Abstract
Purpose
This study aims to explore the causal association between members of the cathepsin family and breast cancer subtypes, with a focus on their expression profiles, prognostic significance, and potential molecular mechanisms in HER2-positive and HER2-negative breast cancer. Additionally, the study includes preliminary in vitro cell experiments to validate the functional relevance of key genes.
Methods
This study employs the two-sample Mendelian randomization (MR) method to analyze the causal relationship between 10 members of the cathepsin family (CTSB, CTSE, CTSF, CTSG, CTSH, CTSL1, CTSL2, CTSO, CTSS, CTSZ) and breast cancer risk. Sensitivity analyses, including MR-Egger, Cochran’s Q test, and MR-PRESSO, are used to validate the robustness of the results. Additionally, the study integrates differential gene expression analysis, prognostic impact analysis, single-cell analysis, protein-protein interaction (PPI) network construction, and targeted cell experiments (qRT-PCR for gene expression, CTSO overexpression model construction, CCK-8 proliferation assay, and wound healing migration assay) to systematically investigate the role of cathepsins in breast cancer.
Results
MR analysis identified significant associations between CTSE and CTSO expression with breast cancer risk. Elevated CTSE expression was positively associated with both HER2-positive (OR = 1.108, P = 0.022) and HER2-negative (OR = 1.099, P = 0.009) breast cancer risk. In contrast, CTSO expression was inversely associated with HER2-negative breast cancer risk (OR = 0.913, P = 0.037). Differential expression analysis confirmed that CTSE was overexpressed in HER2-positive tumors, while CTSO was underexpressed across breast cancer subtypes. Prognostic analysis showed that high CTSE expression was linked to a favorable prognosis in HER2-positive breast cancer patients treated with chemotherapy, whereas high CTSO expression correlated with improved prognosis in HER2-negative patients. Single-cell analysis revealed that CTSO was highly expressed in immune cells and fibroblasts, while CTSE was mainly localized to cancerous epithelial cells. PPI and functional enrichment analyses suggested that CTSE is involved in lysosomal and protein digestion pathways, whereas CTSO is implicated in immune regulation and antigen processing. Cellular experiments revealed that qRT-PCR did not detect quantifiable levels of CTSE in breast cancer cell lines, while CTSO exhibited low expression across all breast cancer cell lines. After overexpressing CTSO in MDA-MB-231 cells, CCK-8 assays showed a significant reduction in cell proliferation, and wound healing assays demonstrated a marked inhibition of cell migration.
Conclusion
This study provides multi-layered evidence for the association between members of the cathepsin family (particularly CTSE and CTSO) and breast cancer subtypes through MR analysis, bioinformatics validation, and cellular experiments. The findings highlight the potential of these markers as subtype-specific biomarkers for precision diagnosis, treatment, and prognosis evaluation in breast cancer. These results deepen our understanding of the biological mechanisms underlying breast cancer and offer both experimental and theoretical support for the development of novel therapeutic strategies targeting the cathepsin family.
Supplementary Information
The online version contains supplementary material available at 10.1007/s12672-025-03787-3.
Keywords: Cathepsins, Mendelian randomization, Breast cancer subtypes, Precision therapy
Introduction
Breast cancer is the most common malignancy among women worldwide, with approximately 2.3 million new cases and 685,000 deaths in 2020. Among the various subtypes of breast cancer, HER2-positive and HER2-negative cancers are the most representative, exhibiting significant differences in molecular characteristics, prognosis, and treatment responses, with distinct therapeutic needs [1, 2]. HER2-positive breast cancer, characterized by amplification or overexpression of the HER2 gene, is typically more aggressive and responds well to targeted therapies. However, despite the significant improvements in prognosis due to targeted treatments, challenges remain, particularly in metastatic disease [3]. In contrast, HER2-negative breast cancer involves more complex biological mechanisms, with considerable variability in treatment response across subtypes, such as triple-negative breast cancer (TNBC) [4]. Therefore, an in-depth understanding of the molecular characteristics of both HER2-positive and HER2-negative breast cancers is crucial not only for advancing our comprehension of the biological basis of the disease but also for providing a theoretical framework for precision therapy.
In recent years, the Cathepsin family, a group of cysteine proteases found in lysosomes, has gained attention in cancer research due to its pivotal role in tumor initiation, progression, and metastasis. Cathepsins influence cancer cell proliferation, migration, invasion, and drug resistance through mechanisms such as extracellular matrix degradation, tumor microenvironment remodeling, and immune evasion, thereby promoting or inhibiting various types of cancer [5–7]. Studies have shown that the abnormal expression of Cathepsins family members in breast cancer is closely associated with tumor proliferation, invasion, and metastasis [8–11]. For example, CST6 inhibits breast cancer bone metastasis by suppressing CTSB activity and disrupting osteoclast maturation [12]. Furthermore, CTSL is involved in the proliferation, invasion, and metastasis of TNBC [13], while CTSV promotes tumor cell proliferation in ER-positive breast cancer, especially in the Luminal A subtype, by regulating histone H3 and H4 stability and the G2/M cell cycle transition [14]. However, the roles of different Cathepsin proteases in various breast cancer subtypes may differ significantly, and studies examining the causal relationships and underlying molecular mechanisms of Cathepsins in HER2-positive versus HER2-negative breast cancer remain limited. Therefore, further investigation into the causal relationship between Cathepsins family members and breast cancer subtype risk is crucial.
Traditional observational studies often rely on correlation analysis to explore the relationship between genes and disease; however, such studies are susceptible to confounding factors and reverse causality, making causal inference difficult. Mendelian randomization (MR), a genetic variant-based instrumental variable method, effectively reduces confounding bias and provides more reliable evidence for causal inference [15]. MR analysis allows for the exploration of causal relationships between the expression of Cathepsins family members and breast cancer risk from a genetic perspective, thereby elucidating the roles of these genes across different breast cancer subtypes.
This study integrates MR, bioinformatics analysis, and targeted cellular experiments to systematically explore the causal relationship between cathepsin family members and breast cancer subtypes. It further examines their expression profiles, prognostic significance, and potential molecular pathways (Fig. 1). The cellular experiments, through gene expression analysis, the construction of a core gene (CTSO) overexpression model, and proliferation/migration phenotype validation, provide direct functional evidence supporting the molecular mechanisms predicted by bioinformatics. The combination of MR and bioinformatics analysis effectively mitigates the confounding biases inherent in traditional observational studies. Through this innovative, multidimensional approach, the study aims to provide a theoretical foundation and experimental evidence for molecular subtyping and precision treatment of breast cancer, while offering new insights into the potential biological functions of the cathepsin family in breast cancer.
Fig. 1.
Flowchart of the study design
Materials and methods
Mendelian randomization
Study design
To explore the potential causal relationship between breast cancer and ten Cathepsin family members (including B, E, F, G, H, L1, L2, O, S, Z), this study employed a two-sample MR approach. MR utilizes single nucleotide polymorphisms (SNPs) as instrumental variables to assess causal relationships between exposure factors (Cathepsins) and outcomes (breast cancer). The key advantage of MR is that genetic variations, as naturally occurring instrumental variables, can effectively avoid confounding factors common in traditional observational studies, providing more reliable evidence for causal inference. To ensure the validity of the selected instrumental variables, the following three assumptions must be met: (1) Relevance assumption: the instrumental variable must be significantly correlated with the exposure; (2) Independence assumption: the instrumental variable must be independent of all potential confounders; and (3) Exclusion restriction assumption: the instrumental variable must affect the outcome (breast cancer) only through the exposure factor, and not through any other pathways. All data were obtained from publicly available GWAS summary statistics, and ethical approval was obtained from the respective institutions; thus, no additional ethical approval was required for this study.
Data sources
The Cathepsins-related GWAS data were sourced from the publicly accessible IEU Open GWAS platform (https://gwas.mrcieu.ac.uk/). This platform provides a range of summary statistics from large-scale genome-wide association studies (GWAS), offering rich genetic data for this study. Specifically, the Cathepsins data were derived from the INTERVAL study, which conducted comprehensive genomic analysis of plasma protein levels in 3,301 European individuals, covering 10.6 million common autosomal variants [16]. Additionally, we used data from GCST90012073, which specifically targets the expression of Cathepsin L. The breast cancer GWAS data were also obtained from the IEU Open GWAS platform. HER2-positive breast cancer data came from the finn-b-C3_BREAST_HERPLUS_EXALLC dataset, which includes 103,530 samples (4,263 cases and 99,267 controls), and HER2-negative breast cancer data were sourced from the finn-b-C3_BREAST_HER2NEG_EXALLC dataset, with 106,676 samples (7,355 cases and 99,321 controls). Given the minimal overlap in the samples across different datasets, potential confounding biases are limited.
Selection of instrumental variables
To ensure the effectiveness of the instrumental variables, SNPs significantly associated with the exposure factor (Cathepsins) were selected. SNPs with a genome-wide significance threshold of P < 5 × 10⁻⁸ were initially chosen. Next, to account for potential linkage disequilibrium, SNPs with an R² value < 0.001 within a 10,000 kb window were removed. Additionally, palindromic variants with incompatible alleles were excluded to avoid erroneous instrument selection. To further mitigate potential confounding effects, all SNPs were screened through the Phenoscanner V2 database to ensure that the selected SNPs were not influenced by other factors. The validity of the selected instrumental variables was assessed by calculating their F-statistic. An F-statistic greater than 10 indicates a valid instrumental variable, and SNPs with an F-statistic < 10 were excluded to ensure the robustness of the results [17].
Mendelian randomization analysis
MR analysis was conducted using the TwoSampleMR R package, with several common methods, including Inverse Variance Weighted (IVW), MR-Egger regression, and Weighted Median [18–20]. The IVW method was used as the primary analysis to estimate the overall effect of the exposure on breast cancer risk. A P-value < 0.05 was considered statistically significant. MR-Egger regression provides robust effect estimates independent of instrument strength and allows for the assessment of potential horizontal pleiotropy. The Weighted Median method is suitable when some instrumental variables may be biased, providing valid estimates even under such conditions. Causal relationships were evaluated using odds ratios (OR) and their 95% confidence intervals (CIs). An OR >1 indicates that the exposure is a risk factor for breast cancer, while an OR < 1 suggests a protective effect.
Sensitivity analysis
To validate the reliability of the instrumental variables and the robustness of the analysis, several sensitivity analyses were performed. First, Cochran’s Q test was used to assess heterogeneity among the SNPs; if significant heterogeneity (P < 0.05) was found, a random-effects model was applied; otherwise, a fixed-effects model was used. The presence of horizontal pleiotropy was evaluated using MR-Egger intercept tests. A P-value < 0.05 indicates potential pleiotropic effects that should be considered in subsequent analyses. To further enhance the robustness of the findings, MR-PRESSO was used to detect and correct for horizontal pleiotropy, especially by removing potential outliers. Lastly, leave-one-out analyses and funnel plots were conducted to identify any single SNPs with extreme effects, ensuring the reliability of the final results [21, 22].
Statistical analysis
All statistical analyses were performed using R software (version 4.2.1) with the following R packages: TwoSampleMR (version 0.5.9), MRPRESSO (version 1.0), and ggplot2 (version 3.4.4). A significance level of P < 0.05 was used for all analyses.
Bioinformatics analysis
Differential expression analysis of genes
We first employed the SangerBox 3.0 (http://sangerbox.com/home.html) online tool to analyze the differential expression of Cathepsins family genes between breast cancer and normal tissues. To further explore the expression patterns of these genes across different breast cancer subtypes, we utilized the NULCAN tool [23, 24]. Additionally, we used the bc-GenExMiner platform to analyze gene expression in HER2-high and HER2-low expressing breast cancers [25].
Prognostic analysis
Prognostic analysis was performed using the Kaplan-Meier Plotter online tool to evaluate the prognostic impact of CTSE and CTSO in both HER2-positive and HER2-negative breast cancer patients [26, 27]. Specifically, we assessed the effects of these genes on patient survival under different treatment modalities (e.g., chemotherapy and hormone therapy) to further explore their potential as prognostic biomarkers.
Single-cell analysis
Single-cell analysis was conducted using the Single Cell Portal platform (https://singlecell.broadinstitute.org/single_cell#), utilizing the dataset titled “A Single-Cell and Spatially Resolved Atlas of Human Breast Cancers.” The primary objective of this analysis was to examine the expression patterns of CTSE and CTSO at the single-cell level in breast cancer tissues. This approach allows for a deeper exploration of the heterogeneity in gene expression within individual tumor cells, providing new insights into the molecular mechanisms underlying breast cancer initiation and progression.
PPI network construction and functional enrichment analysis
The protein-protein interaction (PPI) network was constructed using the GeneMANIA tool to reveal interactions between Cathepsins family members and other proteins. This network was further analyzed through Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses using the ggplot2 R package (version 3.4.4) to explore the molecular functions, biological processes, and potential signaling pathways involved in breast cancer.
Experimental validation
Selection and culture of breast cancer cell lines
To investigate the mRNA expression of CTSE in both HER2-positive and HER2-negative breast cancer cell lines, and to characterize the expression of CTSO in HER2-negative breast cancer cell lines, we selected four breast cancer cell lines representing different cancer subtypes. HER2-positive cell lines included SK-BR-3 and BT-474, while HER2-negative cell lines included MDA-MB-231 and MCF-7. The normal control cell line was MCF-10 A, a human immortalized non-tumorigenic breast epithelial cell line. All cell lines were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China).
The breast cancer cell lines were cultured in RPMI-1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin-streptomycin (Gibco, USA). The MCF-10 A cells were cultured in specialized complete medium containing DMEM/F12 (Gibco, USA) as the base medium, supplemented with 5% horse serum (Gibco, USA), 10 ng/mL epidermal growth factor (EGF, Sigma, USA), 10 µg/mL insulin (Sigma, USA), 0.5 µg/mL hydrocortisone (Sigma, USA), and 100 ng/mL cholera toxin (Sigma, USA), and 1% penicillin-streptomycin. All cells were maintained in a 37 °C, 5% CO₂ humidified incubator. Cells in the logarithmic phase were used for experiments to ensure consistency in cell activity.
Quantitative reverse transcription PCR (qRT-PCR)
RNA extraction was performed using TRIzol reagent (Invitrogen, USA) according to the manufacturer’s protocol. For each sample, 1 × 10⁶ cells were collected, washed twice with pre-chilled PBS, and lysed with 1 mL of TRIzol reagent for 5 min at room temperature. Next, 200 µL chloroform was added, and the mixture was vigorously vortexed for 15 s and incubated at room temperature for 3 min. The mixture was then centrifuged at 12,000×g for 15 min at 4 °C. The aqueous phase was carefully transferred to a new tube, and 500 µL isopropanol was added. After gentle mixing, the mixture was incubated at -20 °C for 30 min and then centrifuged at 12,000×g for 10 min at 4 °C. The supernatant was discarded, and the RNA pellet was washed with 1 mL of 75% ethanol (prepared with DEPC-treated water) and centrifuged at 7,500×g for 5 min at 4 °C. The RNA pellet was air-dried at room temperature for 5–10 min and dissolved in 20 µL DEPC-treated water. RNA concentration and purity were measured using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). A260/A280 ratios between 1.8 and 2.0 were considered acceptable.
Reverse transcription was performed using the PrimeScript™ RT Reagent Kit with gDNA Eraser (Takara, Japan) to remove genomic DNA contamination and convert 1 µg of total RNA into complementary DNA (cDNA). The reaction mixture was 20 µL, containing 4 µL 5×PrimeScript Buffer, 1 µL PrimeScript RT Enzyme Mix, 1 µL gDNA Eraser, 1 µg total RNA, and DEPC-treated water to reach the final volume. The reaction conditions were as follows: 42 °C for 15 min for reverse transcription, followed by 85 °C for 5 s for enzyme inactivation. The cDNA was stored at 4 °C for later use.
Real-time quantitative PCR (qPCR) was performed using the TB Green® Premix Ex Taq™ II Kit (Takara, Japan) on a StepOnePlus Real-Time PCR System (Thermo Fisher Scientific, USA). CTSE expression was assessed in all four breast cancer cell lines and MCF-10 A, while CTSO expression was analyzed in the HER2-negative breast cancer cell lines and MCF-10 A. Primers were designed using PrimerBLAST to target the conserved regions of CTSE, CTSO, and the reference gene GAPDH (for normalization of mRNA levels) (primer sequences are listed in Supplementary Table 1). Each reaction was carried out in a 20 µL volume containing 10 µL of 2×TB Green Premix Ex Taq II, 0.4 µL of each forward and reverse primer (10 µM), 2 µL of cDNA template (1:10 dilution), 0.4 µL ROX Reference Dye II, and 6.8 µL of distilled water. The reaction conditions were: 95 °C for 30 s for pre-denaturation, followed by 40 cycles of 95 °C for 5 s (denaturation) and 60 °C for 30 s (annealing/extension). A melting curve analysis was performed (95 °C for 15 s, 60 °C for 1 min, and 95 °C for 15 s) to verify primer specificity. A single melting peak confirmed the absence of non-specific amplification. Each sample was run in triplicate, and relative expression levels were calculated using the 2^-ΔΔCt method, where ΔCt = Ct of the target gene - Ct of GAPDH, and ΔΔCt = ΔCt of the experimental group - ΔCt of the control group.
Construction and transfection of CTSO overexpression vector
To investigate the effect of CTSO overexpression on HER2-negative breast cancer cell function, we first constructed the CTSO overexpression vector. The full-length coding sequence (CDS) of human CTSO was amplified from human breast tissue cDNA (Clontech, USA). The PCR product was purified using a gel extraction kit (Axygen, USA) and then ligated into the pcDNA3.1(+) eukaryotic expression vector (Invitrogen, USA) using EcoRI/XhoI double digestion. The recombinant vector (pcDNA3.1-CTSO) and the empty vector (pcDNA3.1-NC, used as a negative control) were verified by Sanger sequencing (Sangon Biotech, Shanghai) to ensure that there were no base mutations or frame shifts, ensuring the accuracy of subsequent experimental interventions.
Cell Transfection: MDA-MB-231 cells (HER2-negative, which exhibited relatively high endogenous CTSO expression in preliminary screening) were chosen for transfection to observe the overexpression effects. The cells were seeded at a density of 2 × 10⁵ cells/well in a 6-well plate. When the cells reached 70–80% confluence, transfection was performed using Lipofectamine 3000 reagent (Invitrogen, USA) according to the manufacturer’s instructions. Briefly, 2 µg of pcDNA3.1-CTSO (or pcDNA3.1-NC) was added to 250 µL Opti-MEM medium (Gibco, USA) in one 1.5 mL centrifuge tube. In a second tube, 5 µL Lipofectamine 3000 was mixed with 250 µL Opti-MEM medium. Both solutions were gently mixed and incubated at room temperature for 15 min to form transfection complexes. The complexes were added to the cells in the 6-well plate, and after 6 h, the medium was replaced with fresh complete culture medium. After 48 h of transfection, qPCR was performed to assess CTSO mRNA expression levels. The overexpression efficiency was considered successful if the expression was ≥ 5 times higher than that of the control group, ensuring the effectiveness of subsequent functional experiments.
Cell proliferation assay (CCK-8 method)
To assess the effect of CTSO overexpression on the proliferation of HER2-negative breast cancer cells, transfected MDA-MB-231 cells (divided into pcDNA3.1-CTSO and pcDNA3.1-NC groups) were seeded at a density of 5 × 10³ cells/well in a 96-well plate, with 5 replicates per group to minimize experimental error. Proliferation was measured at 24, 48, and 72 h after seeding. At each time point, 10 µL of CCK-8 reagent (Beyotime, China) was added to each well, followed by incubation at 37 °C for 2 h. Absorbance at 450 nm was measured using a microplate reader (Thermo Fisher Scientific, USA). A cell proliferation curve was plotted with time on the x-axis and absorbance (OD value) on the y-axis.
Cell migration assay
To examine the effect of CTSO overexpression on the migration capacity of HER2-negative breast cancer cells, we performed a wound-healing assay, which simulates the process of in vitro wound repair and reflects the cell migration ability. MDA-MB-231 cells, transfected for 48 h, were divided into two groups: oe-CTSO (CTSO overexpression) and oe-NC (empty vector control). The procedure was as follows:
Cells were seeded at a density of 2 × 10⁵ cells/well in a 6-well plate and cultured at 37 °C, 5% CO₂ until they reached 90% confluence. At this point, the cells were ready for the wound-healing assay. Using a sterile P200 pipette tip, three parallel lines were gently scratched in each well (spacing ≥ 5 mm between lines). The scratched areas were washed three times with serum-free RPMI-1640 medium to remove detached cells and debris, minimizing interference in subsequent observations.
After washing, the cells were cultured in RPMI-1640 medium with 1% FBS (to minimize cell proliferation) and returned to the incubator. Images of the wound areas were captured at 0 h (immediately after scratching) and 24 h (after 24 h of culture) using an optical microscope (Olympus, Japan). For each group, 5 random non-overlapping fields were selected to ensure data representativeness.
The images were analyzed using ImageJ software to quantify the migration rate, which served as an indicator of cell migration capacity.
Statistical analysis
To ensure the reliability and statistical rigor of the experimental results, all experiments were independently repeated three times, with data presented as the mean ± standard deviation (x ± s). Statistical analysis and graphing were performed using GraphPad Prism 9.0 software (GraphPad, USA). Between-group comparisons were conducted using the independent t-test, while comparisons among multiple groups were performed using one-way ANOVA. Differences were considered statistically significant at P < 0.05, ensuring the validity of the experimental conclusions.
Results
Selection of instrumental variables
To ensure the robustness and reliability of our study, a rigorous selection process for genetic instrumental variables was conducted. A total of 155 SNPs associated with Cathepsins were selected, with F-statistics ranging from 20.81 to 1098.94. Since all F-values exceeded the threshold of 10, this confirms that the instrumental variables in this study have sufficient strength, minimizing the risk of significant bias from weak instruments.
MR analysis
To evaluate the causal effects of ten Cathepsins on the two breast cancer subtypes, we performed three different MR methods: IVW, MR-Egger regression, and Weighted Median. The IVW method was used as the primary analysis for the two-sample MR approach. The results revealed that, in assessing the causal relationships between CTSE and CTSO with breast cancer, 10 and 12 SNPs were included as instrumental variables, respectively. IVW analysis showed significant associations (P < 0.05) for both genes, with consistent causal effect directions across all three methods, further confirming these associations (Table 1). Specifically, CTSE expression was positively associated with both HER2-positive (OR = 1.108, 95% CI 1.015–1.211, P = 0.022) and HER2-negative (OR = 1.099, 95% CI 1.024–1.179, P = 0.009) breast cancer risk. In contrast, CTSO expression was significantly associated with a reduced risk of HER2-negative breast cancer (OR = 0.913, 95% CI 0.838–0.995, P = 0.037) (Fig. 2).
Table 1.
Mendelian randomization analysis of the causal relationship between 10 cathepsin genes and thyroid cancer
| Exposure | Nsnp | Outcome | Method | Or | Or_lci95 | Or_uci95 | P value |
|---|---|---|---|---|---|---|---|
| Cathepsin B | 17 | HER-positive | IVW | 0.951740333 | 0.860674209 | 1.052441972 | 0.335085254 |
| MR Egger | 1.026812949 | 0.80934388 | 1.302715519 | 0.830434238 | |||
| Weighted median | 0.989877613 | 0.876113589 | 1.118413984 | 0.870253848 | |||
| HER2-negative | IVW | 1.006440482 | 0.927388497 | 1.092230976 | 0.877751659 | ||
| MR Egger | 1.023709884 | 0.841205656 | 1.245809417 | 0.818216316 | |||
| Weighted median | 0.998181641 | 0.90769951 | 1.097683293 | 0.970053545 | |||
| Cathepsin E | 10 | HER-positive | MR Egger | 1.245961245 | 1.074131214 | 1.445279127 | 0.019753996 |
| IVW | 1.108326098 | 1.014614085 | 1.210693561 | 0.022496005 | |||
| Weighted median | 1.11126036 | 0.979952797 | 1.260162317 | 0.100104063 | |||
| HER2-negative | IVW | 1.098828366 | 1.024047117 | 1.17907053 | 0.008772334 | ||
| MR Egger | 1.179511149 | 1.047907469 | 1.327642557 | 0.025635817 | |||
| Weighted median | 1.102588287 | 1.000302403 | 1.215333409 | 0.04928856 | |||
| Cathepsin F | 12 | HER-positive | MR Egger | 1.261513272 | 1.025586316 | 1.551713113 | 0.052515309 |
| IVW | 1.076503365 | 0.978842676 | 1.183907817 | 0.128691352 | |||
| Weighted median | 1.068047437 | 0.948762157 | 1.20233013 | 0.275924854 | |||
| HER2-negative | IVW | 1.063722754 | 0.966753345 | 1.1704186 | 0.205266545 | ||
| Weighted median | 1.041118561 | 0.943180744 | 1.149226026 | 0.424032799 | |||
| MR Egger | 1.065619579 | 0.84242965 | 1.347940551 | 0.607652188 | |||
| Cathepsin G | 13 | HER-positive | MR Egger | 0.856039132 | 0.713674293 | 1.026803126 | 0.12210011 |
| Weighted median | 0.979160848 | 0.85401647 | 1.122643416 | 0.762768001 | |||
| IVW | 0.991818585 | 0.903158859 | 1.089181705 | 0.863478255 | |||
| HER2-negative | MR Egger | 0.86564269 | 0.748314266 | 1.001367075 | 0.07822617 | ||
| Weighted median | 0.947147138 | 0.846653997 | 1.059568259 | 0.342677073 | |||
| IVW | 0.978420556 | 0.904555119 | 1.058317801 | 0.58594427 | |||
| Cathepsin H | 11 | HER-positive | IVW | 0.990817533 | 0.940287023 | 1.044063525 | 0.729783109 |
| MR Egger | 0.988509968 | 0.917254551 | 1.065300746 | 0.768951216 | |||
| Weighted median | 0.997143371 | 0.941394912 | 1.056193198 | 0.922361761 | |||
| HER2-negative | IVW | 0.983650139 | 0.936656649 | 1.033001364 | 0.509237099 | ||
| Weighted median | 0.988889879 | 0.944518085 | 1.035346182 | 0.633369546 | |||
| MR Egger | 0.988190763 | 0.920674399 | 1.060658345 | 0.749674561 | |||
| Cathepsin L1 | 30 | HER-positive | MR Egger | 1.187298457 | 0.904954697 | 1.557732815 | 0.225574813 |
| IVW | 0.98819866 | 0.876948885 | 1.113561587 | 0.84553478 | |||
| Weighted median | 1.017183353 | 0.853710422 | 1.211958934 | 0.848838793 | |||
| HER2-negative | MR Egger | 1.238366155 | 0.977404219 | 1.569003595 | 0.087497091 | ||
| IVW | 0.94711909 | 0.84600313 | 1.060320628 | 0.345584497 | |||
| Weighted median | 0.933112582 | 0.80013782 | 1.088186398 | 0.377462686 | |||
| Cathepsin L2 | 11 | HER-positive | IVW | 1.028464871 | 0.87999469 | 1.201984516 | 0.724200004 |
| MR Egger | 0.851960791 | 0.555736891 | 1.30608063 | 0.481058049 | |||
| Weighted median | 1.099516091 | 0.920284607 | 1.313654087 | 0.296034277 | |||
| HER2-negative | Weighted median | 1.070612707 | 0.933242396 | 1.22820349 | 0.330121947 | ||
| MR Egger | 0.855859808 | 0.634909777 | 1.153700946 | 0.333648468 | |||
| IVW | 1.050585792 | 0.93596209 | 1.179247021 | 0.402471998 | |||
| Cathepsin O | 12 | HER-positive | IVW | 0.91144026 | 0.818833012 | 1.014521075 | 0.089833095 |
| MR Egger | 0.869143186 | 0.678419943 | 1.113484186 | 0.293156284 | |||
| Weighted median | 0.925248046 | 0.799829534 | 1.070333003 | 0.295829405 | |||
| HER2-negative | IVW | 0.91292126 | 0.837958058 | 0.994590623 | 0.037153116 | ||
| Weighted median | 0.922316775 | 0.817773184 | 1.04022515 | 0.187675905 | |||
| MR Egger | 0.921048775 | 0.749686223 | 1.131581214 | 0.451750232 | |||
| Cathepsin S | 24 | HER-positive | Weighted median | 0.95352102 | 0.861074777 | 1.055892426 | 0.360335493 |
| IVW | 0.975109939 | 0.90594378 | 1.049556732 | 0.50192231 | |||
| MR Egger | 0.979770099 | 0.859776805 | 1.116510053 | 0.762022994 | |||
| HER2-negative | MR Egger | 0.944969263 | 0.861751595 | 1.036223097 | 0.241587852 | ||
| Weighted median | 0.95935705 | 0.885567941 | 1.039294568 | 0.309572566 | |||
| IVW | 0.998016195 | 0.946401325 | 1.052446039 | 0.941572121 | |||
| Cathepsin Z | 13 | HER-positive | MR Egger | 1.096092805 | 0.97297243 | 1.234792889 | 0.159401457 |
| IVW | 1.032759971 | 0.958211036 | 1.113108822 | 0.399072207 | |||
| Weighted median | 1.028465407 | 0.936677144 | 1.129248324 | 0.556216109 | |||
| HER2-negative | MR Egger | 1.030537789 | 0.937177545 | 1.133198443 | 0.547344455 | ||
| IVW | 0.986223586 | 0.928971782 | 1.047003775 | 0.649369819 | |||
| Weighted median | 0.992744911 | 0.915654139 | 1.07632611 | 0.859858124 |
Fig. 2.
Forest plot showing the causal relationship between CTSE, CTSO, and breast cancer risk. Green indicates an odds ratio (OR) greater than 1 with statistical significance; red represents an OR less than 1 with statistical significance; black indicates no significant relationship
Sensitivity analysis
To validate the robustness of the MR analysis, several sensitivity tests were conducted, including the MR-Egger intercept test, MR-PRESSO pleiotropy test, and Cochran’s Q heterogeneity test. Cochran’s Q test was specifically used to quantitatively assess the heterogeneity of causal effects across multiple IVs targeting the same gene, serving as a critical metric to determine whether the IVs exert consistent regulatory effects on the target gene. This is essential to ensure the reliability of the integrated causal effect estimate.
For the core genes of interest, the results were as follows: CTSE displayed a Cochran’s Q value of 6.91 (P = 0.646) in HER2-negative breast cancer and 5.83 (P = 0.756) in HER2-positive breast cancer, while CTSO had a Cochran’s Q value of 10.82 (P = 0.459) in HER2-negative breast cancer (Table 2). These findings suggest that the regulatory effects of IVs on the target genes (CTSE/CTSO) were highly consistent, with no significant heterogeneity. This supports the use of the IVW method for integrating multiple IVs to estimate the overall causal effects, effectively ruling out bias caused by inconsistent IV effects.
Table 2.
Sensitivity analysis of Mendelian randomization
| MR Egger | MR-PRESSO | Cochran Q | |||||
|---|---|---|---|---|---|---|---|
| Egger_intercept | P | RSSobs | P | Q | P | ||
| CTS E | HER-positive | 0.032687281 | 0.09052622 | 7.207017823 | 0.83 | 5.834931805 | 0.75631863 |
| HER2-negative | -0.01981785 | 0.181990083 | 8.121678253 | 0.74 | 6.911522557 | 0.646330795 | |
| CTS O | HER2-negative | -0.001920382 | 0.927272386 | 14.15860133 | 0.48 | 10.81582568 | 0.458819464 |
Similarly, the MR-Egger intercept test and MR-PRESSO pleiotropy test yielded P-values greater than 0.05 (Table 2), further confirming the absence of significant horizontal pleiotropy among the IVs. Together, these results validate the suitability of the selected IVs and the reliability of the identified causal associations.
The scatter plots illustrating these results are shown in Fig. 3A-C. Additionally, leave-one-out analysis showed that most instrumental variable effects were consistent with the overall effect, suggesting that no single instrumental variable exerted a disproportionate influence on the causal relationship, confirming the robustness of our results (Fig. 3D-F).
Fig. 3.
Sensitivity Analysis of MR. A–C Scatter plots from MR analysis confirming the stability and consistency of causal relationships. A Scatter plots for CTSE in HER2-positive breast cancer; B Scatter plots for CTSE in HER2-negative breast cancer; C Scatter plots for CTSO in HER2-negative breast cancer. D–F Leave-one-out analysis validating the consistency of most instrumental variables in the causal relationship. D Leave-One-Out analysis for CTSE in HER2-positive breast cancer; E Leave-One-Out analysis for CTSE in HER2-negative breast cancer; F Leave-One-Out analysis for CTSO in HER2-negative breast cancer
Differential expression analysis
To assess the differential expression of CTSE and CTSO in breast cancer and its subtypes, we performed gene expression analysis. First, using SangerBox, we found that CTSE was highly expressed in breast cancer tissues, while CTSO expression was significantly lower (Fig. 4A, B). Further analysis using NULCAN revealed that CTSE expression was elevated in HER2-positive breast cancer tissues, whereas CTSO was broadly underexpressed across all breast cancer subtypes (Fig. 4C, D). Notably, CTSE expression was significantly higher in HER2-positive tumors and lower in HER2-negative tumors, while CTSO expression exhibited the opposite trend, being lower in HER2-positive tumors and higher in HER2-negative tumors (Fig. 4E, F).
Fig. 4.
Differential expression of CTSE and CTSO in breast cancer tissues. A, B Overall expression of CTSE and CTSO in breast cancer tissues. A CTSE; B CTSO. C-D Expression of CTSE and CTSO across different breast cancer subtypes. C CTSE; D CTSO. E-F Expression trends of CTSE and CTSO in HER2-positive and HER2-negative breast cancer. E CTSE; F CTSO
Prognostic analysis
To investigate the prognostic value of CTSE and CTSO in breast cancer patients, we analyzed their expression levels in relation to patient survival under different treatment regimens (chemotherapy and hormone therapy). In chemotherapy-treated patients, high CTSE expression was associated with better prognosis in HER2-positive breast cancer, although this did not reach statistical significance. In contrast, lower CTSE expression was linked to better prognosis in HER2-negative breast cancer patients. For CTSO, high expression correlated with improved prognosis in HER2-negative patients (Fig. 5A).
Fig. 5.
Prognostic roles of CTSE and CTSO in breast cancer. A Prognostic analysis of CTSE and CTSO expression in chemotherapy-treated patients. B Prognostic analysis of CTSE and CTSO expression in hormone therapy-treated patients
For hormone therapy-treated patients, CTSE low expression was associated with better prognosis in HER2-positive cases, although it did not reach statistical significance, whereas CTSE high expression correlated with better prognosis in HER2-negative patients. Additionally, CTSO low expression was a favorable prognostic indicator for HER2-negative breast cancer (Fig. 5B).
Single-cell analysis
To further explore the expression patterns of CTSE and CTSO in breast cancer and their relationship with tumor cell heterogeneity and the tumor microenvironment, we conducted spatial expression analysis on single-cell transcriptomic data. The results showed that CTSO had a significantly higher overall expression than CTSE, with CTSO widely expressed in both cancer cells (red) and normal cells (green), especially in cancer cells. In contrast, CTSE expression was lower and primarily localized to cancer cells, with minimal expression in normal cells (Figs. 6A-C). Further analysis of single-cell data across breast cancer subtypes indicated that CTSO expression was higher in triple-negative breast cancer (TNBC, green area) compared to HER2-positive breast cancer (HER2+, blue area) (Fig. 6D). Additionally, CTSO expression was elevated in immune cells (including T-cells and B-cells), cancer-associated fibroblasts (CAFs), and some cancerous epithelial cells, while CTSE was almost exclusively expressed in cancerous epithelial cells and a subset of CAFs (Fig. 6E).
Fig. 6.
Single-cell transcriptome analysis of CTSE and CTSO expression patterns in breast cancer. A Gene expression pattern of CTSE. B Gene expression pattern of CTSO. C Clustering analysis based on cancer cells (red) and normal cells (green). D Cluster analysis of breast cancer subtypes. E Cell subtype clustering analysis in breast cancer, showing differential expression of CTSE and CTSO across different cell types
PPI network construction and gene enrichment analysis
To further investigate the functional roles of CTSE and CTSO, we constructed their protein-protein interaction (PPI) networks using GeneMANIA (Figs. 7A, B). The results of the PPI network analysis suggest potential functional relationships between CTSE and CTSO and other key proteins in breast cancer biology. Additionally, GO and KEGG enrichment analyses revealed distinct biological processes and pathways associated with these genes. CTSE was predominantly enriched in biological processes such as protein maturation, digestion, and lysosomal activities, with a major role in lysosomal and protein digestion pathways, indicating its potential involvement in tumor metabolism and invasiveness. On the other hand, CTSO was enriched in processes related to antigen processing and presentation, extracellular matrix degradation, and immune regulation. Its primary function in cysteine-type endopeptidase activity, along with involvement in antigen presentation and apoptosis, suggests its critical role in immune microenvironment regulation and tumor cell survival (Fig. 7C, D).
Fig. 7.
PPI networks and functional enrichment analysis of CTSE and CTSO. A PPI network for CTSE, showing its interactions with other proteins. B PPI network for CTSO, illustrating its potential functional relationships. C Gene Ontology (GO) and KEGG pathway enrichment analysis for CTSE. D GO and KEGG pathway enrichment analysis for CTSO
Experimental validation results
To validate the association between CTSE, CTSO, and breast cancer subtypes, as well as their functional implications suggested by bioinformatics analysis, this study conducted targeted cell-based experiments. These experiments, encompassing gene expression analysis, functional interventions, and phenotypic evaluations, were designed to construct a robust evidence chain and further solidify the reliability of the core conclusions.
Initially, we performed qRT-PCR to assess the expression profiles of CTSE and CTSO in HER2-positive cell lines (SK-BR-3, BT-474), HER2-negative cell lines (MDA-MB-231, MCF-7), and normal breast epithelial cells (MCF-10 A). The results showed that CTSE was not quantifiable in any of the cell lines tested. This finding aligns with the bioinformatics analysis, which indicated CTSE expression to be extremely low in breast cancer, supporting the notion that CTSE expression in breast cancer cells may exhibit significant heterogeneity, or that current detection methods are insufficient to capture its low abundance. Conversely, CTSO showed low expression across all breast cancer cell lines (Fig. 8A). This directly corroborated the bioinformatics conclusion that CTSO is lowly expressed in HER2-negative breast cancer but higher than in HER2-positive subtypes, providing a basis for selecting the MDA-MB-231 cell line for subsequent functional experiments.
Fig. 8.
Results of experimental validation for CTSO function. A Detection of CTSO mRNA expression levels in different breast cancer cell lines; B qRT-PCR results showing the overexpression efficiency of CTSO; C Effect of CTSO overexpression on the proliferation of MDA-MB-231 cells; D Effect of CTSO overexpression on the migration of MDA-MB-231 cells; *P < 0.05, **P < 0.01, ***P < 0.001, #P < 0.0001 (all indicate statistically significant differences compared with the corresponding control group) (both indicate statistically significant differences)
Having established the expression characteristics and subtype specificity of CTSO, we sought to explore its functional impact on HER2-negative breast cancer cells. The MDA-MB-231 cell line, which exhibited the highest endogenous CTSO expression among HER2-negative cell lines, was selected to construct a CTSO overexpression model. Transfection efficiency was assessed 48 h post-transfection, and qRT-PCR results showed a significant increase in CTSO mRNA levels in the pcDNA3.1-CTSO group, as compared to the empty vector control (pcDNA3.1-NC) group (Fig. 8B). This confirmed successful transfection and indicated that the overexpression was sufficient for subsequent functional assays.
Next, we employed the CCK-8 assay to evaluate the effect of CTSO overexpression on the proliferation of MDA-MB-231 cells. The results demonstrated clear intergroup differences in cell proliferation rates. At 24, 48, and 72 h post-culture, the CTSO overexpression group showed significantly lower OD450 values than the empty vector control group. Furthermore, the differences between the groups became more pronounced with extended culture time (Fig. 8C). This result indicates that CTSO overexpression significantly inhibits the proliferative capacity of HER2-negative breast cancer cells, aligning with the causal association observed in meta-regression analysis, which suggested that “increased CTSO expression reduces HER2-negative breast cancer risk” (OR = 0.913, P = 0.037). These findings provide direct cellular evidence supporting the tumor-suppressive role of CTSO.
Finally, we performed a wound-healing assay to assess the effect of CTSO overexpression on the migration capacity of MDA-MB-231 cells. The results showed that CTSO overexpression significantly suppressed the migration ability of HER2-negative breast cancer cells, further supporting the bioinformatics prediction that CTSO may function as a tumor suppressor in HER2-negative breast cancer. These phenotypic results provide further evidence for CTSO’s involvement in tumor progression regulation (Fig. 8D).
Discussion
Molecular heterogeneity in breast cancer is a core factor influencing treatment response and prognosis [28, 29]. For example, tubulin alpha-1b chain (TUBA1B) is a pan-cancer prognostic and immune biomarker, with subtype-specific functions in breast cancer [30]. Significant biological differences exist between HER2-positive and HER2-negative subtypes, and identifying subtype-specific regulatory genes and mechanisms is key to advancing precision medicine. This study systematically investigates the roles of cathepsin family members CTSE and CTSO in breast cancer subtypes through Mendelian randomization (MR), bioinformatics, and cellular experiments, with the following key findings discussed below.
One of the main findings of this study is the significant positive correlation between elevated CTSE expression and breast cancer risk in both HER2-positive (OR = 1.108, P = 0.022) and HER2-negative (OR = 1.099, P = 0.009) subtypes. This causal relationship was further validated through MR-Egger, MR-PRESSO, and other sensitivity analyses, excluding potential confounding by heterogeneity and pleiotropy, thus providing greater reliability than traditional observational studies. Differential expression analysis confirmed that CTSE is significantly overexpressed in breast cancer tissues, with higher expression observed in the HER2-positive subtype. Single-cell analysis revealed that CTSE is primarily localized to malignant epithelial cells, suggesting that it may exert its effects by regulating cancer cell biology. Functional enrichment analysis (with CTSE enriched in lysosomal functions and protein digestion pathways) suggests that CTSE may promote tumor progression by enhancing lysosome-mediated tumor metabolism and extracellular matrix (ECM) degradation, thereby increasing cancer cell invasiveness. It is important to note that CTSE mRNA was not detectable by qPCR in HER2-positive/negative breast cancer cell lines, possibly due to low expression in specific cell subpopulations or limitations in detection sensitivity. However, this does not negate its potential carcinogenic function. Future studies should employ RNA in situ hybridization (RNA ISH) on clinical samples to improve sensitivity and validate CTSE functional expression in cancer epithelial cells through protein activity assays (e.g., fluorescence substrate assays). Furthermore, MR analysis revealed that CTSE could serve as a potential early biomarker for pancreatic and liver cancers [31, 32], and experimental and bioinformatic studies have indicated its role in prognosis and as a potential therapeutic target in cholangiocarcinoma, gastric cancer, colorectal cancer, and bladder cancer [33–35], highlighting its potential as a cross-cancer therapeutic target and reinforcing the need to explore its role in breast cancer.
In contrast to CTSE, CTSO exerts a protective role in HER2-negative breast cancer. MR analysis showed that CTSO expression was negatively correlated with HER2-negative breast cancer risk (OR = 0.913, P = 0.037). Differential expression analysis confirmed that CTSO is generally underexpressed in breast cancer tissues, with the lowest expression observed in the HER2-negative subtype, especially in triple-negative breast cancer. Single-cell analysis revealed that CTSO is primarily localized to immune cells (T cells, B cells) and cancer-associated fibroblasts (CAFs)—a cell-type-specific expression pattern that resolves the apparent contrast between its protective role (MR OR < 1) and low expression in cancerous epithelial cells. Specifically, the MR-derived protective effect of CTSO reflects its “overall regulatory contribution” to HER2-negative breast cancer, which is dominated by its function in immune cells and CAFs rather than its low expression in cancer cells. Combined with the protein-protein interaction (PPI) network results (Fig. 7B), CTSO interacts with key molecules involved in antigen processing (e.g., MHC class I molecules) and immune cell activation (e.g., T cell receptor-associated proteins): in immune cells, its high expression facilitates the processing and presentation of tumor-associated antigens, strengthening the recognition and killing of cancer cells by cytotoxic T cells; in CAFs, it modulates ECM remodeling (consistent with its association with CAF-related gene signatures and ECM pathways [38]) to reduce the formation of an immune-suppressive tumor microenvironment (e.g., inhibiting recruitment of myeloid-derived suppressor cells). This cell-type-specific functional collaboration not only explains why CTSO exerts a protective effect despite low expression in cancer cells but also underscores its role as a mediator of tumor-stromal crosstalk. Functional enrichment analysis (involving antigen processing and immune regulation) suggests that CTSO may exert its anti-cancer effects by remodeling the tumor immune microenvironment. Similarly, cytokine-induced apoptosis inhibitor 1 (CIAPIN1) is closely associated with immune cell infiltration and poor prognosis in invasive breast cancer. Its expression is linked to subtype characteristics such as estrogen receptor (ER), progesterone receptor (PR), and DNA methylation modifications [36]. Previous studies have shown that CAFs promote breast cancer proliferation, angiogenesis, metastasis, and treatment resistance, leading to poor prognosis [37]. As a cysteine protease, CTSO participates in macrophage-mediated matrix remodeling and osteoclast-mediated bone resorption and is part of the CAF-related gene signature in breast cancer, associated with ECM remodeling and immune suppression [38]. Its immune regulatory effects may enhance anti-tumor immune responses, providing a potential direction for HER2-negative breast cancer immunotherapy (e.g., by enhancing CTSO expression/activity to strengthen immune surveillance). Cellular experiments further validated that overexpression of CTSO in HER2-negative MDA-MB-231 cells significantly reduced cell proliferation and migration, directly confirming its inhibitory effect on malignant phenotypes. It is important to note that CTSO’s function is subtype-dependent—its gene variants (e.g., SNP rs10030044) are associated with poor prognosis in hormone receptor-positive breast cancer. Therefore, the development of CTSO-targeted therapies should consider subtype and genetic background differences.
Regarding prognostic value, the effects of CTSE and CTSO are subtype-specific. In HER2-positive chemotherapy patients, high CTSE expression is associated with a better prognosis, while in HER2-negative patients, low CTSE expression and high CTSO expression are both linked to longer survival. These two biomarkers can serve as subtype-specific prognostic indicators (e.g., CTSE could assist in assessing chemotherapy benefit in HER2-positive patients, and CTSO could serve as a prognostic reference for HER2-negative patients). Based on their functional differences, targeted therapeutic strategies can be proposed: HER2-positive breast cancer can target CTSE to inhibit ECM remodeling and tumor invasiveness, while HER2-negative breast cancer could benefit from strategies aimed at enhancing CTSO expression/activity to boost immune surveillance.
In summary, this study adopts a multi-dimensional approach, integrating Mendelian randomization (MR) analysis, bioinformatics, and preliminary experimental validation to comprehensively uncover the causal relationships and potential molecular mechanisms of CTSE and CTSO in breast cancer. However, several limitations must be acknowledged. Firstly, the population representativeness is insufficient. The MR analysis relied on tool variables from European GWAS data, and although CTSO’s expression characteristics were validated using multi-population samples from TCGA, ethnic differences in cathepsin-related SNP allele frequencies and immune cell infiltration patterns may limit the generalizability of the MR causal conclusions to non-European populations. Secondly, there is a gap in experimental validation. The current experiments focus primarily on CTSO, which has been shown to inhibit the proliferation and migration of HER2-negative breast cancer cells, but due to the lack of detectable CTSE mRNA in the selected cell lines via qPCR, functional intervention experiments have not been conducted. Moreover, the mechanism underlying CTSO’s immune-regulatory role has not been fully explored, leaving the molecular details of its suppression of cancer through the immune microenvironment unclear. Furthermore, the MR method has inherent limitations, including an inability to fully exclude unmeasured horizontal pleiotropy, and it can only infer causal associations rather than dynamic changes.
Future research should aim to overcome these limitations by expanding cross-ethnic validation through MR analysis and expression-prognosis studies involving Asian breast cancer cohorts, incorporating GWAS and multi-omics data to investigate ethnic differences in underlying mechanisms. Additionally, experimental validation should be deepened by identifying CTSE-positive cell subpopulations or generating stable transfectants to examine its oncogenic mechanisms using Transwell assays, RNA ISH, and protein activity detection. Transcriptomic sequencing, co-immunoprecipitation (Co-IP), and animal models should also be employed to explore the molecular details of CTSO’s immune microenvironment regulation. Moreover, clinical data from immunotherapy patients should be integrated to validate the predictive value of CTSO for treatment efficacy, with the aim of constructing a combined prognostic model. Finally, research should be expanded to include single-cell spatial transcriptomics and proteomics, enabling the analysis of the spatial regulatory networks of CTSE and CTSO, as well as their downstream substrates, with longitudinal data verifying their functional differences at various stages of tumor progression.
Conclusion
This study is the first to integrate Mendelian randomization, bioinformatics, and preliminary in vitro experiments to explore the potential causal relationships and biological functions of cathepsin family members CTSE and CTSO in different subtypes of breast cancer. Our findings suggest that elevated CTSE expression in both HER2-positive and HER2-negative breast cancer may promote tumor progression, while reduced CTSO expression could potentially facilitate the development of HER2-negative breast cancer through its immune regulatory effects. By combining gene expression analysis, prognostic assessments, and single-cell data, this study identifies promising novel biomarkers for precision therapy and early prognostic evaluation in breast cancer, thereby providing a theoretical foundation for future therapeutic strategies and clinical interventions.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
Not applicable.
Author contributions
LT: Data curation, Validation, Writing–original draft, Writing–review & editing. LX: Data curation, Validation, Writing–original draft, Writing–review & editing. YL: Data curation, Validation, Writing–original draft, Writing–review & editing. XZ: Data curation, Funding acquisition, Resources, Supervision, Validation, Writing–original draft, Writing–review & editing.
Funding
Not applicable.
Data availability
The datasets generated and/or analysed during the current study are available in the IEU Open GWAS platform repository (https://gwas.mrcieu.ac.uk/). Specifically, the cathepsin-related genome-wide association study (GWAS) data consist of two parts: the first is derived from the INTERVAL study, and the second from the GCST90012073 dataset(https://opengwas.io/datasets/ebi-a-GCST90012073). The breast cancer GWAS data were also obtained from the IEU Open GWAS platform. HER2-positive breast cancer data are obtained from the finn-b-C3_BREAST_HERPLUS_EXALLC dataset (https://opengwas.io/datasets/finn-b-C3_BREAST_HERPLUS_EXALLC), while HER2-negative breast cancer data are from the finn-b-C3_BREAST_HER2NEG_EXALLC dataset(https://opengwas.io/datasets/finn-b-C3_BREAST_HER2NEG_EXALLC). All datasets are publicly accessible through the aforementioned platforms.
Declarations
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Ban M, Petrić Miše B, Vrdoljak E. Early HER2-Positive breast cancer: current treatment and novel approaches. Breast Care (Basel Switzerland). 2020;15(6):560–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Marczyk VR, Rosa DD, Maia AL, Goemann IM. Overall survival for HER2-Positive breast cancer patients in the HER2-Targeted era: evidence from a Population-Based study. Clin Breast Cancer. 2022;22(5):418–23. [DOI] [PubMed] [Google Scholar]
- 3.Lodi M, Voilquin L, Alpy F, Molière S, Reix N, Mathelin C, Chenard MP, Tomasetto CL. STARD3: A New Biomarker in HER2-Positive Breast Cancer. Cancers 2023, 15, (2). [DOI] [PMC free article] [PubMed]
- 4.Shikata S, Murata T, Yoshida M, Hashiguchi H, Yoshii Y, Ogawa A, Watase C, Shiino S, Sugino H, Jimbo K, Maeshima A, Iwamoto E, Takayama S, Suto A. Prognostic impact of HER2-low positivity in patients with HR-positive, HER2-negative, node-positive early breast cancer. Sci Rep. 2023;13(1):19669. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Xiao Y, Cong M, Li J, He D, Wu Q, Tian P, Wang Y, Yang S, Liang C, Liang Y, Wen J, Liu Y, Luo W, Lv X, He Y, Cheng DD, Zhou T, Zhao W, Zhang P, Zhang X, Xiao Y, Qian Y, Wang H, Gao Q, Yang QC, Yang Q, Hu G. Cathepsin C promotes breast cancer lung metastasis by modulating neutrophil infiltration and neutrophil extracellular trap formation. Cancer Cell. 2021;39(3):423–e4377. [DOI] [PubMed] [Google Scholar]
- 6.Pranjol MZ, Gutowski N, Hannemann M, Whatmore J. The potential role of the proteases cathepsin D and cathepsin L in the progression and metastasis of epithelial ovarian cancer. Biomolecules. 2015;5(4):3260–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Xu LB, Qin YF, Su L, Huang C, Xu Q, Zhang R, Shi XD, Sun R, Chen J, Song Z, Jiang X, Shang L, Xiao G, Kong X, Liu C, Wong PP. Cathepsin-facilitated invasion of BMI1-high hepatocellular carcinoma cells drives bile duct tumor thrombi formation. Nat Commun. 2023;14(1):7033. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Rochefort H. Cathepsin D in breast cancer. Breast Cancer Res Treat. 1990;16(1):3–13. [DOI] [PubMed] [Google Scholar]
- 9.Vashum Y, Khashim Z, Obesity, Cathepsin K. A complex pathophysiological relationship in breast cancer metastases. Endocr Metab Immune Disord Drug Targets. 2020;20(8):1227–31. [DOI] [PubMed] [Google Scholar]
- 10.Lecaille F, Chazeirat T, Saidi A, Lalmanach G, Cathepsin V. Molecular characteristics and significance in health and disease. Mol Aspects Med. 2022;88:101086. [DOI] [PubMed] [Google Scholar]
- 11.Alhudiri I, Nolan C, Ellis I, Elzagheid A, Green A, Chapman C. Expression of cathepsin D in early-stage breast cancer and its prognostic and predictive value. Breast Cancer Res Treat. 2024;206(1):143–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Li X, Liang Y, Lian C, Peng F, Xiao Y, He Y, Ma C, Wang Y, Zhang P, Deng Y, Su Y, Luo C, Kong X, Yang Q, Liu T, Hu G. CST6 protein and peptides inhibit breast cancer bone metastasis by suppressing CTSB activity and osteoclastogenesis. Theranostics. 2021;11(20):9821–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zhang L, Zhao Y, Yang J, Zhu Y, Li T, Liu X, Zhang P, Cheng J, Sun S, Wei C, Fu J. CTSL, a prognostic marker of breast cancer, that promotes proliferation, migration, and invasion in cells in triple-negative breast cancer. Front Oncol. 2023;13:1158087. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sereesongsaeng N, Burrows JF, Scott CJ, Brix K, Burden RE. Cathepsin V regulates cell cycle progression and histone stability in the nucleus of breast cancer cells. Front Pharmacol. 2023;14:1271435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Burgess S, Davey Smith G, Davies NM, Dudbridge F, Gill D, Glymour MM, Hartwig FP, Kutalik Z, Holmes MV, Minelli C, Morrison JV, Pan W, Relton CL, Theodoratou E. Guidelines for performing Mendelian randomization investigations: update for summer 2023. Wellcome Open Res. 2019;4:186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sun BB, Maranville JC, Peters JE, Stacey D, Staley JR, Blackshaw J, Burgess S, Jiang T, Paige E, Surendran P, Oliver-Williams C, Kamat MA, Prins BP, Wilcox SK, Zimmerman ES, Chi A, Bansal N, Spain SL, Wood AM, Morrell NW, Bradley JR, Janjic N, Roberts DJ, Ouwehand WH, Todd JA, Soranzo N, Suhre K, Paul DS, Fox CS, Plenge RM, Danesh J, Runz H, Butterworth AS. Genomic atlas of the human plasma proteome. Nature. 2018;558(7708):73–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Levin MG, Judy R, Gill D, Vujkovic M, Verma SS, Bradford Y, Ritchie MD, Hyman MC, Nazarian S, Rader DJ, Voight BF, Damrauer SM. Genetics of height and risk of atrial fibrillation: A Mendelian randomization study. PLoS Med 2020, 17, (10), e1003288. [DOI] [PMC free article] [PubMed]
- 18.Mounier N, Kutalik Z. Bias correction for inverse variance weighting Mendelian randomization. Genet Epidemiol. 2023;47(4):314–31. [DOI] [PubMed] [Google Scholar]
- 19.Li P, Wang H, Guo L, Gou X, Chen G, Lin D, Fan D, Guo X, Liu Z. Association between gut microbiota and preeclampsia-eclampsia: a two-sample Mendelian randomization study. BMC Med. 2022;20(1):443. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol. 2017;32(5):377–89. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Gala H, Tomlinson I. The use of Mendelian randomisation to identify causal cancer risk factors: promise and limitations. J Pathol. 2020;250(5):541–54. [DOI] [PubMed] [Google Scholar]
- 22.Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal Pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Chandrashekar DS, Karthikeyan SK, Korla PK, Patel H, Shovon AR, Athar M, Netto GJ, Qin ZS, Kumar S, Manne U, Creighton CJ, Varambally S. UALCAN: an update to the integrated cancer data analysis platform. Neoplasia (New York N Y). 2022;25:18–27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Chandrashekar DS, Bashel B, Balasubramanya SAH, Creighton CJ, Ponce-Rodriguez I, Chakravarthi B, Varambally S. UALCAN: A portal for facilitating tumor subgroup gene expression and survival analyses. Neoplasia (New York N Y). 2017;19(8):649–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Jézéquel P, Gouraud W, Ben Azzouz F, Guérin-Charbonnel C, Juin PP, Lasla H, Campone M. bc-GenExMiner 4.5: new mining module computes breast cancer differential gene expression analyses. Database: the journal of biological databases and curation 2021, 2021. [DOI] [PMC free article] [PubMed]
- 26.Győrffy B. Transcriptome-level discovery of survival-associated biomarkers and therapy targets in non-small-cell lung cancer. Br J Pharmacol. 2024;181(3):362–74. [DOI] [PubMed] [Google Scholar]
- 27.Győrffy B. Integrated analysis of public datasets for the discovery and validation of survival-associated genes in solid tumors. Innov (Cambridge (Mass)). 2024;5(3):100625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wang Y, Li Y, Jing Y, Yang Y, Wang H, Ismtula D, Guo C. Tubulin alpha-1b chain was identified as a prognosis and immune biomarker in pan-cancer combing with experimental validation in breast cancer. Sci Rep. 2024;14(1):8201. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Guo L, Kong D, Liu J, Zhan L, Luo L, Zheng W, Zheng Q, Chen C, Sun S. Breast cancer heterogeneity and its implication in personalized precision therapy. Experimental Hematol Oncol. 2023;12(1):3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Huang X, Wang Y, Wang J, Jing Y, Dilraba E, Li Y, Guo C. Association of DBNDD1 with prognostic and immune biomarkers in invasive breast cancer. Discover Oncol. 2025;16(1):218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Pontious C, Kaul S, Hong M, Hart PA, Krishna SG, Lara LF, Conwell DL, Cruz-Monserrate Z. Cathepsin E expression and activity: role in the detection and treatment of pancreatic cancer. Pancreatology. 2019;19(7):951–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Liu Q, Chen J, Liu Y, Zhang S, Feng H, Wan T, Zhang S, Zhang N, Yang Z. The impact of cathepsins on liver hepatocellular carcinoma: insights from genetic and functional analyses. Gene. 2025;935:149064. [DOI] [PubMed] [Google Scholar]
- 33.Li Y, Xu C, Wang B, Xu F, Ma F, Qu Y, Jiang D, Li K, Feng J, Tian S, Wu X, Wang Y, Liu Y, Qin Z, Liu Y, Qin J, Song Q, Zhang X, Sujie A, Huang J, Liu T, Shen K, Zhao JY, Hou Y, Ding C. Proteomic characterization of gastric cancer response to chemotherapy and targeted therapy reveals new therapeutic strategies. Nat Commun. 2022;13(1):5723. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Chou CL, Chen TJ, Tian YF, Chan TC, Yeh CF, Li WS, Tsai HH, Li CF, Lai HY. CTSE Overexpression Is an Adverse Prognostic Factor for Survival among Rectal Cancer Patients Receiving CCRT. Life (Basel, Switzerland) 2021, 11, (7). [DOI] [PMC free article] [PubMed]
- 35.Wu Y, Zhou W, Yang Z, Li J, Jin Y. miR-185-5p represses cells growth and metastasis of osteosarcoma via targeting cathepsin E. Int J Toxicol. 2022;41(2):115–25. [DOI] [PubMed] [Google Scholar]
- 36.Luo Z, Wang Y, Bi X, Ismtula D, Wang H, Guo C. Cytokine-induced apoptosis inhibitor 1: a comprehensive analysis of potential diagnostic, prognosis, and immune biomarkers in invasive breast cancer. Translational Cancer Res. 2023;12(7):1765–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Hu D, Li Z, Zheng B, Lin X, Pan Y, Gong P, Zhuo W, Hu Y, Chen C, Chen L, Zhou J, Wang L. Cancer-associated fibroblasts in breast cancer: challenges and opportunities. Cancer Commun (London England). 2022;42(5):401–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Li C, Yang L, Zhang Y, Hou Q, Wang S, Lu S, Tao Y, Hu W, Zhao L. Integrating single-cell and bulk transcriptomic analyses to develop a cancer-associated fibroblast-derived biomarker for predicting prognosis and therapeutic response in breast cancer. Front Immunol. 2023;14:1307588. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analysed during the current study are available in the IEU Open GWAS platform repository (https://gwas.mrcieu.ac.uk/). Specifically, the cathepsin-related genome-wide association study (GWAS) data consist of two parts: the first is derived from the INTERVAL study, and the second from the GCST90012073 dataset(https://opengwas.io/datasets/ebi-a-GCST90012073). The breast cancer GWAS data were also obtained from the IEU Open GWAS platform. HER2-positive breast cancer data are obtained from the finn-b-C3_BREAST_HERPLUS_EXALLC dataset (https://opengwas.io/datasets/finn-b-C3_BREAST_HERPLUS_EXALLC), while HER2-negative breast cancer data are from the finn-b-C3_BREAST_HER2NEG_EXALLC dataset(https://opengwas.io/datasets/finn-b-C3_BREAST_HER2NEG_EXALLC). All datasets are publicly accessible through the aforementioned platforms.