BPIFB1 promotes metastasis of hormone receptor‐positive breast cancer via inducing macrophage M2‐like polarization

Abstract

Metastasis is an important factor affecting the prognosis of hormone receptor‐positive breast cancer (BC). However, the molecular basis for migration and invasion of tumor cells remains poorly understood. Here, we identify that bactericidal/permeability‐increasing‐fold‐containing family B member 1 (BPIFB1), which plays an important role in innate immunity, is significantly elevated in breast cancer and associated with lymph node metastasis. High expression of BPIFB1 and its coding mRNA are significantly associated with poor prognosis of hormone receptor‐positive BC. Using enrichment analysis and constructing immune infiltration evaluation, we predict the potential ability of BPIFB1 to promote macrophage M2 polarization. Finally, we demonstrate that BPIFB1 promotes the metastasis of hormone receptor‐positive BC by stimulating the M2‐like polarization of macrophages via the establishment of BC tumor cells/THP1 co‐culture system, qPCR, Transwell assay, and animal experiments. To our knowledge, this is the first report on the role of BPIFB1 as a tumor promoter by activating the macrophage M2 polarization in hormone receptor‐positive breast carcinoma. Together, these results provide novel insights into the mechanism of BPIFB1 in BC.

BPIFB1 plays a pivotal role in promoting macrophage M2 polarization, which enhances the ability of hormone receptor‐positive breast cancer cells to metastasize.

Abbreviations

BC

breast cancer

BHE

BPIFB1 high expression

BLE

BPIFB1 low expression

BMM

bone marrow micrometastases

BP

biological process

BPIFB1

bactericidal/permeability‐increasing‐fold‐containing family B member 1

CC

cellular components

DEGs

differentially expressed genes

EMT

epithelial–mesenchymal transition

ESR1

estrogen receptor 1

ESRRG

estrogen‐related receptor gamma

FFPE

formalin‐fixed, paraffin‐embedding

GO

Gene Ontology

GSEA

gene set enrichment analysis

HR+ BRCA

hormone receptor‐positive breast carcinoma

IHC

immunohistochemical staining

KEGG

Kyoto Encyclopedia of Genes and Genomes

KM

Kaplan–Meier

KMplotter

Kaplan–Meier plotter

LMT

with lymph node metastasis tumor

LPLUNC1

long palate, lung, and nasal epithelium clone

METABRIC

Molecular Taxonomy of Breast Cancer International Consortium

MF

molecular functions

NMT

no lymph node metastasis tumor

OD

optical density

OS

overall survival

qPCR

quantitative real‐time PCR

RFS

relapse‐free survival

RI

resistance index

TAMs

tumor‐associated macrophages

TCGA

The Cancer Genome Atlas

TLR4

Toll‐like receptor 4

TME

tumor microenvironment

1. INTRODUCTION

Female BC has surpassed lung cancer as the most commonly diagnosed cancer with an estimated 2.3 million new cases (11.7%), followed by lung cancer (11.4%). ¹ It is one of the leading causes of cancer‐related deaths in women worldwide. ² Approximately 15.5% of women die from BC, although diagnostic and treatment strategies have improved dramatically in the last few decades. ¹ Among all BC subtypes, HR+ BRCA remains the most numerous. ³ BMM of tumor cells are a distinctive feature of malignancy with poor prognosis. ⁴ The presence of BMM is associated with the occurrence of clinically significant distant metastases and death from cancer‐related causes. ⁵ Lymph node status is significantly associated with BMM ⁵ , ⁶ , ⁷ and may serve as an “incubator” for subsequent metastasis. ⁴ Thus, exploring the underlying molecular mechanisms of lymph node metastasis is essential to identify effective and novel therapeutic strategies.

BPIFB1 (bactericidal/permeability‐increasing‐fold‐containing family B member 1), also known as LPLUNC1 (long palate, lung, and nasal epithelium clone 1), is a member of the BPI/PLUNC superfamily as it contains two BPI domains. ⁸ , ⁹ , ¹⁰ The BPI structural domain is a hydrophobic barrel structure that specifically binds LPS in the cell wall of Gram‐negative bacteria. ¹¹ , ¹² BPIFB1 is aberrantly expressed in cancer tissues such as nasopharyngeal carcinoma, lung cancer, and gastric cancer, regulating chronic infection and inflammation, suggesting that it may play an important role in tumor development. ¹³ , ¹⁴ , ¹⁵ Although, BPIFB1 is reported as an inhibitor in several cancers, ¹³ , ¹⁶ its expression is higher in breast tumor tissues than in normal breast tissues. Its biological function and mechanism in BC is still unknown.

TAMs constitute one of the major tumor‐infiltrating immune cell types and are usually classified into two functionally distinct subtypes, the classically activated M1 macrophages and the alternatively activated M2 macrophages. ¹⁷ , ¹⁸ M2 macrophage has the function of promoting angiogenesis and lymphangiogenesis, ¹⁷ , ¹⁹ as well as promoting tumorigenesis and development. ²⁰ It assists in the migration of tumor cells and tumor stromal cells by secreting MMPs, serine proteases, and histone proteases that disrupt the stromal membrane of endothelial cells and decompose various collagen and other components of the extracellular matrix. ²¹ , ²² Recent studies have shown that the inhibition of the TLR4 signaling pathway promotes macrophage M2 polarization, ²³ , ²⁴ while BPIFB1 exhibits an inhibitory effect on this pathway. ²⁵ Therefore, it is of great interest to explore whether BPIFB1 affects BC by regulating macrophage polarization.

In this study, we found that BPIFB1 was abnormally highly expressed in HR+ BRCA, especially in those with lymph node metastasis. In addition, its high expression was significantly associated with poor prognosis in BC, particularly in the luminal A subtype. Mechanistically, BPIFB1 acted as an activator of macrophage M2 polarization and promoted tumor metastasis both in vitro and in vivo.

2. MATERIALS AND METHODS

2.1. Data sets

Gene expression RNA‐seq datasets and phenotype data for TCGA BRCA cohorts (https://portal.gdc.cancer.gov/) were obtained from the UCSC Xena browser (http://xena.ucsc.edu/). Validation data were from METABRIC (https://www.bccrc.ca/dept/mo/KMplotter, (https://kmplot.com/analysis/). ²⁶

2.2. Identification of DEGs

The limma, edgeR, and DEseq2 R packages were used to perform gene differential analysis. To identify the molecules related to HR+ BRCA lymph node metastasis, 149 samples (estrogen receptors‐positive by immunohistochemistry) from TCGA were divided into two groups, which consisted of 75 NMT samples and 74 withLMT samples. Gene difference analysis was performed between the NMT group and the LMT group. The significant DEGs were selected with the cutoff criteria p value <0.05 and | log2FoldChange| ≥ 1.

2.3. Functional annotation and gene set enrichment analysis

To explore the potential biological processes related to the hub gene, we performed gene enrichment analysis, based on GO (http://geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.kegg.jp/), using the clusterProfier R package. ²⁷ , ²⁸ The GO enrichment analysis was conducted based on three aspects including BP, MF and CC. We also identified the activated or inactivated biological pathways among patients by running the GSEA of the adjusted expression data for all transcripts. The used gene sets were downloaded from the MSigDB database (http://www.gsea‐msigdb.org/gsea/downloads.jsp). ²⁸ , ²⁹ The “c2.all.v7.5.1.entrez” gene set were used to quantify the activity of biological pathways, which was represented by the enrichment score.

2.4. Protein–protein interaction (PPI) network analysis

All identified DEGs were submitted to the STRING database (https://string‐db.org/) for constructing their protein interactions. Then we used the MCODE plugin in Cytoscape to screen significant subnetwork modules from the PPI network. The degree cutoff = 10, node score cutoff = 0.2, k‐core = 2, and max, and results were visualized using Cytoscape software version 3.9.0.

2.5. Tissue microarrays

In total, 151 BRCA tumor tissues and 46 normal breast tissues FFPE specimens were collected at the Second Affiliated Hospital of Harbin Medical University (the 2nd Affiliated Hospital of HMU, Harbin, China). The study was approved by the Joint Ethics Committee of the Harbin Medical University Health Authority and informed consent was obtained from all participants. The diagnoses of all specimens were confirmed via histopathological examination. Tissue chips were constructed in accordance with standard processes. ³⁰ , ³¹ Of the samples from the tissue microarray, 82 specimens that were eligible for both ER‐positive and Her2‐negative were selected for further research. The clinical, pathological and ultrasonographic characteristics of the subjects are shown in Table 1.

TABLE 1.

Clinical characteristics of patients from the 2nd Affiliated Hospital of HMU (n = 82).

Clinical variables	n	%
Age
<50	50	61.0
≥50	32	39.0
Sex
Male	0	0
Female	82	100
ER
Positive	82	100
Negative	0	0
PR
Positive	77	94.0
Negative	5	6.1
Her2
Positive	0	0
Negative	82	100
Ki67
<30	51	62.2
≥30	31	37.8
Elasticity scores
<4	5	6.1
≥4	55	67.1
Unknown	22	26.8
Resistance Index (RI)
<0.7	14	17.1
≥0.7	43	52.4
Unknown	25	30.5
Spiculation
Yes	48	58.5
No	13	15.9
Unknown	21	25.6
T
T1	41	50.0
T2	40	48.8
T3–T4	1	1.2
N
N0	52	63.4
N1–N3	30	36.6
M
M0	82	100
M1	0	0
TNM
I	31	37.8
II	43	52.4
III	7	8.5

Note: M0, no distinct metastasis; M1, with distinct metastasis; N0, no lymph node metastasis; N1–N3, with lymph node metastasis; T1, tumor size <2 cm; T2, tumor size 2–5 cm; T3, tumor size >5 cm.

2.6. Immunohistochemical staining

Immunohistochemistry was performed on the tissue microarrays. Briefly, the tissues were deparaffinized and rehydrated, and the samples were subjected to EDTA‐mediated high‐temperature antigen retrieval; the samples were then incubated overnight at 4°C with the primary anti‐BPIFB1 (HPA024256, Merck KGaA, Darmstadt, Germany). Then the secondary antibody and a DAB kit (K5007, DAKO; Dako REAL™ EnVision™) were applied to the microarrays. The staining was scored according to the staining intensity and the distribution of the stained region. Distribution was evaluated as none (0), ≤10% (1), 10–25% (2), 25–50% (3), and > 50% (4). Intensity was evaluated as none (0), faint (1), moderate (2), strong (3). The sections were reviewed by two pathologists. The product of staining intensity and extent scores was used as the IHC score, with a range of 0 to 12. Scores 0–4 were assigned as negative, and scores 5–12 were assigned as positive. ²⁶

2.7. Evaluation of immune cell fractions

Hormone receptor‐positive BC RNA‐sequencing expression (level 3) profiles were selected from the TCGA dataset for immune score evaluation. To assess the reliable results of immune score evaluation, we used immuneeconv. It is an R software package that integrates six of the latest algorithms, including TIMER, xCell, MCP‐counter, CIBERSORT, EPIC, and quanTIseq. These algorithms have been benchmarked and each has unique advantages. ³²

All the above analysis methods and R package were implemented by R foundation for statistical computing (2022) version 4.2.0 and software packages ggplot2 and pheatmap.

2.8. Cell lines and co‐culture system

Human HR+ BRCA cell lines T47D/MCF‐7, human monocyte cell line THP‐1 and mouse macrophage cell line RAW264.7 were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The indicated cells were cultured in DMEM (Thermo Fisher Scientific) containing 10% FBS and were supplied at 37°C under a humidified 95%:5% (v/v) mixture of air and CO₂. BC cells and THP‐1 cells were co‐cultured in Transwells, with the BC cell suspension added into the inserts and the THP‐1 cell suspension added into the (bottom) wells of the plate. ³³ THP‐1 cells were replaced by RAW264.7 cells and the above experiments were repeated. ³⁴ , ³⁵

2.9. Cell transfection

The overexpression of BPIFB1 in BC cells were performed using transfection of the BPIFB1‐overexpressing plasmids (pcDNA‐BPIFB1). The pcDNA‐BPIFB1 and its negative control (pcDNA) were obtained from Thermo Fisher Scientific. Cell transfection was performed using the Lipofectamine 2000 (Invitrogen). The cells were harvested 48 h after transfection for RNA extraction or functional assays.

2.10. Quantitative real‐time PCR assay

The expression of CD206 and CD163 in cells was detected using qPCR. Total RNA was isolated using TRIzol reagent (Life Technologies) according to the manufacturer's protocol. The PrimeScript® RT reagent kit (TaKaRa) was used to reversely transcript RNAs to cDNAs. SYBR Premix Ex Taq™ Kit (TaKaRa) was used for RT‐PCR analysis on a 7500 RealTime PCR System. GAPDH was measured as the internal control. The sequences of the primers used are shown in Table S1. After completion of the reaction, relative gene expression levels were calculated using the 2^−ΔΔCT method.

2.11. ELISA

The BPIFB1 protein levels in the supernatant of the co‐culture system were measured using an ELISA. Briefly, microtiter 96‐well plates (Nunc) were coated with 1 μg/mL (100 μL/well) of the primary anti‐BPIFB1 (MAB101952, bio‐techne) in 0.1 M bicarbonate‐coated buffer (pH 9.6) overnight at 4°C. Three washes with 200 μL PBS plus 0.05% Tween‐20 (Sigma) were performed between each incubation. Plates were blocked for 1 h at room temperature with 100 μL blocking buffer consisting of PBS 0.05% Tween and 10% FBS. Supernatant with serial dilutions in blocking buffer was added (100 μL/well) and incubated for 1 h. Polyclonal anti‐BPIFB1 (Merk) was added as a detection antibody and incubated for 1 h. Then, horseradish peroxidase (HRP)‐conjugated secondary antibody was added and incubated at 37°C for 1 h. Enzyme activity was detected by incubation with TMB substrate solution (Merk) for 5 min at room temperature in the dark. Finally, a stop solution was added to each well and OD was measured at 450 nm.

2.12. Cell migration and invasion assay

Cell migration and invasion assays were performed on 24‐well Transwell cell culture chambers with 8‐μm sized pores with or without pre‐coated Matrigel (Corning Costar). Next, 2 × 10⁵ BC cells were seeded into the upper chambers, and THP1 (or RAW264.7) cells were added to the lower chambers. Following 48 h of incubation, migrated or invasive cells on the lower surface of filters were fixed and stained with the Diff‐Quik stain (Sysmex), and stained cell nuclei were counted directly in triplicate. ³⁶ Cells were photographed (magnification ×100) and counted in at least five random microscopic fields. We assessed migration and invasion potential by counting the number of cells.

2.13. Animal experiments

Animal studies were permitted by the Joint Ethics Committee of the Harbin Medical University Health Authority (Harbin, China). ³⁷ T47D cell lines expressing luciferase were used for in vivo experiments. Female NOD/SCID mice (Vital Rivers, China) were housed in a specific pathogen‐free barrier facility with a 12‐h light/12‐h dark cycle. We established the lung metastasis model by tail vein injection of BC cells (5 × 10⁵) transfected with BPIFB1 or control into 4‐week‐old mice (three per group). The volume of the tumor and the number of metastatic nodules were recorded weekly. After intraperitoneal injection of D‐luciferin, all mice were imaged by the Xenogen IVIS Spectrum Imaging System (Caliper Life Sciences) before being sacrificed. Tumor volume and the number of lung metastatic nodules were measured. For each tissue, HE, or IHC staining was performed for histological detection. The primary anti‐CD206 (ab300621, abcam) and anti‐CD163 (ab182422, abcam) were used for IHC.

2.14. Statistical analysis

Statistical analyses were done using R software (version 4.2.0), GraphPad Prism v7.00 (GraphPad Software Inc.) and SPSS 28.0. Quantitative data are presented as the mean ± SEM or SD. Categorical variables were assessed using the χ² test. Differences between groups were analyzed using the Wilcox test when there were only two groups or using the Kruskal–Wallis H test when there were more than two groups. The survival data were plotted and analyzed by KM, log‐rank test, and multivariate Cox regression analyses. Spearman's correlation analysis was used to describe the correlation between quantitative variables without a normal distribution. p‐values less than 0.05 were considered statistically significant (*p < 0.05). The plots were constructed using R software or GraphPad Prism.

3. RESULTS

3.1. Identification of DEGs associated with lymph node metastasis in hormone receptor‐positive breast cancer

We collected the RNA‐sequencing expression profiles and clinical information of TCGA and categorized the samples into two groups: samples from patients with no lymph node (N0) metastasis (NMT) and samples from patients with N1–3 lymph node status (LMT). DEGs were identified by using the limma, edgeR, and DEseq2 R packages for gene difference analysis (74 LMT samples vs. 75 NMT samples; Figures 1A and S1A). We obtained 37 common DEGs using diverse R packages (p < 0.05 and |log2FoldChange| ≥ 1). Of these, 22 genes were upregulated and 15 genes were downregulated in the LMT subgroup (Figures 1B and S1B). In addition, 11 of the 37 genes were further selected by directly comparing the expression distribution (LMT group vs. NMT group; Figures 1C and S1C). Among these, BPIFB1, which has not been reported in BC, ¹⁶ was selected as the candidate target for further investigation.

FIGURE 1.

Identifying DEGs related to lymph node metastasis in HR+ BRCA. (A) Volcano plots were constructed using fold change values and P‐adjust. Red dots indicate upregulated genes; blue dots indicate downregulated genes; gray dots indicate insignificant. (B) Venn diagram of the differential genes analyzed using limma, edgeR, and DEseq2 R packages. (C) Violin plots of the 11 most significantly different genes between LMT and NMT samples (ns: not significant, *p < 0.05; **p < 0.01; ***p < 0.001).

3.2. BPIFB1 was upregulated in breast malignancies and associated with lymph node metastasis in hormone receptor‐positive subtype

We next evaluated the clinical significance of BPIFB1 expression in clinical breast tumor samples. TCGA database showed that BPIFB1 was significantly upregulated in the LMT group than the NMT group in HR+ BRCA (Figure S2A). Furthermore, by performing differential expression analysis of mRNA levels of BPIFB1 in BC and normal breast samples (1104 tumors vs. 113 normal samples) from TCGA, we revealed that BPIFB1 was markedly increased in cancer tissues, especially in luminal A breast carcinoma, compared with normal tissues (Figures 2A and S2B). Of note to us was the significantly higher expression of BPIFB1 in luminal A and luminal B subtypes, which were characterized by estrogen receptor positive. Therefore, we used estrogen to stimulate when MCF‐7 cells were cultured and measured BPIFB1 in the culture supernatant of the treated and control groups by ELISA. It indicated that estrogen‐exposed tumor cells had increased BPIFB1 expression (Figure S2C). Moreover, we used TCGA data to analyze the expression connection of BPFIB1 and estrogen receptor‐related genes. We found that the expression of BPIFB1 was positively correlated with two genes, including ESR1 and ESRRG (Figure S2D). Conclusively, we speculated that BPIFB1 expression in BC was probably regulated by the estrogen signaling pathway.

FIGURE 2.

BPIFB1 is significantly upregulated in breast cancer compared with normal breast tissue and associated with lymph node metastasis. (A) Expression distribution of BPIFB1 mRNA (tumors vs. normal samples), shown graphically as box and violin plots. (B) Representative samples of BPIFB1 positive (left) and negative (right) expression (scale bars: 500 μm for the above pictures, and 50 μm for the below pictures). Differences among variables were evaluated by χ².

To further confirm our findings, we made a tissue array containing 82 HR+ BRCA samples and 46 normal samples from the patients of our department and performed IHC staining (Table 1). The semiquantitative IHC scores were used to appraise protein expression levels of BPIFB1, revealing that BPIFB1 was highly expressed in tumor tissues (Figure 2B). Clearly, our data demonstrated that the differential expression of BPIFB1 represented distinct clinical phenotypes. By analyzing the ultrasonographic data from our Imaging Centre, we observed a significant correlation between BPIFB1‐positive status and higher RI (Table 2, Figure S2E). It was indicative of a more abundant blood supply to the tumor. Meanwhile, compared with the BPIFB1‐negative group, patients from the BPIFB1‐positive group trended toward a higher N stage (20/33), validating the correlation between BPIFB1 expression and lymph node metastasis (Table 2). Collectively, these data demonstrated that BPIFB1 was upregulated in HR+ BRCA, which was positively associated with lymph node metastasis tendency.

TABLE 2.

The relationship between BPIFB1 and clinical characteristics of HR+ BRCA patients from our hospital.

Clinical variables	BPIFB1‐Positive (n = 33)		BPIFB1‐Negative (n = 49)		p value
	n	%	n	%
Age
<50	19	57.6	31	63.3	0.604
≥50	14	42.4	18	36.7
Elasticity scores
<4	1	3.0	4	8.2	0.340
≥4	23	69.7	32	65.3
Unknown	9	27.3	13	26.5
RI
<0.7	2	6.1	12	24.5	0.031
≥0.7	20	60.6	23	46.9
Unknown	11	33.3	14	28.6
Spiculation
Yes	19	57.6	29	59.2	0.941
No	5	15.2	8	16.3
Unknown	9	27.3	12	24.5
T
T1	14	42.4	27	55.1	0.260
T2–T4	19	57.6	22	44.9
N
N0	13	39.3	39	79.6	<0.001
N1–N3	20	60.6	10	20.4
TNM
I	6	18.2	25	51.0	0.003
II–III	27	81.8	24	49.0
Ki67
<30	21	63.6	30	61.2	0.825
≥30	12	36.4	19	38.8

Note: The p‐values were determined using the two‐tailed χ² test.

3.3. High BPIFB1 expression predicted poor prognosis of HR+ BRCA patients

To investigate whether BPIFB1 expression was associated with clinical outcomes, we performed KM analysis using survival data with available clinical annotations from TCGA and METABRIC. The inclusion conditions of the samples for conducting the analysis are shown in Figure S3A. Te best cutoff calculated by the survminer R package was used to divide the samples into the BHE group and the BLE group. In the TCGA cohort, the KM estimator showed that OS in the BLE group was significantly higher than that in the BHE group (Figures 3A and S3B). Interestingly, we also observed a significant difference in terms of OS when we analyzed the data of TCGA patients with N0 lymph node status (Figures 3B and S3C). Meanwhile, in the METABRIC cohort, elevated expression of signature BPIFB1 was associated with markedly poorer RFS (Figure S3D). Then we analyzed the OS of the BHE group and the BLE group in different BC subtypes by using the KMplotter online tool. Notably, the results showed that the high expression of BPIFB1 was significantly associated with unfavorable prognosis in the luminal A breast carcinoma, but not in luminal B, HER2‐positive or triple‐negative subtypes (Figure S3E–H).

FIGURE 3.

High expression of BPIFB1 is significantly associated with poor prognosis in hormone receptor‐positive breast cancer. (A, B) Kaplan–Meier survival analysis of BPIFB1 gene using TCGA data. (C, D) Kaplan–Meier survival analysis of BPIFB1 expression using the survival data from our hospital (p < 0.05).

Next, to investigate the relationship between the expression of BPIFB1 protein and patient prognosis, we performed KM analysis using survival data from individuals in our hospital. By validating the BPIFB1 expression in our tissue array, we categorized 33 of 82 patients into the BPIFB1‐positive group and 49 patients into the BPIFB1‐negative group. As expected, significantly better OS and RFS were found in the BPIFB1‐negative group (Figure 3C,D). In conclusion, our findings demonstrated that high BPIFB1 expression might lead to poor prognosis in HR+ BRCA.

3.4. Potential functions and molecular mechanisms of BPIFB1 were predicted

Limited evidence demonstrated the molecular mechanism of BPIFB1 in malignancies, especially in BC. ¹⁶ Therefore, we next performed an enrichment analysis to explore the potential role of BPIFB1 in HR+ BRCA. The DEseq2 R package was used to select the BPIFB1‐related gene set in HR+ BRCA for enrichment analysis (Figure S4A, Table S2). Enriched pathways were identified by GSEA, and the results are shown in Table S3. The critical results suggested that BPIFB1 might correlate with endocrine resistance and epithelial–mesenchymal transition (Figure S4B), which were intimately associated with BC metastasis. ³⁸ , ³⁹

Furthermore, the three GO categories, biological process (GO‐BP), molecular function (GO‐MF) and cellular component (GO‐CC), were tested independently, which indicated that the BPIFB1‐related gene set was significantly enriched in multiple immune‐related pathways (including: humoral immune response, phagocytosis, recognition, immunoglobulin complex, circulating of immunoglobulin complex, antigen binding, and immunoglobulin receptor binding; Figure 4A). Additionally, the results of KEGG pathway enrichment analysis indicated that BPIFB1‐related genes were significantly enriched in the estrogen signaling pathway (Figure 4B), which double confirmed our previous results in Figures S2B–D and S3E. Overall, we hypothesized that BPIFB1 might play an important role in tumor immunity.

FIGURE 4.

Prediction of the potential function and mechanism of BPIFB1 in hormone receptor‐positive breast cancer. (A, B) GO and KEGG enrichment of BPIFB1‐related genes.

3.5. Association of BPIFB1 and tumor microenvironment immune cell infiltration characterization

To further investigate whether BPIFB1 was able to regulate immune cells in the TME, we analyzed the relationship between BPIFB1 expression and immune cell infiltration by using bioinformatics. The 149 HR+ BRCA samples from TCGA were classified into BHE and BLE groups based on the median expression of BPIFB1. To analyze the composition of immune cells in different BPIFB1 subgroups, the immuneeconv R package was used to systematically evaluate the infiltration level of immune cells in each sample (Figure S5A). We used the CIBERSORT abs‐mode algorithm to conduct immune infiltration estimation, which has a better performance in comparing differences in immune cell abundance between groups. ³² We found that M2 macrophages, naive B cell, activated mast cell, plasma B cell, memory B cell, and activated NK cell were more abundant in the BHE subgroup, while resting NK cell was more abundant in the BLE subgroup (Figure 5A). In addition, using the CIBERSORT algorithm, which is commonly used to perform comparisons between different immune cell types of the same samples, ³² we observed the interactions of the components in the immune microenvironment through the association between immune cells (Figure 5B). Thereafter, we verified this result with another credible algorithm quanTIseq. ³² Consistent with our previous results, M2 macrophages were also more abundant in the BHE subgroup (Figure 5C). However, the Spearman correlation of BPIFB1 with four known M2 polarization regulator‐related genes was not statistically significant (Figure S5B). Collectively, the observation that emerged from the data comparison was that BPIFB1 might have an independent role in regulating macrophage polarization in TME.

FIGURE 5.

Immune cell composition of samples from HR+ BRCA patients. (A) Comparison of immune infiltration between high and low BPIFB1 expression groups was performed via CIBERSORT abs mode. (B) Correlations among immune cells in HR+ BRCA was performed via CIBERSORT (blank cells: not significant). (C) Comparison of immune infiltration between high and low BPIFB1 expression groups was performed via quanTIseq algorithm.

3.6. BPIFB1 promoted macrophage M2 polarization and exhibited promoting effects on tumor progression in vitro and in vivo

It was well established that macrophages in the tumor microenvironment could be educated by cancer cells. ⁴⁰ BPIFB1 levels were overexpressed in T47D and MCF‐7 cells. We also used qPCR analysis to confirm the efficiency (Figure 6A). To examine whether cancer cell‐derived BPIFB1 could modulate macrophage M2 activation, we constructed a BC tumor cells/THP1 co‐culture system to detect M2 macrophage markers (CD206, CD163) of THP‐1 cells by qPCR analysis. We first used ELISA assay to confirm that soluble BPIFB1 in the supernatant of the co‐culture system was significantly increased compared to the control (Figure 6B). After co‐culture, as shown in Figure 6C, the expression of both CD206 and CD163 was significantly increased, respectively, by BPIFB1 stimulation in THP‐1. Similar results were observed in experiments using RAW264.7 cells in substitution for THP‐1 cells (Figure S6A). Consistent with the results of the analysis using public databases (Figure 5A,C), we suggested that cancer cell‐derived BPIFB1 might promote macrophage M2 polarization.

FIGURE 6.

BPIFB1 promotes hormone receptor‐positive breast cancer progression via stimulating macrophage M2‐like polarization in vitro and in vivo. (A) BPIFB1 mRNA expression was measured by qPCR analysis. (B) The soluble BPIFB1 protein expression levels in culture supernatants measured by ELISA assay. (C) CD206 and CD163 expression was detected in THP1 using qPCR. (D) Transwell migration and invasion assays. (E) Construction of a mouse lung metastasis model by tail vein injection of T47D tumor cells. (F) Mouse models were used to analyze the growth of tumors. Weekly evaluation of tumor diameter. Right panels: growth curves of xenograft tumors. (G) Quantification of metastatic lung nodules were used to characterize tumor metastasis. H&E‐stained tumor section images showing metastatic nodules. (H) CD206 and CD163 in transplanted tumors. The representative pictures were shown. All experiments included were repeated at least three times, ns: not significant, *p < 0.05; **p < 0.01; ***p < 0.001.

Next, to investigate the tumor‐regulating role of THP1 cells treated with BPIFB1, we performed functional studies in vitro. Compared with the control group, we found that overexpression of BPIFB1 significantly increased the migrative and invasive ability of tumor cells co‐cultured with THP1 (Figure 6D). In the co‐culture system of MCF‐7 and RAW264.7, we also observed promotion of the metastatic ability of tumor cells induced by BPIFB1 overexpression (Figure S6B). Interestingly, we then noticed that deletion of THP‐1 in the co‐culture system inhibited the effect of BPIFB1 overexpression on MCF‐7 cell migration and invasion (Figure S6C). It suggested the significant importance of macrophages in the phenotypic regulation of BC exerted by BPIFB1. To further confirm the above findings in vivo, xenograft mouse models were established (Figure 6E). We observed that overexpression of BPIFB1 in BC tumor cells markedly increased the xenograft tumor size in a lung metastasis model (Figure 6F). Similar to the results of cellular experiments in vitro, mice injected with BPIFB1‐overexpression BC cells exhibited more lung metastatic nodules (Figure 6G). In addition, to further verify whether BPIFB1 also promoted tumor progression in vivo by stimulating macrophage M2 polarization, IHC staining was used to detect the expression of CD206 and CD163 in tissues. We observed higher expression levels of M2 macrophage markers in the tissues of mice models injected with BPIFB1 overexpressing BC cells (Figure 6H). Consistently, as shown in Figure S6A,B, cellular experiments in vitro confirmed that BPIFB1 also modulated the polarization of mouse macrophages RAW264.7 thereby affecting the phenotype of tumor cells. These data collectively confirmed that BPIFB1 promoted tumor cell proliferation and migration in vitro and in vivo, with a possible role in enhancing macrophage M2 polarization.

4. DISCUSSION

BPIFB1, which encodes a protein structural domain with bactericidal and permeation‐enhancing functions, ¹⁶ has previously been shown the ability to inhibit carcinoma in a variety of cancer types including nasopharyngeal carcinoma, non–small‐cell lung cancer, and gastric cancer, but its role in BC has not been reported by other researchers. ¹³ , ¹⁴ , ¹⁶ However, in contrast to the part of other previous studies, KM survival analysis predicts significantly poor survival prognosis in HR+ BRCA patients with high expression of BPIFB1. Compared with normal breast tissue, the expression of BPIFB1 is higher in BC, especially in the hormone receptor‐positive subtype. Our data demonstrate that BPIFB1 expression is stimulated by estrogen and positively correlates with ESR1 and ESRRG. Furthermore, high BPIFB1 expression was significantly associated with poor prognosis in luminal A BC only and its associated gene set was enriched in the estrogen signaling pathway. These data provide further evidence that BPIFB1 might be regulated by the estrogen signaling pathway. We hypothesize that estrogen may regulate BPIFB1 transcription by interacting with related receptors in tumor cells, such as ESR1. The association between BPIFB1 and estrogen signaling pathway in BC is interesting and it will be further investigated in our future work.

The results of GO enrichment analysis suggest that BPIFB1 may participate in the regulation of tumor immunity. Previously, Shin et al. have demonstrated the role of BPFIB1 as a secreted protein engaging innate immunity in the recognition and regulation of microbiota. ²⁵ In recent years, there are emerging lines of evidence that microbes are also part of the tumor tissue itself in a wider range of cancer types beyond colorectal cancer, such as pancreatic, lung, and breast cancers, which were originally thought to be sterile. ⁴¹ , ⁴² The fact that microbiota exists in TME of BC is extremely striking. The presence of these microorganisms even affects the metastasis of tumor cells. ⁴³ BPIFB1, a protein that can interact with bacteria and affect immune cell responses, ²⁵ thus we hypothesize that it may play an important role in the modulation of immune cells in TME.

Hence, to investigate the potential mechanisms that may exist of BPIFB1 in tumor immunity, we construct immune infiltration evaluation via bioinformatics. Perhaps the most significant finding is that high BPIFB1 mRNA expression is remarkably associated with macrophage M2 polarization. Intriguingly, the same results are observed via qPCR assays in vitro and IHC staining in vivo. According to these data, we can infer that BPIFB1 plays an important role in enhancing the polarization of M2 macrophages in HR+ BRCA. Fortunately, many previous studies of BPIFB1 mechanism‐related studies support our findings. Shin et al. indicated that BPIFB1 suppresses TLR4 signaling in response to bacteria LPS. ²⁵ Activation of TLR4 occurs through the binding of LPS. Following LPS binding, TLR4 dimerizes and recruits downstream signaling and/or bridging molecules, leading to the activation of multiple distinct intracellular networks, resulting in a large cellular response in immune and cancer cells. ⁴⁴ Numerous rewarding studies have demonstrated that inhibition of TLR4 signaling promotes M2 polarization in macrophages. ²³ , ²⁴ From our perspective, a possible mechanism by which BPIFB1 promotes M2 polarization is through inhibition of the TLR4 signaling pathway. Although our findings support that BPIFB1 promotes macrophage transition to the M2‐like phenotype, whether it is acting through the TLR4 signaling pathway or other mechanisms has not been verified by further experiments.

The literature reports a close relationship between TAMs and the malignant clinical phenotype of BC, especially the M2 subtype. ⁴⁵ M2 macrophage is commonly described to increase the progression and metastasis of breast carcinoma. Distant metastases induced by M2 macrophages are more immunologically inert. ⁴⁶ Critically, the data from our experiments in vivo and in vitro are consistent with these findings. M2 macrophages can directly affect the endothelial cell stromal membrane by secreting a variety of cytokines, which leads to EMT. ²¹ , ²² EMT is the basis of tumor metastasis. ⁴⁷ This process allows tumor cells to acquire the ability to migrate and endows them with stem cell properties. ⁴⁸ Besides, tumors also rely on secreted factors to regulate macrophage polarization, thus forming a feedback loop between TAMs and EMT. ⁴⁹ BPIFB1, as a secreted protein that regulates innate immunity, is presumed to be involved in this feedback loop. ¹⁶ Together, from our perspective, BPIFB1 promotes the metastasis of HR+ BRCA by promoting macrophage M2 polarization.

In summary, we identified a novel target, BPIFB1, which was significantly associated with HR+ BRCA lymph node metastasis, via bioinformatics analysis. BPIFB1 was associated with various metastasis‐related clinical characteristics and predicted poor prognosis in HR+ BRCA. Mechanistically, it promoted macrophage M2‐like polarization, and consequently enhanced tumor cell migrative and invasive ability. Animal experiments further verified that BPIFB1 promoted the metastasis of hormone receptor‐positive BC in vivo. Collectively, our research provides new insights into the mechanisms of tumor immunity and BPIFB1.

AUTHOR CONTRIBUTIONS

ABH, FM, and BLG devised the study. ABH, YSL, YD, ZYR, YYZ, TW, and YHS completed the data collection and statistical analysis. S.S., ZAC, JYF, YQD, WLC, TSY, and MCL performed the experiments. FM provided technical and material support. ABH, YSL, YLL, and JRZ conceptualized the initial manuscript collectively. BLG and FM edited the manuscript and provided revisions. All authors have read and agreed to the final version of the manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare that they have no conflicts of interest.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Reviewer Board: The study was approved by the Ethics Committee of Harbin Medical University.

Informed Consent: Informed consent for the use of tissues was obtained from the patient prior to the initiation of the study.

Registry and the Registration No. of the study/trial: N/A

Animal Studies: Animal experiments were performed in accordance with the Guidelines for the Management of Laboratory Animal Affairs and were permitted by the Joint Ethics Committee of the Health Administration of Harbin Medical University. The study complied with the ARRIVE guidelines.

Supporting information

Figure S1.

Figure S2.

Figure S3.

Figure S4.

Figure S5.

Figure S6.

Table S1.

Table S2.

Table S3.

Hu A, Liu Y, Zhang H, et al. BPIFB1 promotes metastasis of hormone receptor‐positive breast cancer via inducing macrophage M2‐like polarization. Cancer Sci. 2023;114:4157‐4171. doi: 10.1111/cas.15957

DATA AVAILABILITY STATEMENT

The data sets used and/or analyzed during the current study are available from the Xena website for TCGA database (http://xena.ucsc.edu/, accessed on 11 September 2021) and the METABRIC database (https://www.bccrc.ca/dept/mo/, accessed on 12 January 2022). Other data can be acquired from the corresponding authors on reasonable request.