Breast cancer remains one of the most prevalent types of cancer worldwide and is the leading cause of cancer-related deaths among women. In 2022, 2.3 million new cases and 670,000 deaths from breast cancer were recorded worldwide. The annual rate increased between 1% and 5% in half of the countries included in the Globocan 2022 registry (1). Breast cancer ranks among the first or second most common cancers in women worldwide, and it is associated with high mortality rates, particularly in countries with medium to low Human Development Index (HDI) scores. This cancer is more common in women aged >50 years (71% of new cases and 79% of deaths). By 2050, the number of cases is expected to increase by 38%, and the number of deaths by 68%. Therefore, continued progress in prevention, early diagnosis, and effective treatment is urgently needed to reduce the disease’s impact. Multiple factors, including frequent alcohol consumption, exogenous hormone use, obesity, and not breastfeeding, can contribute to the development of breast cancer (2).

Accurate and early diagnosis is essential to determining the most appropriate prognosis and treatment. Tumor type and size, histological grade, presence of estrogen receptor (ER), progesterone receptor (PR), and human epidermal growth factor receptor 2 (HER2/neu), and lymph node metastasis are some of the factors to consider (3). The 2011 St. Gallen Consensus established guidelines for classifying molecular subtypes of breast cancer into Luminal A (ER+/PR+/HER2-/low Ki-67); Luminal B (ER+/PR+/HER2-/+/high Ki-67); HER2 overexpression (ER-/PR-/HER2+); and triple-negative breast cancers (TNBC) (ER-/PR-/HER2-) (4). The ideal is that the therapeutic approach should be in accordance with the molecular profile of the patient in each case. Therefore, the search for new biomarkers that contribute to obtaining a more complete molecular profile of tumors is a priority, to find the possible association between the expression of miRNAs with their respective target genes, such as those published for TNBCs (5), and to better understand their role in cells.

In this study, we aimed to explore the possible function of Mammaglobin-A, which is encoded by the gene SCGB2A2. It is expressed in tumors derived from the breast, ovary, uterus, and salivary glands and only rarely in tumors derived from other organs, which has made its detection a useful tool to determine the origin of adenocarcinomas, especially in women (6). This protein is a member of the secretoglobin superfamily with an unknown function and is used as a diagnostic marker for breast cancer. Due to its frequent overexpression in breast cancers, it has been proposed as a promising therapeutic target for these tumors (6, 7), even as a breast cancer vaccine target (8). The genes of this family form a cluster located on chromosome 11q13 (9). Mammaglobin-A (SCGB2A2), a protein of 93aa, member of this family, was first identified in 1996 through differential screening of a human breast carcinoma cDNA library, with expression primarily restricted to mammary gland epithelium (10). Positivity rates range from 59% to 100% in lobular breast carcinomas and 25% to 94% in invasive breast carcinomas (10). The presence of Mammaglobin in tumors is associated with low-grade estrogen receptor, in blood to circulating cells, or an established diagnostic tool for recognizing metastatic breast cancer (6). It is considered a promising candidate for a breast cancer vaccine therapy due to its high expression in breast cancers and minimal presence in healthy tissues (8). Little is known about the role of Mammaglobin; however, it is suspected that it can participate in various biological processes, including cell signaling, immune response, and chemotaxis, with potential hormone-binding and transport function (6).

The miRNAs are typically short RNA non-coding sequences of 19-25 nucleotides in length that bind to the untranslated regions (UTRs) of the messenger RNAs, thereby blocking their translation (11), either degrading or preventing access to target mRNAs, leading to reduced protein expression. A single miRNA can regulate the expression of multiple genes, and a gene can be regulated by several miRNAs (11). They regulate approximately 30% of all human genes and play inhibitory or promoting roles in cancer; therefore, they are promising as prognostic and predictive biomarkers in breast cancer, because they can be easily detected on tumor biopsy or exosomes in body fluids, and they are associated with particular breast cancer subtypes (12). Also, miRNAs play a role in development, growth, and metastasis in breast cancer (13).

Analysis of miRNA expression profiles from studies using next-generation sequencing (NGS) techniques, including RNA-seq and microarrays, in tumor and normal tissue samples (13) has enabled the identification of miRNA expression associated with breast cancer subtypes, such as MiR-21. miRNAs co-expressed with cancer-related genes in tumor tissues can serve as potential biomarkers for further experimental validation (11), for example, the detection of exosomal miR-575 in vaginal discharge for ovarian cancer (14).

The role of Mammaglobin remains uncertain, but its undoubted relationship with the presence of breast cancer and micrometastasis, so identifying possible biological pathways in which it participates, remains a priority. As well as the deregulation of miRNA is associated with cancer development and progression, this led us to investigate a possible association of Mammaglobin with the action of some miRNAs. No previous study has evaluated possible relationships between Mammaglobin and miRNA activity. For this reason, the goal of this study was to evaluate the expression profile of miRNAs and their possible correlation with SCGB2A2 gene expression in data extracted from The Cancer Genome Atlas (TCGA), followed by differential expression analysis in breast normal and tumor tissues.

Data acquisition: mRNA (SCGB2A2) and miRNA expression profiling using FireBrowse and TCGABiolink. Gene expression data for breast cancer (BRCA) were obtained from the TCGA cohorts via the Firebrowse platform (http://firebrowse.org/). The dataset comprised 86 matched patient cases, each with paired samples of normal solid tissue and primary breast tumor. The miRNA expression data were downloaded using the TCGAbiolinks (15–17), an R/Bioconductor package for integrative analysis with GDC data (15), following a methodology adapted from Noushmehr et al. (18) based on the same patient case codes found in the miRNA search.

The process began by acquiring miRNA and mRNA expression data from matched patient samples in the TCGA database (Figure 1). Specifically, SCGB2A2 mRNA levels were retrieved via the Firebrowse platform, while miRNA expression profiles were obtained using the TCGAbiolinks R package. Both raw count datasets were analyzed for differential expression with the DESeq2 package from Bioconductor. The resulting data were then normalized with a variance-stabilizing transformation (VST) or a regularized log transformation for downstream analyses. This approach identified differentially expressed miRNAs and determined the differential expression status of SCGB2A2 itself. Subsequently, a correlation analysis was conducted between the significant miRNAs and SCGB2A2 expression to compute the R² value and its associated p-value, evaluating a potential regulatory relationship. Finally, an enrichment analysis was performed on the filtered miRNAs and their predicted target genes. These target genes were identified, and their expression was validated within the same dataset. The resulting gene set was then analyzed to identify enriched cancer-related pathways.

Analysis of mRNA expression. Raw mRNA count data for SCGB2A2 and all other genes were used in this study. To ensure data quality and reliability of downstream analyses, two filtering steps were applied prior to normalization and differential expression analysis. First, genes with low expression across the majority of samples were removed. Specifically, genes with fewer than 10 counts in at least 90% of the samples were excluded, as such low-expressed genes are prone to technical noise and can skew normalization and statistical testing (19, 20). Second, genes with low variability across samples were filtered out based on the coefficient of variation (CV). The CV was calculated for each gene as the ratio of the standard deviation to the mean expression across all samples. Genes with a CV below a predefined threshold (CV <1) were deemed to have low variability and were removed, as they provide little discriminatory information and can destabilize downstream multivariate analyses. This CV-based filtering approach is a standard and reproducible method in transcriptomic studies.

After filtering, normalization was performed using the DESeq2 package (21). Size factors were estimated to account for differences in library depth and composition across samples. The raw counts were then normalized using the VST or regularized logarithm (rlog) transformation, which place the data on a log₂ scale and stabilize the variance across the range of expression values. This transformation attenuates the log₂ fold change (LFC) of genes with low counts, minimizing noise from random fluctuations that could otherwise overshadow biologically significant signals. The normalization process improves overall data stability and enhances the reliability of downstream analyses (19, 20).

Following normalization, SCGB2A2 expression levels were extracted for further analysis. Differential expression analysis between tumor and normal tissues was performed using the Wald test implemented in DESeq2 (21), with empirical Bayes shrinkage to moderate log₂ fold changes and improve stability for genes with low counts. Genes with an adjusted *p*-value (false discovery rate, FDR) <0.05 and an absolute log₂ fold change >0.584 were considered statistically significant.

Analysis of miRNA expression. The miRNA expression data were processed following the same workflow as described for mRNA. Raw miRNA count data were processed (15-17), and an initial quality control assessment was performed to identify potential outliers and assess overall data quality.

The same filtering criteria applied to mRNA data were used: miRNAs with fewer than 10 counts in at least 90% of the samples were removed to eliminate low-expressed features (19, 20), and miRNAs with a CV below 0.1 were excluded to filter out those with insufficient variability for downstream analyses.

The filtered raw count data were then normalized using DESeq2 (21) with size factor estimation and VST, identical to the mRNA processing pipeline. Differential expression analysis was performed using the same thresholds: an adjusted p-value <0.05 and an absolute log₂ fold change >0.584 were considered statistically significant.

Identification and profiling of differentially expressed miRNAs and SCGB2A2 mRNA expression. DESeq2 analysis was conducted using the Wald test to identify differentially expressed miRNAs (DEmiRs) and compare tumor and normal tissues. This analysis was performed in RStudio using the generalized linear model (GLM) approach proposed by Love et al. (19). This method identifies the differentially expressed genes (DEGs) in the data matrices. DEmiRs were defined as miRNAs with an adjusted false discovery rate (padj) of less than 0.05 and a log 2-fold change greater than 0.584 or less than -0.584 relative to normal tissue. By setting a value of 0.584, it is possible to ensure that the genes showing biologically significant changes, which is especially relevant when working with large datasets, in which both magnitude and significance are important for robust conclusions (20).

The empirical Bayes shrinkage approach was employed to enhance the DEmiR identification accuracy. This technique provides a more precise dispersion estimation, thereby improving the reliability of the differential expression analysis (19).

Correlation analysis of miRNAs and SCGBs mRNA counts in normal and tumor tissues. A correlation analysis was conducted using Spearman’s rank correlation test to examine the relationship between miRNA counts and SCGB2A2 mRNA expression in normal and tumor tissues separately. The statistical significance of each correlation was determined using the asymptotic t approximation implemented in the Spearman test, which generates the corresponding p-value. These p-values were then used as a filter, retaining only correlations with p<0.05 for subsequent analysis.

Selection of miRNAs coexpressed with SCGB2A2. miRNAs were selected as co-expressed with SCGB2A2 if they met the following criteria: differential expression between normal and tumor tissues, a statistically significant Spearman correlation (p-value <0.05) with SCGB2A2 expression, and a correlation coefficient (r) >0.3 for positive correlation or <-0.3 for negative correlation.

Target gene selection. Target mRNAs of each filtered miRNA were identified using the TarBase platform. The filtered set included only those miRNAs showing a correlation with SCGB2A2 expression greater than 0.3 or less than –0.3, with a p-value below 0.05. Subsequently, the expression profiles of the same patients from whom SCGB2A2 counts were obtained were searched for the corresponding target mRNAs.

Enrichment analysis. Enrichment analysis of the final set of miRNAs and target genes was performed using miRNA Enrichment and Annotation (miEAA) and Database for Annotation, Visualization, and Integrated Discovery (DAVID) software. Pathways and processes with a p-value less than 0.05 were selected following the upload of the miRNA and target gene lists into both programs.

Kaplan-Meier survival analysis. Kaplan-Meier survival analysis was conducted to determine the biomarker candidates associated with the prognosis of patients with breast cancer. Analysis was performed using the mirPower-Kaplan-Meier plotter web server (www.kmplot.com) (21). The TCGA and METABRIC breast cancer datasets were selected for analysis. A p-value of <0.05 is considered statistically significant, and the best cutoff was set.

miRNAs and SCGB2A2 mRNA data. A total of 172 samples were included in the study, consisting of 86 normal breast tissue samples and 86 breast carcinoma samples. None of the patients had received neoadjuvant therapy before sample collection, ensuring that the expression profiles reflected untreated tissue conditions.

The dataset comprised 1,882 miRNAs with detectable expression levels, which were retained after quality control filtering to remove features with low counts or poor annotation. Quality assessment confirmed the absence of outlier samples and consistent data distribution across the cohort, supporting the reliability of downstream analyses.

Differential expression analysis. A comparative transcriptome analysis between breast tumor tissues and matched normal controls identified 413 DEmiRs using DESeq2 (Figure 2). These DEmiRs were selected based on an adjusted p-value <0.05 and an absolute log₂ fold change greater than 0.584. A parallel analysis of the mRNA expression data, focusing on the secretoglobin family, revealed that three members were differentially expressed. Among these, SCGB2A2 showed significant dysregulation, confirming its altered expression in the tumor compared with normal breast tissue. In addition to the secretoglobins, a DESeq2 analysis of the remaining genes identified 1261 differentially expressed genes.

Correlation between miRNA expression and SCGB expression levels. Two scatterplots were created to illustrate the correlation between miRNA expression and differentially expressed SCGBs, using R² values and p-values (Figure 3). The scatterplots for miRNA hsa-mir-184 and SCGB2A2, with a Spearman R² value of 0.32 and a p-value of 0.0029 (Figure 3A), and for miRNA hsa-mir-190b, with a Spearman R² value of 0.33 and a p-value of 0.0017 (Figure 3B).

Both plots reveal a correlation in the data, as indicated by the R² values, with significant differences in p-values, indicating increased miRNA expression as SCGB2A2 expression rises in tumor tissues.

The differential expression and correlation data for the miRNAs are presented in Table I. For the miRNAs that have log₂ fold change values greater than 3.2, the gene expression under tumor conditions is approximately 8 times higher than under reference conditions. The logarithmic scale was used to better manage large differences in expression levels and to provide a more intuitive interpretation of relative changes (19, 22, 23).

Final miRNAs selection. The two identified miRNAs that passed all applied filters showed significantly different expression levels between normal and tumor tissues (Figure 4). The miRNAs hsa-mir-184 (Figure 4A) and hsa-mir-190b (Figure 4B) were expressed at lower levels in the normal tissue than in the tumor tissue.

Enrichment analysis miRNA coexpressed with SCGB2A2. The processes and pathways in which just one miRNA participates, filtered by miEAA, are presented in Table II, along with their respective p-values. The highlighted processes include cell death, cancer, and apoptosis.

Functional enrichment analysis revealed that both hsa-miR-184 and hsa-miR-190b are strongly associated with processes central to cancer biology. Among the most significant functional categories were angiogenesis, epithelial-to-mesenchymal transition (EMT), apoptosis, and inflammation—all recognized as hallmarks of tumor progression. Notably, angiogenesis showed the highest enrichment (Fold >3×10⁹, p<0.001), involving both miRNAs, suggesting their coordinated participation in vascular remodeling mechanisms that sustain tumor growth and metastasis. In addition, hsa-miR-184 displayed a robust association with keratinocyte apoptosis (p=4.8×10⁻⁷, FDR=3.27×10⁻⁴), reinforcing its potential regulatory role in epithelial cell survival and turnover.

Disease enrichment further highlighted the relevance of these miRNAs in specific cancer types. hsa-miR-184 was found to be downregulated in breast, renal, and oral squamous cell carcinomas, supporting a possible tumor-suppressive function in these tissues. Conversely, its upregulation in glioma, hepatocellular, and nasopharyngeal carcinomas points to a context-dependent oncogenic role. hsa-miR-190b exhibited the strongest association with breast neoplasms (UP; p=4.8×10⁻⁶, FDR=1.22×10⁻³), indicating its potential function as an oncomiR in this context, while its downregulation in gastric carcinoma suggests tissue-specific regulatory mechanisms.

Together, these results suggest that hsa-miR-184 and hsa-miR-190b participate in the modulation of key oncogenic pathways, particularly those controlling angiogenesis, apoptosis, and epithelial plasticity. Their contrasting expression patterns across different carcinoma types highlight their dual roles as potential tumor suppressors or oncogenic miRNAs, depending on the cellular environment, emphasizing their value as candidate biomarkers and regulatory nodes in cancer-related networks.

The miRNA hsa-mir-190b was significantly associated with specific cancer types. It showed a strong association with gastric carcinoma (p-value=1E-300) and was also significantly linked to breast neoplasms (p-value=0.0056).

Correlation between miRNA counts and their targets within the mRNA dataset. From the filtered miRNAs, 671 target genes were identified, with 280 corresponding to the miRNA hsa-mir-190b and 391 to the miRNA hsa-mir-184. Thus, there are multiple interactions between genes and miRNAs. To visualize these interactions, Cytoscape was used to construct the interaction network between the filtered miRNAs and their target genes. Additionally, a correlation analysis was performed between the normalized counts of both miRNAs and their predicted target genes, using the same dataset comprising 20,531 genes. The expression counts of SCGB2A2 were extracted directly from this dataset based on patient identifiers. The results are summarized in Figure 5, where ANXA6 and SNX9 were filtered using the same normalization and selection criteria applied to the miRNA counts. Both target genes, ANXA6 and SNX9, exhibited significant correlations (p<0.05) with the analyzed miRNAs in breast tumor tissue. In normal tissue, the correlation between hsa-miR-184 and ANXA6 was negative (R=–0.6862, p=0.0001) (Figure 5A). In contrast, in tumor tissue, hsa-miR-184 and ANXA6 showed a moderate negative correlation (R=–0.5311, p=2.30×10⁻⁷) (Figure 5B).

A similar trend was observed for hsa-miR-190b and SNX9, where a strong negative correlation was detected in normal tissue (R=–0.6145, p<0.001) (Figure 5C), and a weaker but still significant correlation was found in tumor tissue (R=–0.3727, p=4.48×10⁻⁴) (Figure 5D). Likewise, the correlation between hsa-miR-184 and ANXA6 in normal tissue was negative (R=–0.5136, p=6.35×10⁻⁷) (Figure 5E), while in tumor tissue, hsa-miR-184 and SNX9 displayed a similar inverse relationship (R=–0.5010, p=1.28×10⁻⁶) (Figure 5F).

Finally, scatter plots of hsa-miR-184 versus SNX9 in normal and tumor tissues confirmed this pattern. In normal tissue, a moderate negative correlation was observed [R²(Spearman)=–0.5217, p=3.98×10⁻⁷] (Figure 5G), while in tumor tissue, this inverse association was stronger [R²(Spearman)=–0.6253, p<0.001] (Figure 5H). Collectively, these findings highlight a consistent negative relationship between hsa-miR-184 and SNX9 expression across both tissue types, suggesting a potential regulatory interaction that may be reinforced in tumor tissue.

ANXA6 and SNX9 expression in cancer and normal tissue. The two target genes, ANXA6 and SNX9, showed significantly different expression levels between normal and tumor tissues (Figure 6). Both SNX9 (A) and ANXA6 (B) exhibit a marked expression decrease in tumor tissues compared to normal tissue, indicating a possible downregulation associated with tumor progression.

Enrichment of the target genes. Thirty-three cancer-related terms involving the target genes were identified, with some having a p-value less than 0.05 (Figure 7). The graph displays the terms with the most associated genes: cell cycle and breast cancer. These findings suggest that both miRNAs target proto-oncogenes to become oncogenes.

miRNAs survival analysis. Both miRNAs were evaluated to predict the overall survival. High expression of hsa-mir-190b was associated with the best prognosis evaluated with TCGA and METABRIC databases (Figure 8).

In the present study, we identified that hsa-mir-184 and hsa-mir-190b were co-expressed with SCGB2A2 in breast tumor tissue, and among the analyzed miRNAs, they yielded significant results in the enrichment analysis. Where those miRNAs are involved in breast cancer and other neoplasms. In addition, we described that both miRNAs are also co-expressed in breast cancer tissue, while their predicted target genes are downregulated. This study took a similar approach to that used by Waszczykowska et al. (24), who identified Wnt-associated molecular signatures as potential therapeutic targets, evaluated through bioinformatics analysis data obtained from TCGA repository for BC patients, breast invasive carcinoma, and literature review (24).

SCGB2A2 exhibits a differential expression pattern between tumor and normal tissues, with its expression increasing in tumor tissue. The expression of Mammaglobin as a biomarker of the presence of breast cancer has been evaluated in bone marrow (25) and sentinel lymph nodes. The expression of Mammaglobin as a biomarker of the presence of breast cancer has been assessed in bone marrow, sentinel lymph nodes (26), and serum (27) from patients with breast cancer.

On the other hand, several studies have identified hsa-miR-190b as a diagnostic and prognostic biomarker of breast cancer. Its expression can differentiate tumor grades and receptor status, such as in ER-positive breast cancer, which is overexpressed (28). In other cancers, like endometrial and sarcoma, it is a potential candidate biomarker for diagnosis (29, 30). Another study underscored the role of ‘hsa-mir-190b’ as a down-regulator of apoptosis and several molecular pathways related to cancer cell proliferation and migration (28). In the same way, hsa-mir-184 miRNA plays a regulatory role in promoting cell proliferation and apoptosis (31). In HCC, this miRNA forms a regulatory loop LncRNA SNHG4/miR-184/AGO2, which enhances cell proliferation, migration, invasion, autophagy and inhibits apoptosis. SNHG serves as a sponge for miR-184 and regulates AGO2 expression (32). Another study identified hsa-mir-184 as being associated with the expression of one-third of protein-coding genes in regions of genomic instability in HCC (33). Notably, both miRNAs have been consistently reported in the literature as differentially expressed in several carcinomas, including breast carcinoma, supporting their potential relevance in tumor biology.

The target genes of those miRNAs, SNX9 and ANXA6, are overrepresented in pathways related to the pathogenesis of breast cancer, and both support a scaffold that regulates various cellular processes that modulate the behavior of cancer cells (34, 35). SNX9 negatively regulates invadopodia formation in breast cancer by increasing the internalization of membrane-type 1 matrix metalloproteinase (MT1-MMP) from the plasma membrane (36). In primary breast cancer tumors, it presents low expression, while in metastatic tumors, it is overexpressed (37). The mechanisms that regulate its expression are not well known; here, we propose the possible participation of miRNAs as a complement to other proposed mechanisms. Besides, in MDA-MB-231 cells, SNX9 is required to maintain their ability to metastasize in chicken embryo models; reason why it is considered a regulator of metastasis (35). One possible mechanism involves the activation of RhoA and Cdc42 GTPases, which act in the RhoA-ROCK and N-WASP pathway to modulate cell motility. Other mechanisms could be through direct binding either to the invadopodial protein dedicator of cytokenesis 1 (DOCK1) or to the proteins involved in vesicular trafficking, such as ADAM9 and ADAM15 (36). In addition to routine markers, detection of SNX9 and histone–lysine N-methyltransferase (SETD2), by RT-PCR and western blotting, has been useful to complete the classification and prognosis of breast tumors; they have also been considered as possible therapeutic targets in breast cancer (38).

SNX9 interacts with proteins such as SH3GL3 and SH3GL1, both of which play roles in endocytosis, a process that contributes to multidrug resistance in cancer. Also, with epidermal growth factor-related proteins, such as EPS15 and GRB2, which are involved in the regulation of cell growth in cancer. Alterations in SNX9 have also been linked to functions in cell division and invasion, highlighting its implications in human diseases, including cancer.

In relation to ANXA6, in a model of endothelial cells treated with anti-ANXA6 antibodies, a partial blockage of Ca2-dependent adhesion in migrating cells was observed (39). Suppression of ANXA6 in BT-549 triple-negative breast cancer cells reduced cell adhesion to fetuin-A, an important factor in serum that is thought to promote cell adhesion of tumor cells; on the contrary, it increases proliferation, acting as a tumor suppressor. In the same cells, the adhesion to fibronectin and laminin was not affected; it was also detected that ANXA6 was secreted into exosomes by exocytosis as a contribution to focal adhesion (40). In other cells, such as A431 cells, BCCs, and HNSCCs, as well as ER-negative BCCs, the antiproliferative activity of ANXA6 was mediated by the interaction with PKCa- and p120GAP-mediated EGFR and H-Ras downregulation in EGFR overexpressing (40). In patients with the most aggressive type of breast cancer, the TNBC, low expression of ANXA6 is correlated with tumor suppressor activity and poor survival. There is increasing evidence to suggest that the detection of changes in ANXA6 expression is useful as a biomarker of progression and as a therapeutic target against cancer, and may be useful in the diagnosis, prognosis, and prediction of patient response to therapy (40). The ANXA6 protein exhibited a highly significant interaction in the PPI enrichment analysis (p-value <1.0e-16) with proteins involved in the innate immune response and regulation of inflammation, as well as processes such as signal transduction, membrane aggregation, phagocytosis, proliferation, differentiation, and apoptosis pathways. Notably, these interactions included ANXA1 and ANXA4 (41, 42).

Dual roles of ANXA6 and SNX9 in breast cancer progression. ANXA6 and SNX9 have emerged as key regulators of invasive and metastatic processes in breast cancer (Figure 9), particularly in the TNBC. ANXA6 is a Ca²⁺ dependent protein that belongs to a family of proteins with multiple membrane-related functions, such as spatiotemporal cholesterol homeostasis by direct interaction, membrane repair, formation of exosomes and endosomes, cell migration, and signal transduction pathways It can be detected in the extracellular matrix (ECM), microdomains in the cell membrane, in late endosomes, interacting with L-type Ca channels, Na+/Ca2 exchangers, and SERCA channels of the sarcoplasmic reticulum.

The interaction of AnxA6 with ECM proteins may affect the adhesive and migratory properties of cells. Translocation of ANXA6 to the plasma membrane, triggered by an increase in intracellular Ca2, provides scaffolding functions that lead to rearrangements of the lipid and actin cytoskeleton, likewise cholesterol homeostasis. Several studies show a decrease in cholesterol availability caused by direct interaction with ANXA6, disrupting signal transduction events, molecular trafficking, dynamics of focal adhesions, caveosome formation, integrin recruitment, FN-dependent cell migration, and Rac/Rho GTPase activation. This cellular cholesterol imbalance interferes with multiple Cav-1 and SNARE functions required for cell movement and can indicate that ANXA6 inhibits cell migration. On the other hand, the membrane binding of ANXA6 contributes to the inactivation of RTKs such as EGFR in cancer cell lines, including A431 epidermal carcinoma cells, HeLa, and head and neck cancer cell lines, and also participates in membrane fusion events interacting with S100A8 and S100A9 (34).

The interaction of ANXA6 with clathrin promotes vesicular transport during exocytosis and endocytosis, as well as in the uptake, propagation, and release of viruses. ANXA6 also plays a role in signal transduction by interacting with PKC, Ras GTPase, leading to the activation of p120, RasGAP, H-Ras, Raf-1 type proteins, and transcription factors such as p65 and NF-β, as well as with Rab GTPase activators, recognized mechanisms that lead to tumor growth, motility, and differentiation. In cancer, ANXA6 can act as a potential tumor suppressor and play a proinvasive role in TNBC (41). Loss of ANXA6 is associated with a highly proliferative phenotype and poor prognosis in basal-like tumors, whereas its overexpression may provide adaptive advantages by promoting drug resistance, metabolic reprogramming, and the secretion of extracellular vesicles enriched in lipids and cytokines

As for SNX9, it is a ubiquitously expressed member of the SNX family of proteins involved in membrane remodeling and endosomal sorting, which has a BAR domain (Bin-Amphiphysin-Rvs) linked to the stabilization or induction of membrane curvature during clathrin-mediated endocytosis (CME). They bind to plasma membrane phospholipids, AP2, and dynamin. They interact with membrane receptors such as GPCR and proteases to facilitate their endocytosis and also act as an adaptor and link phosphorylated pCXCR4 for CME (37). Recently, SNX9 was identified as a novel angiogenic factor involved in the recycling pathway of integrin β1. SNX9 acts as a multifunctional scaffold that integrates cellular processes such as RhoGTPase activity, endocytic trafficking, and actin remodeling via ROCK and N-WASP, all of which are related to cancer cell invasion and metastasis (35). Also, plays a role in promoting cell migration by inhibiting the RhoA-Rho-associated protein kinase pathway SNX9 is required for clathrin-independent endocytosis, but RhoGTPase-dependent, and can functionally substitute for the bona fide Rho GAP, GTPase regulator associated with focal adhesion kinase (GRAF1). Together, these mechanisms act by increasing the metastasis of cancer cells. SNX9 expression is modified in numerous tumors; overexpression promotes invasive plasticity and metastatic dissemination by modulating RhoA, Cdc42, and β1-integrin activity. Increased expression of SNX9 has been reported in metastatic human breast cancer compared to primary tumors. SNX9 expression in MDA-MB-231 breast cancer cells was shown to be necessary for their metastatic capacity in a chick embryo model and it is speculated that low expression in primary tumors could increase RhoA activation and confer greater aggressiveness to the tumor. In breast cancer cells, SNX9 binds the invadopodia marker TKS5 and negatively regulates invadopodia formation and function, also increasing MT1-MMP surface exposure, thereby facilitating extracellular matrix degradation.

Thus, both ANXA6 and SNX9 act as regulatory hubs at the membrane–cytoskeleton interface that can function as suppressors in early stages but acquire pro-tumorigenic properties in advanced disease or under therapeutic pressure, highlighting their potential value as contextual biomarkers and therapeutic targets in breast cancer.

Therefore, the positive co-expression of SCGB2A2 with hsa-mir-184 and hsa-mir-190b provides initial insight into its potential function in cancer. Based on our results, which link these miRNAs to the ANXA6 and SNX9 genes, we hypothesize that Mammaglobin-A participates in common cellular pathways. In normal physiology, it may be involved in fundamental processes such as vesicular transport, a function consistent with the known roles of ANXA6 and SNX9. In a tumor context, this involvement could be co-opted to facilitate cancer hallmarks such as invasion and metastasis. A key mechanistic question arising from this work is whether Mammaglobin-A directly interacts with the ANXA6-SNX9 protein network or shares a common regulatory scaffold. Future studies are warranted to investigate the existence of such direct protein-protein interactions. The possible collaboration among Mammaglobin (SCGB2A2), ANXA6, and SNX9 highlights an additional layer of functional relevance. Mammaglobin, primarily known for its association with breast tissue and carcinomas, could cooperate with ANXA6, a protein involved in membrane trafficking and signal transduction, as well as with SNX9, a key regulator of endocytosis and vesicular dynamics (35). Together, these interactions suggest that vesicular transport and proliferative pathways may represent converging points where these proteins act in concert. Such cooperative mechanisms could contribute to cellular reprogramming in cancer, linking post-transcriptional regulation with altered transport and growth processes. Another possibility is that Mammaglobin uses the systems controlled by SNX9 and ANXA6 to mobilize outside the cell.

Regarding the potential roles of hsa-miR-190b and hsa-miR-184 as modulators of SNX9 and ANXA6 gene expression in breast cancer, there is currently limited direct evidence linking these specific miRNAs to these genes in the context of the disease. However, both SNX9 and ANXA6 are involved in critical cellular processes, such as membrane trafficking and cell signaling, which are frequently disrupted in cancer, making them plausible targets for miRNA regulation (40).

Mammaglobin-A interacts with proteins of the same family, SCGB3A1, SCGB3A2, SCGB1, SCGB2A, and SCGB1D2, which are often implicated in the regulation of the immune response, inflammation, and epithelial cell signaling (6). Dysregulation of these proteins has been linked to cancer progression, particularly in breast and lung cancer (10, 43). Additionally, Mammaglobin-A could interact with mucin family proteins such as MUC1, MUCL1, and MUC7, (44) which promote cell proliferation, metastasis, and immune evasion (45). In addition, Mammaglobin could be related to the transcription factors ERBB2 (HER2) and ETV4. ERBB2 is a key oncogene in breast cancer, and its overexpression is linked to aggressive tumor growth (46). ETV4 is involved in promoting cancer cell invasion and metastasis, particularly in breast cancer (47).

The target genes of hsa-mir-184 and hsa-mir-190b are implicated in the control of the expression of several genes related to cancer pathways and processes, such as proliferation, membrane trafficking, apoptosis, and cell signaling.

Therefore, positive correlations were observed between the miRNAs hsa-mir-184, hsa-mir-190b, and SCGB2A2 mRNA, indicating that increased expression of these miRNAs is associated with elevated SCGB2A2 levels. This finding provides clues to elucidate the function of SCGB2A2, through the convergence in the cell membrane of pathways related to cell trafficking, signal transduction, and cell adhesion, which are altered in different types of cancer, including breast cancer.

The processing methods employed in this study enabled the identification of two types of miRNAs associated with various carcinomas, including breast cancer.

The Authors declare that they have no competing interests.

César Payán-Gómez and Sandra Ramírez-Clavijo conceived and designed the study, supervised the project, and revised the manuscript critically for important intellectual content. Juan Felipe Abella-Duque performed the bioinformatic analyses, interpreted the data, and drafted the initial version of the manuscript. Liliana López-Kleine contributed to the data analysis and helped with the interpretation of results. Rafael Parra-Medina participated in the conceptualization of the study, performed additional analyses, and contributed to the writing and review of the manuscript. All Authors read, revised, and approved the final version of the manuscript.

The research group sincerely thanks Méderi Hospital in Bogotá, Colombia, for funding this study and CIMED for their support.

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.