSyn software version 1.0 (ComboSyn Inc., Paramus, NJ, USA) to generate isobolograms. Drug interactions were quantified based on the combination index (CI) as a function of the fraction affected (Fa), where CI = 1 indicates an additive effect, CI > 1 indicates antagonism, and CI < 1 indicates synergism.

4.7. RNA Extraction, Reverse Transcription, and qPCR

Total RNA from ESCC cell lines was extracted using the TRIzol reagent (Invitrogen), according to the manufacturer’s instructions. For reverse transcription, 750 ng of total RNA was converted into cDNA using the GoScript kit (Promega, Madison, WI, USA), and mRNA expression levels were detected by qPCR using the QuantiNova SYBR Green PCR Kit (Qiagen, Hilden, Germany) with gene-specific primers. GAPDH or β-Actin was used as an endogenous control for qPCR normalization. Relative gene expression was calculated using the 2^{^−ΔΔC} method. qPCR primer sequences follow:

GAPDH Sense: 5′-CAACAGCCTCAAGATCATCAGCAA-3′ and Antisense: 5′-AGTGATGGCATGGACTGTGGTCA-3′; ACTB Sense: 5′-CGCCAACACAGTGCTGTCT-3′ and Antisense: 5′-CACGGAGTACTTGCGCTCAG-3′ SNAI1 Sense: 5′-AATACTGCAACAAGGAATACCTCAGCCT-3′ and Antisense: 5′-GGACAGGAGAAGGGCTTCTCGCCAGTG-3′; SNAI2 Sense: 5′-CTTCCTGGTCAAGAAGCA-3′ and Antisense: 5′-GGGAAATAATCACTGTATGTGTG-3′; TWIST1 Sense: 5′-TGTCCGCGTCCCACTAGC-3′ and Antisense: 5′-TGTCCATTTTCTCCTTCTCTGGA-3′;CDH1 Sense: 5′-GAATGACAACAAGCCCGAAT-3′ and Antisense: 5′-GACCTCCATCACAGAGGTTCC-3′; HMGA2 Sense: 5′-GCGCCTCAGAAGAGAGGAC-3′ and Antisense: 5′-GGTCTCTTAGGAGAGGGCTCA-3′.

4.8. Immunofluorescence

Cells grown on sterile glass coverslips were fixed with 4% paraformaldehyde for 10 min at room temperature, washed with PBS, and subjected to antigen retrieval with NH₄Cl (50 mM) for 30 min. Cells were then permeabilized with 0.2% Triton™ X-100 (Sigma-Aldrich, X100) for 5 min and blocked with 5% BSA for 30 min. Primary antibody against vimentin (working dilution 1:50, Molecular Probes—Invitrogen—V2258) was diluted in PBS containing 3% BSA and incubated overnight at 4 °C in a humidified chamber. After washing, cells were incubated with FITC-conjugated secondary antibodies (1:100; Alexa Fluor^® 488, Molecular Probes—Invitrogen—A11001) for 2 h at room temperature, protected from light.

F-actin was stained with ActinRed 555 ReadyProbes Reagent (Invitrogen—R37112) for 30 min. After washing, nuclei were stained with DAPI (Invitrogen, D1306) for 20 min, and coverslips were mounted using N-propyl gallate mounting medium. Fluorescence images were acquired using the EVOS M5000 Imaging System (Thermo Fisher Scientific, Waltham, MA, USA).

Image quantification was performed in Image J version 1.54 (National Institutes of Health, Bethesda, MD, USA). Images were converted to 8-bit, and the mean fluorescence intensity of the vimentin channel was measured for each field. Values are expressed in arbitrary units (A.U.)

4.9. Western Blotting

Stably modulated HMGA2 cell lines were seeded at 2.5 × 10⁵ cells/well in 6-well plates. After 48 h, cells were harvested and lysed for 30 min in RIPA-like buffer (50 mM Tris-HCl pH 7.4, 250 mM NaCl, 0.1% SDS, 2 mM DTT, 0.5% NP-40) supplemented with Complete Mini EDTA-free protease inhibitor (Roche Diagnostics, Mannheim, Germany; 4693159001). Protein concentration was determined by Bradford assay (Bio-Rad Laboratories, Hercules, CA, USA; #500-0006) against a spectrophotometric standard curve. Thirty micrograms of protein were mixed with Laemmli buffer (200 mM Tris-HCl pH 6.8, 8% SDS, 40% glycerol, 0.01% bromophenol blue, 10% β-mercaptoethanol), denatured at 95 °C for 5 min, resolved by SDS-PAGE, and transferred onto nitrocellulose membranes using the iBlot Dry Blotting System (Invitrogen).

Membranes were blocked for 1 h in TBS-T (0.1% Tween-20) containing 5% non-fat dry milk and incubated overnight at 4 °C with anti-HMGA2 (rabbit, 1:1000; Invitrogen, PA5-21320) or anti-Lamin A/C (mouse, 1:1000; Cell Signaling Technology, Danvers, MA, USA; 4777S). Then, membranes were incubated for 2 h at room temperature with HRP-conjugated anti-rabbit (goat, 1:5000; Cell Signaling Technology, 7074S) or anti-mouse (goat, 1:5000; Cell Signaling, 7076) secondary antibodies. Following six 10 min washes, signals were detected by chemiluminescence (ECL Western Blotting Substrate, Promega, W1001).

4.10. In Silico Analyses

Gene expression data from patients with ESCC were obtained from the Gene Expression Omnibus (GEO) database under accession code GSE53625. Differential gene expression (DGE) analysis of microarray data was conducted using the limma package (version 3.66.0) in the R environment (R version 4.5.3; R Foundation for Statistical Computing, Vienna, Austria). Initially, linear models were fitted, and p-values were corrected for multiple testing using the Benjamini–Hochberg (BH) method to control the false discovery rate (FDR). The statistical significance of each gene was evaluated based on adjusted p-values (padj), while log2 fold change (log2FC) values were used to assess differential gene expression between tumor and paired non-tumor samples.

The EMT pathway was evaluated using the Molecular Signatures Database (MSigDB) and Gene Set Enrichment Analysis (GSEA), with genes ranked according to log2FC values, using the clusterProfiler package (version 4.18.4) in the R environment. The correlation between HMGA2 expression and genes associated with the EMT pathway was assessed using Spearman’s correlation analysis. Overall survival analyses were performed after stratifying patients according to scores generated by single-sample Gene Set Enrichment Analysis (ssGSEA). The optimal cutoff point was determined using the survminer package (version 2.4.9). Survival curves were estimated using the Kaplan–Meier method, and differences between groups were assessed using the log-rank test.

4.11. Statistical Analysis

For in vitro assays, the statistical analyses were performed using GraphPad Prism software version 10.3.1 (GraphPad Software, Boston, MA, USA). For the functional assay results, a two-way ANOVA test was applied using mean and standard deviation (SD) values. For gene expression experiments, the Kolmogorov–Smirnov test was used to assess data normality, followed by the Mann–Whitney test, using data derived from three independent experiments. The significance threshold was set at p < 0.05. For the in silico analyses, expression data were transformed into log₂ counts per million (log₂CPM) to evaluate non-parametric correlations. Spearman’s correlation analysis was then performed to assess the association between HMGA2 expression levels and other genes of interest.

Author Contributions

Conceptualization, A.P.J.; Data Curation, L.d.J.L.; Formal Analysis, L.d.J.L., M.L.-C., M.L.B.W., A.P.J., M.S.R., N.M.D.C. and L.M.R.T.; Funding Acquisition, A.P.J. and L.E.N.; Investigation, L.d.J.L., I.P.R.d.O., M.L.-C. and A.R.M.A.; Methodology, L.d.J.L., M.L.-C., M.S.R., L.M.R.T. and A.P.J.; Project Administration, A.P.J.; Resources, A.P.J.; Supervision, L.F.R.P., L.E.N. and A.P.J.; Validation, A.P.J., M.S.R. and L.M.R.T.; Visualization, L.d.J.L. and M.L.-C.; Writing—Original Draft Preparation, L.d.J.L.; Writing—Review and Editing, A.P.J., L.F.R.P., M.S.R., L.M.R.T., N.M.D.C. and M.L.-C. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding Statement

We would like to thank Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq–Brazil), Fundação de Amparo à Pesquisa Carlos Chagas Filho (FAPERJ-Brazil), Swiss Bridge Foundation and Programa de Oncobiologia/Fundação André Frauzino de Pesquisas para o Câncer.