In-depth analysis unveils mutations within MCU genes in breast cancer

Our initial exploration sought to ascertain the frequency and nature of MCU mutations in BRCA, leveraging TCGA’s BRCA dataset. The cBioPortal dataset revealed that MCU genes exhibit mutations in 2% of cancer cases, as illustrated in Fig. 1A. Subsequently, we delved into the amplification patterns of highly invasive breast cancers compared to general breast cancers within the TCGA dataset (Fig. 1B). Expanding our investigation, we extracted data on MCU expression in BRCA patients from the TCGA database and depicted it using a waterfall plot, specifically highlighting the top 25 affected genes (Fig. 1C). To deepen our understanding of MCU mutations in BRCA, we scrutinized their associations with other commonly observed cancer progression genes such as PIK3CA, TP53, and CDH1 (Fig. 1D). A comprehensive assessment of the dependence of 45 breast cancer cell lines on MCU ensued, with the MCU dependence of these cell lines presented in a ranked fashion based on increased MCU dependence (Fig. 1E). Further examination involved evaluating MCU expression across various clinical stages, revealing higher MCU expression in tumor tissues compared to non-tumor tissues (Fig. 1F). The impact of MCU on overall survival and disease-free survival in BRCA patients was analyzed using Kaplan–Meier plots, indicating an association between high MCU performance and poor prognosis (Fig. 1G).

Fig. 1

The illustrates the frequency and functional enrichment analysis of MCU alterations in breast cancer. A Examination of diverse mutations in the MCU gene across various cancer types. B Utilizing cBioPortal cancer genomics analysis to determine the frequency of MCU gene alterations in different cancer types. C Fisher's exact test compared mutation frequencies between MCU-high and MCU-low groups, with the right panel displaying mutation types, driving mutation types, and respective groups. D Investigation into the relationship between MCU and the six highly mutated genes in BRCA, with mutation sites highlighted by red lines. E Significance of MCU dependency in 45 BRCA cell lines based on the CRISPR screen. F Violin plots depicting MCU gene expression from RNA-sequencing data. G Kaplan–Meier survival analysis illustrating overall survival (OS) based on low/high expression of MCU. ***p < 0.001

Assessing the link between MCU expression and clinicopathological parameters in BRCA

To investigate the correlation between MCU performance and clinicopathological parameters in Breast Cancer (BRCA), we conducted analyses utilizing the bc-GenExMiner dataset. Both DNA microarray (Fig. 2A) and RNA sequencing data (Fig. 2B) consistently affirmed the elevated representation of MCU mRNA in Estrogen Receptor-positive (ER +) cases (ER− > ER + , p < 0.0001). Moreover, in the DNA microarray database, MCU mRNA exhibited a significantly higher representation in the Progesterone Receptor-negative (PR−) group compared to the Progesterone Receptor-positive (PR +) group (PR −  > PR + , p < 0.001). In the same database, MCU mRNA was notably upregulated in the Human Epidermal Growth Factor Receptor 2-positive (HER2 +) group in contrast to the HER2-negative (HER−) group (HER− > HER + , p < 0.001). Analysis specific to Triple-Negative Breast Cancer (TNBC) revealed a significant association between elevated TNBC levels and increased MCU transcript levels in both DNA microarray and RNA sequencing data. Across various breast cancer subtypes, MCU expression was consistently lower in normal tissue compared to other subtypes. In summary, these findings collectively underscore the prognostic significance of clinicopathological parameters in breast cancer.

Fig. 2

Bee swarm representation of differential expression in breast cancer patients based on various classified parameters. A–B Illustrate MCU mRNA expression levels in breast cancer patients using bee swarm plots in DNA microarray datasets and RNA-sequencing datasets. (ER estrogen receptor, PR progesterone receptor, HER2 human epidermal growth factor receptor 2, TNBC triple-negative breast cancer)

To confirm the accuracy of the multicomponent analysis, we then explored the protein levels of MCU in breast cancer through the Human Protein Atlas (HPA) database, which showed higher levels in late stages than in early stages (Fig. 3A). Further validation was performed using 59 samples obtained from BRCA patients for Tissue Microarray (TMA) analysis. Immunohistochemistry (IHC) analysis of MCU demonstrated a significant increase in MCU levels as breast cancer advanced and became more malignant with increasing stages (Fig. 3B). The quantitative results from HPA data underscored a substantial elevation in MCU expression (Fig. 3C). Corresponding H-score response results demonstrated abundant MCU expression across benign and malignant tumors, as well as various cancer stages (Fig. 3D, E). Moreover, our investigation included the analysis of 21 breast cancer biopsy samples from the human biobank, affirming that MCU levels were notably higher in tumor tissues compared to non-tumor tissues (Fig. 3F). Subsequently, to illustrate MCU's involvement in BRCA progression, we conducted migration and invasion assays post-MCU knockdown in the MDA-MB-231 cell line (Fig. 3G). The findings revealed that MCU suppression impeded MDA-MB-231 breast cancer cell migration, as confirmed by the wound healing assay (Fig. 3H). Furthermore, MCU deficiency decelerated the invasion of breast cancer cells (Fig. 3I). These findings collectively suggest that MCU functions as a tumor promoter by facilitating BRCA cell migration and invasion. In order to delve deeper into the potential clinical significance of MCU, we examined its expression across various cancer types and assessed its prognostic relevance. Similar expression patterns of MCU were observed in multiple cancer types, including Breast Invasive Carcinoma (BRCA), Cholangiocarcinoma (CHOL), Esophageal Carcinoma (ESCA), Uterine Corpus Endometrial Carcinoma (UCEC), Kidney Renal Papillary Cell Carcinoma (KIRP), Liver Hepatocellular Carcinoma (LIHC), Lung Adenocarcinoma (LUAD), Pancreatic Adenocarcinoma (PAAD), Stomach Adenocarcinoma (STAD), Thyroid Carcinoma (THCA), and similar BRCA-like cancers (Fig. 3J).

Fig. 3

Examination of the pathological alterations of MCU in clinical breast cancer. A Immunohistochemical assessment of MCU protein expression levels in BRCA tissue samples from diverse patients based on the Human Protein Atlas. B Depicts representative images of MCU expression in breast cancer tissues at distinct staining stages. C MCU expression levels in breast cancer were evaluated in tumor and non-tumor tissues using data from the Human Protein Atlas. C MCU expression levels in breast cancer were depicted in benign and malignant violin plots. E Comparative analysis of MCU expression in BRCA, with box plots illustrating expression levels across different stages of the disease. F qPCR analysis of MCU in 21 paired BRCA and non-tumor tissues, denoted as N and T for non-tumor and tumor tissues, respectively. G Quantitative PCR demonstrates a significant decrease in MCU expression in breast cancer cells transfected with siMCU. H Assessment of wound healing in the MDA-MB-231 cell line through a wound healing assay. I Evaluation of breast cancer cell invasion using the MDA-MB-231 cell line in a Transwell assay. (J) MCU expression levels in different cancer types, with comparisons between tumor and normal tissues, highlighting statistical significance using asterisks. *p < 0.05, **p < 0.01, ***p < 0.001. Scale bar = 500 µm

Examining the MCU transcriptome through single-cell RNA sequencing libraries

To scrutinize the transcriptome of MCU at the single-cell level within gastric cancer and investigate the diversity of cell types within the gastric cancer microenvironment, we conducted an analysis utilizing an available single-cell RNA sequencing database from breast cancer (EMTAB8107) (Fig. 4A). This database originates from samples of human breast cancer patients and encompasses 11 distinct cell types. Following the elimination of batch effects and quality control, we employed UMAP plots to identify 11 crucial cell populations (Fig. 4B). Utilizing the highest differentially expressed genes in each cluster, we identified cell type-specific markers, crucial for subsequent cell type classification (Fig. 4C, D). Upon scrutinizing the expression and distribution of MCU in these single-cell RNA-sequencing databases, we observed heightened MCU expression in regions corresponding to the inflammatory response model (Fig. 4E). Furthermore, the activation of TGFβ Signaling, Interferon γ Response, and PI3K/AKT/mTOR Signaling pathways was generally observed (Fig. 4F–H). To explore the potential association between BRCA and MCU, we delved into the gene regulation of cell clusters by various transcription factors and pinpointed those strongly linked to CD8 T cells (Fig. 4I, J). It is important to note that T cells exhibit tissue residency and display either pro- or anti-inflammatory functions.

Fig. 4

Utilizing single-cell RNA sequencing analysis for the identification of immune cell populations. A–B Illustrate the relative proportions of each cell type in the public dataset and the integrated immune cell proportions in the EMTAB8107 dataset. C–D Employ the unified flow approximation and projection (UMAP) technique to visually represent BRCA cells, color-coded based on main cell types and malignancy. E Visualizes the expression clusters of MCU through UMAP plots, subsequently subjected to gene set enrichment analysis (GSEA) for F TGFβ, G interferon γ, and H PI3K/AKT/mTOR signaling. I The heatmap depicts the expression of the MCU gene and cell type biomarkers in different single-cell type clusters of tissues. J Highlights CD8Tex cells as key regulators of transcription factors in this cell cluster

MCU upregulation is associated with immune cell infiltration

MCU overexpression appears to be intricately linked to immune infiltration within the tumor microenvironment of Breast Cancer (BRCA), exhibiting a significant positive correlation with T cell CD8 + (rho = 0.344, p < 0.001; Fig. 5A). In-depth comparative analyses of immune-related functions between the MCU low expression group and the MCU high expression group were conducted. The outcomes indicated no statistically significant differences in TIL regulatory fraction (Fig. 5B). Additionally, a positive correlation was observed between MCU expression and TCR Shannon, TCR richness, Th1 cells, and Th2 cells, while a negative correlation was noted with Th17 cells (Fig. 5C–G). Notably, patients with both high MCU performance and T cell CD8 + infiltration exhibited shorter survival times compared to those with high gene expression alone (Fig. 5H). Further exploration of the correlation between MCU and T cell CD8 + in various breast cancer subtypes also revealed a positive association (Fig. 6A). The investigation extended to studying the T cell CD8 + biomarker, revealing a strong positive correlation between MCU and AHR, FOXP3, ID2, IL10, IL21, IRF4, TGFB1, and STAT4 (Fig. 6B, C). To ascertain the correlation between MCU and T cells in BRCA, we conducted quadruple immunofluorescence labeling, encompassing DAPI, MCU, FOXP3, and TGFb1, across entire BRCA sections in TMA samples. These samples were categorized into non-tumor, early-stage, and late-stage groups, and their fluorescence intensities were assessed using panoramic tissue scanning. As depicted in Fig. 6D, the fluorescence intensities of MCU, FOXP3, and TGFb1 were notably lower in the non-tumor group. Conversely, the late-stage group exhibited a heightened fluorescence overlap ratio of MCU/FOXP3. Similarly, the fluorescence overlap ratio of MCU/TGFb1 was elevated in the late-stage group compared to both the non-tumor and early-stage groups (Fig. 6E).

Fig. 5

Correlation of MCU expression with immune infiltration level in BRCA. A The correlation between MCU expression level and immune infiltration. B–G Displays changes in transcript levels among different MCU levels and immune cells. H Kaplan–Meier plots illustrating the survival differences of macrophages based on different MCU expression levels.*P < 0.05, **P < 0.01, ***P < 0.001

Fig. 6

Investigating the association between MCU and immune infiltration in BRCA. A Assessing the relationship between MCU levels and T cell CD8 + . B–C Investigating the correlation between MCU and genes related to T cell CD8 + . D Investigating the association between MCU and T cell biomarkers in breast cancer biopsies. Utilizing comprehensive immunofluorescent labeling on tissue microarrays (TMA) with DAPI, MCU, FOXP3, and TGFb1, followed by panoramic tissue scanning. Pearson's correlation coefficient was employed to depict the degree of co-localization between MCU and FOXP3 E as well as TGFb1 F fluorescent signals. **P < 0.01, ***P < 0.001

The clinical and pathological significance of MCU and its prognostic relevance in patients

To map transcriptomic signatures onto H&E-stained histological sections of human breast cancer tumors (PMID: STDS0000049) (Fig. 7A), we employed spatial transcriptomics techniques. This innovative approach utilizes the sequencing of spatially localized barcodes to directly correlate transcriptomic signatures with histological images. Our analysis encompassed a total of 2518 counts, each possessing its unique expression signature, superimposed on local barcode-based tumor histology images. Employing an unsupervised clustering approach, we categorized points based on their gene expression, with each cluster denoting a specific cell type identified through known marker genes and underlying histology. Comparing Mitochondrial Calcium Uniporter (MCU) with other genes associated with T cell regulation, MCU exhibited higher expression levels in tumor areas compared to non-tumor areas (Fig. 7B). The analysis further revealed that MCU, alongside genes commonly mutated in breast cancer, demonstrated elevated expression levels comparable to key T cell regulators such as AHR, ID2, and TBFb1 (Fig. 7B). MCU expression was notably concentrated in the tumor areas across all 15 clusters, exhibiting highly significant differences in expression, second only to TBFb1. Additionally, the expression levels of these three genes in the high-magnification area were examined to emphasize histological features (Fig. 7C). The Space Ranger algorithm generated 15 unsupervised clusters aligning with known marker genes for each tumor microenvironment cell, superimposed on histological features. Dot plots illustrate the normalized, log-transformed, and variance-scaled expression of various cell clusters (y-axis) and signature genes (x-axis) in Breast Cancer (BRCA) single-cell RNA sequencing data (Fig. 7D).

Fig. 7

Gene expression in BRCA defined by spatial transcriptomics. A Utilizing spatial transcriptomics, tissue sections were analyzed to identify clusters, accurately aligning them with morphological features observed in hematoxylin and eosin staining and cluster mapping. Malignant areas are outlined by yellow dotted lines, and magnified images of gene variants in boxes i and ii are presented. B–C Spatial distribution and genetic changes involving MCU, AHR, FOXP3, ID2, IL10, IL21, IRF4, TGFb1, and STAT4 in different tissue sections were visualized using 10 × Visium Spatial Gene Expression. Bars indicate tumor versus non-tumor transcript levels. D Dot plots illustrate gene expression levels across various clusters. *P < 0.05

CellChat analysis unveiled a continuum of crucial signals associated with cell lineage

To delve deeper, we meticulously identified key senders, receivers, mediators, and influencers within the five signaling networks in cells. This was achieved by computing various network centrality scores for each cell group (Fig. 8A–E). Our focus was on the top five signaling pathways (TGFb1, CXCL, CCL, IL16, MIF) that play pivotal roles in intercellular signaling between different cell populations. It is noteworthy that the CD8 + T-Nai cell cluster emerged as the most significant regulator among these five signal types, with the T-Exh tumor cluster standing out as the primary influencer and recipient of these signals. Specifically, within the TGFb1 signaling pathway, the primary senders and influencers were identified as the B-Nai, CD8 + T-Nai, Macrophage, and T-Exh clusters. Simultaneously, the critical signal receivers were recognized as the B-Nai and Macrophage clusters.

Fig. 8

Signaling pathway network associated with T cell CD8 + in the spatial transcriptome of BRCA. A–E Showcase interactions of the TGFβ, CXCL, CCL, IL16, and MIF signaling pathways with different ovarian cell types. Cell–cell communication is depicted, illustrating interactions between cells, where line thickness reflects the strength of these connections. Chord diagrams for six signaling pathways reveal intricate interactions between cell populations. Heatmaps display network centrality scores for the signaling pathways

Pharmacogenomic prediction of effective drugs targeting MCU

To identify potential therapeutic agents targeting Breast Cancer (BRCA), we conducted a thorough analysis of drug responses and the effects of Mitochondrial Calcium Uniporter (MCU) knockdown using shRNA in BRCA cells. A total of 489 drugs were scrutinized, revealing four compounds that displayed significant alterations in potency (Fig. 9A). Notably, BRCA cell lines with low shMCU efficiency exhibited heightened responsiveness to NSC319126 (Fig. 9B), RU-SKI 43 (Fig. 9C), OSI-930 (Fig. 9D), and MG-132 (Fig. 9E). These findings collectively suggest that these drugs hold promise as potential anticancer agents targeting MCU to modulate the growth of BRCA. Further investigations are warranted to explore the therapeutic potential of these drugs and their specific mechanisms of action in the context of BRCA. In the final phase of our investigation, we conducted an intricate analysis to evaluate the response of diverse breast cancer cell lines to the administered drug regimens, meticulously determining the IC50 values post-administration (Fig. 9F). This comprehensive assessment revealed notable inhibitory effects on cell proliferation elicited by both drugs. Furthermore, upon specifically targeting the MCU gene in MCF7 and MDA-MB-231 cells, we observed a striking increase in cancer cell viability subsequent to treatment with the same drug dosage (Fig. 9G). These findings underscore the intricate interplay between drug sensitivity and MCU gene expression in breast cancer cells, shedding light on potential mechanisms underlying treatment response variations among different breast cancer subtypes (Additional file 1).

Fig. 9

Evaluation of drug sensitivity and cytotoxicity in breast cancer cells. A The scatter plot depicts cross-association scores of predictivity and descriptivity, employed to identify potent drugs with efficacy against BRCA cells. To unveil gene signatures and potential drugs, we queried the pharmacogenetics database for the MCU gene. Subsequently, we examined the drug sensitivity of the shMCU gene to various chemical drugs in BRCA cell lines. The boxplots B–E illustrate the logarithm of the half maximal inhibitory concentration (IC50) values for four drugs, namely NSC319126, RU-SKI 43, OSI-930, and MG-132, displaying altered potency. F Evaluation of drug responsiveness across various breast cancer cell lines. G Analysis of drug sensitivity following inactivation of the MCU gene. **P < 0.01, ***P < 0.001