Proteome profiling of serum reveals PSMD6 as a biomarker in breast cancer metastasis

Proteome profiling of serum reveals PSMD6 as a biomarker in breast cancer metastasis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Proteome profiling of serum reveals PSMD6 as a biomarker in breast cancer metastasis Chen Ding, Yue Meng, Minjing Huang, Ganfei Xu, Xinwei Li, Bing Gu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3634466/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Breast cancer (BC) has the highest mortality rate and prevalence among cancers in females worldwide. Here, we performed proteomic profiling of 322 serum samples from the discovery cohort [56 healthy controls (HCs), 112 benign breast tumor (BBT) patients, and 154 BC patients] and a prospective validation cohort [27 HCs, 29 BBT patients and 57 BC patients]. Integrated proteomic analysis of tissue and serum samples revealed highly specific tumor biomarkers and demonstrated that the serum proteome can distinguish the different pathological substages in BC progression. We also identified PSMD6 as a potential metastatic breast cancer (MBC) biomarker. Comprehensive analysis of the multicenter independent validation cohort, which included retrospective and prospective cohorts including 61 HCs, 72 BBT patients, and 247 BC patients, indicated that PSMD6 overexpression was an important cause of BC metastasis and an indicator of poor prognosis. Further study revealed that the CLTA-PSMD6-neutrophil axis promotes the transition from invasive ductal carcinoma (IDC) to MBC. Importantly, CLTA amplification might be a potential therapeutic target for MBC patients. We also developed a highly accurate predictive model (accuracy = 0.87) to differentiate benign and malignant tumors and validated its good performance in the prospective validation cohort. Collectively, this study demonstrates the elaborate BC serum proteomic landscape and provides valuable information regarding serum biomarkers, which could reveal novel therapeutic targets and provide opportunities for MBC treatment. Biological sciences/Cancer/Breast cancer Biological sciences/Biochemistry/Proteomics/Protein–protein interaction networks Breast cancer Serum proteome Biomarkers Metastasis PSMD6 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 INTRODUCTION Breast cancer (BC) is one of the most commonly diagnosed cancers in women 1 and presents high heterogeneity in its morphology, molecular expression profile and clinical course 2 . Breast ductal carcinoma (BRDC) is the most common type of BC and has unique clinical and pathological features. BC progression involves a complex multistep process with intermediate transition mechanisms that are difficult to monitor. Histopathologically, the progression of human BRDC is a linear multistep process that begins with a lesion, which progresses to ductal carcinoma in situ (DCIS), develops into microinvasion carcinoma (DCIS-MI), evolves into invasive ductal carcinoma (IDC) and ultimately develops into potentially metastatic breast cancer (MBC) 3,4 . In addition, fibroadenoma and papilloma are common clinically detected benign breast tumors (BBTs). Although BC has characteristic imaging features, BBT and BC have similar presentations, leading to difficulty in making an accurate diagnosis through imaging. Thus, early-stage precision diagnosis and evaluation of metastasis status have significant value for prognosis evaluation and therapy monitoring for BC 5 . Currently, tissue biopsy is the gold standard for clinical diagnosis 6 . However, tissue biopsy has several limitations: surgery is invasive; biopsy can only provide information on local tumor tissue with high heterogeneity; biopsy does not provide sufficient tissue to generate multiple tissue sections; and there is resistance for seeking such services on a yearly basis 7 . Different types of noninvasive diagnostic techniques have been developed to comprehensively assess the characteristics of different tumors. For example, endoscopy is commonly used for early diagnosis of gastrointestinal (GI) cancer; cystoscopy techniques have been used to improve the accuracy of tumor detection for urinary system malignancies; and methods such as dermoscopy have been used as additional clinical diagnostic aids for skin carcinoma 8–10 . One noninvasive approach for BC diagnosis and evaluation of metastasis status is imaging examination 9 . Imaging examinations include mammography, magnetic resonance imaging (MRI), ultrasound, computerized tomography (CT) and positron emission tomography (PET-CT) 11,12 . With improvements in breast imaging, mammography, ultrasound and minimally invasive interventions, the detection rate of early breast cancer, noninvasive cancers, lesions of uncertain malignant potential, and benign lesions has increased 13,14 . However, with improved diagnostic capabilities, there is a substantial risk of false-positive findings for benign tumors and, conversely, false-negative findings for malignant tumors 15,16 . Today, noninvasive methods for sampling materials for biomarkers are being intensively developed 17,18,19 . Liquid biopsy tests showed promise for early cancer detection, tumor classification, and monitoring treatment response 20,21 . Compared with traditional detection methods for tumor diagnosis, liquid biopsy possesses many advantages; it is less invasive and easier to perform and has fewer complications and stronger dynamic monitoring ability, and a higher acceptance rate 18 . Serum, a bodily fluid, could represent an essential component of the liquid biopsy test 22 . Serum protein assessment is a straightforward method that can be performed routinely and frequently 23,24 . Therefore, there has been much interest in the development and validation of serum-based biomarkers for the early detection, risk stratification, and prognosis prediction for breast cancer 25,26,27 . Serum is one of the predominant sample types used for diagnostic analyses in clinical practice, and serum samples from thousands of clinical studies are available in biobanks 28,29 . The serum proteome contains a set of tissue proteomes 29 . Serum proteins secreted by tumors are involved in various biological functions and are an important source of cancer biomarkers 29,30,31 . Thus, the levels of serum proteins and/or their changes between conditions provide information about the physical condition and health status of patients and can be used to track disease progression 32,33 . Several conventional blood biomarkers that are used in the clinic, such as carcinoembryonic antigen (CEA), cancer antigen 153 (CA153), carbohydrate antigen 125 (CA125), carbohydrate antigen 199 (CA199), prostate-specific antigen (PSA), and alpha fetoprotein (AFP), have been reported to be helpful for tumor detection 34,35,36,37,38,39,40,41,42 . Specifically, CEA is an important marker for colon cancer and some other carcinomas 34,35 ; CA199 is present in high concentrations in the serum of patients with pancreatic carcinoma 39,40 ; and AFP is a highly specific and sensitive marker of hepatocellular carcinoma 41,42 . However, there is a lack of published articles on serum biomarkers for BC clinical diagnosis, and knowledge of the mechanisms underlying BC is limited. According to the 2020 IARC survey, BC has become the leading cause of cancer-related death for female patients 43 . Metastasis is mainly responsible for treatment failure and is the cause of most BC-related deaths 44,45,46 . As reported, metastatic breast cancer (MBC) cells acquire aggressive characteristics through several mechanisms, including epithelial–mesenchymal transition (EMT), tumor angiogenesis and metabolic programming 47,48,49,50,51,52 . Characterizing these mechanisms thoroughly may help to stop the development of tumor metastasis and to provide strategies for precise treatment. However, although the transition from IDC to MBC is central to the poor prognosis, little is known about the time of onset or the triggering mechanism by which invasive BC becomes metastatic BC in humans. Moreover, there is a lack of serum biomarkers for MBC diagnosis. This study aimed to uncover the relevant molecular mechanisms of breast cancer metastasis and to explore potential serum biomarkers that may be used for highly sensitive and rapid MBC diagnosis 53,54 . In this study, we collected 813 samples (322 serum samples from the discovery cohort and 365 serum samples and 126 tissue samples from the validation cohort). We performed a quantitative proteomic approach with data-independent acquisition (DIA) on 322 serum samples that consisted of 56 healthy control samples, 91 BBT-fibroadenoma samples, 21 BBT-papilloma samples, and 154 BC samples. The integrated tissue-serum proteomic approach identified tumor biomarkers for identifying patients with BC. Furthermore, we established a predictive model based on 24 features that distinguished HC, BBT and BC samples with good performance. The robustness of this predictive model was further validated in the independent validation cohort. In addition, we found that the CLTA-PSMD6-neutrophil axis promoted BC metastasis. Collectively, this study revealed the BC serum proteomic landscape in what may be the largest cohort to date and provided valuable information on serum biomarkers, which could facilitate the improvement of the clinical diagnosis and management of BC. RESULTS Overall outline of the serum proteome profiling study in BC patients To investigate the proteomic expression patterns of BC, we collected 332 serum samples from the Guangzhou 0 discovery cohort composed of three independent cohorts, including the HC cohort (n = 56), BBT cohort (n = 112: 91 BBT-fibroadenoma patients and 21 BBT-papilloma patients), and BC cohort (n = 154). Notably, the BC subtypes included ductal carcinoma in situ (DCIS, n = 25), ductal carcinoma in situ with microinvasion (DCIS-MI, n = 16), invasive ductal carcinoma (IDC, n = 68), and metastatic breast cancer (MBC, n = 45) (Fig. 1A). We acquired serum proteome profiles of all samples using a data-independent acquisition (DIA) strategy 55 on a Q Exactive HF-X Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Fisher Scientific, Rockford, IL, USA) coupled with a high-performance liquid chromatography system (EASY nLC 1200, Thermo Fisher Scientific) (Fig. 1A; STAR Methods). The overall workflow of this study is shown in Figure. 1A. The demographic and clinical data of all the study participants are summarized in Table S1. Additionally, the clinical information of 306 female patients, including age, histological stage, degree of differentiation, TNM stage (AJCC cancer staging system 8th edition), and clinical subtype, is summarized in Table S1. To monitor the liquid chromatography-tandem mass spectrometry (LC‒MS/MS) platform instrument stability, a mixture of all serum samples from patients with BC was assessed every 20 samples; this approach is generally used in proteomic studies. The quality control (QC) samples were analyzed using the same method and conditions used for our cohort serum samples 56,57,58 . The average Pearson’s correlation coefficient, calculated for all quality control runs of QC samples, was 0.98 (range from 0.93 to 0.99), demonstrating the consistent stability of the MS platform (Figure S1C). A total of 8,944 protein groups were identified in all the serum samples (Figure S1A), with an average of 1,881, 1,908 and 1,889 protein groups per BC, BBT and HC samples, respectively (Fig. 1B and 1C). Proteome quantification was conducted using the intensity-based absolute quantification (iBAQ) algorithm, followed by fraction of total (FOT) normalization as reported previously 59 . In addition, the proteome was highly dynamic, spanning approximately eight orders of magnitude, as indicated by the protein abundance (FOT) values (Fig. 1D). The distribution of log2-transformed FOT values of identified proteins in 322 serum samples is shown in Figure S1B, and the consistency among the samples further indicated the stability of our mass spectrometry platform. Additionally, as shown in Fig. 1E, the number of proteins identified in BC serum that were annotated as serum proteins was not significantly different from that in fibroma, papilloma and normal serum. Interestingly, between BC samples and other serum samples, a slightly higher number of proteins were annotated as cancer-related proteins, proteins highly expressed in the breast, CD markers and proteins associated with drugs approved by the US Food and Drug Administration (FDA) were identified in BC. Our study has thus far established a comprehensive serum proteomic landscape of breast carcinoma. Serum proteomic profiles differ between BC and non-BC samples To investigate the proteome-level differences between BC and non-BC, we performed principal component analysis (PCA) among BC samples and non-BC samples. PCA of BC proteomes and non-BC proteomes showed a relatively obvious separation between BC samples and non-BC samples (including HC, BBT-fibroadenoma, and BBT-papilloma), which indicated that protein variation between the two kinds of samples exceeded the variation between individuals (Fig. 2A). Interestingly, in the non-BC samples, we observed that the two BBT types (BBT-papilloma and BBT-fibroadenoma) clustered together, while the HCs were somewhat distant from them. These results showed a clear distinction inside the proteome of non-BC, revealing a significant molecular difference between the proteomes of HCs and BBTs. To further elucidate serum proteomic expression patterns among the two different samples, we compared the serum proteome profiles of patients with BC to those of non-BC patients. The results revealed a dramatic shift in protein expression profiles; specifically, there were 853 significantly differentially expressed proteins (DEPs), of which 447 DEPs were upregulated and 391 DEPs were downregulated (Student’s t test, p value 2 or < 0.5) (Fig. 2B; Table S2). Pathway enrichment analysis showed that the BC-specific proteins were mainly involved in neutrophil degranulation (Fig. 2D). Furthermore, protein‒protein interaction (PPI) analysis showed that PSMD6, which was upregulated in BC, was associated with neutrophil degranulation, further supporting these enrichment results (Fig. 2E). PSMD6 is a component of the 26S proteasome and shows a correlation with poor prognosis based on an external data source 60 (Tang et al.,2018). We also compared the serum proteome profiles of patients with BC to the BC tissue proteome (Xu et al) and reasoned that ideal biomarkers should be commonly overexpressed in the tumor tissue (upregulated in the tumor tissue sample) and released into the serum (upregulated in the serum samples). To further explore biomarkers for breast cancer, an additional tissue cohort from BC treatment-naïve female patients was assessed in this research. We compared the serum proteome profiles of patients with BC to that of the BC tissue proteome (Fig. 2E). In total, 8,462 proteins were commonly quantified in the BC tissue and serum samples (Fig. 2E). We also confirmed the significant positive correlation between the BC tissue proteome and serum proteome (Fig. 2F, Spearman’s rho = 0.323, p value < 1E-204). Moreover, comparing the two different proteome profiles between BC and HC, 1,436 proteins that were upregulated in BC in both the serum proteome and the tissue proteome were enriched in neutrophil degranulation, etc. (Figs. 2G–- 2H). In a previous report, tumor-associated neutrophils (TANs) were shown to participate in tumor-promoting inflammation by driving angiogenesis, extracellular matrix remodeling, metastasis and immunosuppression. We suggest that a certain relationship exists between BC and neutrophils. We further performed supervised analysis to filter out significant DEPs among neutrophil degranulation at the serum proteome level and obtained 22 significant DEPs (Wilcoxon signed-rank test, p value < 0.05), in which PSMD6 overlapped with previous BC PPI network analysis (Figs. 2I-2K). PSMD6 was overexpressed in BC in both serum and tissue samples, which suggested that PSMD6 might be generated by BC cells and released into the serum. In conclusion, our proteomic analysis revealed significant differences between BC and non-BC samples, which manifested as changes in the abundance of proteins in the serum proteome profiles of patients as well as the enrichment of biological processes directly related to the disease phenotype. Pathway enrichment analysis showed that tumor-induced neutrophil degranulation was enriched in BC and revealed PSMD6 as a potential serum marker associated with tumor progression. Serum proteomic profiling distinguishes patients with BBT from those with BC Clinically, there is often confusion and misdiagnosis regarding BBT and BC differentiation via early screening methods. To further elucidate serum proteomic expression patterns among HC, BBT (including BBT-fibroadenoma and BBT-papilloma), and BC patients in our cohort, we set criteria for defining tumor-specific serum proteins: tumor-specific serum protein expression values in one tumor group should be at least 1.5-fold higher than those in any of the other 3 groups (Kruskal-Willis test, p value < 0.05). Consequently, we identified 1,130 significantly DEPs among the four groups (healthy control: 405 proteins, BBT-fibroadenoma: 130 proteins, BBT-papilloma: 336 proteins, and BC: 259 proteins) (Figs. 3A and 3B; Table S3). To investigate the serum proteomic features of these breast tumor groups, we performed pathway enrichment analysis for these 1,130 DEPs. Fibroadenoma-specific proteins significantly converged on pathways including TP53-regulated metabolic genes, tight junction, RhoG GTPase cycle, etc.; papilloma-specific proteins were enriched in adherens junction, vesicle-mediated transport, cell cycle, etc.; BC-specific proteins participated in VEGF, histidine metabolism, immune system, tyrosine metabolism, fructose and mannose metabolism, innate immune system, proteasome degradation, etc.. Compared with the other 3 groups (HC, BBT-fibroadenoma, and BBT-papilloma), neutrophil degranulation was mainly enriched in BC samples as previously shown (Fig. 3C). These results further indicated that a certain relationship exists between neutrophils and tumorigenesis. Notably, PSMD6, which is upregulated in BC, is a significant factor in the neutrophil degranulation pathway (Figure S2A). Furthermore, we employed the eXtreme Gradient Boosting algorithm based on the differentially expressed proteins and routine blood indexes (ANOVA test, p value < 0.05) to identify a subset of features (top features: OXCT1, HDGFL3, DDX39B, ACO1, SART3, COG1, ACTR2, QDPR, UMOD, DENND4C, CSTA, AOC2, RGN, MVP, TRAP1, UBE2L5, PTGFRN, SMARCC2, FKBP15, OXSR1, PLXNB1, TTR, PSMB3, and NEUT) that discriminated HC, BBT, and BC samples (named the HC/BBT/BC-sig) (Fig. 3D; STAR Methods). To train and subsequently test the classifier, samples from the Guangzhou 0 discovery cohort were grouped based on their type, and 70% and 30% were used as the training and testing sets, respectively. Based on the HC/BBT/BC-sig, 10-fold cross-validation in the training samples (70% of the cohort) yielded a predictive model with 0.91 accuracy and 0.94 precision for distinguishing HC, BBT and BC samples (Fig. 3D). When applied to the testing samples (30% of the cohort), the predictive model achieved high accuracy of 0.91. To further evaluate the accuracy of the predictive signatures for discriminating BBT and BC, we recruited a prospective follow-up validation cohort, named Guangzhou 1, and these patients provided 103 serum samples, including 27 HC, 29 BBT, and 57 BC samples. Based on the DIA approach, we performed comparative analysis of signature proteins among HC, BBT, and BC. As a result, we obtained the significantly differentially expressed genes of these signature features among HC, BBT, and BC. We observed that 9 features (OXCT1, HDGFL3, DDX39B, ACO1, QDPR, UMOD, DENND4C, and SMARCC2) were significantly increased in the HC group compared with the BC and BBT groups (Kruskal‒Wallis test, p value < 0.05). We observed that 4 features (AOC2, RGN, TRAP1, and UBE2L5) were significantly upregulated in BBT samples compared with BC and HC samples (Kruskal‒Wallis test, p value < 0.05). Twelve signature features (SART3, COG1, ACTR2, CSTA, MVP, PTGFRN, FKBP15, OXSR1, PLXNB1, TTR, PSMB3, and NEUT) were significantly overexpressed in BC samples compared with BBT and HC samples (Kruskal‒Wallis test, p value < 0.05). The molecular expression trends for the BBT/BC/HC-sig were consistent between the Guangzhou 0 discovery cohort and the Guangzhou 1 prospective validation cohort. In addition, pathway enrichment analysis showed that the BC-specific pathways in the validation cohort (Kruskal‒Wallis test, p value < 0.05) were consistent with those in the Guangzhou 0 discovery cohort, such as neutrophil degranulation. The predictive model, which was constructed based on the Guangzhou 0 discovery cohort, had high accuracy and good prediction performance (accuracy = 0.87) in the Guangzhou 1 prospective validation cohort (Fig. 3E). The heatmaps showed a clear separation among HC, BBT and BC samples in both the Guangzhou 0 discovery cohort and the Guangzhou 1 prospective validation cohort (Figs. 3F–- 3G). Collectively, the predictive power of the signature proteins in different breast tumors (including BBT-fibroadenoma, BBT-papilloma, and BC) was validated in the Guangzhou 1 prospective validation cohort, and the results indicated that the predictive models in the discovery cohort exhibited robustness, accuracy, and stability in the Guangzhou 1 prospective cohort. Taken together, our proteomic analysis showed neutrophil degranulation as the key signaling pathway in tumorigenesis; PSMD6 was defined as a core tumor-related protein in this pathway. The classifier derived from the Guangzhou 0 discovery cohort could be a potential predictive model for distinguishing BBT and BC samples and achieved good performance in the Guangzhou 1 validation set. Proteomic kinetic changes in BC progression In the past decade, many studies have characterized the multiomics landscape of certain stages of the progression of BC. In our study, we also observed that the serum proteome distinguished 4 pathological phases for BC, including DCIS (n = 25), DCIS-MI (n = 16), IDC (n = 68) and MBC (n = 45). To uncover protein patterns associated with tumor progression, we performed further analysis on the different phases of BC progression. We identified 1,313 proteins that were differentially expressed among HCs and at least one phase of tumor progression (Kruskal‒Wallis test, p value 2 or < 0.5) and calculated the Z scored intensities of these altered proteins (Fig. 4A; Table S4). The clustering analysis of these proteins using the fuzzy c-means algorithm identified 6 clusters of protein trajectories that changed with progression with a range of sizes (from 134 to 327 proteins per cluster) (Fig. 4B). Assessing these proteins would provide many potential candidate biomarkers for early BC screening and advance our understanding of the clinical features of BC progression. To explore the biological function of the groups with distinct expression patterns, we performed pathway enrichment analysis of clusters 1–6 (Fig. 4C; Table S4). These results revealed that cluster 1 (HC cluster) was mainly enriched in integrin and thiamine metabolism. Cluster 2 (DCIS cluster) was enriched in the regulation of the actin cytoskeleton and vesicle-mediated transport. Proteins in cluster 3 (DCIS-MI cluster) were mainly involved in IL-7 signaling and VEGF. Cluster 4 (IDC cluster) was enriched in fatty acid oxidation and signaling by receptor tyrosine kinases. Proteins in cluster 5 (MBC cluster) were mainly involved in neutrophil degranulation and beta-alanine metabolism. Cluster 6 (BC-common cluster) was mainly enriched in apoptosis and spliceosomes. Furthermore, protein‒protein interaction (PPI) analysis of neutrophil degranulation showed that PSMD6 was involved in featured pathways in the MBC cluster, implying that both PSMD6 and neutrophils are associated with BC metastasis (Fig. 4D, Figure S2). To further identify potential peripheral serum biomarkers for the four phases of BC progression, we collected 8,462 overlapping proteins of BC between tissue and serum samples. Then, by assessing the expression levels of the above 8,462 proteins, we found that 642 out of 8,462 were upregulated in BC in both the serum proteome and tissue proteome (ratio > 1.5), indicating that these 642 serum proteins might be derived from BC tissues. In particular, 62 of the 642 proteins were significantly overexpressed in tissue samples (Kruskal‒Wallis test, p value < 0.05). Furthermore, we identified 30 phase-specific proteins (DCIS: n = 7; DCIS-MI: n = 12; IDC: n = 7; MBC: n = 4) among these 62 proteins as hub tumor biomarkers (Kruskal‒Wallis test, p value 2) (Fig. 4E). Interestingly, we identified that PSMD6 was significantly overexpressed in MBC, consistent with the findings of the PPI analysis. We identified PSMD6 as a potential candidate biomarker for the MBC phase. Taken together, these findings suggested that PSMD6 could not only serve as a serum biomarker for BC metastasis but also be involved in the MBC stage and that neutrophils are also involved in this stage. The CLTA-PSMD6-neutrophil axis promotes breast cancer metastasis Aiming to mine the correlation between clinical indicators and pathological stages, we performed weighted gene coexpression network analysis (WGCNA) (STAR Methods), which is an unsupervised method to identify groups of coregulated proteins and their association with clinical variables 22 . (Fig. 5A; Table S5). Separating the proteomic profiles into eigengene modules revealed a group of modules positively associated with tumor metastasis (p value < 0.0001, rho = 0.93). WGCNA identified 7 protein modules, and the number of proteins in different modules ranged from 49 proteins in the brown module related to DCIS-MI to 110 proteins in the turquoise module related to fibroadenomas. Moreover, the findings demonstrated an extensive connection between modules and tumor progression. For instance, the green module (DNM1, NANS, EIF2S1, PSMD6, etc.) was significantly correlated with MBC (Pearson’s rho = 0.95, p value < 1E-136), and it contained all MBC hub proteins (Fig. 5B, Fig. 5C). Furthermore, the green module was strongly associated with the neutrophil ratio, indicating that there was a certain relationship between the neutrophil and the MBC stage (Pearson’s rho = 0.2, p value < 0.0001). Moreover, we found that the PSMD6 expression level was highly correlated with a poor prognosis (Tang et al., 2018, Fig. 5D). To clarify the relationship among PSMD6, neutrophils and tumor metastasis, we performed cell type deconvolution analysis using CIBERSORT ( https://cibersortx.stanford.edu ) to infer the relative abundance level of different cell types in the tumor microenvironment and then evaluated the correlation between PSMD6 and neutrophil score (Table S5). PSMD6 was upregulated in BC in both the serum proteome and tissue proteome, implying that PSMD6 might be produced by tumor tissue and secreted into the serum. Furthermore, to elucidate the mechanism underlying the metastasis ability of MBC and how this ability is affected by changes in PSMD6, which may occur due to genomic alterations, we assessed a set of tissues (Xu et al.) in this study (Fig. 5E). From this analysis, we observed cis -effects for 9 CNA-affected proteins between IDC and MBC, and these proteins were significantly upregulated in MBC (Student’s t test, p value 2) (Figs. 5F-5G). Among these proteins, CLTA was the most highly expressed protein, and it could promote protein transmembrane transport. These findings indicated that the cis -effect of CLTA could account for the difference between IDC and MBC. To further explore how CLTA is involved in tumor metastasis, we performed gene set enrichment analysis (GSEA) for pathway enrichment analysis and also performed correlation analysis. The results showed that the enrichment score of protein secretion signaling had a significant positive correlation with the CLTA protein expression level (Spearman’s rho = 0.57, p value < 0.00001) (Fig. 5H). These findings prompted us to conclude that the cis -effect of CLTA would give rise to tumor metastasis by facilitating protein secretion. Through correlation analysis, we observed 48 DEPs significantly positively correlated with CLTA that were secreted proteins between the IDC and MBC phases (Spearman’s rho > 0.2, p value < 0.05) (Fig. 5K). PSMD6, the top-ranked protein, was notably associated with the neutrophil score (Figs. 5L– 5 M; rho = 0.322, p value = 9E-3). Compared with IDC, MBC displayed a higher neutrophil score (Student’s t test, p value < 0.01, MBC/IDC ratio = 3.9) (Fig. 5N). The neutrophil score between IDC and MBC was also strongly related to angiogenesis, which was enriched in MBC (Figs. 5O-5P). In summary, through integrated analysis of the tissue proteome and serum proteome, we found that the CLTA-PSMD6-neutrophil axis promotes tumor metastasis. Specifically, the results demonstrated that the cis-effect of CLTA probably promotes intracellular to extracellular transport of PSMD6; then, BC cell-derived PSMD6 possibly activates neutrophils; finally, neutrophils might induce tumor cell metastasis by stimulating angiogenesis (Fig. 5Q). Overexpression of PSMD6 synergistically enhances angiogenesis and NET formation by activating neutrophils Proteomic analysis revealed that PSMD6 was one of the MBC-specific proteins (Fig. 6A), and the expression level was highly positively correlated with the hazard ratio (Fig. 5D). To validate the clinical value of PSMD6, we enrolled a second prospective follow-up cohort, named Guangzhou 2. The Guangzhou 2 cohort included 61 HCs, 72 BBT patients, and 121 BC patients, who provided 252 serum samples. The level of PSMD6 in serum was detected by ELISA in the Guangzhou 2 validation set, which suggested that the serum PSMD6 level was significantly higher in patients with metastasis than in healthy controls and patients with benign and invasive breast cancer (Fig. 6B). Then, the amplified expression of PSMD6 in breast cancer tissues was confirmed by IHC staining in a subsequent step (Fig. 6C), which indicated that the expression level of PSMD6 was higher in patients with lymph node metastasis and AJCC stage III or IV disease (Table S6). Moreover, as multicenter cohorts are supposed to be representative of the general population, we recruited a third validation cohort (named Shanghai) from another center, which could make our study clearer results more convincing and more broadly accepted; this cohort provided 126 BC tissue samples. Kaplan–Meier survival analysis and Cox proportional hazards analysis of survival data based on the Shanghai validation set showed that breast cancer patients with high PSMD6 expression levels had significantly shorter overall survival times, indicating that high PSMD6 levels are independently associated with poor outcomes in breast cancer patients (Fig. 6D) (Table S6). Next, we explored how PSMD6 was transferred from the cytoplasm to the extracellular space. Our results indicated that CLTA was involved in tumor metastasis and that the enrichment score of the protein secretion signaling gene set had a significant positive correlation with the CLTA protein expression level (Figs. 5E-5K). Immunofluorescence analysis showed that PSMD6 colocalized with CLTA (Fig. 6E), and knockdown of CLTA by three kinds of shRNA significantly inhibited the PSMD6 protein levels in MCF-7 and MDA-MB-231 cells and their culture medium (Figs. 6E-6H). Therefore, these findings prompted us to conclude that CLTA promotes the transfer of PSMD6 from the cytoplasm to the extracellular space. Then, we explored the functional role of PSMD6 in breast metastasis. We observed that overexpression of PSMD6 could promote cell metastasis in a transwell assay, while knockdown of PSMD6 could inhibit cell metastasis (Figs. 6I-6K). Moreover, PSMD6 overexpression in MDA-MB-231-LM2 cells, a cell line with moderate endogenous PSMD6 expression (Figure S4A), significantly exacerbated the lung metastatic burden after intravenous inoculation of cancer cells into mice (Figs. 6L-6N). These findings demonstrated a prometastatic role of PSMD6 in breast cancer. To explore how serum PSMD6 regulates lung metastasis, we first investigated its effect on tumor migration and invasion in vitro . However, we found that the cell culture supernatants of the PSMD6-overexpressing cell line (PSMD6 OE cells) or the shPSMD6 cell line (shPSMD6 cells) did not affect cell invasion and metastasis, suggesting a microenvironment-dependent role of PSMD6 in metastasis. After coculturing the neutrophils from healthy volunteers with PSMD6 OE cell culture medium or shPSMD6 cell culture medium, the cell culture supernatants promoted or inhibited control breast cancer cell invasion and metastasis in vitro (Figs. 7A-7B, Figure S4B). In addition, our results suggested that the supernatants from PSMD6 OE cells could induce more neutrophil-like cells (dHL-60) to transform into CD66b + CD11b + cells, which means that dHL-60 cells were activated by the supernatants from PSMD6 OE cells, while supernatants from shPSMD6 cells could inhibit neutrophil-like transformation (Figs. 7C-7E). Importantly, flow cytometry analyses also indicated that PSMD6 overexpression increased, while PSMD6 knockdown decreased, the percentages of CD11b + Ly6G + neutrophils, which are tumor-associated neutrophils, in mouse lung metastases (Fig. 7F, Figure S4C). Thus, tumoral PSMD6 may regulate lung metastasis by activating neutrophils. Then, we investigated how PSMD6 affects neutrophils to promote breast cancer metastasis. Previous studies have reported that metastatic cancer cells can induce neutrophils to form metastasis-supporting NETs in the absence of infection 17 . Interestingly, we found that neutrophils cultured in PSMD6-overexpressing LM2 cell culture medium (PSMD6-OE CM) formed more extensive NET structures than those cultured in control medium, as evidenced by IF staining of citrullinated histone H3, a hallmark of chromatin decondensation and extrusion, and the granule protein myeloperoxidase (Figs. 7G-7H, Figure S4D). In addition, the number of free NETs in neutrophils cultured in PSMD6-OE CM was also higher than that in neutrophils cultured in control CM, while it was lower in neutrophils cultured in PSMD6 shRNA MDA-MB-231 cell culture medium (shPSMD6-CM) (Fig. 7I). Importantly, PSMD6 overexpression in 4T1 cells led to enhancement of NETosis near cancer cells in the lungs and after intravenous inoculation of cancer cells, while PSMD6 knockdown in 4T1 cells had the opposite effect (Figs. 7J-7K, Figures S4E-S4F). In addition, angiogenesis is necessary at the beginning of metastasis. Tumor cells must gain access to the vasculature from the primary tumor, survive travel through the circulation, settle in the microvasculature of the target organ, escape from the vasculature into the target organ, and induce angiogenesis in the target organ. Researchers have shown that tumor-associated neutrophils can participate in tumor-promoting inflammation by driving angiogenesis [ 61 ] . Our results showed that the expression of angiogenesis-related proteins in neutrophils after coculture with PSMD6-OE CM was also upregulated, while it was downregulated after coculture with shPSMD6-CM (Figs. 7L-7M; Table S7). Consistently, after coculturing PSMD6-OE CM with neutrophils, breast cancer cell tube formation was stimulated (Figs. 7N-7Q). In summary, PSMD6 could affect neutrophils to promote breast cancer metastasis by regulating NET formation and angiogenesis-related protein expression. Next, we further investigated the molecular mechanism by which PSMD6 acts on neutrophils in tumor cell medium. To determine the protein of neutrophils that interact with PSMD6, recombinant PSMD6 protein with a GST tag was used to pull down the plasma protein of neutrophils from breast cancer patients and healthy volunteers. Complexes recovered from the beads were analyzed by MS and western blotting, and the results suggested that PR3 and CD177 are proteins that may interact with PSMD6 (Figs. 7R-7T, Figures S4G-S4H; Table S7). PR3, also called myeloblastin, is a neutrophil serine protease 62 . It has been shown that the glycosylphosphatidylinositol (GPI)-anchored neutrophil-specific receptor NB1 (CD177) presents PR3 on the membrane of a neutrophil subset 63 , raising the possibility that tumor-derived PSMD6 might directly regulate neutrophil membrane-bound PR3. To demonstrate that PSMD6 could activate neutrophils by regulating neutrophil membrane-bound PR3, the enzymatic activity of membrane-bound PR3 was assessed after treating cells with CM from PSMD6-overexpressing cancer cells. CM from PSMD6-overexpressing cancer cells could activate the membrane-bound PR3 of human primary neutrophils, while the inhibitor of PR3 (sivelestat) could diminish the activation effect of CM from PSMD6-overexpressing cancer cells (Fig. 7U). In addition, previous studies showed that membrane-bound PR3 of human neutrophils could regulate neutrophil chemotaxis [ 64 ] . In our study, we showed that CM from PSMD6-overexpressing cancer cells could significantly enhance angiogenesis-related protein expression, while the inhibitor of PR3 (sivelestat) could inhibit PR3 expression, diminish the effect of PSMD6 and inhibit angiogenesis-related protein expression in neutrophils cocultured with PSMD6-OE CM (Figs. 7V-7W; Table S7). Therefore, our results indicated that PSMD6 could be transferred from the cytoplasm to the extracellular space with the aid of CLTA and that PSMD6 could activate neutrophils to release angiogenesis-related proteins by interacting with the PR3-CD177 axis, which ultimately promotes breast cancer cell metastasis (Fig. 7X). DISCUSSION The World Health Organization (WHO) has proposed that millions of cancer patients could be saved from premature death if early detection and treatment methods were available 65 . Finding the tumor at an early stage when it is still localized and possibly even before clinical symptoms develop is one important application of specific biomarkers 65 . Apart from early diagnosis, biomarkers could also provide physicians with actionable information leading to evidence-based selection of the optimal therapy and to improved and more precise methods for detecting disease progression 65 . Ideally, protein biomarkers should be found via minimally invasive liquid biopsy, such as by taking a simple blood sample 65 . Currently, although a few biomarkers — for example, CA125 for ovarian cancer, CA19-9 for pancreatic cancer and PSA for prostate cancer — have been proposed to be useful for longitudinal disease monitoring, there is still a lack of breast cancer-specific biomarkers that are associated with clinical problems, rather than just the differentiation of cancer patients from healthy individuals. Here, we portrayed the serum proteomic landscape and explored proteomic signatures associated with the progression of BC. Our study included a discovery stage and three validation stages involving a total of 813 samples. The discovery stage was designed as a case‒control study and involved 322 samples, including 54 HCs, 112 BBTs and 156 BCs. The BC samples included those of various pathological stages to represent BC progression; specifically, they included 25 DCIS samples, 16 DCIS-MI samples, 68 IDC samples, and 45 MBC samples. Furthermore, we conducted three-step validation based on three independent cohorts (Guangzhou 1, Guangzhou 2, and Shanghai). For the Guangzhou 1 validation cohort, a prospective follow-up cohort study was conducted to verify the effectiveness of the BC diagnostic classifier. This cohort involved 113 serum samples, including 27 HCs, 29 BBTs, and 57 BCs. For the Guangzhou 2 validation cohort, a prospective follow-up cohort study was conducted to verify PSMD6 as a potential specific serum biomarker for BC metastasis. This cohort provided 252 serum samples, including 61 HC samples, 72 BBT samples, and 121 BC samples. For the Shanghai validation cohort, a retrospective cohort study was conducted to verify PSMD6 as a potential specific biomarker of BC metastasis associated with poor prognosis. The Shanghai validation cohort provided 126 BC tissue samples. To our knowledge, this is the largest study cohort to comprehensively explore the dynamic changes in the serum proteome and protein signature in BC progression. Currently, ultrasonography and mammography are used together with histopathological confirmation as the gold standard for BC diagnosis 66 . There are, however, several disadvantages of the abovementioned modalities, which makes it reasonable to continue research and development in the area of alternative methods for diagnosing BC 66 . Since blood-derived proteins are readily available and perform vital activities, blood-derived protein biomarkers introduce significant potential in BC diagnosis as a complementary and adjunctive modality to the current clinical gold standard. In this study, we observed distinct serum proteomic profiles between subjects with BCs and non-BCs (HCs and BBTs). The 24-feature classifier was further constructed for distinguishing between BCs and non-BCs based on a machine learning method, and the model showed 94% sensitivity and 94% specificity in the diagnosis of BCs. In addition, to validate the predictive strength of our analysis, we entered the validation stage, and the model showed 80% sensitivity and 80% specificity in our Guangzhou 1 validation cohort. Our results showed that the 24-feature classifier can distinguish breast cancer patients from healthy individuals and can also distinguish benign breast tumors from malignant breast tumors. This study may provide a reference value for differentiating BC and non-BC using serum in the future. The 26S proteasome is an important protease in eukaryotic cells and is composed of a 20S core particle (CP) and one or two 19S regulatory particles (RPs) capping one or both ends of the 20S CP 67 . Some studies have shown that the 26S proteasome plays a significant part in tumor progression. For example, Türkoğlu et al. reported that PSMD4 promotes cell proliferation via regulation of the PTEN/Akt pathways in hepatocellular carcinoma (HCC) cells 68 . Okumura et al. reported that PSMD1 is a gene associated with acquired tamoxifen resistance that may contribute to the proliferation of breast cancer cells putatively through modulating the p53 pathway 69 . In this study, we disclosed the serum proteomic features of various histopathological subtypes of BC progression and found that PSMD6 was the top MBC-specific protein that was considered a risk factor for BC metastasis. High expression of PSMD6 was associated with poor BC prognosis in our Shanghai validation cohort. Integrated analysis of Xu et al.’s BC tissue proteomics data (n = 402) and our serum proteomics data showed that PSMD6 was also upregulated in MBC tissue samples. The amplified expression of PSMD6 in MBC tissues was confirmed in our Shanghai validation cohort by IHC staining in a subsequent step. Our findings suggested that serum-derived PSMD6 may be secreted by tissues and interact with the microenvironment to promote BC metastasis, suggesting that it may be a potential target for the treatment of metastatic BC. The tumor microenvironment plays a pivotal role in the tumorigenesis, progression, and metastasis of many cancers, including breast cancer 70 . There is now increasing evidence to support the observations that the multiple interactions between breast cancer cells and neighboring cells in the tumor microenvironment coordinate to regulate metastasis 70,71 . For example, Banerjee et al. reported that BC cells can upregulate IL-6 expression in adipocytes, which in turn promotes angiogenesis, tumor cell proliferation and survival via the JAK/STAT3 signaling pathway 72 . In this study, we found that PSMD6 acts on neutrophils through its interaction with the neutrophil serine protease PR3 and promotes BC metastasis by regulating NET formation and the expression of angiogenesis-related proteins. Based on the role of the PSMD6-neutrophil axis in BC metastasis, we identified the potential therapeutic agent sivelestat, which targets the PSMD6-neutrophil axis and demonstrated its inhibitory effect on tumor cells. These results revealed the role of the PSMD6-neutrophil axis in promoting tumor cell metastasis and showed that this axis can be targeted with inhibitors, providing a potential therapeutic option for patients with metastatic BC with PSMD6 overexpression. In general, gene amplification is associated with increased tumor aggression, metastasis, and resistance to chemotherapy 73 . For example, in breast and lung tumors, ERBB2 amplification induces overexpression of the protein in the cell membrane 73,74,75,76 , which has been associated with a poor prognosis 73,77,78 , while overexpression in gastric tumors is related to the presence of metastases 73,79,80,81 and evolution to the gastric intestinal type 73,82 . In this study, we analyzed BC tissue multiomics data from Xu et al. and found that, from IDC to distant metastasis, the increase in CLTA copy number showed a cis effect on its protein. The overexpression of CLTA promoted the transfer of PSMD6 from the cytoplasm to the extracellular space. Our findings suggested that genetic variants may alter the tumor microenvironment by altering the expression of proteins and pathways, thereby creating conditions for tumor metastasis. In addition, these results suggested that clinical testing of CLTA amplification may be considered a way to assess BC risk in the future. In conclusion, in this study, we captured the changes in the serum proteome of BC patients and showed that these changes are clearly linked to the underlying disease manifestations and clinical observations. We demonstrated that serum proteome profiling enables the discovery of better biomarkers, which could have a major impact on important aspects of disease management: (i) The 24-feature classifier will enable the diagnosis of BC and allow clinicians to distinguish between BBT and BC, (ii) the progression clock will provide information about the progression of the disease, and (iii) the potential therapeutic opportunity revealed herein could provide benefits for MBC patients with CLTA amplification. STAR METHODS KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Antibodies anti-PSMD6 HUABIO Cat#: ER64500 anti-CLTA anti-PSMD6 anti-CiH3 anti-MPO anti-MPO anti-a-tublin anti-CD177 anti-PRTN3 anti-GPR78 anti-Na+/K+ ATPase anti-GAPDH APC anti-mouse Ly6G antibody [1A8] FITC anti-mouse CD45 antibody [30-F11] PE Anti-Mouse/Human CD11b antibody [M1/70] APC Mouse IgM, κ Isotype Control [MM-30] PE Mouse IgG1, κ Isotype Control PE Anti-Human CD11b Antibody [ICRF44] APC Anti-Human CD66b Antibody [G10F5] Proteintech Group Santa Cruz HUABIO Proteintech Group HUABIO SAB SAB Proteintech Group Elabscience HUABIO Proteintech Group Elabscience Elabscience Elabscience Elabscience Elabscience Elabscience Elabscience Cat#: 10852-1-AP Cat#: sc-393580 Cat#: M1306-4 Cat#: 22225-1-AP Cat#: ET1703-21 Cat#: 21581 Cat#: 41689 Cat#: 67030-1 Cat#: 40588 Cat#: ET1609-76 Cat#: 60004-1 Cat#: E-AB-F1108E Cat#: E-AB-F1136C Cat#: E-AB-F1081D Cat#: E-AB-F09782E Cat#: E-AB-F09792D Cat#: E-AB-F1146D Cat#: E-AB-F1267E Biological Samples Serum samples (n=322) from a cohort of 306 breast cancer patients or mammary benign disease patients Human breast cancer tissues array Serum samples (n=255) from a cohort of breast cancer or mammary benign disease patients Guangdong Provincial People’s Hospital Shanghai Outdo Biotech.Co., LTD Guangdong Provincial People’s Hospital This paper hBreD132Su07 This paper Critical Commercial Assay BCA protein assay kit Beyotime Biotechnology Cat#: P0011 Bradford protein assay ThermoFisher Scientific Cat# 23236 Immunohistochemistry kit Tumor dissociation kit, mouse Mouse tumor infiltrating tissues kit ELISA kit MinuteTM Plasma Membrane Protein Isolation and Cell Fractionation Kit Quant-iT™ PicoGreen™ dsDNA mRNA qRT‒PCR starter kit Zhongshan Jinqiao Miltenyi Biotec Solarbio OmnimAbs Invent Biotechnologies Thermo TIANGEN Cat#: SP-9000 Cat#:130-096-730 Cat#: P2430 Cat#: OM487066 Cat#: 89881 Cat#: P7581 Cat#: KR116-01 Chemicals, Peptides, and Recombinant Proteins Trypsin Promega Cat#: V528A PMSF Sigma Cat#: P-7626 Protease inhibitor cocktail Roche Cat#: 04693159001 Phosphatase inhibitor cocktail Roche Cat#: 04906837001 HPLC-grade water J.T. Baker Cat#: 4218-03 Acetonitrile J.T. Baker Cat#: 9829-03 Formic acid Sigma Cat#: F0507 Methanol J.T. Baker Cat#: 9830-03 C18 resin Dikma Technologies Cat#: 85252 SepPark C18 cartridges Waters Cat#: WAT054960 Xbridge C18 column PR3-specific substrate (Abz)-VADnorVADRQ-(EDDnp) Sivelestat Phorbol 12-myristate 13-acetate (PMA) Waters Cayman MedChemExpress Selleck Cat#: 186003576 Cat#: 9002021-1 Cat#: 201677-61-4 Cat#: S7791 Software and Algorithms Source Identifier(i.e., links) Firmiana platform (Feng et al., 2017) https://phenomics.fudan.edu.cn/firmiana/gardener/ R (version 3.5.1) https://bioconductor.org/packages/release/bioc/html/CopywriteR.html KEGG database Reactome database https://www.phosphosite.org/homeAction https://reactome.org DAVID https://david. ncifcrf.gov xCell (Aran et al., 2017) https://xcell.ucsf.edu ImageJ software (version 1.51j) National Institutes of Health https://imagej.nih.gov/ij/ Deposited Data Proteomics data This paper iProx: IPX0007188000 LEAD CONTACT AND MATERIALS AVAILABILITY RESOURCE AVAILABILITY Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Chen Ding ( [email protected] ). This study did not generate new unique reagents. Material availability This study did not generate new unique reagents. Data and Code Availability Proteomics raw datasets are available through the iProx Consortium ( https://www.iprox.org/ ) with the subproject ID (IPX0007188000) or the firmiana platform ( http://www.firmiana.org/login/ ). Sample annotation and processed and normalized data files are provided in Tables S1-S2. The software and code used in this study are referenced in their corresponding STAR Method sections and the Key Resource Table. EXPERIMENTAL MODEL AND SUBJECT DETAILS Patient samples The serum samples used in this study were obtained from the Guangdong Provincial People’s Hospital. Serum samples were collected from patients or healthy controls. Breast cancer was determined by the attending doctors based on the clinical diagnostic guidelines of the Chinese Health Commission ( 6t h edition) and previous studies, which revealed the clinical courses of 306 patients. Blood samples (≤ 3 mL) from patients were collected over the course of their disease at intervals of 3–5 days. The clinical information of 306 patients, including tumor type, sex, age, tumor node metastasis (TNM) staging, and biochemical indicators, is listed in Table S1. The study was approved by the Research Ethics Committees of Zhongshan Hospital (No.), and written, informed consent was provided by all patients. The tissue chips are commercial and come from a company named OUTDO BIOTECH CO., LTD in Shanghai. Study cohorts Our studies had a discovery stage and three validation stages involving a total of 813 samples. The discovery stage was designed as a case‒control study and involved 322 samples, including 54 HC samples, 112 BBT samples and 156 BC samples. Among them, the BC samples included those reflecting various pathological stages of BC progression, such as 25 DCIS samples, 16 DCIS-MI samples, 68 IDC samples, and 45 MBC samples. Furthermore, to ensure the uniformity of serum samples for mass spectrometry experiments and a sufficient sample size for ELISA experiments (two parallel experiments for each sample, one experiment dosage of 200 µl), we conducted three-step validation based on three independent cohorts from Guangzhou 1, Guangzhou 2, and Shanghai. For the Guangzhou 1 validation cohort, a prospective follow-up cohort study was conducted to verify the effectiveness of the BC diagnostic classifier. This cohort involved 103 serum samples, including 17 HCs, 29 BBTs, and 57 BCs. For the Guangzhou 2 validation cohort, a prospective follow-up cohort study was conducted to verify PSMD6 as a potential specific serum biomarker for BC metastasis. This cohort involved 252 serum samples, including 61 HCs, 72 BBTs, and 121 BCs. For the Shanghai validation cohort, a retrospective cohort study was conducted to verify PSMD6 as a potential specific biomarker of BC metastasis associated with poor prognosis. Methods Details Proteomic Workflow Serum protein extraction and trypsin digestion Serum samples were mixed with 100 µL 50 mM ammonium bicarbonate (ABC) buffer, and the proteins were inactivated at 95°C for 5 min. The samples were cooled to room temperature and digested using trypsin at an enzyme to protein mass ratio of 1:25 for 17 hours in a 37°C incubator. Then, 5 µL of aqueous ammonia was added to each tube and vortexed to quench the digestion reaction, and the supernatant was subsequently dried using a 60°C vacuum drier (SpeedVac, Eppendorf). Then, the peptides were dissolved in 100 µL 0.1% formic acid (FA), vortexed for 3 min, and then sedimentation for 3 min (12,000 ×g). The supernatant was picked into a new tube and then desalinated. Before desalination, the activation of pillars with 2 slides of 3 M C18 disk is needed, and the lipid is as follows: 90 µL 100% acetonitrile (ACN) twice, 90 µL 50% and 80% ACN once in turn, and then 90 µL 50% ACN once. After pillar balance with 90 µL 0.1% FA twice, the supernatant of the tubes was loaded into the pillar twice and decontamination with 90 µL 0.1% FA twice. Finally, 90 µL elution buffer (0.1% FA in 50% ACN) was added to the pillar fir elution twice, and only the effluent was collected for MS. Finally, the collected peptides were dried using a 60°C vacuum drier. Nano-LC‒MS/MS The acquisition of samples was randomized to avoid bias. Samples were measured using LC‒MS instrumentation consisting of an EASY-nLC 1200 ultrahigh-pressure system (Thermo Fisher Scientific) coupled via a nano-electrospray ion source (Thermo Fisher Scientific) to a Q Exactive HF-X Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). Peptides, redissolved in Solvent A (0.1% formic acid in water), were loaded onto a 2-cm self-packed trap column (100-µm inner diameter, 3-µm ReproSil-Pur C18-AQ beads, Dr. Maisch GmbH) using Solvent A, separated on a 150-µm-inner-diameter column with a length of 8 cm (1.9-µm ReproSil-Pur C18-AQ beads, Dr. Maisch GmbH) with 6–95% mobile phase B (80% ACN and 0.1% formic acid) at 600 nL/min for 8.2 min, held constant at 95% solvent B at 800 nL/min for 4.1 min and then returned to 3% B for an additional 2.7 min to equilibrate the column. The eluted peptides were ionized under 2 kV and introduced into the mass spectrometer. The MS analysis was performed in a data-independent acquisition (DIA) mode. The DIA method consisted of MS1 Spectra full scan with m/z ranging from 300 to 1,400 at a high resolution of 30,000 with an automatic gain control (AGC) target 3E + 06. The maximal ion injection time was 20 ms. Then, 30 DIA segments were acquired at 15,000 resolution with an AGC target of 1E + 06 for maximal injection time. The setting “inject ions for all available parallelizable time” was enabled. HCD fragmentation was set to a normalized collision energy of 27%. The spectra were recorded in profile mode. The default charge state for the DIA was set to 3. All data were acquired using Xcalibur software v2.2 (Thermo Fisher Scientific). Peptide identification and protein quantification All data were processed using Firmiana 83 . DIA data were searched against the UniProt human protein database (updated on 2019.12.17, 20406 entries) using FragPipe (v12.1) with MSFragger (2.2) 84 . The mass tolerances were 20 ppm for precursor and 50 mmu for product ions. Up to two missed cleavages were allowed. The search engine was set with cysteine carbamidomethylation as a fixed modification and N-acetylation and oxidation of methionine as variable modifications. Precursor ion score charges were limited to + 2, +3, and + 4. The data were also searched against a decoy database so that protein identifications were accepted at a false discovery rate (FDR) of 1%. The results of DDA data were combined into spectral libraries. A total of 327 libraries were used as reference spectra libraries. DIA data were analyzed using DIA-NN (v1.7.0) 85 . The default settings were used for DIA-NN (Precursor FDR: 1%, Log lev: 1, Mass accuracy: 20 ppm, MS1 accuracy: 10 ppm, Scan window: 30, Implicit protein group: genes, Quantification strategy: robust LC (high accuracy)). Quantification of identified peptides was calculated as the average of chromatographic fragment ion peak areas across all reference spectra libraries. Label-free protein quantifications were calculated using a label-free, intensity-based absolute quantification (iBAQ) approach 86 . We calculated the peak area values as parts of the corresponding proteins. The fraction of total (FOT) was used to represent the normalized abundance of a particular protein across samples. FOT was defined as a protein’s iBAQ divided by the total iBAQ of all identified proteins within a sample. The FOT values were multiplied by 10 5 for ease of presentation, and missing values were imputed with 10 − 5 . Missing value imputation Before performing any downstream statistical analyses, proteome datasets of plasma were filtered for 50% valid values across all samples. Missing values were subjected to KNN imputation on the data using the “impute.knn” function from the “impute” R package 87 . Proteome Data Preprocessing Mass spectrometry platform QC For quality control of the MS performance during serum sample detection, we mixed all 322 samples into a serum pool as QC standards. The QC standards were analyzed using the same method and conditions as our serum cohort. Pearson’s correlation coefficient was calculated for QC standards. The average correlation coefficient of the QC standards was 0.98. The minimum and maximum correlation coefficients were 0.93 and 0.99, respectively, which demonstrated the stability of the mass spectrometry platform. Preprocessing of DIA proteomic data and batch correction Considering the balance between the confidence of protein identification and sample heterogeneity, we selected proteins by a specific threshold and then imputed them. First, the analysis in this study focused on the proteins identified in more than 50% of samples for each sample type (3 tumor subtypes and 1 healthy control). Second, we performed KNN imputation separately on the data for each sample type using the “impute.knn” function from the “impute” R package. We then combined the imputed data across all 4 sample types and obtained 8,944 proteins in total. Because the missing proteins of each tumor subtype are different, we then filled the empty value by 10 − 5 . Finally, we applied the R tool Combat, with the tumor type as a covariate to remove batch effects 88 . Gene set score for a single sample To functionally characterize NMF cluster results by single-sample gene set enrichment analysis (ssGSEA), we calculated the normalized enrichment score of each sample based on four classes of gene sets: GOBP, KEGG, hallmark, and reactome gene sets. We utilized the R package GSVA 89 with the following parameters: min.sz = 10, max.sz = 300, and other parameters were set to default values. Construction and validation of predictive models to distinguish between BC, BBT and normal samples Logistic regression analysis was used to distinguish between BCs, BBTs and the normal prediction model based on the significantly differentially expressed proteins in BCs, BBTs and normal serum samples using Python software v3.10.0. The backward stepwise method was utilized for feature selection. Samples were randomly divided into the training set (n = 217) and the testing set (n = 94). Moreover, the diagnostic value of this model was verified using scikit-learn analysis (Version 1.3.1). Sensitivity, specificity, accuracy, and AUC were used to determine the predictive value of the model. The predictive value of the model was validated in the validation cohort. Fuzzy C-means clustering Proteins were grouped into different clusters using the Mfuzz package in R with the fuzzy c-means algorithm 90 . WGCNA Weighted gene coexpression network analysis (WGCNA) 91 was applied to the proteins from plasma tumor samples using R code implemented in R software. The parameters were TOMType = ‘unsigned’, corType = ‘unsigned’, mingene = 50, and the rest were default. Spearman correlation analysis was conducted between the protein modules and the tumor types. Immune scores The levels of different cell types within tumor tissues were computed via CIBERSORT using protein expression values 92 . Table S5 contains the final score computed by CIBERSORT of different cell types for tumor samples. Correlation between tumor types and clinical features To estimate the correlations between tumor types and clinical features, the chi-square test was used for categorical variables, and the Kruskal-Willis test was used for continuous variables. Screening potential druggable targets To screen potential druggable targets, the following criteria need to be met: 1) candidates for core tumor markers and 2) drug targets annotated by the HPA 93 and DrugBank databases (version 5.1.5) ( http://www.drugbank.ca/ ). Cell lines The cell lines generated in this study (MDA-MB-231, MDA-MB-231-LM2, MCF7, MCF10A and BT474) were grown in DMEM with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. 4T1 cells were grown in RPMI-1640 with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. HL-60 cells were grown in IMDM with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. Cell cultures were tested for mycoplasma contamination every week. Human breast cancer tissues and blood samples A human breast cancer tissue (n = 126) array (Shanghai Outdo Biotech Co., LTD) with follow-up information was used for Kaplan‒Meier analysis of disease-free survival and overall survival. Univariate Cox and multivariate Cox analyses were also conducted. Whole blood samples for neutrophil isolation and ELISA (n = 255) were collected from Guangdong Provincial People’s Hospital under exemption approval of the Guangdong Provincial People’s Hospital Institutional Review Board. FACS analysis of immunocytes in lung metastases and molecular expression of dHL-60 Lung metastases were picked. Tumor Dissociation Kit, mouse (Cat, 130-096-730, Miltenyi Biotec) was used to generate single cell suspensions. Neutrophils were isolated from mouse tumor infiltrating tissues using a kit (Cat, P2430, Solarbio). HL-60 cells were treated with 1 µM ATRA for 5 days, which could transform HL-60 to dHL-60 cells. dHL-60 cells were treated with culture medium from PSMD6 OE LM2 cells or sh-PSMD6 MDA-MB-231 cells. Cells were incubated for 30 min with 0.5% FBS to block FcR before antibody staining. APC anti-mouse Ly6G antibody [1A8], FITC anti-mouse CD45 antibody [30-F11], PE anti-mouse/human CD11b antibody [M1/70], APC anti-human CD66b antibody [G10F5], PE anti-human CD11b antibody [ICRF44], PE mouse IgG1, κ isotype control, APC mouse IgM, and κ isotype control [MM-30] were used for staining (Elabscience). Flow cytometry was performed by a CANTO II (BD) FACS system and quantified by FlowJo V10 software. IHC staining The tissue microarrays were stained with immunohistochemistry. The sections were deparaffinized in xylene and dehydrated through alcohol changes. The sections were stained with a PSMD6 rabbit polyclonal antibody (HUABIO). Antibodies were prediluted by the manufacturer, and staining was performed following the manufacturer’s protocols. Two pathologists independently reviewed the pathological specimens. Scoring was assessed according to a previous study description 94 . ELISA Plasma was collected from healthy volunteers or patients who were diagnosed with breast cancer or benign tumors and stored at -80°C°C. A standard curve of PSMD6 was established with standard samples in the kit (OmnimAbs). Fifty microliters of sample was added to the appropriate well of the antibody precoated microtiter plate and gently mixed. Incubate for 45 min at 37°C. The liquid was removed, the plate was dried by swing, and washing buffer was added to every well for 30 seconds; the buffer was then removed, and this process was repeated 4 times. Diluted biotinylated anti-IgG (50 µl) was added to the sample wells and incubated for 30 min at 37°C. The sample was washed and dried. Then, 50 µl of streptavidin-HRP was added to all wells and gently mixed. The sample was incubated for 30 min at 37°C. After washing three times, signals were detected using TMB solution and read at 450 nm. Isolation of neutrophil plasma membrane proteins Plasma membrane proteins were isolated from human peripheral blood-derived neutrophils with the MinuteTM Plasma Membrane Protein Isolation and Cell Fractionation Kit (Invent Biotechnologies, 89881) as previously described 95 Briefly, cells were first sensitized by buffer A before passing through the proprietary filter in a zigzag manner when high-speed centrifugal force was applied, resulting in a cell lysate containing ruptured cell membranes and intact nuclei. As a result, nuclear contamination was virtually eliminated. The plasma membrane was further separated from the cell lysate (a mixture of crude membranes, intact nuclei, cytosolic proteins and organelles) by subsequent differential and density centrifugation with a regular tabletop microcentrifuge. Tubule formation assay A 96-well plate was coated with 50 µl of Matrigel per well, HUVECs were seeded in a Matrigel-coated 96-well plate at a density of 1.5 × 10 4 cells per well, and medium from neutrophils pretreated with cancer cell CM for 12 h (NCM) was added. After 6 h, pictures were taken with a light microscope (Olympus, Tokyo, Japan), and the Angiogenesis Analyzer plugin of ImageJ was used to count the number of branches and junctions. Enzymatic activity assays Enzyme activity assays were conducted as previously reported (Korkmaz et al., 2008). The PR3 enzymatic activity was quantified by detecting the rate of hydrolysis of the PR3-specific substrate (Abz)-VADnorVADRQ-(EDDnp) by cell suspension. To analyze membrane-bound PR3 activity, neutrophils were cultured in cancer cell CM or nonconditioned medium and treated with DMSO or sivelestat (10 mM) for 30–45 min at 37°C. Then, the cells were suspended in activity buffer (5×10 6 cells/ml, PBS, 4 mM EGTA, pH 7.4) with 20 mM PR3-specific substrate, and the kinetics of hydrolysis were determined by measuring the fluorescence at lex = 320 nm and lem = 420 nm. GST-pull down The GST pull-down procedure was conducted as previously reported (Nature methods.,2004). Fifty microliters of glutathione-magbeads were remixed with 300 µg GST-PSMD6 recombinant protein or GST protein for 2 hours. Then, the plasma member proteins of neutrophils were added and incubated for 16 hours at 4°C. Subsequently, the magnetic beads were washed with lysis buffer, and the proteins bound to the magnetic beads were analyzed by western blotting, MS and Coomassie electrophoresis staining. Immunofluorescence (IF) staining For murine tissue, tissues were perfused with 4% PFA for 24 h and washed once in 1×PBS for 30 mins, followed by 30–95% alcohol and n-butanol prior to being embedded in paraffin. Tissues were sectioned to 4 mm thickness, washed twice with PBS, permeabilized in 0.2% Triton X-100 for 15 min, and blocked in PBS containing 5% BSA for 45 min. CiH3 antibody (HUABIO, M1306-4) and MPO antibody (Proteintech Group, 22225-1-AP) were used for IF staining. For cancer cell IF staining, CLTA knockdown cells and control cells were seeded on coverslips coated with poly-L-lysine (WHB-24-CS-LC, WHB) in 24-well plates. After 24 h at 37 ℃, the cells were fixed with 4% PFA for 10 min at room temperature, washed three times with PBS and permeabilized in 0.1% Triton X-100 for 10 min. The cells were blocked in PBS containing 5% BSA for 30 min and then incubated with anti-CLTA (Proteintech Group, 10852-1-AP) and anti-PSMD6 (Santa Cruz, sc-393580) in blocking buffer overnight at 4°C. After three washes in PBS, the cells were incubated with fluorochrome-conjugated secondary antibodies (1:500, BIOESN) for 1 h and then counterstained with DAPI (ZSGB-BIO, ZLI-9557). Observation and photographing were performed with the confocal microscopy Cell Observer (Zeiss, Germany), and image processing and analysis were performed with Zen blue edition software (Zeiss, Germany). The Plugin-Colocalization Finder of ImageJ was used in the colocalization quantitative analysis. Pearson’s correlation coefficient and overlap coefficient according to Manders 96 were used to quantitatively evaluate the colocalization results. Mouse experiments For the tail vein metastasis assay, 2 × 10 6 human breast cancer cells (MDA-MB-231 PSMD6 knockdown cells, mock cells, MDA-MB-231-LM2 PSMD6 overexpression cells, and normal control cells) were injected into the tail vein of 6-week-old nude mice (n = 4, 5 for each group). After 6 weeks, mice were killed by cervical dislocation, and the lungs were removed for fixation with 4% PFA. To observe neutrophil infiltration and NETs in vivo , 1×10 5 murine breast cancer cells (4T1-PSMD6 knockdown cells, mock cells, 4T1-PSMD6 overexpression cells, and normal control cells) were injected into the tail vein of 6-week-old BALB/c mice (n = 4, 5 for each group). After 1–2 weeks, the mice were killed by cervical dislocation, and lung or lung metastasis nodules were prepared as single-cell suspensions or fixed in 4% PFA. RNAi and cell transfection Lentivirus packaging was carried out by Shanghai Obio (China). For the knockdown of PSMD6 and CLTA, one validated hairpin (human PSMD6 target sequences: GAATGCCGTTACTCTGTTT, murine psmd6 target sequences: AGAGTTCTGTGTTTCTAAA), three validated hairpins (CLTA target sequences: GGAGCTAGAAGAATGGTAT, GAGCAGCTACAGAAAACAA, GAAGCAGAGTGGAAAGAAA) targeting the PSMD6 and CLTA transcripts were cloned and inserted into the pSLenti-U6-shRNA(PSMD6)-CMV-F2A-Puro-WPRE vector. For the overexpression of PSMD6, the full length of their transcripts was cloned and inserted into the CMV-MCS-3FLAG-SV40-puromycin vector. RNA extraction and quantitative real-time polymerase chain reaction (qRT‒PCR) Total RNA was isolated from breast cancer cell lines with an RNA easy fast cell kit (TIANGEN, Beijing). Reverse transcription (RT) of complementary DNA (cDNA) was carried out by using the TIANGEN mRNA qRT‒PCR starter kit (TIANGEN, KR116-01). SYBR Green PCR Master Mix was used to amplify cDNA aliquots. GAPDH served as an endogenous control. The sequences of the sense and antisense primers were as follows: IL-6-F: CCTCTCTCTAATCAGCCCTCTG, IL-6-R: GAGGACCTGGGAGTAGATGAG, MMP9-R: GGCAGGGACAGTTGCTTCT, MMP9-F: TGTACCGCTATGGTTACACTCG, VEGF-R: AGGGTCTCGATTGGATGGCA, VEGF-F: AGGGCAGAATCATCACGAAGT, IL-8-F: TTTTGCCAAGGAGTGCTAAAGA, IL-8-R: AACCCTCTGCACCCAGTTTTC. Neutrophil isolation To isolate neutrophils from murine lung metastasis nodules, lung metastasis nodules from 8-week-old BALB/C mice were harvested. Preparation of single-cell suspensions from lung metastasis tissues was performed with a gentleMACS Dissociator, followed by neutrophil isolation with a mouse tumor infiltrating tissue neutrophil isolation kit (P2430, Solarbio). Human neutrophils were isolated from the peripheral blood of healthy female volunteers with a human peripheral blood neutrophil isolation kit (P9040, Solarbio). Neutrophils were cultured in RPMI 1640 medium containing 0.2% BSA. NET analysis To analyze NET formation, neutrophils (1×10 5 cells) were seeded on coverslips coated with poly-L-lysine (WHB-24-CS-LC, WHB) in 24-well plates for 30 min before adding 50% cancer cell CM. After 6 h at 37°C, neutrophils were subjected to IF staining. Anti-histone H3 (1:200, M1306-4, HUABIAO) and anti-MPO (1:200, ET1703-21, HUABIAO) were used to stain the sections. NET area quantification was performed using a previously published method (Cardiovascular Research (2022) 118, 2179–2195). Briefly, NET-positive area (%) = (the colocalized area of MPO fl and CitH3 fl/the area of the tissue in each microscopic field using the 20x objective) *100%. Two-chamber migration assays The two-chamber migration assay procedure was previously described (Zhang et al. Molecular Cancer (2018) 17:146). Briefly, 1×10 5 MDA-MB-231 cells in DMEM were added to the upper chamber (11965092, Gibco), and a 1:1 mixture of DMEM and cancer cell CM, or medium from neutrophils cultured in cancer cell CM, was added to the lower chamber as the chemoattractant. The migrated cells in the lower chamber were counted after 12 h. Western blotting Western blotting was performed as described previously. In summary, total proteins were extracted from cells via RIPA buffer supplemented with protease and phosphatase inhibitors (Beyotime, Beijing), and aliquots of these proteins were separated by SDS/PAGE and visualized with Millipore Immobilon Western HRP substrate. The antibodies used in this assay included anti-PSMD6 (HUABIO, ER64500), anti-CLTA (Proteintech Group, 10852-1-AP), anti-a-tubulin (SAB,21581), anti-CD177 (SAB, 41689), anti-PRTN3 (Proteintech Group, 67030-1), anti-GPR78 (Elabscience, 40588), anti-Na + /K + ATPase (HUABIO, ET1609-76), and anti-GAPDH (Proteintech Group, 60004-1). Detection of circulating NETs dHL-60 cells were induced from HL-60 by 1 µM ATRA for 5 days. dHL-60 cells were cultured in culture medium from PSMD6-overexpressing cells or PSMD6-knockdown cells for 6 hours. The culture medium of dHL-60 cells was detected by Quant-iT™ PicoGreen™ dsDNA (Thermo). Statistical analyses Data analyses were performed using GraphPad Prism 9.0 (GraphPad Software, La Jolla, USA). The data presentation and statistical analyses are described in the figure legends. P values < 0.05 were considered statistically significant. The in vitro experiments were repeated independently multiple times with similar results, as indicated in the figure legends. Quantification and statistical analysis Quantification methods and statistical analysis methods for plasma proteomic analyses were described or referenced in the respective Methods Details subsections. Additionally, standard statistical tests were used to analyze the data, including but not limited to Student’s t test, rank sums test, ANOVA test, Kruskal‒Wallis test, Fisher’s exact test, and chi-square test. Statistical significance was considered when the p value < 0.05. All analyses of plasma proteomic data were performed in R, Python, and GraphPad Prism. Declarations Funding This work is supported by the National Key R&D Program of China (2022YFA1303200, 2022YFA1303201); the National Natural Science Foundation of China (32330062, 31972933, 82003149); the Program of Shanghai Academic/Technology Research Leader (22XD1420100); the Major Project of Special Development Funds of Zhangjiang National Independent Innovation Demonstration Zone (ZJ2019-ZD-004); the Shanghai Municipal Science and Technology Major Project (2017SHZDZX01); and the Fudan Original Research Personalized Support Project. Ethics approval and consent to participate This study was approved by the Ethics Committee of Guangdong Provincial People’s Hospital. This animal study was approved by the Institutional Animal Care and Use Committee of Guangdong Provincial People’s Hospital (KY2020-042-02). References Siegel, R.L., Miller, K.D., Fuchs, H.E., and Jemal, A. (2022). Cancer statistics, 2022. CA Cancer J Clin 72 , 7–33. Huang, T., Bao, H., Meng, Y.H., Zhu, J.L., Chu, X.D., Chu, X.L., and Pan, J.H. (2022). Tumour budding is a novel marker in breast cancer: the clinical application and future prospects. Ann Med 54 , 1303–1312. Tower H, Ruppert M, Britt K. The Immune Microenvironment of Breast Cancer Progression. Cancers (Basel). 2019;11(9):1375. Bombonati, A., and Sgroi, D.C. (2011). The molecular pathology of breast cancer progression. J Pathol 223 , 307–317. Wang, L. (2017). Early Diagnosis of Breast Cancer. Sensors (Basel) 17. 10.3390/s17071572 . Zubor P, Kubatka P, Kajo K, Dankova Z, Polacek H, Bielik T, Kudela E, Samec M, Liskova A, Vlcakova D, Kulkovska T, Stastny I, Holubekova V, Bujnak J, Laucekova Z, Büsselberg D, Adamek M, Kuhn W, Danko J, Golubnitschaja O. Why the Gold Standard Approach by Mammography Demands Extension by Multiomics? Application of Liquid Biopsy miRNA Profiles to Breast Cancer Disease Management. Int J Mol Sci. 2019;20(12):2878. Ilié M, Hofman P. Pros: Can tissue biopsy be replaced by liquid biopsy? Transl Lung Cancer Res. 2016;5(4):420–3. Hu X, Li G, Wu S. Advances in Diagnosis and Therapy for Bladder Cancer. Cancers (Basel). 2022;14(13):3181. Kardynal A, Olszewska M. Modern non-invasive diagnostic techniques in the detection of early cutaneous melanoma. J Dermatol Case Rep. 2014;8(1):1–8. Liang J, Jiang Y, Abboud Y, Gaddam S. Role of Endoscopy in Management of Upper Gastrointestinal Cancers. Diseases. 2022;11(1):3. Oeffinger, K. C., Fontham, E. T. H., Etzioni, R., Herzig, A., Michaelson, J. S., Shih, Y.-C. T., et al. (2015). Breast Cancer Screening for Women at Average Risk. Jama 314 (15), 1599–1614. Bleyer, A., and Welch, H. G. (2012). Effect of Three Decades of Screening Mammography on Breast-Cancer Incidence. N. Engl. J. Med. 367 (21), 1998–2005. Zanzonico PB. Benefits and Risks in Medical Imaging. Health Phys. 2019;116(2):135–137. Richardson, R.R. (2017). Imaging Modalities: Advantages and Disadvantages. In: Atlas of Acquired Cardiovascular Disease Imaging in Children. Springer, Cham. https://doi.org/10.1007/978-3-319-44115-3_1 Bleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med. 2012;367(21):1998–2005. Hussain S, Mubeen I, Ullah N, Shah SSUD, Khan BA, Zahoor M, Ullah R, Khan FA, Sultan MA. Modern Diagnostic Imaging Technique Applications and Risk Factors in the Medical Field: A Review. Biomed Res Int. 2022;2022:5164970. Crowley, E., Di Nicolantonio, F., Loupakis, F. & Bardelli, A. Liquid biopsy: monitoring cancer-genetics in the blood. Nat Rev Clin Oncol 10, 472–484 (2013). Li, Joshua J.X.; Tse, Gary M. (2020). Core needle biopsy diagnosis of fibroepithelial lesions of the breast: a diagnostic challenge. Pathology ,2020.06.005. Łukasiewicz, S.; Czeczelewski, M.; Forma, A.; Baj, J.; Sitarz, R.; Stanisławek, A. Breast Cancer—Epidemiology, Risk Factors, Classification, Prognostic Markers, and Current Treatment Strategies—An Updated Review. Cancers 2021, 13, 4287. Ahmed M. Kabel, Tumor markers of breast cancer: New prospectives, Journal of Oncological Sciences, Volume 3, Issue 1, 2017, Pages 5–11, ISSN 2452–3364. Tóthová, Csilla, Mihajlovičová, Xénia, Nagy, Oskar. "The Use of Serum Proteins in the Laboratory Diagnosis of Health Disorders in Ruminants". Ruminants - The Husbandry, Economic and Health Aspects, edited by Muhammad Abubakar, IntechOpen, 2017. 10.5772/intechopen.72154 . Tang Z, Gu Y, Shi Z, Min L, Zhang Z, Zhou P, Luo R, Wang Y, Cui Y, Sun Y, Wang X. Multiplex immune profiling reveals the role of serum immune proteomics in predicting response to preoperative chemotherapy of gastric cancer. Cell Rep Med. 2023;4(2):100931. Li, J., Qi, G. & Liu, Y. Proteomics analysis of serum from thymoma patients. Sci Rep 13, 5117 (2023). Fernandez-Garcia D, Hills A, Page K, Hastings RK, Toghill B, Goddard KS, et al. Plasma cell-free DNA (cfDNA) as a predictive and prognostic marker in patients with metastatic breast cancer. Breast Cancer Res. 2019;21:8. Alix-Panabières C, Pantel K. Clinical applications of circulating tumor cells and circulating tumor DNA as liquid biopsy. Cancer Discov. 2016;6(5):479–91. Núñez C. Blood-based protein biomarkers in breast cancer. Clin Chim Acta. 2019;490:113–27. Issaq HJ, Xiao Z, Veenstra TD. Serum and plasma proteomics. Chem Rev. 2007;107(8):3601–20. Harbeck, N. & Gnant, M. Breast cancer. Lancet 389, 1134–1150 (2017). Marrugo-Ramírez, J., Mir, M. & Samitier, J. Blood-Based Cancer Biomarkers in Liquid Biopsy: A Promising Non-Invasive Alternative to Tissue Biopsy. Int J Mol Sci 19 (2018). Harris L, Fritsche H, Mennel R, Norton L, Ravdin P, Taube S, et al. American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J Clin Oncol. 2007;25(33):5287–312. Brandi J, Manfredi M, Speziali G, Gosetti F, Marengo E, Cecconi D. Proteomic approaches to decipher cancer cell secretome. Semin Cell Develop Biol. 2018;78:93–101. Pavlou MP, Diamandis EP. The cancer cell secretome: A good source for discovering biomarkers? J Proteom. 2010;73(10):1896–906. D. E. Misek and E. H. Kim. Protein Biomarkers for the Early Detection of Breast Cancer International Journal of Proteomics 2011 Vol. 2011 Pages 343582. Hammarström, S. The carcinoembryonic antigen (CEA) family: structures, suggested functions and expression in normal and malignant tissues. Semin Cancer Biol 9, 67–81 (1999). Hao, C., Zhang, G., and Zhang, L. (2019). Serum CEA levels in 49 different types of cancer and noncancer diseases. Prog Mol Biol Transl Sci 162 , 213–227. 10.1016/bs.pmbts.2018.12.011 . Felder, M. et al. MUC16 (CA125): tumor biomarker to cancer therapy, a work in progress. Mol Cancer 13, 129 (2014). Dochez, V. et al. Biomarkers and algorithms for diagnosis of ovarian cancer: CA125, HE4, RMI and ROMA, a review. J Ovarian Res 12, 28 (2019). Manuali, E., De Giuseppe, A., Feliziani, F. et al. CA 15–3 cell lines and tissue expression in canine mammary cancer and the correlation between serum levels and tumour histological grade. BMC Vet Res 8, 86 (2012). Tempero, M.A. et al. Relationship of carbohydrate antigen 19 – 9 and Lewis antigens in pancreatic cancer. Cancer Res 47, 5501–5503 (1987). He, M., Wu, C., Xu, J., Guo, H., Yang, H., Zhang, X., Sun, J., Yu, D., Zhou, L., Peng, T., et al. (2014). A genome wide association study of genetic loci that influence tumour biomarkers cancer antigen 19 – 9, carcinoembryonic antigen and α fetoprotein and their associations with cancer risk. Gut 63 , 143–151. 10.1136/gutjnl-2012-303434 . Kim, D.Y., Toan, B.N., Tan, C.K., Hasan, I., Setiawan, L., Yu, M.L., Izumi, N., Huyen, N.N., Chow, P.K., Mohamed, R., et al. (2023). Utility of combining PIVKA-II and AFP in the surveillance and monitoring of hepatocellular carcinoma in the Asia-Pacific region. Clin Mol Hepatol 29 , 277–292. 10.3350/cmh.2022.0212 . Taketa, K. Alpha-fetoprotein: reevaluation in hepatology. Hepatology 12, 1420–1432 (1990). Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209–249. Hanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell. 2011;144(5):646–74. Gonzalez H, Hagerling C, Werb Z. Roles of the immune system in cancer: from tumor initiation to metastatic progression. Genes Dev. 2018;32(19–20):1267–84. Joyce JA, Pollard JW. Microenvironmental regulation of metastasis. Nat Rev Cancer. 2009;9(4):239–52. Robinson JL, Feizi A, Uhlén M, Nielsen J. A systematic investigation of the malignant functions and diagnostic potential of the cancer secretome. Cell Rep. 2019;26(10):2622–35. Valastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell. 2011;147(2):275–92. Giancotti FG. Mechanisms governing metastatic dormancy and reactivation. Cell. 2013;155(4):750–64. Redig AJ, McAllister SS. Breast cancer as a systemic disease: a view of metastasis. J Intern Med. 2013;274(2):113–26. Ghajar CM. Metastasis prevention by targeting the dormant niche. Nat Rev Cancer. 2015;15(4):238–47. Benoy IH, Salgado R, Dam PV, Geboers K, Marck EV, Scharpé S, et al. Increased serum interleukin-8 in patients with early and metastatic breast cancer correlates with early dissemination and survival. Clin Cancer Res. 2004;10(21):7157–62. Noman AS, Uddin M, Chowdhury AA, Nayeem MJ, Raihan Z, Rashid MI, et al. Serum sonic hedgehog (SHH) and interleukin-(IL-6) as dual prognostic biomarkers in progressive metastatic breast cancer. Sci Rep. 2017;7:89. Endo M, Yamamoto Y, Nakano M, Masuda T, Odagiri H, Horiguchi H, et al. Serum ANGPTL2 levels reflect clinical features of breast cancer patients: implications for the pathogenesis of breast cancer metastasis. Int J Biol Markers. 2014;29(3):239–45. Guo, T., Kouvonen, P., Koh, C.C., Gillet, L.C., Wolski, W.E., Rost, H.L., Rosenberger, G., Collins, B.C., Blum, L.C., Gillessen, S., et al. (2015). Rapid mass spectrometric conversion of tissue biopsy samples into permanent quantitative digital proteome maps. Nat Med 21, 407–413. 10.1038/nm.3807 . Bittremieux, W. et al. Quality control in mass spectrometry-based proteomics. Mass Spectrom Rev 37, 697–711, doi: 10.1002/mas.21544 (2018). Jiang, Y., Sun, A.H., Zhao, Y., Ying, W.T., Sun, H.C., Yang, X.R., Xing, B.C., Sun, W., Ren, L.L., Hu, B., et al. (2019). Proteomics identifies new therapeutic targets of early- stage hepatocellular carcinoma. Nature 567, 257-+. 10.1038/s41586-019-0987-8 . Xu, J.Y., Zhang, C.C., Wang, X., Zhai, L.H., Ma, Y.M., Mao, Y.S., Qian, K., Sun, C.Q., Liu, Z.W., Jiang, S.W., et al. (2020). Integrative Proteomic Characterization of Human Lung Adenocarcinoma. Cell 182, 245-+. 10.1016/j.cell.2020.05.043 . Schwanhäusser, B. et al. Global quantification of mammalian gene expression control. Nature 473, 337–342 (2011). Tang, W., Zhou, M., Dorsey, T.H., Prieto, D.A., Wang, X.W., Ruppin, E., Veenstra, T.D., and Ambs, S. (2018). Integrated proteotranscriptomics of breast cancer reveals globally increased protein-mRNA concordance associated with subtypes and survival. Genome Med 10 , 94. 10.1186/s13073-018-0602-x . Jaillon S, Ponzetta A, Di Mitri D, Santoni A, Bonecchi R, Mantovani A. (2020). Neutrophil diversity and plasticity in tumour progression and therapy. Nat Rev Cancer. 20 , 485–503. 10.1038/s41568-020-0281-y . Cheung, P., Schaffert, S., Chang, S.E., Dvorak, M., Donato, M., Macaubas, C., Foecke, M.H., Li, T.M., Zhang, L., Coan, J.P., et al. (2021). Repression of CTSG, ELANE and PRTN3-mediated histone H3 proteolytic cleavage promotes monocyte-to-macrophage differentiation. Nat Immunol 22, 711–722. 10.1038/s41590-021-00928-y . Marino, S.F., Jerke, U., Rolle, S., Daumke, O., and Kettritz, R. (2022). Competitively disrupting the neutrophil-specific receptor-autoantigen CD177:proteinase 3 membrane complex reduces anti-PR3 antibody-induced neutrophil activation. J Biol Chem 298, 101598. 10.1016/j.jbc.2022.101598 . Xiao, Y., Cong, M., Li, J., He, D., Wu, Q., Tian, P., et al. (2021). Cathepsin C promotes breast cancer lung metastasis by modulating neutrophil infiltration and neutrophil extracellular trap formation. Cancer Cell. 39,(3):423–437.e7. doi: 10.1016/j.ccell.2020.12.012 . Borrebaeck CA. Precision diagnostics: moving towards protein biomarker signatures of clinical utility in cancer. Nat Rev Cancer. 2017;17(3):199–204. Biernat M, Trzyna M, Byszek A, et al. Liquid crystal foil for the detection of breast cancer[C]//Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2016. SPIE, 2016, 10031: 873–882. Zhang JY, Shi KZ, Liao XY, Li SJ, Bao D, Qian Y, Li DJ. The Silence of PSMC6 Inhibits Cell Growth and Metastasis in Lung Adenocarcinoma. Biomed Res Int. 2021;2021:9922185. AydoĞan TÜrkoĞlu S, Dayi G, KÖÇkar F. Upregulation of PSMD4 gene by hypoxia in prostate cancer cells. Turk J Biol. 2020;44(5):275–283. Okumura T, Ikeda K, Ujihira T, Okamoto K, Horie-Inoue K, Takeda S, Inoue S. Proteasome 26S subunit PSMD1 regulates breast cancer cell growth through p53 protein degradation. J Biochem. 2018;163(1):19–29. Terceiro LEL, Edechi CA, Ikeogu NM, et al. The Breast Tumor Microenvironment: A Key Player in Metastatic Spread. Cancers. 2021;13(19):4798. Monteran L, Erez N. The Dark Side of Fibroblasts: Cancer-Associated Fibroblasts as Mediators of Immunosuppression in the Tumor Microenvironment. Front Immunol. 2019;10:1835. Banerjee, K.; Resat, H. Constitutive activation of STAT3 in breast cancer cells: A review. Int. J. Cancer 2015, 138, 2570–2578 Bizari, L., Silva, A.E., & Tajara, E.H. (2006). Gene amplification in carcinogenesis. Genetics and Molecular Biology, 29, 1–7. Ohta JI, Miyoshi Y, Uemura H, Fujinami K, Mikata K, Hosaka M, Tokita Y and Kubota Y (2001) Fluorescence in situ hybridization evaluation of c-erbB-2 gene amplification and chromosomal anomalies in bladder cancer. Clin Cancer Res7:2463–2467. Takehana T, Kunitomo K, Kono K, Kitahara F, Iizuka H, Matsumoto Y and Fujino MA (2002) Status of c-erbB-2 in gastric adenocarcinoma: A comparative study of immunohistoche mistry, fluorescence in situ hybridization and enzymelinked immuno-sorbent assay. Int J Cancer 98:833–837. Lear-Kaul KC, Yoon HR, Kleinschmidt-DeMasters BK, Mc Gavran L and Singh M (2003) Her-2/neu status in breast cancer metastases to the central nervous system. Arch Pathol Lab Med 127:1451–1457 Cianciulli AM, Guadagni F, Benevolo M, Merola R, Giannarelli D, Marandino F, Vocaturo G, Mariani L and Mottelese M (2003) HER-2/neu oncogene amplification and chromosome 17 aneusomy in endometrial carcinoma: correlation with oncoprotein expression and conventional pathol Hirsch FR, Scagliotti GV, Langer CJ, Varella-Garcia M and Franklin WA (2003) Epidermal growth factor family of receptors in preneoplasia and lung cancer: Perspectives for target therapies. Lung Cancer 41:29–42. Nessling M, Solinas-Toldo S, Wilgenbus KK, Borchard F and Lichter P (1998) Mapping of chromosomal imbalances in gastric adenocarcinoma revealed amplified protooncogenes MYCN, MET, WNT2, and ERBB2. Genes Chromosomes Cancer 23:307–316. Vidgren V, Varis A, Kokkola A, Monni O, Puolakkainen P, Nordling S, Forozan F, Kallioniemi A, Vakkari ML, Kivilaakso E and Knuutila S (1999) Concomitant gastrin and ERBB2 gene amplifications at 17q12-q21 in the intestinal type of gastric cancer. Genes Chromosomes Cancer 24:24–29. Varis A, Wolf M, Monni O, Vakkari ML, Kokkola A, Moskaluk C, Frierson H Jr, Powell SM, Knuutila S, Kallioniemi A and El-Rifai W (2002) Targets of gene amplification and overexpression at 17q in gastric cancer. Cancer Res 62:2625–2629. Becker KF, Keller G and Hoefler H (2000) The use of molecular biology in diagnosis and prognosis of gastric cancer. Surg Oncol 9:5–11. Feng, J. et al. Firmiana: towards a one-stop proteomic cloud platform for data processing and analysis. Nat Biotechnol 35, 409–412 (2017). Kong, A.T., Leprevost, F.V., Avtonomov, D.M., Mellacheruvu, D. & Nesvizhskii, A.I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics. Nat Methods 14, 513–520 (2017). Demichev, V., Messner, C.B., Vernardis, S.I., Lilley, K.S. & Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat Methods 17, 41–44 (2020). Zhang, W. et al. LFQuant: a label-free fast quantitative analysis tool for high-resolution LC-MS/MS proteomics data. Proteomics 12, 3475–3484, doi: 10.1002/pmic.201200017 (2012). Hastie, T., Tibshirani, R., Sherlock, G., Eisen, M., Brown, P. and Botstein, D., Imputing Missing Data for Gene Expression Arrays, Stanford University Statistics Department Technical report (1999), http://www-stat.stanford.edu/~hastie/Papers/missing.pdf Olga Troyanskaya, Michael Cantor, Gavin Sherlock, Pat Brown, Trevor Hastie, Robert Tibshirani, David Botstein and Russ B. Altman, Missing value estimation methods for DNA microarrays BIOINFORMATICS Vol. 17 no. 6, 2001 Pages 520–525 Johnson, W.E., Li, C. & Rabinovic, A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics (Oxford, England) 8, 118–127 (2007). Hänzelmann, S., Castelo, R. & Guinney, J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 14, 7 (2013) Ruspini, E.H.; Bezdek, J.C.; Keller, J.M. Fuzzy Clustering: A Historical Perspective. IEEE Comput. Intell. Mag.2019, 14, 45–55. Langfelder, P. & Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9, 559 (2008). Newman AM, Steen CB, Liu CL, Gentles AJ, Chaudhuri AA, Scherer F, Khodadoust MS, Esfahani MS, Luca BA, Steiner D, Diehn M, Alizadeh AA. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol. 2019;37(7):773–782. doi: 10.1038/s41587-019-0114-2 . Epub 2019 May 6. Chen YM, Zheng Y, Yu Y, Wang Y, Huang Q, Qian F, Sun L, Song ZG, Chen Z, Feng J, An Y, Yang J, Su Z, Sun S, Dai F, Chen Q, Lu Q, Li P, Ling Y, Yang Z, Tang H, Shi L, Jin L, Holmes EC, Ding C, Zhu TY, Zhang YZ. Blood molecular markers associated with COVID-19 immunopathology and multi-organ damage. EMBO J. 2020;39(24):e105896. doi: 10.15252/embj.2020105896 . Epub 2020 Dec 14. PMID: 33140861; PMCID: PMC7737620. Wang, L., Meng, Y., Xu, J.J., and Zhang, Q.Y. (2018). The Transcription Factor AP4 Promotes Oncogenic Phenotypes and Cisplatin Resistance by Regulating LAPTM4B Expression. Mol Cancer Res 16 , 857–868. 10.1158/1541–7786.Mcr-17-0519 . Gu, J., Ke, P., Guo, H., Liu, J., Liu, Y., Tian, X., Huang, Z., Xu, X., Xu, D., Ma, Y., Wang, X., and Xiao, F. (2023). KCTD13-mediated ubiquitination and degradation of GluN1 regulates excitatory synaptic transmission and seizure susceptibility. Cell Death Differ 30 , 1726–1741. 10.1038/s41418-023-01174-5 . Manders, E.M.M., Verbeek, F.J., and Aten, J.A. (1993). Measurement of co-localization of objects in dual-colour confocal images. J Microsc 169 , 375–382. 10.1111/j.1365-2818.1993.tb03313.x . Additional Declarations There is NO Competing Interest. Supplementary Files TableS1.xlsx Table 1 TableS2.xlsx Table 2 TableS3.xlsx Table 3 TableS4.xlsx Table 4 TableS5.xlsx Table 5 TableS6.xlsx Table 6 sFigure1.pdf Figure S1. Overall outline of the serum proteome profiling study of BC patients, related to Figure 1 (A) Longitudinal quality control of mass spectrometry using tryptic digest of HEK293T cells. The bottom left half of the panel represents the pairwise Pearson’s correlation coefficients of the samples. (B) The number of proteins identified in each sample. (C) Distribution of protein levels in BC, BBT and normal serum samples by a density plot. All of the samples passed proteomic quality control. (D) The number of proteins identified in BC, BBT and normal serum samples. (E) Overview of the proteomics profile of 322 samples. The dynamics of protein expression levels identified in BC, BBT and normal serum samples are shown. sFigure2.pdf Figure S2. Serum proteomic profiles distinguish between patients with BBT and those with BC, related to Figure 3 (A) The protein expression levels of neutrophil degranulation signaling. (B) The feature importance levels in the multiclassifier. sFigure3.pdf Figure S3. Proteomic analysis indicated that the CLTA-PSMD6-neutrophil axis promotes breast cancer metastasis, related to Figure 5 (A) Cluster dendrogram obtained from WGCNA. (B) Eigengene adjacency heatmap obtained from WGCNA. (C) Scale independence obtained from WGCNA. (D) Mean connectivity obtained from WGCNA. sFigure4.pdf Figure S4. Overexpression of PSMD6 synergistically enhanced angiogenesis and NET formation by activating neutrophils, related to Figure 7 (A) Western blot assays showed the PSMD6 level in murine breast cancer cells. (B) Migration and invasion of MDA-MB-231 cells recruited by CM from LM2 cells with PSMD6 overexpression or MDA-MB-231 cells with PSMD6 knockdown or by medium from neutrophils pretreated with cancer cell CM for 12 h (NCM). (C) western blot analysis showing the PSMD6 level in human breast cancer cells. (D) Neutrophils from healthy volunteers were treated with CM from LM2 cells overexpressing PSMD6 or MDA-MB-231 cells with PSMD6 knockdown for 6 h. (E-F) Quantification of MPO+ and Ci-H3+ tumor cell percentages in lung tissue slices of mice after inoculation of 4T1 cells with PSMD6 overexpression or 4T1 cells with PSMD6 knockdown; n =3 mice per group. (G) GST pull-down assay. GST-magbeads bound to GST-PSMD6 or GST proteins were incubated with neutrophil plasma membrane proteins. One set of samples was subjected to western blot analysis (left panel). To visualize the interaction on a Coomassie gel, the released proteins were also subjected to SDS‒PAGE followed by Coomassie staining (right panel). For the Coomassie-stained gel, ‘*’ indicates PR3, ‘**’ indicates CD177, ‘☆☆’ indicates GST-PSMD6 and ‘☆’ indicates GST protein alone. H. The expression of PR3 and CD177 in different cell fractions of neutrophils from healthy volunteers. Na/K ATPase, GRP78, and β-actin were used as markers of various fractions. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3634466","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":257480447,"identity":"97a6b248-ff28-4927-b7e6-5f3b3737d125","order_by":0,"name":"Chen Ding","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAyElEQVRIiWNgGAWjYDACCSDmqZBjYDgA4rERreWMMalaeNtI0cI/u/nYg7fzDBL7jp89wPCh7DBQpIGAJXeOpRvO3WaQOPNMXgLjjHOHgSIH8GsxkMgxk+bd9idxw4EcA2betsNAkQRCWvK/SfPOMUjccP6NAfNf4rTksEnzNgC13ADawkiMFokbaWaSc44ZGM+88cbgYM+5dB6JGwS08M9IfibxpsZAtu98juGDH2XWcvwzCGhBAQeAmIcE9aNgFIyCUTAKcAEAiINESzUP7sUAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0001-8673-3464","institution":"Fudan University","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Chen","middleName":"","lastName":"Ding","suffix":""},{"id":257480449,"identity":"38df573f-69b3-4a23-b3e6-57df84e912df","order_by":1,"name":"Yue Meng","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Yue","middleName":"","lastName":"Meng","suffix":""},{"id":257480450,"identity":"82812c80-8507-488f-9566-2bc3c1dcd749","order_by":2,"name":"Minjing Huang","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Minjing","middleName":"","lastName":"Huang","suffix":""},{"id":257480451,"identity":"bb8a13ed-599b-43ec-bbdd-74385c21913c","order_by":3,"name":"Ganfei Xu","email":"","orcid":"https://orcid.org/0000-0003-4707-9804","institution":"Fudan University","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ganfei","middleName":"","lastName":"Xu","suffix":""},{"id":257480452,"identity":"86231a29-328b-4767-a82b-781ec9a1a2fb","order_by":4,"name":"Xinwei Li","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Xinwei","middleName":"","lastName":"Li","suffix":""},{"id":257480453,"identity":"6a482d93-5ba0-4b36-9c85-e0f47eeb76e2","order_by":5,"name":"Bing Gu","email":"","orcid":"","institution":"Guangdong Provincial People's Hospital","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Bing","middleName":"","lastName":"Gu","suffix":""}],"badges":[],"createdAt":"2023-11-19 12:00:40","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3634466/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3634466/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":47988131,"identity":"8e6e9751-1ed1-47b8-97b8-9b08afb83e9d","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":484836,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverall outline of the serum proteome profiling study of BC patients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Mass spectrometry (MS)-based proteomics technology used to analyze serum samples from 154 BC patients, 112 BBT patients and 56 healthy controls. (B) The number of proteins identified in each sample. (C) The cumulative number of protein identifications of 322 samples. (D) Dynamic range of protein levels observed in BC, BBT and normal serum samples. (E) The bar plot shows the proportion of proteins in BCs, BBTs and healthy controls.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/202dfb389a3245c5b9e49ec8.png"},{"id":47988124,"identity":"e513fab1-478f-4708-a718-bda3169ab621","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1130989,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSerum proteomic profiles differ between BC and non-BC samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) PCA of proteomic data (1,782 proteins) among normal, fibroma, papilloma and BC samples. (B-D) Volcano plot of differentially expressed proteins (B), heatmap of differentially expressed proteins (C) and their associated biological pathways (D). (E) The interaction diagram of proteins in neutrophil degranulation signaling. (F) Spearman correlation between serum and tissue proteomes (Spearman's correlation test). (G-H) Ratio of serum and tissue protein levels in BC and non-BC samples (G) and enrichment levels of pathways related to specific DEPs (H). (I-K) Strategy for using serum biomarkers to distinguish between BC and non-BC samples (I), ratio of differentially expressed proteins involved in neutrophil degranulation between BC and non-BC serum samples (J), and heatmap of the selected proteins expressed in BC and non-BC serum samples (K).\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/704cef87396cc7b4834288f4.png"},{"id":47988126,"identity":"39d6e3c6-e885-437e-9269-7501e5d1bd88","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1538583,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSerum proteomic profiles distinguish between BBT patients and BC patients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-C) Heatmap of differentially expressed proteins among fibroma, papilloma, BC and normal serum samples (A), line plots of differentially expressed proteins among fibroma, papilloma, BC and normal serum samples (B), and different pathways among fibroma, papilloma, BC and normal serum samples (C). (D-E) The ROC curves of diagnostic biomarkers showing their abilities to distinguish BC and BBT patients from healthy controls (D) and their predictive performance in the validation cohort (E). (F-G) Heatmap showing differential expression of predictive model signatures.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/fe839de489eb76340bfa7c76.png"},{"id":47988969,"identity":"6c1a3bf1-52e8-435c-8f39-9760bef6f49e","added_by":"auto","created_at":"2023-12-11 15:07:16","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":940321,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProteomic changes in BC progression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Bioinformatics pipeline for identifying differentially expressed proteins (DEPs) followed by clustering and cluster analyses. (B) Fuzzy c-means clustering identified six distinct temporal patterns of protein expression. The x-axis represents the tumor progression, while the y-axis represents the z score-normalized intensity. (C) Heatmap showing the gene ontology (GO) and KEGG terms enriched in each of the 6 clusters. (D) The interaction diagram of proteins in neutrophil degranulation signaling and heatmap of differentially expressed proteins. (E-F) Strategy for using serum biomarkers to distinguish between DCIS, DCIS-MI, IDC and MBC samples (E), heatmap of the selected proteins expressed in BC and normal plasma samples (F).\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/51c5424392ebc5ed46980897.png"},{"id":47988138,"identity":"896fec95-8a23-4859-8eda-3727f6c8bee3","added_by":"auto","created_at":"2023-12-11 14:59:17","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":638665,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eProteomic analysis indicated that the CLTA-PSMD6-neutrophil axis promotes breast cancer metastasis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Heatmap showing the correlation between modules obtained from WGCNA and clinical outcomes. (B-D) Venn diagram showing differentially expressed proteins identified in the metastasis stage in the ME green module (B). Boxplots showing the DEPs in different stages of BC (C). Volcano plot showing the hazard ratio of the DEPs. Significantly expressed proteins are colored in red (D). (E) Venn diagram depicting the proteins with \u003cem\u003ecis\u003c/em\u003e-effects between IDC and MBC. Volcano plot showing the proteins with \u003cem\u003ecis-effect\u003c/em\u003e expression levels between IDC and MBC. (F) Significantly expressed proteins are colored red. (G) PSMD6 was upregulated in the MBC stage. (H) Volcano plot showing the correlation between CLTA and the pathway scores according to ssGSEA. The red point indicates protein secretion. (I) Protein secretion was upregulated in the MBC stage. (J) Venn diagram showing CLTA-correlated serum proteins with a \u003cem\u003ecis\u003c/em\u003e-effect between CNA and protein. (K) Volcano plot depicting the CLTA-correlated serum proteins with \u003cem\u003ecis-effect\u003c/em\u003e expression levels between IDC and MBC. (L) The correlation between the protein abundance of PSMD6 and the neutrophil score. (M) The CIBERSORT score of neutrophils significantly differed between IDC and MBC stages. (N-O) Volcano plot showing the correlation between the protein expression level of the neutrophil score and the pathway scores by ssGSEA. The red points show the correlated pathways. The bar graph shows the ratio of correlated pathways in IDC and MBC (N). Angiogenesis was upregulated in the MBC stage (O). (P) A brief model depicting the functional impact of CLTA \u003cem\u003ecis-effect\u003c/em\u003e-associated PSMD6 overexpression in MBC cells.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/8ddcf0a6b5816c01d5b12729.png"},{"id":47988130,"identity":"71b2c2db-cc12-4e5a-ae30-940364d4a616","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":2268604,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePSMD6 promotes breast cancer metastasis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Mass spectrometry analysis revealed elevated levels of PSMD6 in breast cancer tissues and serum samples from patients. (B) Serological levels of PSMD6 were measured in healthy volunteers and patients at different stages of the disease. (C-D) Tissue expression levels of PSMD6 were assessed in human breast cancer tissue arrays, and overall survival analyses were performed. (E) Immunofluorescence staining demonstrated the cellular localization of both PSMD6 and CLTA proteins. (F-H) Western blot analysis confirmed the protein expression levels of PSMD6 and CLTA in sh-CLTA MDA-MB-231 cells compared to control cells. (I-K) Transwell assays were conducted using PSMD6-overexpressing LM2 cells, mock LM2 cells, sh-PSMD6 MDA-MB-231 cells, and control MDA-MB-231 cells to evaluate their migratory potential. (L-N) Intravenous injection of either PSMD6-overexpressing LM2 cells or sh-PSMD6 MDA-MB-231 cells resulted in lung metastasis nodule formation (n = 5 mice per group). (O) Schematic representation illustrating the role of the CLTA-PSMD6 axis in breast cancer metastasis. P values (*\u0026lt;0.05, **\u0026lt;0.01, ***\u0026lt;0.001, ****\u0026lt;0.0001) were determined using two-tailed unpaired t test (A, B), log rank test (D), or repeated measures two-way ANOVA (J, K, M, N). Data are presented as the mean ± SD with a scale bar indicating 10 μm.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/7d563f38bd7f88ac9d79c601.png"},{"id":47988134,"identity":"c280c8f4-d11d-4fff-8b53-6d3c3daa06ae","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":1500989,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eOverexpression of PSMD6 synergistically enhances angiogenesis and NET formation by activating neutrophils\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A-B) Migration and invasion of MDA-MB-231 cells recruited by conditioned medium (CM) from LM2 cells with PSMD6 overexpression or MDA-MB-231 cells with PSMD6 knockdown or by medium from neutrophils pretreated with CM from these cancer cells for 12 hours (NCM). (C-E) Expression of CD11b and CD66b on the surface of dHL-60 cells after coculturing them with CM from LM2 cells with PSMD6 overexpression or MDA-MB-231 cells with PSMD6 knockdown. (F) Flow cytometry analysis shows the percentages of CD11b+Ly6G+ neutrophils in CD45+ immune cells in lung metastases after inoculation of 4T1 cells with either PSMD6 overexpression or PSMD6 knockdown; n =5 mice per group. Boxplots display values for minimum, first quartile, median, third quartile, and maximum. (G-H) Neutrophils from healthy volunteers were treated with CM from LM2 cells overexpressing PSMD6 or MDA-MB-231 cells with PSMD6 knockdown for 6 hours. The NET-positive area (%) was calculated as the colocalized area of MPO fluorescence and CitH3 fluorescence divided by the area occupied by the cells in each microscopic field using a 20 objective lens multiplied by 100%. (I) dHL-60 cells were treated with BSA or CM from LM2 cells with PSMD6 overexpression or MDA-MB-231 cells with PSMD6 knockdown for 6 hours or PMA for 3 hours. Free DNA was detected via PicoGreen staining after digesting chromosomes using DNaseI. (J-K) Quantification of the percentage of MPO+ and Ci-H3+ tumor cells in lung tissue slices from mice following inoculation with 4T1 cells overexpressing PSMD6 or 4T1 cells with PSMD6 knockdown was performed, with a total of three mice per group. (L-M) Neutrophils isolated from healthy volunteers were treated with conditioned media (CM) derived from LM2 cells overexpressing PSMD6 or MDA-MB-231 cells with PSMD6 knockdown, followed by mass spectrometry (MS) analysis. (N-Q) Human umbilical vein endothelial cells (HUVECs) were exposed to medium obtained from neutrophils pretreated with CM derived from these cancer cell lines for 12 hours (NCM), and the Angiogenesis Analyzer plugin of ImageJ was utilized to quantify the number of branches and junctions. (R-T) The results obtained through GST pull-down assays were analyzed using mass spectrometry. (U) Membrane-bound PR3 activity of human neutrophils cultured in LM2 (control or PSMD6 overexpression) CM and inhibitor was quantified (n = 3). (V-W) Neutrophils cultured in LM2 (control or PSMD6 overexpression) CM and inhibitor were subjected to mass spectrometry analysis. (X) Schematic diagram illustrating the CLTA-PSMD6-neutrophil axis in breast cancer metastasis. P values (*\u0026lt;0.05, **\u0026lt;0.01, ***\u0026lt;0.001, ****\u0026lt;0.0001) were obtained by repeated measures two-way ANOVA. Data are shown as the mean ± SD. Scale bar, 200 μm.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/705f276a31d1d987839745e0.png"},{"id":104400598,"identity":"f31b8a2b-4a03-418a-b9e8-2cbc7267939b","added_by":"auto","created_at":"2026-03-11 12:10:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":9996276,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/e254d80c-9654-4e35-9726-ebdce1050c53.pdf"},{"id":47988141,"identity":"394d8887-8120-4341-ab4c-735d8cf84020","added_by":"auto","created_at":"2023-12-11 14:59:17","extension":"xlsx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":23402792,"visible":true,"origin":"","legend":"Table 1","description":"","filename":"TableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/d4179535849c92ef6e441450.xlsx"},{"id":47988972,"identity":"c698162f-bc54-442b-afa0-c91009f2ba33","added_by":"auto","created_at":"2023-12-11 15:07:17","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":437480,"visible":true,"origin":"","legend":"\u003cp\u003eTable 2\u003c/p\u003e","description":"","filename":"TableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/10346c504e347e51005209f9.xlsx"},{"id":47988125,"identity":"dd0002f5-361f-49a6-b1ab-10e66143fba3","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":227025,"visible":true,"origin":"","legend":"\u003cp\u003eTable 3\u003c/p\u003e","description":"","filename":"TableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/421abaa036652bafd92db5fa.xlsx"},{"id":47988970,"identity":"3106bd2d-dd1a-41fd-92cb-73cfeeaa0e81","added_by":"auto","created_at":"2023-12-11 15:07:16","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":161666,"visible":true,"origin":"","legend":"Table 4","description":"","filename":"TableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/a4993e4fbe707728304042c3.xlsx"},{"id":47988135,"identity":"3882a496-5e3a-4ab3-b91d-674ab34a9ec7","added_by":"auto","created_at":"2023-12-11 14:59:17","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":50723,"visible":true,"origin":"","legend":"Table 5","description":"","filename":"TableS5.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/3a9cee587639ad7cdae8c8ff.xlsx"},{"id":47988129,"identity":"8fb210df-8af3-40a4-a73f-0012ce7645cc","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":14372,"visible":true,"origin":"","legend":"Table 6","description":"","filename":"TableS6.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/95ba3d38b676b441675d4134.xlsx"},{"id":47988132,"identity":"7824868a-0129-4006-a98f-02bd15c4dc77","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"pdf","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":1319693,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure S1. Overall outline of the serum proteome profiling study of BC patients, related to Figure 1\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Longitudinal quality control of mass spectrometry using tryptic digest of HEK293T cells. The bottom left half of the panel represents the pairwise Pearson’s correlation coefficients of the samples. (B) The number of proteins identified in each sample. (C) Distribution of protein levels in BC, BBT and normal serum samples by a density plot. All of the samples passed proteomic quality control. (D) The number of proteins identified in BC, BBT and normal serum samples. (E) Overview of the proteomics profile of 322 samples. The dynamics of protein expression levels identified in BC, BBT and normal serum samples are shown.\u003c/p\u003e","description":"","filename":"sFigure1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/1557225e84bc98eda40d4eb5.pdf"},{"id":47988142,"identity":"6ba29697-3085-4f8e-8f11-fce7add56d08","added_by":"auto","created_at":"2023-12-11 14:59:17","extension":"pdf","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":262575,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure S2. Serum proteomic profiles distinguish between patients with BBT and those with BC, related to Figure 3\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) The protein expression levels of neutrophil degranulation signaling. (B) The feature importance levels in the multiclassifier.\u003c/p\u003e","description":"","filename":"sFigure2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/6770529b697e5e7effe1c35b.pdf"},{"id":47988133,"identity":"d8879996-cc0b-4053-b342-3b06be7e8761","added_by":"auto","created_at":"2023-12-11 14:59:16","extension":"pdf","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":939822,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure S3. Proteomic analysis indicated that the CLTA-PSMD6-neutrophil axis promotes breast cancer metastasis, related to Figure 5\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Cluster dendrogram obtained from WGCNA. (B) Eigengene adjacency heatmap obtained from WGCNA. (C) Scale independence obtained from WGCNA. (D) Mean connectivity obtained from WGCNA.\u003c/p\u003e","description":"","filename":"sFigure3.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/54bc02aeda3d299a6aec7e5a.pdf"},{"id":47988140,"identity":"9abcac17-d411-4dea-b899-271467dd7fe3","added_by":"auto","created_at":"2023-12-11 14:59:17","extension":"pdf","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":6367234,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure S4. Overexpression of PSMD6 synergistically enhanced angiogenesis and NET formation by activating neutrophils, related to Figure 7\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Western blot assays showed the PSMD6 level in murine breast cancer cells. (B) Migration and invasion of MDA-MB-231 cells recruited by CM from LM2 cells with PSMD6 overexpression or MDA-MB-231 cells with PSMD6 knockdown or by medium from neutrophils pretreated with cancer cell CM for 12 h (NCM). (C) western blot analysis showing the PSMD6 level in human breast cancer cells. (D) Neutrophils from healthy volunteers were treated with CM from LM2 cells overexpressing PSMD6 or MDA-MB-231 cells with PSMD6 knockdown for 6 h. (E-F) Quantification of MPO+ and Ci-H3+ tumor cell percentages in lung tissue slices of mice after inoculation of 4T1 cells with PSMD6 overexpression or 4T1 cells with PSMD6 knockdown; n =3 mice per group. (G) GST pull-down assay. GST-magbeads bound to GST-PSMD6 or GST proteins were incubated with neutrophil plasma membrane proteins. One set of samples was subjected to western blot analysis (left panel). To visualize the interaction on a Coomassie gel, the released proteins were also subjected to SDS‒PAGE followed by Coomassie staining (right panel). For the Coomassie-stained gel, ‘*’ indicates PR3, ‘**’ indicates CD177, ‘☆☆’ indicates GST-PSMD6 and ‘☆’ indicates GST protein alone. H. The expression of PR3 and CD177 in different cell fractions of neutrophils from healthy volunteers. Na/K ATPase, GRP78, and β-actin were used as markers of various fractions.\u003c/p\u003e","description":"","filename":"sFigure4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3634466/v1/a8e1abf5bfc1c57e2cf03f4c.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Proteome profiling of serum reveals PSMD6 as a biomarker in breast cancer metastasis","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eBreast cancer (BC) is one of the most commonly diagnosed cancers in women\u003csup\u003e1\u003c/sup\u003e and presents high heterogeneity in its morphology, molecular expression profile and clinical course\u003csup\u003e2\u003c/sup\u003e. Breast ductal carcinoma (BRDC) is the most common type of BC and has unique clinical and pathological features. BC progression involves a complex multistep process with intermediate transition mechanisms that are difficult to monitor. Histopathologically, the progression of human BRDC is a linear multistep process that begins with a lesion, which progresses to ductal carcinoma \u003cem\u003ein situ\u003c/em\u003e (DCIS), develops into microinvasion carcinoma (DCIS-MI), evolves into invasive ductal carcinoma (IDC) and ultimately develops into potentially metastatic breast cancer (MBC)\u003csup\u003e3,4\u003c/sup\u003e. In addition, fibroadenoma and papilloma are common clinically detected benign breast tumors (BBTs). Although BC has characteristic imaging features, BBT and BC have similar presentations, leading to difficulty in making an accurate diagnosis through imaging. Thus, early-stage precision diagnosis and evaluation of metastasis status have significant value for prognosis evaluation and therapy monitoring for BC\u003csup\u003e5\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCurrently, tissue biopsy is the gold standard for clinical diagnosis\u003csup\u003e6\u003c/sup\u003e. However, tissue biopsy has several limitations: surgery is invasive; biopsy can only provide information on local tumor tissue with high heterogeneity; biopsy does not provide sufficient tissue to generate multiple tissue sections; and there is resistance for seeking such services on a yearly basis\u003csup\u003e7\u003c/sup\u003e. Different types of noninvasive diagnostic techniques have been developed to comprehensively assess the characteristics of different tumors. For example, endoscopy is commonly used for early diagnosis of gastrointestinal (GI) cancer; cystoscopy techniques have been used to improve the accuracy of tumor detection for urinary system malignancies; and methods such as dermoscopy have been used as additional clinical diagnostic aids for skin carcinoma\u003csup\u003e8\u0026ndash;10\u003c/sup\u003e. One noninvasive approach for BC diagnosis and evaluation of metastasis status is imaging examination\u003csup\u003e9\u003c/sup\u003e. Imaging examinations include mammography, magnetic resonance imaging (MRI), ultrasound, computerized tomography (CT) and positron emission tomography (PET-CT)\u003csup\u003e11,12\u003c/sup\u003e. With improvements in breast imaging, mammography, ultrasound and minimally invasive interventions, the detection rate of early breast cancer, noninvasive cancers, lesions of uncertain malignant potential, and benign lesions has increased\u003csup\u003e13,14\u003c/sup\u003e. However, with improved diagnostic capabilities, there is a substantial risk of false-positive findings for benign tumors and, conversely, false-negative findings for malignant tumors\u003csup\u003e15,16\u003c/sup\u003e. Today, noninvasive methods for sampling materials for biomarkers are being intensively developed\u003csup\u003e17,18,19\u003c/sup\u003e. Liquid biopsy tests showed promise for early cancer detection, tumor classification, and monitoring treatment response\u003csup\u003e20,21\u003c/sup\u003e. Compared with traditional detection methods for tumor diagnosis, liquid biopsy possesses many advantages; it is less invasive and easier to perform and has fewer complications and stronger dynamic monitoring ability, and a higher acceptance rate\u003csup\u003e18\u003c/sup\u003e. Serum, a bodily fluid, could represent an essential component of the liquid biopsy test\u003csup\u003e22\u003c/sup\u003e. Serum protein assessment is a straightforward method that can be performed routinely and frequently\u003csup\u003e23,24\u003c/sup\u003e. Therefore, there has been much interest in the development and validation of serum-based biomarkers for the early detection, risk stratification, and prognosis prediction for breast cancer\u003csup\u003e25,26,27\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eSerum is one of the predominant sample types used for diagnostic analyses in clinical practice, and serum samples from thousands of clinical studies are available in biobanks\u003csup\u003e28,29\u003c/sup\u003e. The serum proteome contains a set of tissue proteomes\u003csup\u003e29\u003c/sup\u003e. Serum proteins secreted by tumors are involved in various biological functions and are an important source of cancer biomarkers\u003csup\u003e29,30,31\u003c/sup\u003e. Thus, the levels of serum proteins and/or their changes between conditions provide information about the physical condition and health status of patients and can be used to track disease progression\u003csup\u003e32,33\u003c/sup\u003e. Several conventional blood biomarkers that are used in the clinic, such as carcinoembryonic antigen (CEA), cancer antigen 153 (CA153), carbohydrate antigen 125 (CA125), carbohydrate antigen 199 (CA199), prostate-specific antigen (PSA), and alpha fetoprotein (AFP), have been reported to be helpful for tumor detection \u003csup\u003e34,35,36,37,38,39,40,41,42\u003c/sup\u003e. Specifically, CEA is an important marker for colon cancer and some other carcinomas\u003csup\u003e34,35\u003c/sup\u003e; CA199 is present in high concentrations in the serum of patients with pancreatic carcinoma\u003csup\u003e39,40\u003c/sup\u003e; and AFP is a highly specific and sensitive marker of hepatocellular carcinoma\u003csup\u003e41,42\u003c/sup\u003e. However, there is a lack of published articles on serum biomarkers for BC clinical diagnosis, and knowledge of the mechanisms underlying BC is limited.\u003c/p\u003e \u003cp\u003eAccording to the 2020 IARC survey, BC has become the leading cause of cancer-related death for female patients\u003csup\u003e43\u003c/sup\u003e. Metastasis is mainly responsible for treatment failure and is the cause of most BC-related deaths\u003csup\u003e44,45,46\u003c/sup\u003e. As reported, metastatic breast cancer (MBC) cells acquire aggressive characteristics through several mechanisms, including epithelial\u0026ndash;mesenchymal transition (EMT), tumor angiogenesis and metabolic programming\u003csup\u003e47,48,49,50,51,52\u003c/sup\u003e. Characterizing these mechanisms thoroughly may help to stop the development of tumor metastasis and to provide strategies for precise treatment. However, although the transition from IDC to MBC is central to the poor prognosis, little is known about the time of onset or the triggering mechanism by which invasive BC becomes metastatic BC in humans. Moreover, there is a lack of serum biomarkers for MBC diagnosis. This study aimed to uncover the relevant molecular mechanisms of breast cancer metastasis and to explore potential serum biomarkers that may be used for highly sensitive and rapid MBC diagnosis\u003csup\u003e53,54\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIn this study, we collected 813 samples (322 serum samples from the discovery cohort and 365 serum samples and 126 tissue samples from the validation cohort). We performed a quantitative proteomic approach with data-independent acquisition (DIA) on 322 serum samples that consisted of 56 healthy control samples, 91 BBT-fibroadenoma samples, 21 BBT-papilloma samples, and 154 BC samples. The integrated tissue-serum proteomic approach identified tumor biomarkers for identifying patients with BC. Furthermore, we established a predictive model based on 24 features that distinguished HC, BBT and BC samples with good performance. The robustness of this predictive model was further validated in the independent validation cohort. In addition, we found that the CLTA-PSMD6-neutrophil axis promoted BC metastasis. Collectively, this study revealed the BC serum proteomic landscape in what may be the largest cohort to date and provided valuable information on serum biomarkers, which could facilitate the improvement of the clinical diagnosis and management of BC.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eOverall outline of the serum proteome profiling study in BC patients\u003c/h2\u003e \u003cp\u003eTo investigate the proteomic expression patterns of BC, we collected 332 serum samples from the Guangzhou 0 discovery cohort composed of three independent cohorts, including the HC cohort (n\u0026thinsp;=\u0026thinsp;56), BBT cohort (n\u0026thinsp;=\u0026thinsp;112: 91 BBT-fibroadenoma patients and 21 BBT-papilloma patients), and BC cohort (n\u0026thinsp;=\u0026thinsp;154). Notably, the BC subtypes included ductal carcinoma \u003cem\u003ein situ\u003c/em\u003e (DCIS, n\u0026thinsp;=\u0026thinsp;25), ductal carcinoma \u003cem\u003ein situ\u003c/em\u003e with microinvasion (DCIS-MI, n\u0026thinsp;=\u0026thinsp;16), invasive ductal carcinoma (IDC, n\u0026thinsp;=\u0026thinsp;68), and metastatic breast cancer (MBC, n\u0026thinsp;=\u0026thinsp;45) (Fig.\u0026nbsp;1A). We acquired serum proteome profiles of all samples using a data-independent acquisition (DIA) strategy\u003csup\u003e55\u003c/sup\u003e on a Q Exactive HF-X Hybrid Quadrupole-Orbitrap Mass Spectrometer (Thermo Fisher Scientific, Rockford, IL, USA) coupled with a high-performance liquid chromatography system (EASY nLC 1200, Thermo Fisher Scientific) (Fig.\u0026nbsp;1A; STAR Methods). The overall workflow of this study is shown in Figure. 1A. The demographic and clinical data of all the study participants are summarized in Table S1. Additionally, the clinical information of 306 female patients, including age, histological stage, degree of differentiation, TNM stage (AJCC cancer staging system 8th edition), and clinical subtype, is summarized in Table S1.\u003c/p\u003e \u003cp\u003eTo monitor the liquid chromatography-tandem mass spectrometry (LC‒MS/MS) platform instrument stability, a mixture of all serum samples from patients with BC was assessed every 20 samples; this approach is generally used in proteomic studies. The quality control (QC) samples were analyzed using the same method and conditions used for our cohort serum samples\u003csup\u003e56,57,58\u003c/sup\u003e. The average Pearson\u0026rsquo;s correlation coefficient, calculated for all quality control runs of QC samples, was 0.98 (range from 0.93 to 0.99), demonstrating the consistent stability of the MS platform (Figure S1C).\u003c/p\u003e \u003cp\u003eA total of 8,944 protein groups were identified in all the serum samples (Figure S1A), with an average of 1,881, 1,908 and 1,889 protein groups per BC, BBT and HC samples, respectively (Fig.\u0026nbsp;1B and 1C). Proteome quantification was conducted using the intensity-based absolute quantification (iBAQ) algorithm, followed by fraction of total (FOT) normalization as reported previously\u003csup\u003e59\u003c/sup\u003e. In addition, the proteome was highly dynamic, spanning approximately eight orders of magnitude, as indicated by the protein abundance (FOT) values (Fig.\u0026nbsp;1D). The distribution of log2-transformed FOT values of identified proteins in 322 serum samples is shown in Figure S1B, and the consistency among the samples further indicated the stability of our mass spectrometry platform.\u003c/p\u003e \u003cp\u003eAdditionally, as shown in Fig.\u0026nbsp;1E, the number of proteins identified in BC serum that were annotated as serum proteins was not significantly different from that in fibroma, papilloma and normal serum. Interestingly, between BC samples and other serum samples, a slightly higher number of proteins were annotated as cancer-related proteins, proteins highly expressed in the breast, CD markers and proteins associated with drugs approved by the US Food and Drug Administration (FDA) were identified in BC. Our study has thus far established a comprehensive serum proteomic landscape of breast carcinoma.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eSerum proteomic profiles differ between BC and non-BC samples\u003c/h2\u003e \u003cp\u003eTo investigate the proteome-level differences between BC and non-BC, we performed principal component analysis (PCA) among BC samples and non-BC samples. PCA of BC proteomes and non-BC proteomes showed a relatively obvious separation between BC samples and non-BC samples (including HC, BBT-fibroadenoma, and BBT-papilloma), which indicated that protein variation between the two kinds of samples exceeded the variation between individuals (Fig.\u0026nbsp;2A). Interestingly, in the non-BC samples, we observed that the two BBT types (BBT-papilloma and BBT-fibroadenoma) clustered together, while the HCs were somewhat distant from them. These results showed a clear distinction inside the proteome of non-BC, revealing a significant molecular difference between the proteomes of HCs and BBTs.\u003c/p\u003e \u003cp\u003eTo further elucidate serum proteomic expression patterns among the two different samples, we compared the serum proteome profiles of patients with BC to those of non-BC patients. The results revealed a dramatic shift in protein expression profiles; specifically, there were 853 significantly differentially expressed proteins (DEPs), of which 447 DEPs were upregulated and 391 DEPs were downregulated (Student\u0026rsquo;s t test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, BC/non-BC ratio\u0026thinsp;\u0026gt;\u0026thinsp;2 or \u0026lt;\u0026thinsp;0.5) (Fig.\u0026nbsp;2B; Table S2). Pathway enrichment analysis showed that the BC-specific proteins were mainly involved in neutrophil degranulation (Fig.\u0026nbsp;2D). Furthermore, protein‒protein interaction (PPI) analysis showed that PSMD6, which was upregulated in BC, was associated with neutrophil degranulation, further supporting these enrichment results (Fig.\u0026nbsp;2E). PSMD6 is a component of the 26S proteasome and shows a correlation with poor prognosis based on an external data source\u003csup\u003e60\u003c/sup\u003e (Tang et al.,2018).\u003c/p\u003e \u003cp\u003eWe also compared the serum proteome profiles of patients with BC to the BC tissue proteome (Xu et al) and reasoned that ideal biomarkers should be commonly overexpressed in the tumor tissue (upregulated in the tumor tissue sample) and released into the serum (upregulated in the serum samples). To further explore biomarkers for breast cancer, an additional tissue cohort from BC treatment-na\u0026iuml;ve female patients was assessed in this research. We compared the serum proteome profiles of patients with BC to that of the BC tissue proteome (Fig.\u0026nbsp;2E). In total, 8,462 proteins were commonly quantified in the BC tissue and serum samples (Fig.\u0026nbsp;2E). We also confirmed the significant positive correlation between the BC tissue proteome and serum proteome (Fig.\u0026nbsp;2F, Spearman\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.323, p value\u0026thinsp;\u0026lt;\u0026thinsp;1E-204). Moreover, comparing the two different proteome profiles between BC and HC, 1,436 proteins that were upregulated in BC in both the serum proteome and the tissue proteome were enriched in neutrophil degranulation, etc. (Figs.\u0026nbsp;2G\u0026ndash;- 2H). In a previous report, tumor-associated neutrophils (TANs) were shown to participate in tumor-promoting inflammation by driving angiogenesis, extracellular matrix remodeling, metastasis and immunosuppression. We suggest that a certain relationship exists between BC and neutrophils. We further performed supervised analysis to filter out significant DEPs among neutrophil degranulation at the serum proteome level and obtained 22 significant DEPs (Wilcoxon signed-rank test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05), in which PSMD6 overlapped with previous BC PPI network analysis (Figs.\u0026nbsp;2I-2K). PSMD6 was overexpressed in BC in both serum and tissue samples, which suggested that PSMD6 might be generated by BC cells and released into the serum.\u003c/p\u003e \u003cp\u003eIn conclusion, our proteomic analysis revealed significant differences between BC and non-BC samples, which manifested as changes in the abundance of proteins in the serum proteome profiles of patients as well as the enrichment of biological processes directly related to the disease phenotype. Pathway enrichment analysis showed that tumor-induced neutrophil degranulation was enriched in BC and revealed PSMD6 as a potential serum marker associated with tumor progression.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eSerum proteomic profiling distinguishes patients with BBT from those with BC\u003c/h2\u003e \u003cp\u003eClinically, there is often confusion and misdiagnosis regarding BBT and BC differentiation via early screening methods. To further elucidate serum proteomic expression patterns among HC, BBT (including BBT-fibroadenoma and BBT-papilloma), and BC patients in our cohort, we set criteria for defining tumor-specific serum proteins: tumor-specific serum protein expression values in one tumor group should be at least 1.5-fold higher than those in any of the other 3 groups (Kruskal-Willis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Consequently, we identified 1,130 significantly DEPs among the four groups (healthy control: 405 proteins, BBT-fibroadenoma: 130 proteins, BBT-papilloma: 336 proteins, and BC: 259 proteins) (Figs.\u0026nbsp;3A and 3B; Table S3). To investigate the serum proteomic features of these breast tumor groups, we performed pathway enrichment analysis for these 1,130 DEPs. Fibroadenoma-specific proteins significantly converged on pathways including TP53-regulated metabolic genes, tight junction, RhoG GTPase cycle, etc.; papilloma-specific proteins were enriched in adherens junction, vesicle-mediated transport, cell cycle, etc.; BC-specific proteins participated in VEGF, histidine metabolism, immune system, tyrosine metabolism, fructose and mannose metabolism, innate immune system, proteasome degradation, etc.. Compared with the other 3 groups (HC, BBT-fibroadenoma, and BBT-papilloma), neutrophil degranulation was mainly enriched in BC samples as previously shown (Fig.\u0026nbsp;3C). These results further indicated that a certain relationship exists between neutrophils and tumorigenesis. Notably, PSMD6, which is upregulated in BC, is a significant factor in the neutrophil degranulation pathway (Figure S2A).\u003c/p\u003e \u003cp\u003eFurthermore, we employed the eXtreme Gradient Boosting algorithm based on the differentially expressed proteins and routine blood indexes (ANOVA test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) to identify a subset of features (top features: OXCT1, HDGFL3, DDX39B, ACO1, SART3, COG1, ACTR2, QDPR, UMOD, DENND4C, CSTA, AOC2, RGN, MVP, TRAP1, UBE2L5, PTGFRN, SMARCC2, FKBP15, OXSR1, PLXNB1, TTR, PSMB3, and NEUT) that discriminated HC, BBT, and BC samples (named the HC/BBT/BC-sig) (Fig.\u0026nbsp;3D; STAR Methods). To train and subsequently test the classifier, samples from the Guangzhou 0 discovery cohort were grouped based on their type, and 70% and 30% were used as the training and testing sets, respectively. Based on the HC/BBT/BC-sig, 10-fold cross-validation in the training samples (70% of the cohort) yielded a predictive model with 0.91 accuracy and 0.94 precision for distinguishing HC, BBT and BC samples (Fig.\u0026nbsp;3D). When applied to the testing samples (30% of the cohort), the predictive model achieved high accuracy of 0.91.\u003c/p\u003e \u003cp\u003eTo further evaluate the accuracy of the predictive signatures for discriminating BBT and BC, we recruited a prospective follow-up validation cohort, named Guangzhou 1, and these patients provided 103 serum samples, including 27 HC, 29 BBT, and 57 BC samples. Based on the DIA approach, we performed comparative analysis of signature proteins among HC, BBT, and BC. As a result, we obtained the significantly differentially expressed genes of these signature features among HC, BBT, and BC. We observed that 9 features (OXCT1, HDGFL3, DDX39B, ACO1, QDPR, UMOD, DENND4C, and SMARCC2) were significantly increased in the HC group compared with the BC and BBT groups (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). We observed that 4 features (AOC2, RGN, TRAP1, and UBE2L5) were significantly upregulated in BBT samples compared with BC and HC samples (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Twelve signature features (SART3, COG1, ACTR2, CSTA, MVP, PTGFRN, FKBP15, OXSR1, PLXNB1, TTR, PSMB3, and NEUT) were significantly overexpressed in BC samples compared with BBT and HC samples (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). The molecular expression trends for the BBT/BC/HC-sig were consistent between the Guangzhou 0 discovery cohort and the Guangzhou 1 prospective validation cohort. In addition, pathway enrichment analysis showed that the BC-specific pathways in the validation cohort (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) were consistent with those in the Guangzhou 0 discovery cohort, such as neutrophil degranulation. The predictive model, which was constructed based on the Guangzhou 0 discovery cohort, had high accuracy and good prediction performance (accuracy\u0026thinsp;=\u0026thinsp;0.87) in the Guangzhou 1 prospective validation cohort (Fig.\u0026nbsp;3E). The heatmaps showed a clear separation among HC, BBT and BC samples in both the Guangzhou 0 discovery cohort and the Guangzhou 1 prospective validation cohort (Figs.\u0026nbsp;3F\u0026ndash;- 3G). Collectively, the predictive power of the signature proteins in different breast tumors (including BBT-fibroadenoma, BBT-papilloma, and BC) was validated in the Guangzhou 1 prospective validation cohort, and the results indicated that the predictive models in the discovery cohort exhibited robustness, accuracy, and stability in the Guangzhou 1 prospective cohort.\u003c/p\u003e \u003cp\u003eTaken together, our proteomic analysis showed neutrophil degranulation as the key signaling pathway in tumorigenesis; PSMD6 was defined as a core tumor-related protein in this pathway. The classifier derived from the Guangzhou 0 discovery cohort could be a potential predictive model for distinguishing BBT and BC samples and achieved good performance in the Guangzhou 1 validation set.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eProteomic kinetic changes in BC progression\u003c/h2\u003e \u003cp\u003eIn the past decade, many studies have characterized the multiomics landscape of certain stages of the progression of BC. In our study, we also observed that the serum proteome distinguished 4 pathological phases for BC, including DCIS (n\u0026thinsp;=\u0026thinsp;25), DCIS-MI (n\u0026thinsp;=\u0026thinsp;16), IDC (n\u0026thinsp;=\u0026thinsp;68) and MBC (n\u0026thinsp;=\u0026thinsp;45).\u003c/p\u003e \u003cp\u003eTo uncover protein patterns associated with tumor progression, we performed further analysis on the different phases of BC progression. We identified 1,313 proteins that were differentially expressed among HCs and at least one phase of tumor progression (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, ratio\u0026thinsp;\u0026gt;\u0026thinsp;2 or \u0026lt;\u0026thinsp;0.5) and calculated the Z scored intensities of these altered proteins (Fig.\u0026nbsp;4A; Table S4). The clustering analysis of these proteins using the fuzzy c-means algorithm identified 6 clusters of protein trajectories that changed with progression with a range of sizes (from 134 to 327 proteins per cluster) (Fig.\u0026nbsp;4B). Assessing these proteins would provide many potential candidate biomarkers for early BC screening and advance our understanding of the clinical features of BC progression.\u003c/p\u003e \u003cp\u003eTo explore the biological function of the groups with distinct expression patterns, we performed pathway enrichment analysis of clusters 1\u0026ndash;6 (Fig.\u0026nbsp;4C; Table S4). These results revealed that cluster 1 (HC cluster) was mainly enriched in integrin and thiamine metabolism. Cluster 2 (DCIS cluster) was enriched in the regulation of the actin cytoskeleton and vesicle-mediated transport. Proteins in cluster 3 (DCIS-MI cluster) were mainly involved in IL-7 signaling and VEGF. Cluster 4 (IDC cluster) was enriched in fatty acid oxidation and signaling by receptor tyrosine kinases. Proteins in cluster 5 (MBC cluster) were mainly involved in neutrophil degranulation and beta-alanine metabolism. Cluster 6 (BC-common cluster) was mainly enriched in apoptosis and spliceosomes. Furthermore, protein‒protein interaction (PPI) analysis of neutrophil degranulation showed that PSMD6 was involved in featured pathways in the MBC cluster, implying that both PSMD6 and neutrophils are associated with BC metastasis (Fig.\u0026nbsp;4D, Figure S2).\u003c/p\u003e \u003cp\u003eTo further identify potential peripheral serum biomarkers for the four phases of BC progression, we collected 8,462 overlapping proteins of BC between tissue and serum samples. Then, by assessing the expression levels of the above 8,462 proteins, we found that 642 out of 8,462 were upregulated in BC in both the serum proteome and tissue proteome (ratio\u0026thinsp;\u0026gt;\u0026thinsp;1.5), indicating that these 642 serum proteins might be derived from BC tissues. In particular, 62 of the 642 proteins were significantly overexpressed in tissue samples (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Furthermore, we identified 30 phase-specific proteins (DCIS: n\u0026thinsp;=\u0026thinsp;7; DCIS-MI: n\u0026thinsp;=\u0026thinsp;12; IDC: n\u0026thinsp;=\u0026thinsp;7; MBC: n\u0026thinsp;=\u0026thinsp;4) among these 62 proteins as hub tumor biomarkers (Kruskal‒Wallis test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, ratio\u0026thinsp;\u0026gt;\u0026thinsp;2) (Fig.\u0026nbsp;4E). Interestingly, we identified that PSMD6 was significantly overexpressed in MBC, consistent with the findings of the PPI analysis. We identified PSMD6 as a potential candidate biomarker for the MBC phase.\u003c/p\u003e \u003cp\u003eTaken together, these findings suggested that PSMD6 could not only serve as a serum biomarker for BC metastasis but also be involved in the MBC stage and that neutrophils are also involved in this stage.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eThe CLTA-PSMD6-neutrophil axis promotes breast cancer metastasis\u003c/h2\u003e \u003cp\u003eAiming to mine the correlation between clinical indicators and pathological stages, we performed weighted gene coexpression network analysis (WGCNA) (STAR Methods), which is an unsupervised method to identify groups of coregulated proteins and their association with clinical variables\u003csup\u003e22\u003c/sup\u003e. (Fig.\u0026nbsp;5A; Table S5). Separating the proteomic profiles into eigengene modules revealed a group of modules positively associated with tumor metastasis (p value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001, rho\u0026thinsp;=\u0026thinsp;0.93). WGCNA identified 7 protein modules, and the number of proteins in different modules ranged from 49 proteins in the brown module related to DCIS-MI to 110 proteins in the turquoise module related to fibroadenomas. Moreover, the findings demonstrated an extensive connection between modules and tumor progression. For instance, the green module (DNM1, NANS, EIF2S1, PSMD6, etc.) was significantly correlated with MBC (Pearson\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.95, p value\u0026thinsp;\u0026lt;\u0026thinsp;1E-136), and it contained all MBC hub proteins (Fig.\u0026nbsp;5B, Fig.\u0026nbsp;5C). Furthermore, the green module was strongly associated with the neutrophil ratio, indicating that there was a certain relationship between the neutrophil and the MBC stage (Pearson\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.2, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.0001). Moreover, we found that the PSMD6 expression level was highly correlated with a poor prognosis (Tang et al., 2018, Fig.\u0026nbsp;5D).\u003c/p\u003e \u003cp\u003eTo clarify the relationship among PSMD6, neutrophils and tumor metastasis, we performed cell type deconvolution analysis using CIBERSORT (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cibersortx.stanford.edu\u003c/span\u003e\u003cspan address=\"https://cibersortx.stanford.edu\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to infer the relative abundance level of different cell types in the tumor microenvironment and then evaluated the correlation between PSMD6 and neutrophil score (Table S5). PSMD6 was upregulated in BC in both the serum proteome and tissue proteome, implying that PSMD6 might be produced by tumor tissue and secreted into the serum.\u003c/p\u003e \u003cp\u003eFurthermore, to elucidate the mechanism underlying the metastasis ability of MBC and how this ability is affected by changes in PSMD6, which may occur due to genomic alterations, we assessed a set of tissues (Xu et al.) in this study (Fig.\u0026nbsp;5E). From this analysis, we observed \u003cem\u003ecis\u003c/em\u003e-effects for 9 CNA-affected proteins between IDC and MBC, and these proteins were significantly upregulated in MBC (Student\u0026rsquo;s t test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05, MBC/IDC ratio\u0026thinsp;\u0026gt;\u0026thinsp;2) (Figs.\u0026nbsp;5F-5G). Among these proteins, CLTA was the most highly expressed protein, and it could promote protein transmembrane transport. These findings indicated that the \u003cem\u003ecis\u003c/em\u003e-effect of CLTA could account for the difference between IDC and MBC.\u003c/p\u003e \u003cp\u003eTo further explore how CLTA is involved in tumor metastasis, we performed gene set enrichment analysis (GSEA) for pathway enrichment analysis and also performed correlation analysis. The results showed that the enrichment score of protein secretion signaling had a significant positive correlation with the CLTA protein expression level (Spearman\u0026rsquo;s rho\u0026thinsp;=\u0026thinsp;0.57, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.00001) (Fig.\u0026nbsp;5H). These findings prompted us to conclude that the \u003cem\u003ecis\u003c/em\u003e-effect of CLTA would give rise to tumor metastasis by facilitating protein secretion. Through correlation analysis, we observed 48 DEPs significantly positively correlated with CLTA that were secreted proteins between the IDC and MBC phases (Spearman\u0026rsquo;s rho\u0026thinsp;\u0026gt;\u0026thinsp;0.2, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05) (Fig.\u0026nbsp;5K). PSMD6, the top-ranked protein, was notably associated with the neutrophil score (Figs.\u0026nbsp;5L\u0026ndash; 5 M; rho\u0026thinsp;=\u0026thinsp;0.322, p value\u0026thinsp;=\u0026thinsp;9E-3). Compared with IDC, MBC displayed a higher neutrophil score (Student\u0026rsquo;s t test, p value\u0026thinsp;\u0026lt;\u0026thinsp;0.01, MBC/IDC ratio\u0026thinsp;=\u0026thinsp;3.9) (Fig.\u0026nbsp;5N). The neutrophil score between IDC and MBC was also strongly related to angiogenesis, which was enriched in MBC (Figs.\u0026nbsp;5O-5P).\u003c/p\u003e \u003cp\u003eIn summary, through integrated analysis of the tissue proteome and serum proteome, we found that the CLTA-PSMD6-neutrophil axis promotes tumor metastasis. Specifically, the results demonstrated that the \u003cem\u003ecis-effect\u003c/em\u003e of CLTA probably promotes intracellular to extracellular transport of PSMD6; then, BC cell-derived PSMD6 possibly activates neutrophils; finally, neutrophils might induce tumor cell metastasis by stimulating angiogenesis (Fig.\u0026nbsp;5Q).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eOverexpression of PSMD6 synergistically enhances angiogenesis and NET formation by activating neutrophils\u003c/h2\u003e \u003cp\u003eProteomic analysis revealed that PSMD6 was one of the MBC-specific proteins (Fig.\u0026nbsp;6A), and the expression level was highly positively correlated with the hazard ratio (Fig.\u0026nbsp;5D). To validate the clinical value of PSMD6, we enrolled a second prospective follow-up cohort, named Guangzhou 2. The Guangzhou 2 cohort included 61 HCs, 72 BBT patients, and 121 BC patients, who provided 252 serum samples. The level of PSMD6 in serum was detected by ELISA in the Guangzhou 2 validation set, which suggested that the serum PSMD6 level was significantly higher in patients with metastasis than in healthy controls and patients with benign and invasive breast cancer (Fig.\u0026nbsp;6B). Then, the amplified expression of PSMD6 in breast cancer tissues was confirmed by IHC staining in a subsequent step (Fig.\u0026nbsp;6C), which indicated that the expression level of PSMD6 was higher in patients with lymph node metastasis and AJCC stage III or IV disease (Table S6). Moreover, as multicenter cohorts are supposed to be representative of the general population, we recruited a third validation cohort (named Shanghai) from another center, which could make our study clearer results more convincing and more broadly accepted; this cohort provided 126 BC tissue samples. Kaplan\u0026ndash;Meier survival analysis and Cox proportional hazards analysis of survival data based on the Shanghai validation set showed that breast cancer patients with high PSMD6 expression levels had significantly shorter overall survival times, indicating that high PSMD6 levels are independently associated with poor outcomes in breast cancer patients (Fig.\u0026nbsp;6D) (Table S6).\u003c/p\u003e \u003cp\u003eNext, we explored how PSMD6 was transferred from the cytoplasm to the extracellular space. Our results indicated that CLTA was involved in tumor metastasis and that the enrichment score of the protein secretion signaling gene set had a significant positive correlation with the CLTA protein expression level (Figs.\u0026nbsp;5E-5K). Immunofluorescence analysis showed that PSMD6 colocalized with CLTA (Fig.\u0026nbsp;6E), and knockdown of CLTA by three kinds of shRNA significantly inhibited the PSMD6 protein levels in MCF-7 and MDA-MB-231 cells and their culture medium (Figs.\u0026nbsp;6E-6H). Therefore, these findings prompted us to conclude that CLTA promotes the transfer of PSMD6 from the cytoplasm to the extracellular space.\u003c/p\u003e \u003cp\u003eThen, we explored the functional role of PSMD6 in breast metastasis. We observed that overexpression of PSMD6 could promote cell metastasis in a transwell assay, while knockdown of PSMD6 could inhibit cell metastasis (Figs.\u0026nbsp;6I-6K). Moreover, PSMD6 overexpression in MDA-MB-231-LM2 cells, a cell line with moderate endogenous PSMD6 expression (Figure S4A), significantly exacerbated the lung metastatic burden after intravenous inoculation of cancer cells into mice (Figs.\u0026nbsp;6L-6N). These findings demonstrated a prometastatic role of PSMD6 in breast cancer.\u003c/p\u003e \u003cp\u003eTo explore how serum PSMD6 regulates lung metastasis, we first investigated its effect on tumor migration and invasion \u003cem\u003ein vitro\u003c/em\u003e. However, we found that the cell culture supernatants of the PSMD6-overexpressing cell line (PSMD6 OE cells) or the shPSMD6 cell line (shPSMD6 cells) did not affect cell invasion and metastasis, suggesting a microenvironment-dependent role of PSMD6 in metastasis. After coculturing the neutrophils from healthy volunteers with PSMD6 OE cell culture medium or shPSMD6 cell culture medium, the cell culture supernatants promoted or inhibited control breast cancer cell invasion and metastasis \u003cem\u003ein vitro\u003c/em\u003e (Figs.\u0026nbsp;7A-7B, Figure S4B). In addition, our results suggested that the supernatants from PSMD6 OE cells could induce more neutrophil-like cells (dHL-60) to transform into CD66b\u0026thinsp;+\u0026thinsp;CD11b\u0026thinsp;+\u0026thinsp;cells, which means that dHL-60 cells were activated by the supernatants from PSMD6 OE cells, while supernatants from shPSMD6 cells could inhibit neutrophil-like transformation (Figs.\u0026nbsp;7C-7E). Importantly, flow cytometry analyses also indicated that PSMD6 overexpression increased, while PSMD6 knockdown decreased, the percentages of CD11b\u0026thinsp;+\u0026thinsp;Ly6G\u0026thinsp;+\u0026thinsp;neutrophils, which are tumor-associated neutrophils, in mouse lung metastases (Fig.\u0026nbsp;7F, Figure S4C). Thus, tumoral PSMD6 may regulate lung metastasis by activating neutrophils.\u003c/p\u003e \u003cp\u003eThen, we investigated how PSMD6 affects neutrophils to promote breast cancer metastasis. Previous studies have reported that metastatic cancer cells can induce neutrophils to form metastasis-supporting NETs in the absence of infection\u003csup\u003e17\u003c/sup\u003e. Interestingly, we found that neutrophils cultured in PSMD6-overexpressing LM2 cell culture medium (PSMD6-OE CM) formed more extensive NET structures than those cultured in control medium, as evidenced by IF staining of citrullinated histone H3, a hallmark of chromatin decondensation and extrusion, and the granule protein myeloperoxidase (Figs.\u0026nbsp;7G-7H, Figure S4D). In addition, the number of free NETs in neutrophils cultured in PSMD6-OE CM was also higher than that in neutrophils cultured in control CM, while it was lower in neutrophils cultured in PSMD6 shRNA MDA-MB-231 cell culture medium (shPSMD6-CM) (Fig.\u0026nbsp;7I). Importantly, PSMD6 overexpression in 4T1 cells led to enhancement of NETosis near cancer cells in the lungs and after intravenous inoculation of cancer cells, while PSMD6 knockdown in 4T1 cells had the opposite effect (Figs.\u0026nbsp;7J-7K, Figures S4E-S4F). In addition, angiogenesis is necessary at the beginning of metastasis. Tumor cells must gain access to the vasculature from the primary tumor, survive travel through the circulation, settle in the microvasculature of the target organ, escape from the vasculature into the target organ, and induce angiogenesis in the target organ. Researchers have shown that tumor-associated neutrophils can participate in tumor-promoting inflammation by driving angiogenesis\u003csup\u003e[\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e]\u003c/sup\u003e. Our results showed that the expression of angiogenesis-related proteins in neutrophils after coculture with PSMD6-OE CM was also upregulated, while it was downregulated after coculture with shPSMD6-CM (Figs.\u0026nbsp;7L-7M; Table S7). Consistently, after coculturing PSMD6-OE CM with neutrophils, breast cancer cell tube formation was stimulated (Figs.\u0026nbsp;7N-7Q). In summary, PSMD6 could affect neutrophils to promote breast cancer metastasis by regulating NET formation and angiogenesis-related protein expression.\u003c/p\u003e \u003cp\u003eNext, we further investigated the molecular mechanism by which PSMD6 acts on neutrophils in tumor cell medium. To determine the protein of neutrophils that interact with PSMD6, recombinant PSMD6 protein with a GST tag was used to pull down the plasma protein of neutrophils from breast cancer patients and healthy volunteers. Complexes recovered from the beads were analyzed by MS and western blotting, and the results suggested that PR3 and CD177 are proteins that may interact with PSMD6 (Figs.\u0026nbsp;7R-7T, Figures S4G-S4H; Table S7). PR3, also called myeloblastin, is a neutrophil serine protease\u003csup\u003e62\u003c/sup\u003e. It has been shown that the glycosylphosphatidylinositol (GPI)-anchored neutrophil-specific receptor NB1 (CD177) presents PR3 on the membrane of a neutrophil subset\u003csup\u003e63\u003c/sup\u003e, raising the possibility that tumor-derived PSMD6 might directly regulate neutrophil membrane-bound PR3. To demonstrate that PSMD6 could activate neutrophils by regulating neutrophil membrane-bound PR3, the enzymatic activity of membrane-bound PR3 was assessed after treating cells with CM from PSMD6-overexpressing cancer cells. CM from PSMD6-overexpressing cancer cells could activate the membrane-bound PR3 of human primary neutrophils, while the inhibitor of PR3 (sivelestat) could diminish the activation effect of CM from PSMD6-overexpressing cancer cells (Fig.\u0026nbsp;7U). In addition, previous studies showed that membrane-bound PR3 of human neutrophils could regulate neutrophil chemotaxis\u003csup\u003e[\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e]\u003c/sup\u003e. In our study, we showed that CM from PSMD6-overexpressing cancer cells could significantly enhance angiogenesis-related protein expression, while the inhibitor of PR3 (sivelestat) could inhibit PR3 expression, diminish the effect of PSMD6 and inhibit angiogenesis-related protein expression in neutrophils cocultured with PSMD6-OE CM (Figs.\u0026nbsp;7V-7W; Table S7). Therefore, our results indicated that PSMD6 could be transferred from the cytoplasm to the extracellular space with the aid of CLTA and that PSMD6 could activate neutrophils to release angiogenesis-related proteins by interacting with the PR3-CD177 axis, which ultimately promotes breast cancer cell metastasis (Fig.\u0026nbsp;7X).\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThe World Health Organization (WHO) has proposed that millions of cancer patients could be saved from premature death if early detection and treatment methods were available\u003csup\u003e65\u003c/sup\u003e. Finding the tumor at an early stage when it is still localized and possibly even before clinical symptoms develop is one important application of specific biomarkers\u003csup\u003e65\u003c/sup\u003e. Apart from early diagnosis, biomarkers could also provide physicians with actionable information leading to evidence-based selection of the optimal therapy and to improved and more precise methods for detecting disease progression\u003csup\u003e65\u003c/sup\u003e. Ideally, protein biomarkers should be found via minimally invasive liquid biopsy, such as by taking a simple blood sample\u003csup\u003e65\u003c/sup\u003e. Currently, although a few biomarkers \u0026mdash; for example, CA125 for ovarian cancer, CA19-9 for pancreatic cancer and PSA for prostate cancer \u0026mdash; have been proposed to be useful for longitudinal disease monitoring, there is still a lack of breast cancer-specific biomarkers that are associated with clinical problems, rather than just the differentiation of cancer patients from healthy individuals.\u003c/p\u003e \u003cp\u003eHere, we portrayed the serum proteomic landscape and explored proteomic signatures associated with the progression of BC. Our study included a discovery stage and three validation stages involving a total of 813 samples. The discovery stage was designed as a case‒control study and involved 322 samples, including 54 HCs, 112 BBTs and 156 BCs. The BC samples included those of various pathological stages to represent BC progression; specifically, they included 25 DCIS samples, 16 DCIS-MI samples, 68 IDC samples, and 45 MBC samples. Furthermore, we conducted three-step validation based on three independent cohorts (Guangzhou 1, Guangzhou 2, and Shanghai). For the Guangzhou 1 validation cohort, a prospective follow-up cohort study was conducted to verify the effectiveness of the BC diagnostic classifier. This cohort involved 113 serum samples, including 27 HCs, 29 BBTs, and 57 BCs. For the Guangzhou 2 validation cohort, a prospective follow-up cohort study was conducted to verify PSMD6 as a potential specific serum biomarker for BC metastasis. This cohort provided 252 serum samples, including 61 HC samples, 72 BBT samples, and 121 BC samples. For the Shanghai validation cohort, a retrospective cohort study was conducted to verify PSMD6 as a potential specific biomarker of BC metastasis associated with poor prognosis. The Shanghai validation cohort provided 126 BC tissue samples. To our knowledge, this is the largest study cohort to comprehensively explore the dynamic changes in the serum proteome and protein signature in BC progression.\u003c/p\u003e \u003cp\u003eCurrently, ultrasonography and mammography are used together with histopathological confirmation as the gold standard for BC diagnosis\u003csup\u003e66\u003c/sup\u003e. There are, however, several disadvantages of the abovementioned modalities, which makes it reasonable to continue research and development in the area of alternative methods for diagnosing BC\u003csup\u003e66\u003c/sup\u003e. Since blood-derived proteins are readily available and perform vital activities, blood-derived protein biomarkers introduce significant potential in BC diagnosis as a complementary and adjunctive modality to the current clinical gold standard. In this study, we observed distinct serum proteomic profiles between subjects with BCs and non-BCs (HCs and BBTs). The 24-feature classifier was further constructed for distinguishing between BCs and non-BCs based on a machine learning method, and the model showed 94% sensitivity and 94% specificity in the diagnosis of BCs. In addition, to validate the predictive strength of our analysis, we entered the validation stage, and the model showed 80% sensitivity and 80% specificity in our Guangzhou 1 validation cohort. Our results showed that the 24-feature classifier can distinguish breast cancer patients from healthy individuals and can also distinguish benign breast tumors from malignant breast tumors. This study may provide a reference value for differentiating BC and non-BC using serum in the future.\u003c/p\u003e \u003cp\u003eThe 26S proteasome is an important protease in eukaryotic cells and is composed of a 20S core particle (CP) and one or two 19S regulatory particles (RPs) capping one or both ends of the 20S CP\u003csup\u003e67\u003c/sup\u003e. Some studies have shown that the 26S proteasome plays a significant part in tumor progression. For example, T\u0026uuml;rkoğlu et al. reported that PSMD4 promotes cell proliferation via regulation of the PTEN/Akt pathways in hepatocellular carcinoma (HCC) cells\u003csup\u003e68\u003c/sup\u003e. Okumura et al. reported that PSMD1 is a gene associated with acquired tamoxifen resistance that may contribute to the proliferation of breast cancer cells putatively through modulating the p53 pathway\u003csup\u003e69\u003c/sup\u003e. In this study, we disclosed the serum proteomic features of various histopathological subtypes of BC progression and found that PSMD6 was the top MBC-specific protein that was considered a risk factor for BC metastasis. High expression of PSMD6 was associated with poor BC prognosis in our Shanghai validation cohort. Integrated analysis of Xu et al.\u0026rsquo;s BC tissue proteomics data (n\u0026thinsp;=\u0026thinsp;402) and our serum proteomics data showed that PSMD6 was also upregulated in MBC tissue samples. The amplified expression of PSMD6 in MBC tissues was confirmed in our Shanghai validation cohort by IHC staining in a subsequent step. Our findings suggested that serum-derived PSMD6 may be secreted by tissues and interact with the microenvironment to promote BC metastasis, suggesting that it may be a potential target for the treatment of metastatic BC.\u003c/p\u003e \u003cp\u003eThe tumor microenvironment plays a pivotal role in the tumorigenesis, progression, and metastasis of many cancers, including breast cancer\u003csup\u003e70\u003c/sup\u003e. There is now increasing evidence to support the observations that the multiple interactions between breast cancer cells and neighboring cells in the tumor microenvironment coordinate to regulate metastasis\u003csup\u003e70,71\u003c/sup\u003e. For example, Banerjee et al. reported that BC cells can upregulate IL-6 expression in adipocytes, which in turn promotes angiogenesis, tumor cell proliferation and survival via the JAK/STAT3 signaling pathway\u003csup\u003e72\u003c/sup\u003e. In this study, we found that PSMD6 acts on neutrophils through its interaction with the neutrophil serine protease PR3 and promotes BC metastasis by regulating NET formation and the expression of angiogenesis-related proteins. Based on the role of the PSMD6-neutrophil axis in BC metastasis, we identified the potential therapeutic agent sivelestat, which targets the PSMD6-neutrophil axis and demonstrated its inhibitory effect on tumor cells. These results revealed the role of the PSMD6-neutrophil axis in promoting tumor cell metastasis and showed that this axis can be targeted with inhibitors, providing a potential therapeutic option for patients with metastatic BC with PSMD6 overexpression.\u003c/p\u003e \u003cp\u003eIn general, gene amplification is associated with increased tumor aggression, metastasis, and resistance to chemotherapy\u003csup\u003e73\u003c/sup\u003e. For example, in breast and lung tumors, \u003cem\u003eERBB2\u003c/em\u003e amplification induces overexpression of the protein in the cell membrane\u003csup\u003e73,74,75,76\u003c/sup\u003e, which has been associated with a poor prognosis\u003csup\u003e73,77,78\u003c/sup\u003e, while overexpression in gastric tumors is related to the presence of metastases\u003csup\u003e73,79,80,81\u003c/sup\u003e and evolution to the gastric intestinal type\u003csup\u003e73,82\u003c/sup\u003e. In this study, we analyzed BC tissue multiomics data from Xu et al. and found that, from IDC to distant metastasis, the increase in \u003cem\u003eCLTA\u003c/em\u003e copy number showed a \u003cem\u003ecis\u003c/em\u003e effect on its protein. The overexpression of \u003cem\u003eCLTA\u003c/em\u003e promoted the transfer of PSMD6 from the cytoplasm to the extracellular space. Our findings suggested that genetic variants may alter the tumor microenvironment by altering the expression of proteins and pathways, thereby creating conditions for tumor metastasis. In addition, these results suggested that clinical testing of \u003cem\u003eCLTA\u003c/em\u003e amplification may be considered a way to assess BC risk in the future.\u003c/p\u003e \u003cp\u003eIn conclusion, in this study, we captured the changes in the serum proteome of BC patients and showed that these changes are clearly linked to the underlying disease manifestations and clinical observations. We demonstrated that serum proteome profiling enables the discovery of better biomarkers, which could have a major impact on important aspects of disease management: (i) The 24-feature classifier will enable the diagnosis of BC and allow clinicians to distinguish between BBT and BC, (ii) the progression clock will provide information about the progression of the disease, and (iii) the potential therapeutic opportunity revealed herein could provide benefits for MBC patients with CLTA amplification.\u003c/p\u003e"},{"header":"STAR METHODS","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\n \u003ch2\u003eKEY RESOURCES TABLE\u003c/h2\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eREAGENT or RESOURCE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSOURCE\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eIDENTIFIER\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eAntibodies\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eanti-PSMD6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eHUABIO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: ER64500\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eanti-CLTA\u003c/p\u003e\n \u003cp\u003eanti-PSMD6\u003c/p\u003e\n \u003cp\u003eanti-CiH3\u003c/p\u003e\n \u003cp\u003eanti-MPO\u003c/p\u003e\n \u003cp\u003eanti-MPO\u003c/p\u003e\n \u003cp\u003eanti-a-tublin\u003c/p\u003e\n \u003cp\u003eanti-CD177\u003c/p\u003e\n \u003cp\u003eanti-PRTN3\u003c/p\u003e\n \u003cp\u003eanti-GPR78\u003c/p\u003e\n \u003cp\u003eanti-Na+/K+ ATPase\u003c/p\u003e\n \u003cp\u003eanti-GAPDH\u003c/p\u003e\n \u003cp\u003eAPC anti-mouse Ly6G antibody [1A8]\u003c/p\u003e\n \u003cp\u003eFITC anti-mouse CD45 antibody [30-F11]\u003c/p\u003e\n \u003cp\u003ePE Anti-Mouse/Human CD11b antibody [M1/70]\u003c/p\u003e\n \u003cp\u003eAPC Mouse IgM, \u0026kappa; Isotype Control [MM-30]\u003c/p\u003e\n \u003cp\u003ePE Mouse IgG1, \u0026kappa; Isotype Control\u003c/p\u003e\n \u003cp\u003ePE Anti-Human CD11b Antibody [ICRF44]\u003c/p\u003e\n \u003cp\u003eAPC Anti-Human CD66b Antibody [G10F5]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eProteintech Group\u003c/p\u003e\n \u003cp\u003eSanta Cruz\u003c/p\u003e\n \u003cp\u003eHUABIO\u003c/p\u003e\n \u003cp\u003eProteintech Group\u003c/p\u003e\n \u003cp\u003eHUABIO\u003c/p\u003e\n \u003cp\u003eSAB\u003c/p\u003e\n \u003cp\u003eSAB\u003c/p\u003e\n \u003cp\u003eProteintech Group Elabscience\u003c/p\u003e\n \u003cp\u003eHUABIO\u003c/p\u003e\n \u003cp\u003eProteintech Group\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eElabscience\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 10852-1-AP\u003c/p\u003e\n \u003cp\u003eCat#: sc-393580\u003c/p\u003e\n \u003cp\u003eCat#: M1306-4\u003c/p\u003e\n \u003cp\u003eCat#: 22225-1-AP\u003c/p\u003e\n \u003cp\u003eCat#: ET1703-21\u003c/p\u003e\n \u003cp\u003eCat#: 21581\u003c/p\u003e\n \u003cp\u003eCat#: 41689\u003c/p\u003e\n \u003cp\u003eCat#: 67030-1\u003c/p\u003e\n \u003cp\u003eCat#: 40588\u003c/p\u003e\n \u003cp\u003eCat#: ET1609-76\u003c/p\u003e\n \u003cp\u003eCat#: 60004-1\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F1108E\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F1136C\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F1081D\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F09782E\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F09792D\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F1146D\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: E-AB-F1267E\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eBiological Samples\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eSerum samples (n=322) from a cohort of 306 breast cancer patients or mammary benign disease patients\u003c/p\u003e\n \u003cp\u003eHuman breast cancer tissues array\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eSerum samples (n=255) from a cohort of breast cancer or mammary benign disease patients\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eGuangdong Provincial People\u0026rsquo;s Hospital Shanghai Outdo Biotech.Co., LTD\u003c/p\u003e\n \u003cp\u003eGuangdong Provincial People\u0026rsquo;s Hospital\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eThis paper\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003ehBreD132Su07\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eThis paper\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eCritical Commercial Assay\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eBCA protein assay kit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eBeyotime Biotechnology\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: P0011\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eBradford protein assay\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eThermoFisher Scientific\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat# 23236\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eImmunohistochemistry kit\u003c/p\u003e\n \u003cp\u003eTumor dissociation kit, mouse\u003c/p\u003e\n \u003cp\u003eMouse tumor infiltrating tissues kit\u003c/p\u003e\n \u003cp\u003eELISA kit\u003c/p\u003e\n \u003cp\u003eMinuteTM Plasma Membrane Protein Isolation and Cell Fractionation Kit\u003c/p\u003e\n \u003cp\u003eQuant-iT\u0026trade; PicoGreen\u0026trade; dsDNA\u003c/p\u003e\n \u003cp\u003emRNA qRT‒PCR starter kit\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eZhongshan Jinqiao\u003c/p\u003e\n \u003cp\u003eMiltenyi Biotec\u003c/p\u003e\n \u003cp\u003eSolarbio\u003c/p\u003e\n \u003cp\u003eOmnimAbs\u003c/p\u003e\n \u003cp\u003eInvent Biotechnologies\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eThermo\u003c/p\u003e\n \u003cp\u003eTIANGEN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: SP-9000\u003c/p\u003e\n \u003cp\u003eCat#:130-096-730\u003c/p\u003e\n \u003cp\u003eCat#: P2430\u003c/p\u003e\n \u003cp\u003eCat#: OM487066\u003c/p\u003e\n \u003cp\u003eCat#: 89881\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: P7581\u003c/p\u003e\n \u003cp\u003eCat#: KR116-01\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" colspan=\"3\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eChemicals, Peptides, and Recombinant Proteins\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eTrypsin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003ePromega\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: V528A\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003ePMSF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eSigma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: P-7626\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eProtease inhibitor cocktail\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eRoche\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 04693159001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003ePhosphatase inhibitor cocktail\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eRoche\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 04906837001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eHPLC-grade water\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eJ.T. Baker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 4218-03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eAcetonitrile\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eJ.T. Baker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 9829-03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eFormic acid\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eSigma\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: F0507\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eMethanol\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eJ.T. Baker\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 9830-03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eC18 resin\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eDikma Technologies\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 85252\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eSepPark C18 cartridges\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eWaters\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: WAT054960\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eXbridge C18 column\u003c/p\u003e\n \u003cp\u003ePR3-specific substrate (Abz)-VADnorVADRQ-(EDDnp)\u003c/p\u003e\n \u003cp\u003eSivelestat\u003c/p\u003e\n \u003cp\u003ePhorbol 12-myristate 13-acetate (PMA)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eWaters\u003c/p\u003e\n \u003cp\u003eCayman\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eMedChemExpress\u003c/p\u003e\n \u003cp\u003eSelleck\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eCat#: 186003576\u003c/p\u003e\n \u003cp\u003eCat#: 9002021-1\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eCat#: 201677-61-4\u003c/p\u003e\n \u003cp\u003eCat#: S7791\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSoftware and Algorithms\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eSource\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eIdentifier(i.e., links)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eFirmiana platform\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e(Feng et al., 2017)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://phenomics.fudan.edu.cn/firmiana/gardener/\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eR (version 3.5.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://bioconductor.org/packages/release/bioc/html/CopywriteR.html\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003eKEGG database\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eReactome database\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" rowspan=\"2\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://www.phosphosite.org/homeAction\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"100%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://reactome.org\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eDAVID\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://david. ncifcrf.gov\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003exCell\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e(Aran et al., 2017)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://xcell.ucsf.edu\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eImageJ software (version 1.51j)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eNational Institutes of Health\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003ehttps://imagej.nih.gov/ij/\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cstrong\u003eDeposited Data\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003e\u003cu\u003e\u0026nbsp;\u003c/u\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd width=\"41.04882459312839%\" valign=\"top\"\u003e\n \u003cp\u003eProteomics data\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"27.305605786618443%\" valign=\"top\"\u003e\n \u003cp\u003eThis paper\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd width=\"31.645569620253166%\" valign=\"top\"\u003e\n \u003cp\u003eiProx: IPX0007188000\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\n \u003ch2\u003eLEAD CONTACT AND MATERIALS AVAILABILITY\u003c/h2\u003e\n \u003cdiv id=\"Sec13\" class=\"Section3\"\u003e\n \u003ch2\u003eRESOURCE AVAILABILITY\u003c/h2\u003e\n \u003cdiv id=\"Sec14\" class=\"Section4\"\u003e\n \u003ch2\u003eLead contact\u003c/h2\u003e\n \u003cp\u003eFurther information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Chen Ding ([email protected]). This study did not generate new unique reagents.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\n \u003ch2\u003eMaterial availability\u003c/h2\u003e\n \u003cp\u003eThis study did not generate new unique reagents.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\n \u003ch2\u003eData and Code Availability\u003c/h2\u003e\n \u003cp\u003eProteomics raw datasets are available through the iProx Consortium (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.iprox.org/\u003c/span\u003e\u003c/span\u003e) with the subproject ID (IPX0007188000) or the firmiana platform (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.firmiana.org/login/\u003c/span\u003e\u003c/span\u003e). Sample annotation and processed and normalized data files are provided in Tables S1-S2. The software and code used in this study are referenced in their corresponding STAR Method sections and the Key Resource Table.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec17\" class=\"Section2\"\u003e\n \u003ch2\u003eEXPERIMENTAL MODEL AND SUBJECT DETAILS\u003c/h2\u003e\n \u003cdiv id=\"Sec18\" class=\"Section3\"\u003e\n \u003ch2\u003ePatient samples\u003c/h2\u003e\n \u003cp\u003eThe serum samples used in this study were obtained from the Guangdong Provincial People\u0026rsquo;s Hospital. Serum samples were collected from patients or healthy controls. Breast cancer was determined by the attending doctors based on the clinical diagnostic guidelines of the Chinese Health Commission (\u003csup\u003e6t\u003c/sup\u003eh edition) and previous studies, which revealed the clinical courses of 306 patients. Blood samples (\u0026le;\u0026thinsp;3 mL) from patients were collected over the course of their disease at intervals of 3\u0026ndash;5 days. The clinical information of 306 patients, including tumor type, sex, age, tumor node metastasis (TNM) staging, and biochemical indicators, is listed in Table S1. The study was approved by the Research Ethics Committees of Zhongshan Hospital (No.), and written, informed consent was provided by all patients. The tissue chips are commercial and come from a company named OUTDO BIOTECH CO., LTD in Shanghai.\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e\n \u003ch2\u003eStudy cohorts\u003c/h2\u003e\n \u003cp\u003eOur studies had a discovery stage and three validation stages involving a total of 813 samples. The discovery stage was designed as a case‒control study and involved 322 samples, including 54 HC samples, 112 BBT samples and 156 BC samples. Among them, the BC samples included those reflecting various pathological stages of BC progression, such as 25 DCIS samples, 16 DCIS-MI samples, 68 IDC samples, and 45 MBC samples.\u003c/p\u003e\n \u003cp\u003eFurthermore, to ensure the uniformity of serum samples for mass spectrometry experiments and a sufficient sample size for ELISA experiments (two parallel experiments for each sample, one experiment dosage of 200 \u0026micro;l), we conducted three-step validation based on three independent cohorts from Guangzhou 1, Guangzhou 2, and Shanghai. For the Guangzhou 1 validation cohort, a prospective follow-up cohort study was conducted to verify the effectiveness of the BC diagnostic classifier. This cohort involved 103 serum samples, including 17 HCs, 29 BBTs, and 57 BCs. For the Guangzhou 2 validation cohort, a prospective follow-up cohort study was conducted to verify PSMD6 as a potential specific serum biomarker for BC metastasis. This cohort involved 252 serum samples, including 61 HCs, 72 BBTs, and 121 BCs. For the Shanghai validation cohort, a retrospective cohort study was conducted to verify PSMD6 as a potential specific biomarker of BC metastasis associated with poor prognosis.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec20\" class=\"Section2\"\u003e\n \u003ch2\u003eMethods Details\u003c/h2\u003e\n \u003cdiv id=\"Sec21\" class=\"Section3\"\u003e\n \u003ch2\u003eProteomic Workflow\u003c/h2\u003e\n \u003cdiv id=\"Sec22\" class=\"Section4\"\u003e\n \u003ch2\u003eSerum protein extraction and trypsin digestion\u003c/h2\u003e\n \u003cp\u003eSerum samples were mixed with 100 \u0026micro;L 50 mM ammonium bicarbonate (ABC) buffer, and the proteins were inactivated at 95\u0026deg;C for 5 min. The samples were cooled to room temperature and digested using trypsin at an enzyme to protein mass ratio of 1:25 for 17 hours in a 37\u0026deg;C incubator. Then, 5 \u0026micro;L of aqueous ammonia was added to each tube and vortexed to quench the digestion reaction, and the supernatant was subsequently dried using a 60\u0026deg;C vacuum drier (SpeedVac, Eppendorf). Then, the peptides were dissolved in 100 \u0026micro;L 0.1% formic acid (FA), vortexed for 3 min, and then sedimentation for 3 min (12,000 \u0026times;g). The supernatant was picked into a new tube and then desalinated. Before desalination, the activation of pillars with 2 slides of 3 M C18 disk is needed, and the lipid is as follows: 90 \u0026micro;L 100% acetonitrile (ACN) twice, 90 \u0026micro;L 50% and 80% ACN once in turn, and then 90 \u0026micro;L 50% ACN once. After pillar balance with 90 \u0026micro;L 0.1% FA twice, the supernatant of the tubes was loaded into the pillar twice and decontamination with 90 \u0026micro;L 0.1% FA twice. Finally, 90 \u0026micro;L elution buffer (0.1% FA in 50% ACN) was added to the pillar fir elution twice, and only the effluent was collected for MS. Finally, the collected peptides were dried using a 60\u0026deg;C vacuum drier.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e\n \u003ch2\u003eNano-LC‒MS/MS\u003c/h2\u003e\n \u003cp\u003eThe acquisition of samples was randomized to avoid bias. Samples were measured using LC‒MS instrumentation consisting of an EASY-nLC 1200 ultrahigh-pressure system (Thermo Fisher Scientific) coupled via a nano-electrospray ion source (Thermo Fisher Scientific) to a Q Exactive HF-X Hybrid Quadrupole-Orbitrap mass spectrometer (Thermo Fisher Scientific). Peptides, redissolved in Solvent A (0.1% formic acid in water), were loaded onto a 2-cm self-packed trap column (100-\u0026micro;m inner diameter, 3-\u0026micro;m ReproSil-Pur C18-AQ beads, Dr. Maisch GmbH) using Solvent A, separated on a 150-\u0026micro;m-inner-diameter column with a length of 8 cm (1.9-\u0026micro;m ReproSil-Pur C18-AQ beads, Dr. Maisch GmbH) with 6\u0026ndash;95% mobile phase B (80% ACN and 0.1% formic acid) at 600 nL/min for 8.2 min, held constant at 95% solvent B at 800 nL/min for 4.1 min and then returned to 3% B for an additional 2.7 min to equilibrate the column. The eluted peptides were ionized under 2 kV and introduced into the mass spectrometer. The MS analysis was performed in a data-independent acquisition (DIA) mode. The DIA method consisted of MS1 Spectra full scan with m/z ranging from 300 to 1,400 at a high resolution of 30,000 with an automatic gain control (AGC) target 3E\u0026thinsp;+\u0026thinsp;06. The maximal ion injection time was 20 ms. Then, 30 DIA segments were acquired at 15,000 resolution with an AGC target of 1E\u0026thinsp;+\u0026thinsp;06 for maximal injection time. The setting \u0026ldquo;inject ions for all available parallelizable time\u0026rdquo; was enabled. HCD fragmentation was set to a normalized collision energy of 27%. The spectra were recorded in profile mode. The default charge state for the DIA was set to 3. All data were acquired using Xcalibur software v2.2 (Thermo Fisher Scientific).\u003c/p\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec24\" class=\"Section2\"\u003e\n \u003ch2\u003ePeptide identification and protein quantification\u003c/h2\u003e\n \u003cp\u003eAll data were processed using Firmiana\u003csup\u003e83\u003c/sup\u003e. DIA data were searched against the UniProt human protein database (updated on 2019.12.17, 20406 entries) using FragPipe (v12.1) with MSFragger (2.2)\u003csup\u003e84\u003c/sup\u003e. The mass tolerances were 20 ppm for precursor and 50 mmu for product ions. Up to two missed cleavages were allowed. The search engine was set with cysteine carbamidomethylation as a fixed modification and N-acetylation and oxidation of methionine as variable modifications. Precursor ion score charges were limited to +\u0026thinsp;2, +3, and +\u0026thinsp;4. The data were also searched against a decoy database so that protein identifications were accepted at a false discovery rate (FDR) of 1%. The results of DDA data were combined into spectral libraries. A total of 327 libraries were used as reference spectra libraries.\u003c/p\u003e\n \u003cp\u003eDIA data were analyzed using DIA-NN (v1.7.0)\u003csup\u003e85\u003c/sup\u003e. The default settings were used for DIA-NN (Precursor FDR: 1%, Log lev: 1, Mass accuracy: 20 ppm, MS1 accuracy: 10 ppm, Scan window: 30, Implicit protein group: genes, Quantification strategy: robust LC (high accuracy)). Quantification of identified peptides was calculated as the average of chromatographic fragment ion peak areas across all reference spectra libraries. Label-free protein quantifications were calculated using a label-free, intensity-based absolute quantification (iBAQ) approach\u003csup\u003e86\u003c/sup\u003e. We calculated the peak area values as parts of the corresponding proteins. The fraction of total (FOT) was used to represent the normalized abundance of a particular protein across samples. FOT was defined as a protein\u0026rsquo;s iBAQ divided by the total iBAQ of all identified proteins within a sample. The FOT values were multiplied by 10\u003csup\u003e5\u003c/sup\u003e for ease of presentation, and missing values were imputed with 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e.\u003c/p\u003e\n \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e\n \u003ch2\u003eMissing value imputation\u003c/h2\u003e\n \u003cp\u003eBefore performing any downstream statistical analyses, proteome datasets of plasma were filtered for 50% valid values across all samples. Missing values were subjected to KNN imputation on the data using the \u0026ldquo;impute.knn\u0026rdquo; function from the \u0026ldquo;impute\u0026rdquo; R package\u003csup\u003e87\u003c/sup\u003e.\u003c/p\u003e\n \u003c/div\u003e\n \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e\n \u003ch2\u003eProteome Data Preprocessing\u003c/h2\u003e\n \u003cdiv id=\"Sec27\" class=\"Section4\"\u003e\n \u003ch2\u003eMass spectrometry platform QC\u003c/h2\u003e\n \u003cp\u003eFor quality control of the MS performance during serum sample detection, we mixed all 322 samples into a serum pool as QC standards. The QC standards were analyzed using the same method and conditions as our serum cohort. Pearson\u0026rsquo;s correlation coefficient was calculated for QC standards. The average correlation coefficient of the QC standards was 0.98. The minimum and maximum correlation coefficients were 0.93 and 0.99, respectively, which demonstrated the stability of the mass spectrometry platform.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/div\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec28\" class=\"Section2\"\u003e\n \u003ch2\u003ePreprocessing of DIA proteomic data and batch correction\u003c/h2\u003e\n \u003cp\u003eConsidering the balance between the confidence of protein identification and sample heterogeneity, we selected proteins by a specific threshold and then imputed them. First, the analysis in this study focused on the proteins identified in more than 50% of samples for each sample type (3 tumor subtypes and 1 healthy control). Second, we performed KNN imputation separately on the data for each sample type using the \u0026ldquo;impute.knn\u0026rdquo; function from the \u0026ldquo;impute\u0026rdquo; R package. We then combined the imputed data across all 4 sample types and obtained 8,944 proteins in total. Because the missing proteins of each tumor subtype are different, we then filled the empty value by 10\u003csup\u003e\u0026minus;\u0026thinsp;5\u003c/sup\u003e. Finally, we applied the R tool Combat, with the tumor type as a covariate to remove batch effects\u003csup\u003e88\u003c/sup\u003e.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec29\" class=\"Section2\"\u003e\n \u003ch2\u003eGene set score for a single sample\u003c/h2\u003e\n \u003cp\u003eTo functionally characterize NMF cluster results by single-sample gene set enrichment analysis (ssGSEA), we calculated the normalized enrichment score of each sample based on four classes of gene sets: GOBP, KEGG, hallmark, and reactome gene sets. We utilized the R package GSVA\u003csup\u003e89\u003c/sup\u003e with the following parameters: min.sz\u0026thinsp;=\u0026thinsp;10, max.sz\u0026thinsp;=\u0026thinsp;300, and other parameters were set to default values.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eConstruction and validation of predictive models to distinguish between BC, BBT and normal samples\u003c/h3\u003e\n\u003cp\u003eLogistic regression analysis was used to distinguish between BCs, BBTs and the normal prediction model based on the significantly differentially expressed proteins in BCs, BBTs and normal serum samples using Python software v3.10.0. The backward stepwise method was utilized for feature selection. Samples were randomly divided into the training set (n\u0026thinsp;=\u0026thinsp;217) and the testing set (n\u0026thinsp;=\u0026thinsp;94). Moreover, the diagnostic value of this model was verified using scikit-learn analysis (Version 1.3.1). Sensitivity, specificity, accuracy, and AUC were used to determine the predictive value of the model. The predictive value of the model was validated in the validation cohort.\u003c/p\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003eFuzzy C-means clustering\u003c/h2\u003e \u003cp\u003eProteins were grouped into different clusters using the Mfuzz package in R with the fuzzy c-means algorithm\u003csup\u003e90\u003c/sup\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eWGCNA\u003c/h2\u003e \u003cp\u003eWeighted gene coexpression network analysis (WGCNA)\u003csup\u003e91\u003c/sup\u003e was applied to the proteins from plasma tumor samples using R code implemented in R software. The parameters were TOMType = \u0026lsquo;unsigned\u0026rsquo;, corType = \u0026lsquo;unsigned\u0026rsquo;, mingene\u0026thinsp;=\u0026thinsp;50, and the rest were default. Spearman correlation analysis was conducted between the protein modules and the tumor types.\u003c/p\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003eImmune scores\u003c/h2\u003e \u003cp\u003eThe levels of different cell types within tumor tissues were computed via CIBERSORT using protein expression values \u003csup\u003e92\u003c/sup\u003e. Table S5 contains the final score computed by CIBERSORT of different cell types for tumor samples.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section3\"\u003e \u003ch2\u003eCorrelation between tumor types and clinical features\u003c/h2\u003e \u003cp\u003eTo estimate the correlations between tumor types and clinical features, the chi-square test was used for categorical variables, and the Kruskal-Willis test was used for continuous variables.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eScreening potential druggable targets\u003c/h3\u003e\n\u003cp\u003eTo screen potential druggable targets, the following criteria need to be met: 1) candidates for core tumor markers and 2) drug targets annotated by the HPA\u003csup\u003e93\u003c/sup\u003e and DrugBank databases (version 5.1.5) (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.drugbank.ca/\u003c/span\u003e\u003cspan address=\"http://www.drugbank.ca/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eCell lines\u003c/h3\u003e\n\u003cp\u003eThe cell lines generated in this study (MDA-MB-231, MDA-MB-231-LM2, MCF7, MCF10A and BT474) were grown in DMEM with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. 4T1 cells were grown in RPMI-1640 with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. HL-60 cells were grown in IMDM with 10% v/v FBS and 100 mg/ml penicillin/streptomycin. Cell cultures were tested for mycoplasma contamination every week.\u003c/p\u003e \u003cdiv id=\"Sec37\" class=\"Section2\"\u003e \u003ch2\u003eHuman breast cancer tissues and blood samples\u003c/h2\u003e \u003cp\u003eA human breast cancer tissue (n\u0026thinsp;=\u0026thinsp;126) array (Shanghai Outdo Biotech Co., LTD) with follow-up information was used for Kaplan‒Meier analysis of disease-free survival and overall survival. Univariate Cox and multivariate Cox analyses were also conducted. Whole blood samples for neutrophil isolation and ELISA (n\u0026thinsp;=\u0026thinsp;255) were collected from Guangdong Provincial People\u0026rsquo;s Hospital under exemption approval of the Guangdong Provincial People\u0026rsquo;s Hospital Institutional Review Board.\u003c/p\u003e \u003cdiv id=\"Sec38\" class=\"Section3\"\u003e \u003ch2\u003eFACS analysis of immunocytes in lung metastases and molecular expression of dHL-60\u003c/h2\u003e \u003cp\u003eLung metastases were picked. Tumor Dissociation Kit, mouse (Cat, 130-096-730, Miltenyi Biotec) was used to generate single cell suspensions. Neutrophils were isolated from mouse tumor infiltrating tissues using a kit (Cat, P2430, Solarbio). HL-60 cells were treated with 1 \u0026micro;M ATRA for 5 days, which could transform HL-60 to dHL-60 cells. dHL-60 cells were treated with culture medium from PSMD6 OE LM2 cells or sh-PSMD6 MDA-MB-231 cells. Cells were incubated for 30 min with 0.5% FBS to block FcR before antibody staining. APC anti-mouse Ly6G antibody [1A8], FITC anti-mouse CD45 antibody [30-F11], PE anti-mouse/human CD11b antibody [M1/70], APC anti-human CD66b antibody [G10F5], PE anti-human CD11b antibody [ICRF44], PE mouse IgG1, κ isotype control, APC mouse IgM, and κ isotype control [MM-30] were used for staining (Elabscience). Flow cytometry was performed by a CANTO II (BD) FACS system and quantified by FlowJo V10 software.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec39\" class=\"Section2\"\u003e \u003ch2\u003eIHC staining\u003c/h2\u003e \u003cp\u003eThe tissue microarrays were stained with immunohistochemistry. The sections were deparaffinized in xylene and dehydrated through alcohol changes. The sections were stained with a PSMD6 rabbit polyclonal antibody (HUABIO). Antibodies were prediluted by the manufacturer, and staining was performed following the manufacturer\u0026rsquo;s protocols. Two pathologists independently reviewed the pathological specimens. Scoring was assessed according to a previous study description\u003csup\u003e94\u003c/sup\u003e.\u003c/p\u003e \u003cdiv id=\"Sec40\" class=\"Section3\"\u003e \u003ch2\u003eELISA\u003c/h2\u003e \u003cp\u003ePlasma was collected from healthy volunteers or patients who were diagnosed with breast cancer or benign tumors and stored at -80\u0026deg;C\u0026deg;C. A standard curve of PSMD6 was established with standard samples in the kit (OmnimAbs). Fifty microliters of sample was added to the appropriate well of the antibody precoated microtiter plate and gently mixed. Incubate for 45 min at 37\u0026deg;C. The liquid was removed, the plate was dried by swing, and washing buffer was added to every well for 30 seconds; the buffer was then removed, and this process was repeated 4 times. Diluted biotinylated anti-IgG (50 \u0026micro;l) was added to the sample wells and incubated for 30 min at 37\u0026deg;C. The sample was washed and dried. Then, 50 \u0026micro;l of streptavidin-HRP was added to all wells and gently mixed. The sample was incubated for 30 min at 37\u0026deg;C. After washing three times, signals were detected using TMB solution and read at 450 nm.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIsolation of neutrophil plasma membrane proteins\u003c/b\u003e \u003c/p\u003e \u003cp\u003ePlasma membrane proteins were isolated from human peripheral blood-derived neutrophils with the MinuteTM Plasma Membrane Protein Isolation and Cell Fractionation Kit (Invent Biotechnologies, 89881) as previously described\u003csup\u003e95\u003c/sup\u003e Briefly, cells were first sensitized by buffer A before passing through the proprietary filter in a zigzag manner when high-speed centrifugal force was applied, resulting in a cell lysate containing ruptured cell membranes and intact nuclei. As a result, nuclear contamination was virtually eliminated. The plasma membrane was further separated from the cell lysate (a mixture of crude membranes, intact nuclei, cytosolic proteins and organelles) by subsequent differential and density centrifugation with a regular tabletop microcentrifuge.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTubule formation assay\u003c/b\u003e \u003c/p\u003e \u003cp\u003eA 96-well plate was coated with 50 \u0026micro;l of Matrigel per well, HUVECs were seeded in a Matrigel-coated 96-well plate at a density of 1.5 \u0026times; 10\u003csup\u003e4\u003c/sup\u003e cells per well, and medium from neutrophils pretreated with cancer cell CM for 12 h (NCM) was added. After 6 h, pictures were taken with a light microscope (Olympus, Tokyo, Japan), and the Angiogenesis Analyzer plugin of ImageJ was used to count the number of branches and junctions.\u003c/p\u003e \u003cp\u003e \u003cb\u003eEnzymatic activity assays\u003c/b\u003e \u003c/p\u003e \u003cp\u003eEnzyme activity assays were conducted as previously reported (Korkmaz et al., 2008). The PR3 enzymatic activity was quantified by detecting the rate of hydrolysis of the PR3-specific substrate (Abz)-VADnorVADRQ-(EDDnp) by cell suspension. To analyze membrane-bound PR3 activity, neutrophils were cultured in cancer cell CM or nonconditioned medium and treated with DMSO or sivelestat (10 mM) for 30\u0026ndash;45 min at 37\u0026deg;C. Then, the cells were suspended in activity buffer (5\u0026times;10\u003csup\u003e6\u003c/sup\u003e cells/ml, PBS, 4 mM EGTA, pH 7.4) with 20 mM PR3-specific substrate, and the kinetics of hydrolysis were determined by measuring the fluorescence at lex\u0026thinsp;=\u0026thinsp;320 nm and lem\u0026thinsp;=\u0026thinsp;420 nm.\u003c/p\u003e \u003cp\u003e \u003cb\u003eGST-pull down\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe GST pull-down procedure was conducted as previously reported (Nature methods.,2004). Fifty microliters of glutathione-magbeads were remixed with 300 \u0026micro;g GST-PSMD6 recombinant protein or GST protein for 2 hours. Then, the plasma member proteins of neutrophils were added and incubated for 16 hours at 4\u0026deg;C. Subsequently, the magnetic beads were washed with lysis buffer, and the proteins bound to the magnetic beads were analyzed by western blotting, MS and Coomassie electrophoresis staining.\u003c/p\u003e \u003cp\u003e \u003cb\u003eImmunofluorescence (IF) staining\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFor murine tissue, tissues were perfused with 4% PFA for 24 h and washed once in 1\u0026times;PBS for 30 mins, followed by 30\u0026ndash;95% alcohol and n-butanol prior to being embedded in paraffin. Tissues were sectioned to 4 mm thickness, washed twice with PBS, permeabilized in 0.2% Triton X-100 for 15 min, and blocked in PBS containing 5% BSA for 45 min. CiH3 antibody (HUABIO, M1306-4) and MPO antibody (Proteintech Group, 22225-1-AP) were used for IF staining. For cancer cell IF staining, CLTA knockdown cells and control cells were seeded on coverslips coated with poly-L-lysine (WHB-24-CS-LC, WHB) in 24-well plates. After 24 h at 37 ℃, the cells were fixed with 4% PFA for 10 min at room temperature, washed three times with PBS and permeabilized in 0.1% Triton X-100 for 10 min. The cells were blocked in PBS containing 5% BSA for 30 min and then incubated with anti-CLTA (Proteintech Group, 10852-1-AP) and anti-PSMD6 (Santa Cruz, sc-393580) in blocking buffer overnight at 4\u0026deg;C. After three washes in PBS, the cells were incubated with fluorochrome-conjugated secondary antibodies (1:500, BIOESN) for 1 h and then counterstained with DAPI (ZSGB-BIO, ZLI-9557). Observation and photographing were performed with the confocal microscopy Cell Observer (Zeiss, Germany), and image processing and analysis were performed with Zen blue edition software (Zeiss, Germany). The Plugin-Colocalization Finder of ImageJ was used in the colocalization quantitative analysis. Pearson\u0026rsquo;s correlation coefficient and overlap coefficient according to Manders\u003csup\u003e96\u003c/sup\u003e were used to quantitatively evaluate the colocalization results.\u003c/p\u003e \u003cp\u003e \u003cb\u003eMouse experiments\u003c/b\u003e \u003c/p\u003e \u003cp\u003eFor the tail vein metastasis assay, 2 \u0026times; 10\u003csup\u003e6\u003c/sup\u003e human breast cancer cells (MDA-MB-231 PSMD6 knockdown cells, mock cells, MDA-MB-231-LM2 PSMD6 overexpression cells, and normal control cells) were injected into the tail vein of 6-week-old nude mice (n\u0026thinsp;=\u0026thinsp;4, 5 for each group). After 6 weeks, mice were killed by cervical dislocation, and the lungs were removed for fixation with 4% PFA. To observe neutrophil infiltration and NETs \u003cem\u003ein vivo\u003c/em\u003e, 1\u0026times;10\u003csup\u003e5\u003c/sup\u003e murine breast cancer cells (4T1-PSMD6 knockdown cells, mock cells, 4T1-PSMD6 overexpression cells, and normal control cells) were injected into the tail vein of 6-week-old BALB/c mice (n\u0026thinsp;=\u0026thinsp;4, 5 for each group). After 1\u0026ndash;2 weeks, the mice were killed by cervical dislocation, and lung or lung metastasis nodules were prepared as single-cell suspensions or fixed in 4% PFA.\u003c/p\u003e \u003cp\u003e \u003cb\u003eRNAi and cell transfection\u003c/b\u003e \u003c/p\u003e \u003cp\u003eLentivirus packaging was carried out by Shanghai Obio (China). For the knockdown of PSMD6 and CLTA, one validated hairpin (human \u003cem\u003ePSMD6\u003c/em\u003e target sequences: GAATGCCGTTACTCTGTTT, murine psmd6 target sequences: AGAGTTCTGTGTTTCTAAA), three validated hairpins (CLTA target sequences: GGAGCTAGAAGAATGGTAT, GAGCAGCTACAGAAAACAA, GAAGCAGAGTGGAAAGAAA) targeting the PSMD6 and CLTA transcripts were cloned and inserted into the pSLenti-U6-shRNA(PSMD6)-CMV-F2A-Puro-WPRE vector. For the overexpression of PSMD6, the full length of their transcripts was cloned and inserted into the CMV-MCS-3FLAG-SV40-puromycin vector.\u003c/p\u003e \u003cp\u003e \u003cb\u003eRNA extraction and quantitative real-time polymerase chain reaction (qRT‒PCR)\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTotal RNA was isolated from breast cancer cell lines with an RNA easy fast cell kit (TIANGEN, Beijing). Reverse transcription (RT) of complementary DNA (cDNA) was carried out by using the TIANGEN mRNA qRT‒PCR starter kit (TIANGEN, KR116-01). SYBR Green PCR Master Mix was used to amplify cDNA aliquots. GAPDH served as an endogenous control. The sequences of the sense and antisense primers were as follows:\u003c/p\u003e \u003cp\u003eIL-6-F: CCTCTCTCTAATCAGCCCTCTG,\u003c/p\u003e \u003cp\u003eIL-6-R: GAGGACCTGGGAGTAGATGAG,\u003c/p\u003e \u003cp\u003eMMP9-R: GGCAGGGACAGTTGCTTCT,\u003c/p\u003e \u003cp\u003eMMP9-F: TGTACCGCTATGGTTACACTCG,\u003c/p\u003e \u003cp\u003eVEGF-R: AGGGTCTCGATTGGATGGCA,\u003c/p\u003e \u003cp\u003eVEGF-F: AGGGCAGAATCATCACGAAGT,\u003c/p\u003e \u003cp\u003eIL-8-F: TTTTGCCAAGGAGTGCTAAAGA,\u003c/p\u003e \u003cp\u003eIL-8-R: AACCCTCTGCACCCAGTTTTC.\u003c/p\u003e \u003cp\u003e \u003cb\u003eNeutrophil isolation\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo isolate neutrophils from murine lung metastasis nodules, lung metastasis nodules from 8-week-old BALB/C mice were harvested. Preparation of single-cell suspensions from lung metastasis tissues was performed with a gentleMACS Dissociator, followed by neutrophil isolation with a mouse tumor infiltrating tissue neutrophil isolation kit (P2430, Solarbio). Human neutrophils were isolated from the peripheral blood of healthy female volunteers with a human peripheral blood neutrophil isolation kit (P9040, Solarbio). Neutrophils were cultured in RPMI 1640 medium containing 0.2% BSA.\u003c/p\u003e \u003cp\u003e \u003cb\u003eNET analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003eTo analyze NET formation, neutrophils (1\u0026times;10\u003csup\u003e5\u003c/sup\u003e cells) were seeded on coverslips coated with poly-L-lysine (WHB-24-CS-LC, WHB) in 24-well plates for 30 min before adding 50% cancer cell CM. After 6 h at 37\u0026deg;C, neutrophils were subjected to IF staining. Anti-histone H3 (1:200, M1306-4, HUABIAO) and anti-MPO (1:200, ET1703-21, HUABIAO) were used to stain the sections. NET area quantification was performed using a previously published method (Cardiovascular Research (2022) 118, 2179\u0026ndash;2195). Briefly, NET-positive area (%) = (the colocalized area of MPO fl and CitH3 fl/the area of the tissue in each microscopic field using the 20x objective) *100%.\u003c/p\u003e \u003cp\u003e \u003cb\u003eTwo-chamber migration assays\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe two-chamber migration assay procedure was previously described (Zhang et al. Molecular Cancer (2018) 17:146). Briefly, 1\u0026times;10\u003csup\u003e5\u003c/sup\u003e MDA-MB-231 cells in DMEM were added to the upper chamber (11965092, Gibco), and a 1:1 mixture of DMEM and cancer cell CM, or medium from neutrophils cultured in cancer cell CM, was added to the lower chamber as the chemoattractant. The migrated cells in the lower chamber were counted after 12 h.\u003c/p\u003e \u003cp\u003e \u003cb\u003eWestern blotting\u003c/b\u003e \u003c/p\u003e \u003cp\u003eWestern blotting was performed as described previously. In summary, total proteins were extracted from cells via RIPA buffer supplemented with protease and phosphatase inhibitors (Beyotime, Beijing), and aliquots of these proteins were separated by SDS/PAGE and visualized with Millipore Immobilon Western HRP substrate. The antibodies used in this assay included anti-PSMD6 (HUABIO, ER64500), anti-CLTA (Proteintech Group, 10852-1-AP), anti-a-tubulin (SAB,21581), anti-CD177 (SAB, 41689), anti-PRTN3 (Proteintech Group, 67030-1), anti-GPR78 (Elabscience, 40588), anti-Na\u003csup\u003e+\u003c/sup\u003e/K\u003csup\u003e+\u003c/sup\u003e ATPase (HUABIO, ET1609-76), and anti-GAPDH (Proteintech Group, 60004-1).\u003c/p\u003e \u003cp\u003e \u003cb\u003eDetection of circulating NETs\u003c/b\u003e \u003c/p\u003e \u003cp\u003edHL-60 cells were induced from HL-60 by 1 \u0026micro;M ATRA for 5 days. dHL-60 cells were cultured in culture medium from PSMD6-overexpressing cells or PSMD6-knockdown cells for 6 hours. The culture medium of dHL-60 cells was detected by Quant-iT\u0026trade; PicoGreen\u0026trade; dsDNA (Thermo).\u003c/p\u003e \u003cp\u003e \u003cb\u003eStatistical analyses\u003c/b\u003e \u003c/p\u003e \u003cp\u003eData analyses were performed using GraphPad Prism 9.0 (GraphPad Software, La Jolla, USA). The data presentation and statistical analyses are described in the figure legends. P values\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were considered statistically significant. The \u003cem\u003ein vitro\u003c/em\u003e experiments were repeated independently multiple times with similar results, as indicated in the figure legends.\u003c/p\u003e \u003cp\u003e \u003cb\u003eQuantification and statistical analysis\u003c/b\u003e \u003c/p\u003e \u003cp\u003eQuantification methods and statistical analysis methods for plasma proteomic analyses were described or referenced in the respective Methods Details subsections.\u003c/p\u003e \u003cp\u003eAdditionally, standard statistical tests were used to analyze the data, including but not limited to Student\u0026rsquo;s t test, rank sums test, ANOVA test, Kruskal‒Wallis test, Fisher\u0026rsquo;s exact test, and chi-square test. Statistical significance was considered when the p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05. All analyses of plasma proteomic data were performed in R, Python, and GraphPad Prism.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work is supported by the National Key R\u0026amp;D Program of China (2022YFA1303200, 2022YFA1303201); the National Natural Science Foundation of China (32330062, 31972933, 82003149); the Program of Shanghai Academic/Technology Research Leader (22XD1420100); the Major Project of Special Development Funds of Zhangjiang National Independent Innovation Demonstration Zone (ZJ2019-ZD-004); the Shanghai Municipal Science and Technology Major Project (2017SHZDZX01); and the Fudan Original Research Personalized Support Project.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of Guangdong Provincial People\u0026rsquo;s Hospital. This animal study was approved by the Institutional Animal Care and Use Committee of Guangdong Provincial People\u0026rsquo;s Hospital (KY2020-042-02).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSiegel, R.L., Miller, K.D., Fuchs, H.E., and Jemal, A. (2022). Cancer statistics, 2022. CA Cancer J Clin \u003cem\u003e72\u003c/em\u003e, 7\u0026ndash;33.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang, T., Bao, H., Meng, Y.H., Zhu, J.L., Chu, X.D., Chu, X.L., and Pan, J.H. (2022). Tumour budding is a novel marker in breast cancer: the clinical application and future prospects. Ann Med \u003cem\u003e54\u003c/em\u003e, 1303\u0026ndash;1312.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTower H, Ruppert M, Britt K. The Immune Microenvironment of Breast Cancer Progression. Cancers (Basel). 2019;11(9):1375.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBombonati, A., and Sgroi, D.C. (2011). The molecular pathology of breast cancer progression. J Pathol \u003cem\u003e223\u003c/em\u003e, 307\u0026ndash;317.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, L. (2017). Early Diagnosis of Breast Cancer. Sensors (Basel) 17. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/s17071572\u003c/span\u003e\u003cspan address=\"10.3390/s17071572\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZubor P, Kubatka P, Kajo K, Dankova Z, Polacek H, Bielik T, Kudela E, Samec M, Liskova A, Vlcakova D, Kulkovska T, Stastny I, Holubekova V, Bujnak J, Laucekova Z, B\u0026uuml;sselberg D, Adamek M, Kuhn W, Danko J, Golubnitschaja O. Why the Gold Standard Approach by Mammography Demands Extension by Multiomics? Application of Liquid Biopsy miRNA Profiles to Breast Cancer Disease Management. Int J Mol Sci. 2019;20(12):2878.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIli\u0026eacute; M, Hofman P. Pros: Can tissue biopsy be replaced by liquid biopsy? Transl Lung Cancer Res. 2016;5(4):420\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHu X, Li G, Wu S. Advances in Diagnosis and Therapy for Bladder Cancer. Cancers (Basel). 2022;14(13):3181.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKardynal A, Olszewska M. Modern non-invasive diagnostic techniques in the detection of early cutaneous melanoma. J Dermatol Case Rep. 2014;8(1):1\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiang J, Jiang Y, Abboud Y, Gaddam S. Role of Endoscopy in Management of Upper Gastrointestinal Cancers. Diseases. 2022;11(1):3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOeffinger, K. C., Fontham, E. T. H., Etzioni, R., Herzig, A., Michaelson, J. S., Shih, Y.-C. T., et al. (2015). Breast Cancer Screening for Women at Average Risk. Jama 314 (15), 1599\u0026ndash;1614.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBleyer, A., and Welch, H. G. (2012). Effect of Three Decades of Screening Mammography on Breast-Cancer Incidence. N. Engl. J. Med. 367 (21), 1998\u0026ndash;2005.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZanzonico PB. Benefits and Risks in Medical Imaging. Health Phys. 2019;116(2):135\u0026ndash;137.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRichardson, R.R. (2017). Imaging Modalities: Advantages and Disadvantages. In: Atlas of Acquired Cardiovascular Disease Imaging in Children. Springer, Cham. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/978-3-319-44115-3_1\u003c/span\u003e\u003cspan address=\"10.1007/978-3-319-44115-3_1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBleyer A, Welch HG. Effect of three decades of screening mammography on breast-cancer incidence. N Engl J Med. 2012;367(21):1998\u0026ndash;2005.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHussain S, Mubeen I, Ullah N, Shah SSUD, Khan BA, Zahoor M, Ullah R, Khan FA, Sultan MA. Modern Diagnostic Imaging Technique Applications and Risk Factors in the Medical Field: A Review. Biomed Res Int. 2022;2022:5164970.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCrowley, E., Di Nicolantonio, F., Loupakis, F. \u0026amp; Bardelli, A. Liquid biopsy: monitoring cancer-genetics in the blood. Nat Rev Clin Oncol 10, 472\u0026ndash;484 (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, Joshua J.X.; Tse, Gary M. (2020). Core needle biopsy diagnosis of fibroepithelial lesions of the breast: a diagnostic challenge. \u003cem\u003ePathology\u003c/em\u003e,2020.06.005.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eŁukasiewicz, S.; Czeczelewski, M.; Forma, A.; Baj, J.; Sitarz, R.; Stanisławek, A. Breast Cancer\u0026mdash;Epidemiology, Risk Factors, Classification, Prognostic Markers, and Current Treatment Strategies\u0026mdash;An Updated Review. \u003cem\u003eCancers\u003c/em\u003e 2021, 13, 4287.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed M. Kabel, Tumor markers of breast cancer: New prospectives, Journal of Oncological Sciences, Volume 3, Issue 1, 2017, Pages 5\u0026ndash;11, ISSN 2452\u0026ndash;3364.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eT\u0026oacute;thov\u0026aacute;, Csilla, Mihajlovičov\u0026aacute;, X\u0026eacute;nia, Nagy, Oskar. \"The Use of Serum Proteins in the Laboratory Diagnosis of Health Disorders in Ruminants\". Ruminants - The Husbandry, Economic and Health Aspects, edited by Muhammad Abubakar, IntechOpen, 2017. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5772/intechopen.72154\u003c/span\u003e\u003cspan address=\"10.5772/intechopen.72154\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang Z, Gu Y, Shi Z, Min L, Zhang Z, Zhou P, Luo R, Wang Y, Cui Y, Sun Y, Wang X. Multiplex immune profiling reveals the role of serum immune proteomics in predicting response to preoperative chemotherapy of gastric cancer. Cell Rep Med. 2023;4(2):100931.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi, J., Qi, G. \u0026amp; Liu, Y. Proteomics analysis of serum from thymoma patients. Sci Rep 13, 5117 (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFernandez-Garcia D, Hills A, Page K, Hastings RK, Toghill B, Goddard KS, et al. Plasma cell-free DNA (cfDNA) as a predictive and prognostic marker in patients with metastatic breast cancer. Breast Cancer Res. 2019;21:8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlix-Panabi\u0026egrave;res C, Pantel K. Clinical applications of circulating tumor cells and circulating tumor DNA as liquid biopsy. Cancer Discov. 2016;6(5):479\u0026ndash;91.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eN\u0026uacute;\u0026ntilde;ez C. Blood-based protein biomarkers in breast cancer. Clin Chim Acta. 2019;490:113\u0026ndash;27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIssaq HJ, Xiao Z, Veenstra TD. Serum and plasma proteomics. Chem Rev. 2007;107(8):3601\u0026ndash;20.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarbeck, N. \u0026amp; Gnant, M. Breast cancer. Lancet 389, 1134\u0026ndash;1150 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarrugo-Ram\u0026iacute;rez, J., Mir, M. \u0026amp; Samitier, J. Blood-Based Cancer Biomarkers in Liquid Biopsy: A Promising Non-Invasive Alternative to Tissue Biopsy. Int J Mol Sci 19 (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarris L, Fritsche H, Mennel R, Norton L, Ravdin P, Taube S, et al. American Society of Clinical Oncology 2007 update of recommendations for the use of tumor markers in breast cancer. J Clin Oncol. 2007;25(33):5287\u0026ndash;312.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrandi J, Manfredi M, Speziali G, Gosetti F, Marengo E, Cecconi D. Proteomic approaches to decipher cancer cell secretome. Semin Cell Develop Biol. 2018;78:93\u0026ndash;101.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePavlou MP, Diamandis EP. The cancer cell secretome: A good source for discovering biomarkers? J Proteom. 2010;73(10):1896\u0026ndash;906.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eD. E. Misek and E. H. Kim. Protein Biomarkers for the Early Detection of Breast Cancer International Journal of Proteomics 2011 Vol. 2011 Pages 343582.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHammarstr\u0026ouml;m, S. The carcinoembryonic antigen (CEA) family: structures, suggested functions and expression in normal and malignant tissues. Semin Cancer Biol 9, 67\u0026ndash;81 (1999).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHao, C., Zhang, G., and Zhang, L. (2019). Serum CEA levels in 49 different types of cancer and noncancer diseases. Prog Mol Biol Transl Sci \u003cem\u003e162\u003c/em\u003e, 213\u0026ndash;227. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/bs.pmbts.2018.12.011\u003c/span\u003e\u003cspan address=\"10.1016/bs.pmbts.2018.12.011\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFelder, M. et al. MUC16 (CA125): tumor biomarker to cancer therapy, a work in progress. Mol Cancer 13, 129 (2014).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDochez, V. et al. Biomarkers and algorithms for diagnosis of ovarian cancer: CA125, HE4, RMI and ROMA, a review. J Ovarian Res 12, 28 (2019).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManuali, E., De Giuseppe, A., Feliziani, F. et al. CA 15\u0026ndash;3 cell lines and tissue expression in canine mammary cancer and the correlation between serum levels and tumour histological grade. BMC Vet Res 8, 86 (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTempero, M.A. et al. Relationship of carbohydrate antigen 19 \u0026ndash; 9 and Lewis antigens in pancreatic cancer. Cancer Res 47, 5501\u0026ndash;5503 (1987).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe, M., Wu, C., Xu, J., Guo, H., Yang, H., Zhang, X., Sun, J., Yu, D., Zhou, L., Peng, T., et al. (2014). A genome wide association study of genetic loci that influence tumour biomarkers cancer antigen 19 \u0026ndash; 9, carcinoembryonic antigen and α fetoprotein and their associations with cancer risk. Gut \u003cem\u003e63\u003c/em\u003e, 143\u0026ndash;151. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1136/gutjnl-2012-303434\u003c/span\u003e\u003cspan address=\"10.1136/gutjnl-2012-303434\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKim, D.Y., Toan, B.N., Tan, C.K., Hasan, I., Setiawan, L., Yu, M.L., Izumi, N., Huyen, N.N., Chow, P.K., Mohamed, R., et al. (2023). Utility of combining PIVKA-II and AFP in the surveillance and monitoring of hepatocellular carcinoma in the Asia-Pacific region. Clin Mol Hepatol \u003cem\u003e29\u003c/em\u003e, 277\u0026ndash;292. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3350/cmh.2022.0212\u003c/span\u003e\u003cspan address=\"10.3350/cmh.2022.0212\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTaketa, K. Alpha-fetoprotein: reevaluation in hepatology. Hepatology 12, 1420\u0026ndash;1432 (1990).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A, Bray F. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021;71(3):209\u0026ndash;249.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanahan D, Weinberg RA. Hallmarks of Cancer: The Next Generation. Cell. 2011;144(5):646\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGonzalez H, Hagerling C, Werb Z. Roles of the immune system in cancer: from tumor initiation to metastatic progression. Genes Dev. 2018;32(19\u0026ndash;20):1267\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJoyce JA, Pollard JW. Microenvironmental regulation of metastasis. Nat Rev Cancer. 2009;9(4):239\u0026ndash;52.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobinson JL, Feizi A, Uhl\u0026eacute;n M, Nielsen J. A systematic investigation of the malignant functions and diagnostic potential of the cancer secretome. Cell Rep. 2019;26(10):2622\u0026ndash;35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eValastyan S, Weinberg RA. Tumor metastasis: molecular insights and evolving paradigms. Cell. 2011;147(2):275\u0026ndash;92.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGiancotti FG. Mechanisms governing metastatic dormancy and reactivation. Cell. 2013;155(4):750\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRedig AJ, McAllister SS. Breast cancer as a systemic disease: a view of metastasis. J Intern Med. 2013;274(2):113\u0026ndash;26.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGhajar CM. Metastasis prevention by targeting the dormant niche. Nat Rev Cancer. 2015;15(4):238\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBenoy IH, Salgado R, Dam PV, Geboers K, Marck EV, Scharp\u0026eacute; S, et al. Increased serum interleukin-8 in patients with early and metastatic breast cancer correlates with early dissemination and survival. Clin Cancer Res. 2004;10(21):7157\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoman AS, Uddin M, Chowdhury AA, Nayeem MJ, Raihan Z, Rashid MI, et al. Serum sonic hedgehog (SHH) and interleukin-(IL-6) as dual prognostic biomarkers in progressive metastatic breast cancer. Sci Rep. 2017;7:89.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEndo M, Yamamoto Y, Nakano M, Masuda T, Odagiri H, Horiguchi H, et al. Serum ANGPTL2 levels reflect clinical features of breast cancer patients: implications for the pathogenesis of breast cancer metastasis. Int J Biol Markers. 2014;29(3):239\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGuo, T., Kouvonen, P., Koh, C.C., Gillet, L.C., Wolski, W.E., Rost, H.L., Rosenberger, G., Collins, B.C., Blum, L.C., Gillessen, S., et al. (2015). Rapid mass spectrometric conversion of tissue biopsy samples into permanent quantitative digital proteome maps. Nat Med 21, 407\u0026ndash;413. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/nm.3807\u003c/span\u003e\u003cspan address=\"10.1038/nm.3807\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBittremieux, W. et al. Quality control in mass spectrometry-based proteomics. Mass Spectrom Rev 37, 697\u0026ndash;711, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/mas.21544\u003c/span\u003e\u003cspan address=\"10.1002/mas.21544\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJiang, Y., Sun, A.H., Zhao, Y., Ying, W.T., Sun, H.C., Yang, X.R., Xing, B.C., Sun, W., Ren, L.L., Hu, B., et al. (2019). Proteomics identifies new therapeutic targets of early- stage hepatocellular carcinoma. Nature 567, 257-+. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-019-0987-8\u003c/span\u003e\u003cspan address=\"10.1038/s41586-019-0987-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu, J.Y., Zhang, C.C., Wang, X., Zhai, L.H., Ma, Y.M., Mao, Y.S., Qian, K., Sun, C.Q., Liu, Z.W., Jiang, S.W., et al. (2020). Integrative Proteomic Characterization of Human Lung Adenocarcinoma. Cell 182, 245-+. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.cell.2020.05.043\u003c/span\u003e\u003cspan address=\"10.1016/j.cell.2020.05.043\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchwanh\u0026auml;usser, B. et al. Global quantification of mammalian gene expression control. Nature 473, 337\u0026ndash;342 (2011).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang, W., Zhou, M., Dorsey, T.H., Prieto, D.A., Wang, X.W., Ruppin, E., Veenstra, T.D., and Ambs, S. (2018). Integrated proteotranscriptomics of breast cancer reveals globally increased protein-mRNA concordance associated with subtypes and survival. Genome Med \u003cem\u003e10\u003c/em\u003e, 94. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13073-018-0602-x\u003c/span\u003e\u003cspan address=\"10.1186/s13073-018-0602-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaillon S, Ponzetta A, Di Mitri D, Santoni A, Bonecchi R, Mantovani A. (2020). Neutrophil diversity and plasticity in tumour progression and therapy. Nat Rev Cancer. \u003cem\u003e20\u003c/em\u003e, 485\u0026ndash;503. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41568-020-0281-y\u003c/span\u003e\u003cspan address=\"10.1038/s41568-020-0281-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheung, P., Schaffert, S., Chang, S.E., Dvorak, M., Donato, M., Macaubas, C., Foecke, M.H., Li, T.M., Zhang, L., Coan, J.P., et al. (2021). Repression of CTSG, ELANE and PRTN3-mediated histone H3 proteolytic cleavage promotes monocyte-to-macrophage differentiation. Nat Immunol 22, 711\u0026ndash;722. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41590-021-00928-y\u003c/span\u003e\u003cspan address=\"10.1038/s41590-021-00928-y\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarino, S.F., Jerke, U., Rolle, S., Daumke, O., and Kettritz, R. (2022). Competitively disrupting the neutrophil-specific receptor-autoantigen CD177:proteinase 3 membrane complex reduces anti-PR3 antibody-induced neutrophil activation. J Biol Chem 298, 101598. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.jbc.2022.101598\u003c/span\u003e\u003cspan address=\"10.1016/j.jbc.2022.101598\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXiao, Y., Cong, M., Li, J., He, D., Wu, Q., Tian, P., et al. (2021). Cathepsin C promotes breast cancer lung metastasis by modulating neutrophil infiltration and neutrophil extracellular trap formation. Cancer Cell. 39,(3):423\u0026ndash;437.e7. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.ccell.2020.12.012\u003c/span\u003e\u003cspan address=\"10.1016/j.ccell.2020.12.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBorrebaeck CA. Precision diagnostics: moving towards protein biomarker signatures of clinical utility in cancer. Nat Rev Cancer. 2017;17(3):199\u0026ndash;204.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiernat M, Trzyna M, Byszek A, et al. Liquid crystal foil for the detection of breast cancer[C]//Photonics Applications in Astronomy, Communications, Industry, and High-Energy Physics Experiments 2016. SPIE, 2016, 10031: 873\u0026ndash;882.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang JY, Shi KZ, Liao XY, Li SJ, Bao D, Qian Y, Li DJ. The Silence of PSMC6 Inhibits Cell Growth and Metastasis in Lung Adenocarcinoma. Biomed Res Int. 2021;2021:9922185.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAydoĞan T\u0026Uuml;rkoĞlu S, Dayi G, K\u0026Ouml;\u0026Ccedil;kar F. Upregulation of PSMD4 gene by hypoxia in prostate cancer cells. Turk J Biol. 2020;44(5):275\u0026ndash;283.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOkumura T, Ikeda K, Ujihira T, Okamoto K, Horie-Inoue K, Takeda S, Inoue S. Proteasome 26S subunit PSMD1 regulates breast cancer cell growth through p53 protein degradation. J Biochem. 2018;163(1):19\u0026ndash;29.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTerceiro LEL, Edechi CA, Ikeogu NM, et al. The Breast Tumor Microenvironment: A Key Player in Metastatic Spread. Cancers. 2021;13(19):4798.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonteran L, Erez N. The Dark Side of Fibroblasts: Cancer-Associated Fibroblasts as Mediators of Immunosuppression in the Tumor Microenvironment. Front Immunol. 2019;10:1835.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanerjee, K.; Resat, H. Constitutive activation of STAT3 in breast cancer cells: A review. Int. J. Cancer 2015, 138, 2570\u0026ndash;2578\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBizari, L., Silva, A.E., \u0026amp; Tajara, E.H. (2006). Gene amplification in carcinogenesis. Genetics and Molecular Biology, 29, 1\u0026ndash;7.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOhta JI, Miyoshi Y, Uemura H, Fujinami K, Mikata K, Hosaka M, Tokita Y and Kubota Y (2001) Fluorescence in situ hybridization evaluation of c-erbB-2 gene amplification and chromosomal anomalies in bladder cancer. Clin Cancer Res7:2463\u0026ndash;2467.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTakehana T, Kunitomo K, Kono K, Kitahara F, Iizuka H, Matsumoto Y and Fujino MA (2002) Status of c-erbB-2 in gastric adenocarcinoma: A comparative study of immunohistoche mistry, fluorescence in situ hybridization and enzymelinked immuno-sorbent assay. Int J Cancer 98:833\u0026ndash;837.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLear-Kaul KC, Yoon HR, Kleinschmidt-DeMasters BK, Mc Gavran L and Singh M (2003) Her-2/neu status in breast cancer metastases to the central nervous system. Arch Pathol Lab Med 127:1451\u0026ndash;1457\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCianciulli AM, Guadagni F, Benevolo M, Merola R, Giannarelli D, Marandino F, Vocaturo G, Mariani L and Mottelese M (2003) HER-2/neu oncogene amplification and chromosome 17 aneusomy in endometrial carcinoma: correlation with oncoprotein expression and conventional pathol\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHirsch FR, Scagliotti GV, Langer CJ, Varella-Garcia M and Franklin WA (2003) Epidermal growth factor family of receptors in preneoplasia and lung cancer: Perspectives for target therapies. Lung Cancer 41:29\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNessling M, Solinas-Toldo S, Wilgenbus KK, Borchard F and Lichter P (1998) Mapping of chromosomal imbalances in gastric adenocarcinoma revealed amplified protooncogenes MYCN, MET, WNT2, and ERBB2. Genes Chromosomes Cancer 23:307\u0026ndash;316.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVidgren V, Varis A, Kokkola A, Monni O, Puolakkainen P, Nordling S, Forozan F, Kallioniemi A, Vakkari ML, Kivilaakso E and Knuutila S (1999) Concomitant gastrin and ERBB2 gene amplifications at 17q12-q21 in the intestinal type of gastric cancer. Genes Chromosomes Cancer 24:24\u0026ndash;29.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaris A, Wolf M, Monni O, Vakkari ML, Kokkola A, Moskaluk C, Frierson H Jr, Powell SM, Knuutila S, Kallioniemi A and El-Rifai W (2002) Targets of gene amplification and overexpression at 17q in gastric cancer. Cancer Res 62:2625\u0026ndash;2629.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBecker KF, Keller G and Hoefler H (2000) The use of molecular biology in diagnosis and prognosis of gastric cancer. Surg Oncol 9:5\u0026ndash;11.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng, J. et al. Firmiana: towards a one-stop proteomic cloud platform for data processing and analysis. Nat Biotechnol 35, 409\u0026ndash;412 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong, A.T., Leprevost, F.V., Avtonomov, D.M., Mellacheruvu, D. \u0026amp; Nesvizhskii, A.I. MSFragger: ultrafast and comprehensive peptide identification in mass spectrometry-based proteomics. Nat Methods 14, 513\u0026ndash;520 (2017).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDemichev, V., Messner, C.B., Vernardis, S.I., Lilley, K.S. \u0026amp; Ralser, M. DIA-NN: neural networks and interference correction enable deep proteome coverage in high throughput. Nat Methods 17, 41\u0026ndash;44 (2020).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, W. et al. LFQuant: a label-free fast quantitative analysis tool for high-resolution LC-MS/MS proteomics data. Proteomics 12, 3475\u0026ndash;3484, doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/pmic.201200017\u003c/span\u003e\u003cspan address=\"10.1002/pmic.201200017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2012).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHastie, T., Tibshirani, R., Sherlock, G., Eisen, M., Brown, P. and Botstein, D., Imputing Missing Data for Gene Expression Arrays, Stanford University Statistics Department Technical report (1999), \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www-stat.stanford.edu/~hastie/Papers/missing.pdf\u003c/span\u003e\u003cspan address=\"http://www-stat.stanford.edu/~hastie/Papers/missing.pdf\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Olga Troyanskaya, Michael Cantor, Gavin Sherlock, Pat Brown, Trevor Hastie, Robert Tibshirani, David Botstein and Russ B. Altman, Missing value estimation methods for DNA microarrays BIOINFORMATICS Vol.\u0026nbsp;17 no. 6, 2001 Pages 520\u0026ndash;525\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohnson, W.E., Li, C. \u0026amp; Rabinovic, A. Adjusting batch effects in microarray expression data using empirical Bayes methods. Biostatistics (Oxford, England) 8, 118\u0026ndash;127 (2007).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eH\u0026auml;nzelmann, S., Castelo, R. \u0026amp; Guinney, J. GSVA: gene set variation analysis for microarray and RNA-seq data. BMC Bioinformatics 14, 7 (2013)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRuspini, E.H.; Bezdek, J.C.; Keller, J.M. Fuzzy Clustering: A Historical Perspective. IEEE Comput. Intell. Mag.2019, 14, 45\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangfelder, P. \u0026amp; Horvath, S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 9, 559 (2008).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNewman AM, Steen CB, Liu CL, Gentles AJ, Chaudhuri AA, Scherer F, Khodadoust MS, Esfahani MS, Luca BA, Steiner D, Diehn M, Alizadeh AA. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol. 2019;37(7):773\u0026ndash;782. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41587-019-0114-2\u003c/span\u003e\u003cspan address=\"10.1038/s41587-019-0114-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2019 May 6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen YM, Zheng Y, Yu Y, Wang Y, Huang Q, Qian F, Sun L, Song ZG, Chen Z, Feng J, An Y, Yang J, Su Z, Sun S, Dai F, Chen Q, Lu Q, Li P, Ling Y, Yang Z, Tang H, Shi L, Jin L, Holmes EC, Ding C, Zhu TY, Zhang YZ. Blood molecular markers associated with COVID-19 immunopathology and multi-organ damage. EMBO J. 2020;39(24):e105896. doi: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.15252/embj.2020105896\u003c/span\u003e\u003cspan address=\"10.15252/embj.2020105896\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Epub 2020 Dec 14. PMID: 33140861; PMCID: PMC7737620.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, L., Meng, Y., Xu, J.J., and Zhang, Q.Y. (2018). The Transcription Factor AP4 Promotes Oncogenic Phenotypes and Cisplatin Resistance by Regulating LAPTM4B Expression. Mol Cancer Res \u003cem\u003e16\u003c/em\u003e, 857\u0026ndash;868. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1158/1541\u0026ndash;7786.Mcr-17-0519\u003c/span\u003e\u003cspan address=\"10.1158/1541\u0026ndash;7786.Mcr-17-0519\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu, J., Ke, P., Guo, H., Liu, J., Liu, Y., Tian, X., Huang, Z., Xu, X., Xu, D., Ma, Y., Wang, X., and Xiao, F. (2023). KCTD13-mediated ubiquitination and degradation of GluN1 regulates excitatory synaptic transmission and seizure susceptibility. Cell Death Differ \u003cem\u003e30\u003c/em\u003e, 1726\u0026ndash;1741. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41418-023-01174-5\u003c/span\u003e\u003cspan address=\"10.1038/s41418-023-01174-5\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eManders, E.M.M., Verbeek, F.J., and Aten, J.A. (1993). Measurement of co-localization of objects in dual-colour confocal images. J Microsc \u003cem\u003e169\u003c/em\u003e, 375\u0026ndash;382. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/j.1365-2818.1993.tb03313.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1365-2818.1993.tb03313.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Breast cancer, Serum proteome, Biomarkers, Metastasis, PSMD6","lastPublishedDoi":"10.21203/rs.3.rs-3634466/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3634466/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eBreast cancer (BC) has the highest mortality rate and prevalence among cancers in females worldwide. Here, we performed proteomic profiling of 322 serum samples from the discovery cohort [56 healthy controls (HCs), 112 benign breast tumor (BBT) patients, and 154 BC patients] and a prospective validation cohort [27 HCs, 29 BBT patients and 57 BC patients]. Integrated proteomic analysis of tissue and serum samples revealed highly specific tumor biomarkers and demonstrated that the serum proteome can distinguish the different pathological substages in BC progression. We also identified PSMD6 as a potential metastatic breast cancer (MBC) biomarker. Comprehensive analysis of the multicenter independent validation cohort, which included retrospective and prospective cohorts including 61 HCs, 72 BBT patients, and 247 BC patients, indicated that PSMD6 overexpression was an important cause of BC metastasis and an indicator of poor prognosis. Further study revealed that the CLTA-PSMD6-neutrophil axis promotes the transition from invasive ductal carcinoma (IDC) to MBC. Importantly, \u003cem\u003eCLTA\u003c/em\u003e amplification might be a potential therapeutic target for MBC patients. We also developed a highly accurate predictive model (accuracy\u0026thinsp;=\u0026thinsp;0.87) to differentiate benign and malignant tumors and validated its good performance in the prospective validation cohort. Collectively, this study demonstrates the elaborate BC serum proteomic landscape and provides valuable information regarding serum biomarkers, which could reveal novel therapeutic targets and provide opportunities for MBC treatment.\u003c/p\u003e","manuscriptTitle":"Proteome profiling of serum reveals PSMD6 as a biomarker in breast cancer metastasis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2023-12-11 14:59:11","doi":"10.21203/rs.3.rs-3634466/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"787365c2-69be-4acc-8ec3-71410b678a81","owner":[],"postedDate":"December 11th, 2023","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":27200727,"name":"Biological sciences/Cancer/Breast cancer"},{"id":27200728,"name":"Biological sciences/Biochemistry/Proteomics/Protein\u0026#x2013;protein interaction networks"}],"tags":[],"updatedAt":"2026-03-03T14:31:46+00:00","versionOfRecord":[],"versionCreatedAt":"2023-12-11 14:59:11","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-3634466","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3634466","identity":"rs-3634466","version":["v1"]},"buildId":"WrCJVZZCHTDjtuVLN7oU0","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}