Skeletal muscle regeneration is a highly choreographed process governed by the dynamic interplay between myogenic progenitors and the inflammatory microenvironment. However, the molecules that integrate early inflammatory signals with myogenic fate transitions remain poorly understood. Here, we identify the long noncoding RNA (lncRNA) Brip1os as a critical sentinel that orchestrates this crosstalk. Brip1os expression is inversely correlated with postnatal skeletal muscle development but is robustly and specifically induced during the early inflammatory phase following muscle acute injury. Functional assays demonstrate that Brip1os sustains the myogenic proliferative pool by promoting myoblast expansion and preventing premature differentiation. Mechanistically, we show that Brip1os regulates a set of IFN-γ-responsive genes and is continuously upregulated in injured muscle cells during the early phase through a positive feedback loop involving M1 macrophages. Notably, loss of Brip1os in vivo leads to a breakdown of this inflammatory-myogenic crosstalk, manifested by impaired M1 macrophage recruitment and defective muscle repair, and exogenous IFN-γ supplementation partially rescues regeneration in Brip1os -deficient mice. Collectively, our findings reveal a Brip1os -mediated regulatory hub that tunes the early inflammatory response to support effective tissue repair, offering a promising therapeutic target for regenerative failure and muscle wasting disorders.

Liver pre-metastatic niches (PMN) formation is a pivotal process in colorectal cancer liver metastasis (CLM). Phosphatase of regenerating liver-3 (PRL-3) has been demonstrated as a key factor in promoting CRC progression (e.g., therapeutic resistance and metastasis), but its role in liver PMN formation remains unknown. Using mouse models and CRC patient samples, we herein reveal that high PRL-3 expression in CRC cells could enhance the recruitment of myeloid-derived suppressor cells (MDSCs) into the liver and impair the hepatic infiltration of CD8+ T cells, thereby promoting the liver PMN formation and CLM. Mechanistically, high PRL-3 expression could activate the Src/STAT3 signaling pathway in CRC cells and thus up-regulate integrin αvβ5 (ITGαvβ5) expression in their secreted exosomes, which could specifically target F4/80+ macrophages in the liver to activate the P38/STAT1 signaling pathway. With this activation of P38/STAT1 pathway, the secretion of C-X-C motif chemokine ligand 12 (CXCL12) from F4/80+ macrophages is significantly improved, which could enhance the recruitment of MDSCs into the liver and impair the hepatic infiltration of CD8+ T cells, ultimately leading to the liver PMN formation and CLM. Taken together, our findings not only uncover the important role of PRL-3 in CLM via promoting the liver PMN formation, but also provide the evidence for the treatment of CLM by targeting PRL-3.

Metastases are a primary cause of cancer-associated mortality; however, the mechanisms underlying aggressive progression have not been clearly elucidated. Genome-wide features of chromatin accessibility through ATAC-seq from HCC primary and metastatic tumors revealed that many distal regulatory elements spreading the genome become accessible during aggressive progression, the changes of which are associated with NFY-family. And NFYB is frequently upregulated in tumor with metastasis. Mechanistically, LINC01137 recruits SMYD3 to enhance H3K4me3 occupancy at IL-1β, CXCL2 and CCL20 promoters by inhibiting lysine ubiquitination to stabilize NFYB, which in turn upregulates IL-1β, CXCL2 and CCL20. HCC-derived cytokine transforms macrophages to the M2 phenotype to foster an inhibitory tumor microenvironment and anti-PDL1 tolerance. Importantly, LINC01137 transcription is activated by the NFYB/KAT2B complex in a feed-forward loop. Notably, treatment with an IL-1β inhibitor enhances the blockade efficacy of PD-L1 in NFYB-overexpressing HCC. Our findings imply an immunosuppressive role of NFYB-LINC01137 signaling during aggressive HCC progression and support the concept of microenvironment engineering in immunotherapy.

Tumorigenesis is a complex biological process, accompanied by cellular dedifferentiation and metabolic reprogramming, which similarities to the metabolic characteristics of embryonic development stages. In our research, we focused particularly on the RNA-binding protein PEG10, which is highly expressed in the placenta and found to be similarly overexpressed in liver cancer cells, playing a key role in the process of aerobic glycolysis in tumor cells. This discovery suggests that there may be a phenomenon of fetal-like metabolic reprogramming in hepatocellular carcinoma (HCC), which is of significant importance for understanding the pathogenesis of HCC and for seeking new therapeutic strategies. In hepatocellular carcinoma (HCC) cells, we found that PEG10 enhances the stability of mRNA for key glycolytic genes hexokinase2 (HK2) and Glucose transporter 1 (GLUT1) by inhibiting the STAU1-mediated RNA degradation pathway, thereby promoting aerobic glycolysis in tumor cells. This process not only promotes the proliferation of tumor cells but also reduces the sensitivity of tumor cells to sorafenib. Furthermore, we observed in clinical samples that high expression of PEG10 is closely associated with poor prognosis in HCC patients, further confirming the important role of PEG10 in the development of HCC. Our research also found that retinoic acid can effectively inhibit aerobic glycolysis in tumor cells by suppressing the transcriptional expression of PEG10, thereby inhibiting the growth of tumor cells. In animal models, mice with liver-specific PEG10 gene knockout showed significant resistance to oncogene-induced liver cancer occurrence and development, providing strong evidence for PEG10 as a therapeutic target. These findings not only reveal the biological link between PEG10 and onco-fetal metabolic reprogramming at the molecular level but also indicate from a clinical perspective that PEG10 may be a potential biomarker and therapeutic target for HCC. The high expression of PEG10 is associated with poor prognosis in HCC patients, and the inhibitory effect of retinoic acid on PEG10 expression provides a new strategy for the treatment of HCC. These results provide new ideas for the future diagnosis and treatment of HCC, especially in the development of targeted therapeutic drugs against PEG10, which has important application prospects. Our research has laid a solid foundation for a deeper understanding of the metabolic reprogramming mechanisms of HCC and the development of new therapeutic methods. Dong Yin, Xinyi Yao, Jingyuan Zhang, Yin Zhang, Jianyou Liao, Kaishun Hu, Jiehua He, Daning Lu. Placental gene PEG10 promotes onco-fetal metabolic reprogramming in hepatocellular carcinoma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5399.

The efficacy of human epidermal growth factor receptor 2 (HER2)-targeting antibody-drug conjugates has underscored the critical need for precise HER2 diagnostics in breast cancer treatment. Despite the clinical importance, variability in immunohistochemical (IHC) staining protocols and interobserver inconsistencies challenge the reliability of HER2 status assessment, which is critical for guiding patient treatment strategies. To investigate the factors affecting HER2 interpretation consistency, tissue microarrays from 1063 breast carcinoma cases underwent 3 distinct IHC protocols, and a novel artificial intelligence (AI) model was developed to standardize HER2-stained images. A total of 5 sets of tissue microarrays (Nordi QC, protocol 1, protocol 2, protocol 1 AI, and protocol 2 AI) were independently reviewed by 8 pathologists. The Fleiss Kappa value and overall agreement rate measured interobserver agreement, with logistic regression analyzing the impact of variables on diagnostic accuracy. Our results showed that the Nordi QC protocol had the highest interobserver agreement (Kappa 0.754). AI-based image normalization notably enhanced consistency, particularly for HER2 low cases, aligning scores toward the Nordi QC standard. Logistic regression analysis indicated that both staining protocol and AI-based image standardization significantly influenced diagnostic accuracy (P < .001). The American Society of Clinical Oncology/College of American Pathologists 2018 binary criteria demonstrated the highest HER2 interobserver consistency (Kappa > 0.95). Compared with the American Society of Clinical Oncology/College of American Pathologists 2023 criteria, the newly proposed null, ultra-low/low, positive criteria, merging HER2 low and ultra-low categories, demonstrated improved reliability and agreement, especially in distinguishing the challenging HER2-ultra-low cases, which showed an exceedingly low interobserver agreement (Kappa < 0.20) across all protocols. Overall, variability in IHC staining protocols and HER2 classification criteria significantly affect the diagnostic consistency among pathologists. The integration of an AI model for image standardization and the adoption of the null, ultra-low/low, positive criteria may refine diagnostic precision and bolster clinical decision-making in breast cancer treatment.

The prognostic significance of tumor-infiltrating lymphocytes (TILs) in Luminal-type breast cancer remains controversial, primarily due to typically low TIL infiltration levels, methodological inconsistencies in assessment, and insufficient consideration of spatial distribution patterns. To overcome these limitations, we developed an advanced artificial intelligence (AI)-driven computational TIL assessment (CTA) system, compliant with international visual assessment guidelines, which enables precise quantification of both TIL abundance (automatic TILs, aTILs) and spatial distribution patterns (aggregated: aTILs-agg; distributed: aTILs-dis) in Luminal-type breast cancer. Our comprehensive analysis suggests that elevated TIL levels were significantly associated with improved overall survival (OS) and progression-free survival (PFS) outcomes. Notably, Luminal B subtype demonstrated significantly higher TIL infiltration compared to Luminal A. In the Luminal A cohort, the aggregated spatial pattern (aTILs-agg) emerged as a favorable prognostic indicator for both OS and PFS, while in Luminal B cases, overall TIL abundance (aTILs) and distributed patterns (aTILs-dis) were associated with enhanced survival outcomes. Multivariate Cox regression analysis confirmed the independent prognostic value of aTILs, aTILs-agg, and aTILs-dis for PFS in Luminal A patients, though no significant associations were observed in the Luminal B subgroup. This study demonstrates the clinical utility of AI-powered TIL assessment as a promising prognostic indicator for predicting clinical outcomes in Luminal breast cancer patients, offering new insights into tumor-immune interactions within this molecular subtype.

Cryosectioned tissues often exhibit artifacts that compromise pathologists’ diagnostic accuracy during intraoperative assessments. These inconsistencies, compounded by variations in frozen section (FS) production across laboratories, highlight the need for improved diagnostic tools. This study aims to develop and validate a deep-learning model that transforms cryosectioned images into formalin-fixed paraffin-embedded (FFPE) images to enhance diagnostic performance in breast lesions. We developed an unpaired image-to-image translation model (AI-FFPE) using the TCGA-BRCA dataset to convert FS images into FFPE-like images. The model employs a modified generative adversarial network (GAN) enhanced with an attention mechanism to correct artifacts and a self-regularization constraint to preserve clinically significant features. For validation, 132 FS whole slide images (WSIs) of breast lesions were collected from three cohorts (SYSUCC, GSPCH, and TCGA). These FS-WSIs were transformed into AI-FFPE-WSIs and independently evaluated by six pathologists for image quality, diagnostic concordance, and confidence in lesion properties and final diagnoses. Diagnostic performance was assessed using a diagnostic score (DS), calculated by multiplying the accuracy index by the confidence level. The dataset included 132 reference diagnoses and 1,584 pathologist reads. The AI-FFPE group showed a significant improvement in image quality compared to the FS group (p < 0.001). Concordance rates for lesion properties (79.9

Hyperactivation of ribosome biogenesis (RiBi) drives cancer progression, yet the role of RiBi-associated proteins (RiBPs) in breast cancer (BC) is underexplored. In this study, we perform a comprehensive multi-omics analysis and reveal that assembly and maturation factors (AMFs), a subclass of RiBPs, are upregulated at both RNA and protein levels in BC, correlating with poor patient outcomes. In contrast, ribosomal proteins (RPs) do not show systematic upregulation across various cancers, including BC. We further demonstrate that the oncogenic activation of a top AMF candidate in BC, DCAF13, enhances Pol I transcription and promotes proliferation in BC cells both in vitro and in vivo. Mechanistically, DCAF13 promotes Pol I transcription activity by facilitating the K63-linked ubiquitination of RPA194. This process stimulates global protein synthesis and cell growth. Our findings uncover a modification of RPA194 that regulates Pol I activity; this modification is dysregulated in BC, contributing to cancer progression.

Perturb-Seq combines CRISPR (clustered regularly interspaced short palindromic repeats)-based genetic screens with single-cell RNA sequencing readouts for high-content phenotypic screens. Despite the rapid accumulation of Perturb-Seq datasets, there remains a lack of a user-friendly platform for their efficient reuse. Here, we developed PerturbDB (http://research.gzsys.org.cn/perturbdb), a platform to help users unveil gene functions using Perturb-Seq datasets. PerturbDB hosts 66 Perturb-Seq datasets, which encompass 4 518 521 single-cell transcriptomes derived from the knockdown of 10 194 genes across 19 different cell lines. All datasets were uniformly processed using the Mixscape algorithm. Genes were clustered by their perturbed transcriptomic phenotypes derived from Perturb-Seq data, resulting in 421 gene clusters, 157 of which were stable across different cellular contexts. Through integrating chemically perturbed transcriptomes with Perturb-Seq data, we identified 552 potential inhibitors targeting 1409 genes, including an mammalian target of rapamycin (mTOR) signaling inhibitor, retinol, which was experimentally verified. Moreover, we developed a ‘Cancer’ module to facilitate the understanding of the regulatory role of genes in cancer using Perturb-Seq data. An interactive web interface has also been developed, enabling users to visualize, analyze and download all the comprehensive datasets available in PerturbDB. PerturbDB will greatly drive gene functional studies and enhance our understanding of the regulatory roles of genes in diseases such as cancer.

Cysteine-rich angiogenic inducer 61 (CYR61), also called CCN1, has long been characterized as a secretory protein. Nevertheless, the intracellular function of CYR61 remains unclear. Here, we found that CYR61 is important for proper cell cycle progression. Specifically, CYR61 interacts with microtubules and promotes microtubule polymerization to ensure mitotic entry. Moreover, CYR61 interacts with PLK1 and accumulates during the mitotic process, followed by degradation as mitosis concludes. The proteolysis of CYR61 requires the PLK1 kinase activity, which directly phosphorylates two conserved motifs on CYR61, enhancing its interaction with the SCF E3 complex subunit FBW7 and mediating its degradation by the proteasome. Mutations of phosphorylation sites of Ser167 and Ser188 greatly increase CYR61's stability, while deletion of CYR61 extends prophase and metaphase and delays anaphase onset. In summary, our findings highlight the precise control of the intracellular CYR61 by the PLK1-FBW7 pathway, accentuating its significance as a microtubule-associated protein during mitotic progression.

Neutrophil extracellular traps (NETs) have been shown to exhibit chemotactic effects on circulating tumor cells at metastatic sites, promoting the progression of colorectal cancer liver metastasis (CRLM). However, the origin and factors contributing to the formation of NETs (NETosis) in the pre-metastatic niche (PMN) of target organs remain unclear. In this study, we investigated the relationship between phosphatase of regenerating liver-3 (PRL-3), myeloid-derived suppressor cells (MDSCs), neutrophils, and NETs through a retrospective clinical cohort study and a mouse model of CRLM. Our clinical findings revealed associations between PRL-3 expression and the infiltration of neutrophils and MDSCs in CRLM patients. Moreover, NETosis emerged as a robust indicator of poor prognosis for overall survival in these patients. In CRLM models, PRL-3 overexpression enhanced both primary tumor growth and liver metastasis. We found that the first week after tumor cell implantation appeared to be a crucial period for PMN formation, with notable infiltration of neutrophils, MDSCs, and NETosis during this time. Notably, co-culturing neutrophils with MDSCs induced NETosis in vitro, particularly with granulocyte-like MDSCs (Gr-MDSCs), the predominant subtype of infiltrating MDSCs. In summary, our findings elucidate the likely origin of NETs in the PMN during CRLM development and underscore the significant influence of PRL-3. These findings may offer potential immunotherapeutic targets for patients at risk of developing CRLM, warranting further investigation in clinical settings.

Matrix Gla protein (MGP) and trichorhinophalangeal syndrome type 1 (TRPS1) have recently emerged as novel breast-specific immunohistochemical (IHC) markers, particularly for triple-negative breast cancer (TNBC) and metaplastic carcinoma. The present study aimed to validate and compare the expression of MGP, TRPS1 and GATA binding protein 3 (GATA3) in metastatic breast carcinoma (MBC), invasive breast carcinoma (IBC) with special features, including special types of invasive breast carcinoma (IBC-STs) and invasive breast carcinoma of no special type with unique features, and mammary and non-mammary salivary gland-type tumours (SGTs). Among all enrolled cases, MGP, TRPS1 and GATA3 had comparable high positivity for ER/PR-positive (p=0.148) and HER2-positive (p=0.310) breast carcinoma (BC), while GATA3 positivity was significantly lower in TNBC (p<0.001). Similarly, the positive rates of MGP and TRPS1 in MBCs (99.4%), were higher than in GATA3 (90.9%, p<0.001). Among the IBC-STs, 98.4% of invasive lobular carcinomas (ILCs) were positive for all three markers. Among neuroendocrine tumours (NTs), all cases were positive for TRPS1 and GATA3, while MGP positivity was relatively low (81.8%, p=0.313). In the neuroendocrine carcinoma (NC) subgroup, all cases were positive for GATA3 and MGP, while one case was negative for TRPS1. All carcinomas with apocrine differentiation (APOs) were positive for GATA3 and MGP, while only 60% of the cases demonstrated moderate staining for TRPS1. Among mammary SGTs, MGP demonstrated the highest positivity (100%), followed by TRPS1 (96.0%) and GATA3 (72.0%). Positive staining for these markers was also frequently observed in non-mammary SGTs. Our findings further validate the high sensitivity of MGP and TRPS1 in MBCs, IBC-STs, and breast SGTs. However, none of these markers are capable of distinguishing between mammary and non-mammary SGTs.

Purpose:To evaluate the diagnostic performance of a deep learning (DL) model for breast US across four hospitals and assess its value to readers with different levels of experience. Materials and Methods:In this retrospective study, a dual attention-based convolutional neural network was built and validated to discriminate malignant tumors from benign tumors by using B-mode and color Doppler US images (n = 45 909, March 2011-August 2018), acquired with 42 types of US machines, of 9895 pathologic analysis-confirmed breast lesions in 8797 patients (27 men and 8770 women; mean age, 47 years ± 12 [SD]). With and without assistance from the DL model, three novice readers with less than 5 years of US experience and two experienced readers with 8 and 18 years of US experience, respectively, interpreted 1024 randomly selected lesions. Differences in the areas under the receiver operating characteristic curves (AUCs) were tested using the DeLong test. Results:The DL model using both B-mode and color Doppler US images demonstrated expert-level performance at the lesion level, with an AUC of 0.94 (95% CI: 0.92, 0.95) for the internal set. In external datasets, the AUCs were 0.92 (95% CI: 0.90, 0.94) for hospital 1, 0.91 (95% CI: 0.89, 0.94) for hospital 2, and 0.96 (95% CI: 0.94, 0.98) for hospital 3. DL assistance led to improved AUCs (P < .001) for one experienced and three novice radiologists and improved interobserver agreement. The average false-positive rate was reduced by 7.6% (P = .08). Conclusion:The DL model may help radiologists, especially novice readers, improve accuracy and interobserver agreement of breast tumor diagnosis using US.Keywords: Ultrasound, Breast, Diagnosis, Breast Cancer, Deep Learning, Ultrasonography Supplemental material is available for this article. © RSNA, 2023.

Abstract Background Bone is one of the most frequent sites for breast cancer metastasis. Breast cancer bone metastasis (BCBM) leads to skeletal morbidities including pain, fractures, and spinal compression, all of which severely impact quality of life. Immunotherapy is a promising therapy for patients with advanced cancer, but whether it may provide benefit to metastatic bone cancer is currently unknown. Thus, a better understanding of the immune landscape of bone-disseminated breast cancers may reveal new therapeutic strategies. In this study, we use histopathological analysis to investigate changes within the immune microenvironment of primary breast cancer and paired BCBM. Methods Sixty-three patients with BCBM, including 31 with paired primary and bone metastatic lesions, were included in our study. The percentage of stroma and stromal tumor-infiltrating lymphocytes (TILs) was evaluated by histopathological analysis. The quantification of stromal TILs (CD4 + and CD8 +), macrophages (CD68 + and HLA-DR +), programmed cell death protein 1 (PD-1), and programmed cell death protein ligand 1 (PD-L1) was evaluated through immunohistochemical (IHC) staining. Statistical analysis was performed with paired t test, Wilcoxon test, spearman correlation test, and univariate and multivariate cox regression. Results Median survival after BCBM pathological diagnosis was 20.5 months (range: 3–95 months). Of the immune parameters measured, none correlated with survival after bone metastasis was diagnosed. Compared to the primary site, bone metastases exhibited more tumor stroma (mean: 58.5% vs 28.87%, p < 0.001) and less TILs (mean: 8.45% vs 14.03%, p = 0.042), as determined by H&E analysis. The quantification of primary vs metastatic tissue area with CD4 + (23.95/mm2 vs 51.69/mm2, p = 0.027 and with CD8 + (18.15/mm2 vs 58.95/mm2, p = 0.004) TILs similarly followed this trend and was reduced in number for bone metastases. The number of CD68 + and HLA-DR + macrophages showed no significant difference between primary sites and bone metastases. PD-1 expression was present in 68.25% of the bone metastasis, while PD-L1 expression was only present in 7.94% of the bone metastasis. Conclusions Our findings suggest that compared to the primary breast cancer site, bone metastases harbor a less active immune microenvironment. Despite this relatively dampened immune landscape, expression of PD-1 and PD-L1 in the bone metastasis indicates a potential benefit from immune checkpoint inhibitors for some BCBM cases.

Background Expression of basal-like markers (BMs) is usually observed in triple-negative breast cancer (TNBC) and HER2-overexpressing breast cancer and is related to poor prognosis. Recently, some have reported that BM is expressed in 7.8%-27.1% of HR+/HER2- breast cancers (basal-like phenotype, BM+ BCs), but their biological behaviours remain controversial. This study aimed to compare the differences in clinicopathologic features, clinical outcomes and benefit from adjuvant regimens between BM+ BCs and nonbasal-like phenotype breast cancer (BM- BCs) and evaluate their tumour immune microenvironment. Methods BM+ BCs were defined as EGFR- and/or CK5/6-positive (n=150). Overall, 180 BM- BCs were selected as the control cohort. Clinicopathological characteristics were retrospectively collected. Univariate and multivariate analyses were used to assess the relationship between disease-free survival (DFS), overall survival (OS) and BM status. Stromal tumour-infiltrating lymphocytes (sTILs) were evaluated on haematoxylin & eosin-stained slides, and CD3, CD8, FOPX3, chemokine (C-X-C motif) ligand 13 (CXCL13), CD68, PD-1 and PD-L1 were stained by IHC. Results Compared to BM- BCs, BM+ BCs were more frequently diagnosed at a younger age and had a higher tumour stage, histological grade and necrosis. No obvious differences were seen in lymphovascular invasion or lymph node metastasis. Lower ER and PR expression and higher Ki-67 expression were observed in BM+ BCs. Overall, 62.8% BM+ BCs had p53 mutations. Shorter DFS and OS were related to BM+ BCs (p=0.021 and 0.018, respectively). In multivariate analysis, CK5/6 and clinical stage were independent prognostic factors for DFS and OS. BM+ BCs had a higher proportion of high sTILs with increased CD3, CD8, FOXP3, CXCL13 and CD68 expression. PD-1 and PD-L1 expression was positively related to BM+ BCs. Conclusion Our study demonstrated that more aggressive morphologic features and worse prognosis were associated with BM+ BCs, and its immune-activated status may support new therapeutic strategies, such as immunotherapy, as a potential treatment.

MEX3A is an RNA-binding protein that mediates mRNA decay through binding to 30 untranslated regions. However, its role and mechanism in clear cell renal cell carcinoma remain unknown. In this study, we found that MEX3A expression was transcriptionally activated by ETS1 and upregulated in clear cell renal cell carcinoma. Silencing MEX3A markedly reduced clear cell renal cell carcinoma cell proliferation in vitro and in vivo. Inhibiting MEX3A induced G1/S cell-cycle arrest. Gene set enrichment analysis revealed that E2F targets are the central downstream pathways of MEX3A. To identify MEX3A targets, systematic screening using enhanced cross-linking and immunoprecipitation sequencing, and RNA-immunoprecipitation sequencing assays were performed. A network of 4,000 genes was identified as potential targets of MEX3A. Gene ontology analysis of upregulation of the cell proliferation pathway was highly enriched. Further assays indicated that MEX3A bound to the CDKN2B 30 untranslated region, promoting its mRNA degradation. This leads to decreased levels of CDKN2B and an uncontrolled cell cycle in clear cell renal cell carcinoma, which was confirmed by rescue experiments. Our findings carcinoma.

Background Metastatic breast carcinoma is commonly considered during differential diagnosis when metastatic disease is detected in females. In addition to the tumor morphology and documented clinical history, sensitive and specific immunohistochemical (IHC) markers such as GCDFP-15, mammaglobin, and GATA3 are helpful for determining breast origin. However, these markers are reported to show lower sensitivity in certain subtypes, such as triple-negative breast cancer (TNBC). Materials and methods Using bioinformatics analyses, we identified a potential diagnostic panel to determine breast origin: matrix Gla protein (MGP), transcriptional repressor GATA binding 1 (TRPS1), and GATA-binding protein 3 (GATA3). We compared MGP, TRPS1, and GATA3 expression in different subtypes of breast carcinoma of ( n = 1201) using IHC. As a newly identified marker, MGP expression was also evaluated in solid tumors ( n = 2384) and normal tissues ( n = 1351) from different organs. Results MGP and TRPS1 had comparable positive expression in HER2-positive (91.2% vs. 92.0%, p = 0.79) and TNBC subtypes (87.3% vs. 91.2%, p = 0.18). GATA3 expression was lower than MGP ( p < 0.001) or TRPS1 ( p < 0.001), especially in HER2-positive (77.0%, p < 0.001) and TNBC (43.3%, p < 0.001) subtypes. TRPS1 had the highest positivity rate (97.9%) in metaplastic TNBCs, followed by MGP (88.6%), while only 47.1% of metaplastic TNBCs were positive for GATA3. When using MGP, GATA3, and TRPS1 as a novel IHC panel, 93.0% of breast carcinomas were positive for at least two markers, and only 9 cases were negative for all three markers. MGP was detected in 36 cases (3.0%) that were negative for both GATA3 and TRPS1. MGP showed mild-to-moderate positive expression in normal hepatocytes, renal tubules, as well as 31.1% (99/318) of hepatocellular carcinomas. Rare cases (0.6–5%) had focal MGP expression in renal, ovarian, lung, urothelial, and cholangiocarcinomas. Conclusions Our findings suggest that MGP is a newly identified sensitive IHC marker to support breast origin. MGP, TRPS1, and GATA3 could be applied as a reliable diagnostic panel to determine breast origin in clinical practice.