SerpinB2 deficiency is associated with delayed mammary tumor development and decreased pro-tumorigenic macrophage polarization

The in vivo functions of SerpinB2 in tumor cells and tumor-associated macrophages (TAMs) during breast cancer development and metastasis remain elusive. SerpinB2-deficient MMTV-PyMT mice (PyMTSB2−/−) were previously produced to explore the biological roles of SerpinB2 in breast cancer. Compared with MMTV-PyMT wild-type (PyMTWT) mice, PyMTSB2−/− mice showed delayed tumor progression and reduced CK8 + tumor cell dissemination to lymph nodes. RNA-Seq data revealed significantly enriched genes associated with inflammatory responses, especially upregulated M1 and downregulated M2 macrophage marker genes in PyMTSB2−/− tumors. Decreased CD206+M2 and increased NOS2+M1 markers were detected in the primary tumors and metastatic lymph nodes of PyMTSB2−/− mice. In an in vitro study, SerpinB2 knockdown decreased the sphere formation and migration of MDA-MB-231 cells and suppressed protumorigenic M2 polarization of RAW264.7 cells. The combination of low SerpinB2, high NOS2, and low CD206 expression was favorable for survival in patients with breast cancer, as assessed in the BreastMark dataset. Our study demonstrates that SerpinB2 deficiency delays mammary tumor development and metastasis in PyMTWT mice, along with reduced sphere formation and migration abilities of tumor cells and decreased macrophage protumorigenic polarization. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-024-12473-6.


Introduction
SerpinB2, a plasminogen activator inhibitor-2 (PAI-2), is a serine protease inhibitor of the serpin superfamily that inactivates tissue plasminogen activator (tPA) and urokinase plasminogen activator (uPA) [1].The role of SerpinB2 in breast cancer growth and metastasis is complex and controversial because SerpinB2 may have both pro-tumor and anti-tumor effects [2].High expression of tumor cell-associated SerpinB2 facilitates breast cancer cell survival and metastasis by protecting breast cancer cells from death signals, promoting breast cancer cell migration ability, and enhancing macrophage recruitment into tumor tissues [3,4].In contrast, Ser-pinB2 expression by breast cancer cells significantly inhibits metastasis by inhibiting extracellular uPA [5,6].There is a significant association between SerpinB2 level and survival; breast cancer cell-associated SerpinB2 is identified as an unfavorable prognostic indicator [3,4].Nevertheless, the precise function of SerpinB2 in vivo regarding breast cancer progression and metastasis remains uncertain.
Among the immune cells recruited to the breast tumor, macrophages are one of the most abundant immune cell types shown to play a role as a facilitator for breast tumor development and progression at all stages [7,8].In macrophages, SerpinB2 is strongly upregulated by many inflammatory and/or stress-related factors and is involved in macrophage immune function regulation [9,10].SerpinB2 has also been implicated in the tumor suppression and promotion functions of TAMs [5].TAMs undergoing different classic (M1) or alternative (M2) phenotypic polarization express the specific M1 marker nitric oxide synthase-2 (NOS2) or the M2 markers CD206 which are believed to hold tumor-preventing or -promoting activities [11].An increase in SerpinB2 protein levels is observed in both M1 and M2 macrophages; notably, SerpinB2 protein levels are higher in M2 than in M1 macrophages [12].SerpinB2 has been reported to belong to the pro-tumorigenic M2-associated gene family [9,[13][14][15].The role of SerpinB2 in controlling TAM immune responses in mammary cancer development and progression has not been well studied.
We previously observed that SerpinB2 deficiency changes the expression of numerous genes, specifically those involved in the immunological and inflammatory response found in mammary tumors of PyMT SB2−/− mice [16].In this study, we explored different TAM polarizations as well as mammary cancer progression and metastasis in PyMT SB2−/− mice.Our study indicated a substantial postponement in the development of primary tumors and the metastasis to lymph nodes (LN) in PyMT SB2−/− mice, accompanied by decreased pro-tumorigenic polarization of TAMs that exhibited low CD206 and high NOS2 expression.Additionally, an in vitro study revealed that knocking down SerpinB2 in MDA-MB-231 breast cancer cells and RAW264.7 macrophages decreased breast cancer cell migration and sphere formation and the pro-tumorigenic polarization of macrophages.We verified the clinical significance of the combination of low SerpinB2, high NOS2, and low CD206 expression as predictive indicators for good patient survival using the online web-based service BreastMark.

Animals
The MMTV-PyMT mice on the C57BL/6 background, initially created by Muller's research group [17], were generously supplied by Dr. Sandra Gendler (Mayo Clinic in Scottsdale, AZ).The SerpinB2-deficient (B6.129S1-Serpinb2tm1Dgi/J) (SB2 − / −) mice in the C57BL/6 background were obtained from the Jackson Laboratory (Bar Harbor, ME) for this study.PyMT SB2−/− mice were generated by mating male PyMT WT mice with female SB2 − / − mice, both in the C57BL/6 background.The genotyping of PyMT WT and PyMT SB2−/− mice was conducted using methods previously described [16].Zoletil (5 mg/kg of mouse body weight) was administrated intramuscularly to excise mammary tumor tissues and axillary lymph nodes from 16-25-week-old PyMT WT and PyMT SB2−/− mice.Subsequently, the mice were transferred to a CO 2 chamber for euthanasia, where CO 2 was gradually introduced to fill 30-70% of the chamber volume per minute.The mice were continuously exposed to the flowing CO 2 for > 1 min following the cessation of breathing to ensure death.The Biomedical Center for Animal Resource Development of Seoul National University (SNU) provided animal care for all mice.All animal care and experimental procedures were conducted according to the National Research Council's guideline approved by the SNU Institutional Animal Care and Use Committee (SNU-150210-3-4).The experimental methods described in this study follow the ARRIVE guidelines (available at https:// arriv eguid elines.org) for reporting.

RNA-sequencing and differential gene functional annotation
For the study of RNA-sequencing (RNA-Seq), tumor tissues were obtained from 8 female mice of PyMT WT and PyMT SB2−/− groups (20 weeks old [n = 1], 22 weeks old [n = 2], and 25 weeks old [n = 5]), because tumor tissues obtained from mice at 16 weeks of age were found to have limited sizes.To visualize the results of differential gene expression (DEGs) between PyMT WT and PyMT SB2−/− tumors, a volcano plot and heat map were created using EdgeR within R (R development Core Team, 2016) through Bioconductor [18].Genes with a p-value below 0.05 and an absolute log2-fold change (FC) exceeding 1.5 were deemed significant.Furthermore, significant Gene Ontology (GO) terms associated with the DEGs were identified.The Benjamini-Hochberg method was employed for multiple testing adjustments, with a threshold of Benjamini < 0.05 considered statistically significant.The resulting GO term lists, specifically focusing on the biological process (BP) category, were compared and matched with each other.

Quantitative real-time RT-PCR
To verify the expression levels of genes identified from RNA-Seq, quantitative real-time RT-PCR (qRT-PCR) was performed using the same samples (n = 8) which used for RNA-Seq and specific primer sets for (Table S1).Lipopolysaccharides (LPS, 100 ng/ml, Sigma-Aldrich, St. Louis, MO, USA) was used to induce the upregulation of SerpinB2 in RAW264.7 cells as a positive control.The results were analyzed using the 2 -∆∆Ct method.

Western blot
For western blots, the protein lysates were extracted from the tumor tissues of PyMT WT and PyMT SB2−/− female mice (25-week-old).The cells and tumor tissues were lysed in RIPA buffer (Sigma-Adrich).The intensity of each band was quantified by ImageJ software (NIH, Bethesda, MD, USA).Relative protein levels were expressed after normalizing their intensities with the intensity of β-actin.The Western blot's raw data is presented in Supplementary Data 1.

Cytokine analysis
Sera of female mice (25-week-old mice (n = 7) per group) were isolated from blood samples for cytokine measurement.The levels of cytokines in the sera were measured using the Bio-Plex200 multiplex array system according to the recommended protocol (Bio-Rad, Hercules, CA, USA).

Sphere formation assay
To facilitate the reformation of spheres, MDA-MB-231 single cells were suspended in serum-free DMEM/F12 medium.The medium was supplemented with antibiotic-antimycotic solution (Invitrogen), B27 (Invitrogen), 10 ng/ml leukemia inhibitory factor (Millipore Ltd., Darmstadt, Germany), 20 ng/ml epidermal growth factor (Invitrogen), and basic fibroblast growth factor (Millipore).The cells were cultured in low attachment plates for 6 days, with medium replenishment every 3 days.

Proliferation and migration assays
In co-culture with RAW264.7 cells, the proliferation activity of MDA-MB-231 cells was assessed using transwell chambers with a 4-μm pore size insert (Costar, Cambridge, MA, USA).Briefly, RAW264.7 cells in the upper trans-well chamber and MDA-MB-231 cells in the bottom of 24 well plate were seeded and cultured for 24-48 h.The MDA-MB-231 cell migration assay was assessed in trans-well chambers with an 8-μm pore size insert (Costar).For MDA-MB-231 cell migration in co-culture with RAW264.7 cells, MDA-MB-231 cells and RAW264.7 cells were seeded in the upper transwell chamber and the bottom of 24 well plate, respectively, and cultured for 24-48 h.

Immunohistochemistry and analysis of LN metastasis
Mammary glands and tumors excised from 16-week-old female mice (n = 5 per group) were fixed with 4% buffered formalin and embedded in paraffin blocks.Next, 4-μm-thick sections were deparaffinized in xylene, rehydrated in a series of graded ethanol and water solutions, and pretreated at 98 °C for 20 min in citrate buffer (pH 6.0) for antigen retrieval.After blocking of endogenous peroxidase activity by peroxidase inhibition buffer and nonspecific binding of immunological reagents by blocking solution (Dako Agilent, CA, USA), each primary antibody was incubated at 4 °C overnight, followed by incubation of HRP-conjugated secondary antibodies, and immunoreaction was visualized using the 3,3′-diaminobenzidine (DAB) chromogen kit (Agilent Technologies, Glostrup, Denmark).Nuclei were counterstained with hematoxylin solution according to the manufacturer's instructions.Hematoxylin and eosin staining were performed.
For analysis of lymph node metastasis in the mammary tumor tissues of PyMT WT and PyMT SB2−/− mice, cancer cells in the lymph nodes were immunostained with an CK8 antibody (Developmental Studies Hybridoma Bank).For the quantification of CK8-immunostained cells, one section per lymph node (n = 5) was chosen, and immunostained cells in five fields at 40 × magnification within each section were quantified as the percentage of brownstained area in each microphotograph using Leica QWin image analysis and image processing software (Leica Imaging Solutions, Cambridge, UK).

Immunofluorescence staining
For double immunofluorescence staining, additional 4-μm slides of breast cancer tissue from 16-week-old mice were utilized.Primary antibodies, either F4/80 with CD206 or F4/80 with NOS2, were mixed and diluted in antibody diluent (Dako Agilent), and allowed to incubate overnight at 4 °C.The primary antibodies were then visualized with secondary antibodies conjugated to different fluorophores (Alexa Fluor 488 or 594) (Thermo Fisher Scientific) at room temperature for 1 h.Tissue slides were subsequently stained with the 4' ,6-diamidino-2-phenylindole dihydrochloride (DAPI) for nuclear counterstaining and covered using Prolong antifade mounting medium (Solarbio, Beijing, China).Immunofluorescence images were captured using an Olympus BX63 fluorescence microscope (Shinjuku-ku, Tokyo, Japan), using identical exposure and gain settings.Mean fluorescence intensity (MFI) of double-stained cells in each tumor section (n = 3) was calculated using ImageJ software.

Analysis of the BreastMark dataset
The prognostic value of the putative genes in patients with breast cancer was determined using the publicly available online tool BreastMark; disease-free survival (DFS) was analyzed by the combination of NOS2, Ser-pinB2, and CD206 expression, and the median was used to dichotomize the data [19].N tells us the number of samples used in this comparison.The Hazard ratio (HR) is generated using Cox regression, and a logrank test is used to assign significance to the HR.

Statistical analysis
All data were expressed as the means ± standard errors (S.E).Statistical significance was determined by unpaired t-test, one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test or a two-tailed Mann-Whitney U-test.Statistical significance was defined as *P < 0.05, ** P < 0.01, *** P < 0.001.GraphPad Prism v9.2.0 (GraphPad Software Inc., La Jolla, CA, USA) was utilized for all statistical analyses.

Results
SerpinB2 expression is not detected in tumors of PyMT SB2−/− mice Our previous studies on mammary tumors in agematched PyMT WT and PyMT SB2-/-mice demonstrated that SerpinB2 loss delayed tumor onset and significantly reduced tumor incidence rate and volume in the 4th-5th glands in PyMT SB2−/− mice compared to PyMT WT mice [16].The patterns and expression levels of SerpinB2 in the mammary tumors and stroma of PyMT WT and PyMT SB2−/− mice were analyzed.The abundant expression of SerpinB2 was observed in epidermal keratinocytes of PyMT WT mouse skin, but the tumor cells did not exhibit strong SerpinB2 staining, showing that tumor cells in PyMT WT -induced mammary tumors express very low levels of SerpinB2 (Fig. 1A).Intriguingly, stromal cells localized at the peritumoral site of PyMT WT tumors displayed strong SerpinB2-positive staining (Fig. 1A).As expected, significant staining for SerpinB2 was absent in tumors of PyMT SB2−/− mice (Fig. 1A).LPS-treated RAW264.7 cells exhibited high SerpinB2 protein and mRNA levels (Fig. 1B and C).In representative western blot images with quantitative analysis, the SerpinB2 protein levels were low in the tumors of PyMT WT mice (0.41 ± 0.28) but were not detected in the PyMT SB2−/− mice (Fig. 1B).Consistent with the previous western blot, the qRT-PCR analysis revealed no expression of SerpinB2 mRNA in all tumors of PyMT SB2−/− mice (Fig. 1C).The SerpinE1 expression levels are higher in PyMT SB2−/− tumors than those in PyMT WT tumors, while there is no significant difference in uPA expression The expression patterns and levels of uPA and SerpinE1 in the mammary tumors of PyMT WT and PyMT SB2−/− mice were analyzed by immunohistochemistry and western blotting.Strong staining of uPA was observed in the tumor cells of both PyMT WT and PyMT SB2−/− tumor tissues (Fig. 2A).SerpinE1 staining was strong in stromal areas, including adipocytes, fibroblasts, and macrophages, but not in tumor cells in both PyMT WT  and PyMT SB2−/− tumor tissues (Fig. 2A). Figure 2B shows a representative western blot for uPA and Ser-pinE1.SerpinE1 protein level were significantly higher in PyMT SB2−/− tumors (0.54 ± 0.09) than in PyMT WT tumors (0.24 ± 0.03) (P = 0.008).In contrast, uPA protein levels exhibited no significant difference between PyMT WT (0.41 ± 0.16) and PyMT SB2−/− (0.40 ± 0.01) tumors (P = 0.94) (Fig. 2B).WT and PyMT SB2−/− tumors After RNA-Seq transcriptional profiling analysis of PyMT WT and PyMT SB2−/− tumors, 23,282 expressed genes were identified.Compared with the PyMT WT tumors, 784 DEGs were identified in PyMT SB2−/− tumors.Of the 784 genes, 156 were upregulated and 628 were downregulated (Fig. 3A).In the top 10 enriched GO terms of biological process in up and downregulated DEGs of PyMT SB2−/− tumors compared to PyMT WT tumors, we found the following GO terms: wound healing, negative regulation of cell proliferation, cell adhesion, inflammatory response and immune system process (Fig. 3B and Table S2).DEGs (CCL8, IRF4, CXCL13, CCL17, CD206, IL23α, CXCL2, NOS2) related to twofold increased M1 macrophage markers and twofold decreased M2 macrophage markers were identified in the primary tumors of PyMT SB2−/− mice relative to those of PyMT WT mice.However, the expression of CXCL13, a marker of M1 macrophages, was decreased fivefold in PyMT SB2−/− mice relative to those of PyMT WT mice (Fig. 3C).

CK8-positive tumor cells and CD206-positive TAMs decrease in metastatic LNs of PyMT SB2−/− mice
To test the intrinsic role of SerpinB2 in LN metastasis, local LN metastasis in 16-week-old PyMT WT and PyMT SB2−/− mice with comparable mammary tumors was analyzed by CK8 immunostaining.The incidence of LN metastasis was 100% in PyMT WT and PyMT SB2−/− mice bearing tumors.Figure 6A (top) shows the H&E staining of LNs in the 4th mammary tumors of PyMT WT and PyMT SB2−/− mice.Representative immunohistochemistry of the epithelial marker CK8 revealed a reduction in number of CK8-positive tumor cells within the trabecular zone of LN sections of PyMT SB2−/− mice (2.95 ± 0.67%) compared with that of PyMT WT mice (6.37 ± 0.98%) (Fig. 6A bottom and Table 1, P = 0.02).Next, the TAM phenotypic polarization status in LN metastasis was investigated by immunohistochemistry of NOS2 and CD206.As shown in Fig. 6B, TAMs were detected in the trabecular zone of LNs, and the NOS2-positive area was increased, but there was no statistically significant difference in the LNs of PyMT SB2−/− mice (9.45 ± 2.03%) compared with that in PyMT WT mice (7.29 ± 2.72%).A significantly decreased CD206positive area was found in the LNs of PyMT SB2−/− mice (3.71 ± 0.66%) relative to that of PyMT WT mice (7.01 ± 1.00%) (Fig. 6B and Table 1, P = 0.025).

Combination of low SerpinB2, high NOS2, and low CD206 expression is a prognostic indicator of favorable survival
The BreastMark website was used to explore the association between SerpinB2, NOS2, and CD206 expression, and DFS in patients with breast cancer.The samples were separated into the combination of low SerpinB2, high NOS2, and low CD206 groups and the combination of high SerpinB2, low NOS2, and high CD206 groups.Kaplan-Meier plots revealed that patients with the combination of low SerpinB2, high NOS2, and low CD206 expression exhibited a significantly favorable DFS (n = 662, HR = 0.6723, P = 0.02469, Fig. 8A).The combination of low SerpinB2, high NOS2, and low CD206 expression was associated with good DFS in patients with breast cancer with LN metastasis (n = 272, HR = 0.5329, P = 0.0341, Fig. 8B).The DFS in patients with breast cancer without LN metastasis (n = 240, HR = 0.6358, P = 0.1912) was not significantly different, but there was still a notable difference in DFS within 132 months (11 years) in patients with breast cancer without LN metastasis (Fig. 8C).The analyses of the combinations of SerpinB2/NOS2, SerpinB2/CD206, and NOS2/CD206 groups, as well as the individual expressions of SerpinB2, NOS2, or CD206 alone did not demonstrate a significant difference in the DFS of patients with breast cancer (Supplementary Data 3).

Discussion
The role of SerpinB2 as a prognostic marker in breast cancer progression or suppression remains controversial.This study represents the first trial to analyze the consequences of SerpinB2 deficiency on tumorigenesis, tumor growth, and metastasis throughout the course of mammary cancer progression.We show that in PyMT SB2−/− mice, SerpinB2 deficiency delays mammary tumor initiation, growth, and LN metastasis, which is accompanied by a decrease in CD206 + M2 TAMs and an increase in NOS2 + M1 TAMs.Although SerpinB2 expression was observed in diverse cell types, the significant role of SerpinB2 expressed in TAMs as well as tumor cells during mammary tumor development and progression remains poorly understood.Several mechanisms have been proposed where SerpinB2 expression by tumor cells might influence tumorigenesis and include the inhibition of apoptosis or uPA signaling [3,20,21].The reports revealed that SerpinB2 expression in tumor cells may function as a survival factor by repressing proapoptotic signal transduction.We previously found that SerpinB2-upregulated MDA-MB-231 cells that were stably overexpressing miR200c promoted lung metastasis and boosted macrophage infiltration in tumor tissues, but the knockdown of SerpinB2 decreased lung metastasis and macrophage infiltration in xenograft mouse models [4].Current study revealed that the knockdown of SerpinB2 in tumor cells led to a Fig. 8 Kaplan-Meier plots of breast cancer patient survival based on the combination of SerpinB2, NOS2, and CD206 expression in the BreastMark dataset.A Disease-free survival (DFS) analysis of all breast cancer patients using the combination of three genes.B-C DFS analysis of breast cancer patients with or without lymph node (LN) metastasis using the combination of three genes notable decrease in the sphere formation and migration of MDA-MB-231 cells and delayed mammary cancer progression and LN metastasis in PyMT SB2−/− mice, suggesting that mammary tumor cell-produced SerpinB2 may contribute to the aggressiveness on of mammary tumor cells.
SerpinE1 and SerpinB2, both serine protease inhibitors, regulate the activity of plasminogen activators, primarily uPA [22].Their role in breast cancer appears to be paradoxical, with both pro-tumorigenic and potential tumor-suppressive effects: they can promote tumor progression by facilitating diverse signaling pathways involved in inflammation, cell proliferation, angiogenesis, and invasion, in certain contexts, whereas they have anti-tumor effects by inhibiting plasminogen activators and reducing metastatic potential [2,22,23].Compensatory mechanisms involving SerpinE1 and SerpinB2 have implications in various diseases, including cancer [24].In a previous study, we found that miR200c/141-transduced MDA-MB-231 cells had higher SerpinB2 but lower SerpinE1 levels [4].In present findings, SerpinB2 deficiency in PyMT SB2−/− tumors led to a significant increase in SerpinE1 expression, but no significant change in uPA expression.This suggests as SerpinB2 expression decreases, SerpinE1 expression may increase to compensate for their role in regulating uPA activity.SerpinE1, which is upregulated in PyMT SB2−/− tumors, may inhibit uPA activity, potentially suppressing uPA-mediated breast cancer development and progression.Therefore, understanding the intricate interplay among SerpinB2, SerpinE1 and uPA is crucial for unraveling their roles in breast cancer biology and developing targeted therapeutic strategies.
Based on co-culture studies with MDA-MB-231 and SerpinB2-knockdown RAW264.7 (RAW264.7-siSB2),we speculate that macrophage-produced SerpinB2 regulates the migration ability of breast cancer cells without substantially impacting proliferative activity, leading to increased metastasis.The following observations support our finding: an increase in NOS2-positive M1 macrophages and a decrease in CD206-positive M2 macrophages in the tumor tissues of PyMT SB2−/− mice and the SerpinB2 knockdown RAW264.7 cells.We previously found that SerpinB2-upregulated MDA-MB-231 cells that were stably overexpressing miR200c promoted lung metastasis and boosted macrophage infiltration in tumor tissues, but the knockdown of SerpinB2 decreased lung metastasis and macrophage infiltration in xenograft mouse models [4], implying that SerpinB2 secreted by MDA-MB-231 cells may contribute to macrophage migration and polarization.Recently, Meng et.al reported that SerpinB2 in MDA-MB-231 cells with miR-200c overexpression increased the secretion of IL-10, resulting in the induction of M2 polarization of RAW264.7 cells [27], which supports our findings.Although we investigated the NOS2 and CD206 known as M1 and M2 polarization markers typically expressed in both human and mouse macrophages in cross-talk experiments between MDA-MB-231 and RAW264.7,there are potential caveats such as a species mismatch between the cell lines and discrepancies in M1 and M2 polarization markers between human and mouse macrophages [28].Taken together, our results suggest that SerpinB2 deficiency leads to macrophages with more M1 and fewer M2 characteristics, resulting in an inhibitory effect on the growth and metastatic potential of mammary cancer.SerpinB2 deficiency is associated with a dysregulation of the Th1-and Th2-promoting cytokine release [9,14,23], which has traditionally been viewed as inhibiting and favoring tumor growth.We found that the circulating levels of Th1 and Th2 cytokines in the peripheral blood of tumor-bearing PyMT SB2−/− mice were lower than those in PyMT WT mice.These findings imply that SerpinB2 deficiency may cause Th1/Th2 immune perturbations in PyMT SB2−/− mice.Further research is needed to fully understand the molecular mechanisms and clinical implications of the interplay between SerpinB2 and Th1/Th2 immune responses in breast cancer.
SerpinB2 function remains paradoxical in breast cancer; SerpinB2 is associated with reduced metastasis and prolonged survival in patients with breast cancer [29,30] and shows increased significance for a favorable prognosis [31][32][33].In contrast, SerpinB2 promotes tumorigenesis and provides metastatic potential by inhibiting apoptosis and fostering vascular co-option advantage [3,4,21].In our previous study, immunohistochemistry revealed that SerpinB2 protein is highly expressed in tumor cells in aggressive breast tumor tissue with the triple-negative subtype, and is associated with a short overall survival in breast cancer patients [4].The BreastMark dataset based on mRNA analysis showed no significant difference in DFS between breast cancer groups with low and high SerpinB2 expression.
The inconsistent results for survival between our previous report and the BreastMark data might be due to differences in experimental analysis; thus, our previous study demonstrated the SerpinB2 expression level by immunohistological analysis in tumor cells.We found that the combination of low SerpinB2, high NOS2, and low CD206 mRNA expression was associated with favorable DFS in patients with breast cancer, suggesting that rather than focusing on a single gene biomarker, a combined multigene marker may have more powerful prognostic or predictive value.

Conclusion
Our findings reveal that in vivo SerpinB2 deficiency delays breast cancer development and metastasis and decreases the differentiation of TAMs toward a protumorigenic M2 phenotype.The combination of Ser-pinB2, NOS2, and CD206 expression in tumor cells as well as TAMs and its clinical significance should be further evaluated in a cohort of patients with breast cancer.

Fig. 1
Fig. 1 SerpinB2 deficiency expression in mammary tumors.A Representative immunostaining images of SerpinB2 in the tumors of PyMT WT and PyMT SB2−/− mice.B-C Analysis of SerpinB2 protein and mRNA levels in the tumor lysates of PyMT WT and PyMT SB2−/− mice using western blot and qRT-PCR.Data were presented as the means ± S.E.from the primary tumors of 3 mice per group.*P < 0.05, ***P < 0.001 using unpaired t-test

Fig. 2
Fig. 2 Immunohistochemistry and western blot analysis of uPA and SerpinE1 in tumor tissues.A Representative immunostaining images of uPA and SerpinE1 in the tumor tissues of PyMT WT and PyMT SB2−/− mice.B Western blot analysis of the uPA and SerpinE1 protein levels in the tumor lysates of PyMT WT and PyMT SB2−/− mice.Data were presented as the means ± S.E.from tumors of 7 mice per group.**P < 0.01 using unpaired t-test

Fig. 3
Fig. 3 Differentially expressed gene (DEG) clusters and Gene Ontology (GO) categorization of all transcriptomes assessed in the RNA-Seq analysis of primary tumors of PyMT WT and PyMT SB2−/− mice.A Volcano plot of the relationship between fold-change and significance between the two groups.B Top 10 enriched GO terms in the significantly up and downregulated DEGs in PyMT SB2−/− tumors relative to PyMT WT tumors.C Heatmap reflecting the expression profiles of DEGs associated with macrophage polarization

Fig. 4 Fig. 6
Fig. 4 SerpinB2 deficiency results in increased NOS2 + M1 and decreased CD206 + M2 macrophage polarization.A Analysis of relative mRNA expressions of M1 polarization markers (NOS2, IL23α, CXCL2, CXCL13) and M2 polarization markers (CD206, CCL17, IRF4, CCL8) in the primary tumors of PyMT WT and PyMT SB2−/− mice using qRT-PCR.Data were presented as means ± S.E. of primary tumors of 8 mice per group.B Western blot analysis of the NOS2 and CD206 protein levels in the tumor lysates of PyMT WT and PyMT SB2−/− mice.Data were presented as the means ± S.E.from tumors of 4-5 mice per group.C Immunostaining analysiss of F4/80, NOS2, and CD206 in the tumor tissues of PyMT WT and PyMT.SB2−/− mice.Data were presented as the means ± S.E.from tumor sections of 5 mice per group.*P < 0.05, **P < 0.01, ***P < 0.001 using unpaired t-test