Integration analysis of single-cell transcriptome reveals SPP1⁺ and TFF3⁺ macrophage subsets contributing to the brain metastasis from lung cancer

Introduction

Lung cancer is a prevalent malignant tumor with a high mortality rate worldwide. The substantial mortality is closely associated with the elevated incidence of brain metastasis in lung cancer, at 20–40% (1). Over 10% of patients with small cell lung cancer (SCLC) present with brain metastases at diagnosis, and more than 50% of metastases occur within 2 years (2). Non-small cell lung cancer (NSCLC) exhibits similar patterns of brain metastasis. Over the past decade, advances in imaging technology have enhanced the detection rate of brain metastasis (3). The combined use of chemotherapy, targeted therapy, and immunotherapy in lung cancer treatment has extended patient survival, concurrently elevating the cumulative incidence of brain metastasis (4). There is an urgent need to enhance the quality of life of lung cancer patients and extend their survival period. Therefore, studying brain metastasis in lung cancer is pressing.

Inflammatory components involved in the progression of lung cancer encompass various immune cells capable of releasing diverse cytokines, chemokines, and cytotoxic mediators, which are recognized as pivotal factors in lung cancer. Among these immune cells, tumor-infiltrating macrophages, or tumor-associated macrophages (TAMs), play a crucial role in the onset and progression of lung cancer. TAMs, involved in shaping the tumor microenvironment, actively contribute to lung cancer cell proliferation, invasion, and metastasis. Guo et al. have demonstrated that M2-type TAMs can boost cancer cell invasion by promoting epithelial-mesenchymal transformation, upregulating CRYAB expression, and subsequently inducing lung cancer metastasis (5). Microglia, immune cells of the central nervous system, function as tumor-related macrophages in the brain. Research indicates that the growth of brain metastases is likely to result in neuronal degeneration or even death, thus activating microglia. Activated microglia contribute to local immunosuppression and facilitate metastatic tumor development by releasing transforming growth factor-β (TGFβ) and interleukin-10 (IL-10). They particularly enhance the proliferation of tumor cells and their infiltration into surrounding normal brain tissues (6). Microglia can migrate to the site before tumor metastasis, inducing local and systemic immunosuppression, promoting tumor angiogenesis, reorganizing surrounding tissues, facilitating matrix remodeling, and encouraging tumor invasion. Microglia play a crucial role in tumor brain metastasis (7). Therefore, gaining a deeper understanding of the underlying mechanisms of TAM function holds potential for developing innovative immune interventions targeting TAMs.

The incidence of brain metastasis in lung cancer patients is high, with poor prognosis and a short natural survival period. One challenge is the blood-brain barrier. Tumor cells must penetrate this complex barrier to form colonies, but its structure makes drug penetration difficult. Additionally, there is molecular target heterogeneity. Tumor cells have immune checkpoints like programmed death-1/programmed death ligand 1 (PD-1/PD-L1) and cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), but their expression levels differ from those in the primary lesion, limiting the effectiveness of immune checkpoint inhibitors like pembrolizumab. Current radiotherapy techniques also have limitations, such as potential cognitive impairment, with whole brain radiotherapy (WBRT) possibly causing hippocampal damage and leading to short-term memory decline. Among the treatment methods for lung cancer with brain metastasis, there are no existing drugs targeting TAM. Since TAM has a carcinogenic effect in lung cancer brain metastasis, it is poised to become an therapeutic target. TAM polarization is regulated by various signaling pathways, and modulating these pathways can effectively change the TAM phenotype. Strategies targeting TAM repolarization, such as exosomes, bacterial therapy, NPs, and CAR-M therapy, have demonstrated potential in treating solid tumors (8). The paradigm shift in targeting TAM polarization in tumor immunotherapy and its significant impact on certain tumors has made it a hot research topic.

Single-cell transcriptome sequencing technology, which enables the interpretation of cell heterogeneity at the single-cell level, has been rapidly applied to oncology, such as tumor heterogeneity, tumor metastasis, tumor microenvironment, and anti-tumor drug development. A recent study at the single-cell level has unveiled the diversity of TAMs across various cancers. Lavin et al. (9) observed distinctions between TAM and lung tissue-resident macrophages in early lung adenocarcinoma. Compared to macrophages in normal tissues, TAMs exhibit higher expression of the immunoregulatory factor PPARγ, elevated levels of CD64, CD14, CD11c, TREM2, CD81, MARCO, APOE, and lower levels of CD86 and CD206. Concurrently, TAMs were found to persistently express interleukin-6 (IL-6), linked to the immunosuppressive mechanism of TAM. Li et al. (10) identified differentially expressed genes related to macrophages using published lung adenocarcinoma (LUAD) single-cell RNA sequencing (scRNA-seq) data. Based on these macrophage-related genes, three distinct immune patterns with varying immune infiltration characteristics were identified in The Cancer Genome Atlas (TCGA)-LUAD cohort. The diversity of macrophage function stems from the variety of its phenotypes. The advancement of single-cell sequencing technology is gradually unraveling the mystery of macrophage heterogeneity. However, there is currently no single-cell study specifically addressing brain metastasis of lung cancer. In this study, scRNA-seq data from lung cancer tissues and brain metastasis tissues are analyzed to understand the subtypes and characteristics of tumor-related macrophages in brain metastasis tissues of lung cancer. This analysis aims to enhance our comprehension of the pathological process, related genes, and potential biological mechanisms underlying brain metastasis of lung cancer. It opens new possibilities for developing mechanisms, diagnosis, and treatment strategies for brain metastasis of lung cancer. We present this article in accordance with the STROBE reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2489/rc).

Methods

Acquisition and collection of single-cell RNA sequencing (scRNA-seq) dataset

The dataset of brain metastases in patients with lung cancer utilized in this study comprised GSE158803, GSE123902, GSE131907, GSE186344, and E-MTAB-8230. ScRNA-seq data for GSE158803, GSE123902, GSE131907, GSE186344 were generated using 10x Genomics and obtained via the Gene Expression Omnibus (GEO) database platform (www.ncbi.nlm.nih.gov/gds). ScRNA-seq data for E-MTAB-8230 was acquired from 10x Genomics through the ArrayExpress platform (www.ebi.ac.uk/biostudies/arrayexpress). These datasets encompassed cells from normal lung tissue, lung cancer tissue, normal brain tissue, and brain metastasis tissue of lung cancer. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Analysis of single-cell data

The raw single-cell data were sourced from previously published studies. The single-cell data was were analyzed utilizing the Seurat package within the R software. The comprehensive procedure was as follows: setting up the Seurat Object, implementing a standard pre-processing workflow, normalizing the data, identifying highly variable features, scaling the data, conducting linear dimensional reduction, determining the “dimensionality” of the dataset, clustering the cells, implementing non-linear dimensional reduction [uniform manifold approximation and projection/t-distributed stochastic neighbor embedding (UMAP/tSNE)], identifying differentially expressed features (cluster biomarkers), assigning cell type identity to clusters, and visualization.

Integration of single-cell transcriptomes

All scRNA-seq matrix counts utilized in this study were obtained from the GEO database. For scRNA-seq integration, analysis, and visualization, we employed R software (version: 4.1.0) along with the “Seurat” package (11,12). Prior to integration, separate procedures were conducted for each dataset, encompassing quality control and cell selection, data normalization, and identification of highly variable features. Following the integration method of the “Seurat” software package, the data were integrated to gain a comprehensive understanding of the molecular mechanisms and cellular component differences in lung cancer brain metastasis. A cross-dataset of cells derived from matched biological states was identified, and technical discrepancies between the datasets were rectified using the specific algorithm of “Seurat”. Subsequently, the two datasets were integrated for further analyses. Human data from different patients and tissues were also integrated to correct sample differences using the same method. The data were then analyzed based on the standard processing workflow of “Seurat”, which includes scaling the data, linear [principal component analysis (PCA)] and non-linear (UMAP) dimensional reductions, and clustering of cells.

Identification of cell type

Building on previous literature and marker genes utilized in original studies, we employed C1QA, C1QB, and C1QC as marker genes for macrophages. These genes were used to identify macrophages in all datasets following dimensionality reduction and clustering.

Visualization of single-cell data

The visualization of data analysis in this study was conducted using “Seurat”, incorporating UMAP scatter plots, violin plots, dot plots, volcano plots, and heatmaps. In the UMAP scatter plot, each cell is distributed along spatial axes based on gene expression, with cells sharing similar expression patterns positioned closer in spatial distance. The violin plot depicts the expression level of a specific gene or set of genes in a particular cell type. In the dot plot, the color of the dots indicates the average expression level of genes in a cell type, with darker colors signifying higher expression levels. The dot size reflects the percentage of positive expression for a gene in a cell type, with a larger diameter indicating a higher positive ratio. The volcano plot illustrates differentially expressed genes in a cell type compared to other cell types, encompassing both upregulated and downregulated genes. The heatmap compares the average expression levels of specific genes across different cell types.

Gene enrichment analysis

The differentially expressed genes across the clusters were identified utilizing Seurat. The up-regulated or down-regulated gene list in the cluster was uploaded to the online tool “David” (https://david.ncifcrf.gov/summary.jsp), a database for annotation, visualization, and integrated discovery (David). This platform offers researchers a comprehensive suite of functional annotation tools to comprehend the biological significance of extensive gene lists. KEGG signaling pathway enrichment analysis was conducted on different cell subsets utilizing the provided gene lists.

Statistical analysis

Single-cell RNA-seq analyses were performed using R (version 4.1.0) with the Seurat package (version 4.0.2) developed by Rahul Satija, Satija Lab and Collaborators (12).

Results

Single-cell transcriptome analysis of lung cancer brain metastasis

To investigate the mechanism of brain metastasis in lung cancer, we identified five single-cell transcriptome sequencing datasets related to brain metastasis of lung cancer. GSE123902, GSE186344, GSE158803, and GSE131907 were sourced from the GEO database of NCBI, while E-MTAB-8230 came from the ArrayExpress database. The GSE123902 dataset involved the transcriptional profiling of 41,384 single cells from 17 patients with various stages of lung adenocarcinoma progression, providing a comprehensive cellular profile of normal lung, primary tumors, and metastases (13). The GSE186344 dataset collected fresh brain metastasis (BrM) specimens from 15 patients, including 3 cases of lung cancer (14). We extracted data from these 3 patients with brain metastasis from lung cancer, as well as single-cell transcription data of infiltrated T cells and macrophages. The GSE158803 dataset integrated >500,000 cells from 217 patients and 13 cancer types (15). We selected the cells of a patient with NSCLC brain metastasis for scRNA-seq. Single-cell RNA sequencing (scRNA-seq) was conducted on 208,506 cells from 58 lung adenocarcinoma samples of 44 patients in the GSE131907 dataset, providing a cell map that included cancer cells, interstitial cells, and immune cells in the surrounding tumor microenvironment (16). We retrieved the myeloid cells from the GSE131907 dataset for further examination. Additionally, scRNA-seq was performed on three healthy adult brain cell types, including brain cells, interstitial cells, and glial cells, in the E-MTAB-8230 dataset (17). Within these datasets, we specifically selected tumor specimens from 56 patients with brain metastases from lung cancer, encompassing tumor cells, stromal cells, and immune cells. All these single-cell transcriptome data were generated using the 10x Genomics platform.

Initially, we utilized the Seurat package in the R language to conduct standard single-cell analysis on each of the five single-cell datasets separately. Following dimensionality reduction and clustering, cells in the five datasets were categorized into numerous subsets. UMAP maps depict the cluster distribution of all subsets in the GSE123902, GSE186344, GSE158803, GSE131907, and E-MTAB-8230 datasets, respectively. Simultaneously, to isolate macrophages from each dataset, we employed C1QA, C1QB, and C1QC to identify macrophage subsets. Dot plots illustrate the expression levels of these three typical macrophage marker genes across all subsets in each dataset (Figure 1A-1E).

Figure 1 Integrated analysis of single-cell transcriptome of brain metastasis from lung cancer UMAP plot and dot plot show the cluster distribution of all subsets and C1QA, C1QB, and C1QC expression of all subsets in GSE123902 (A), GSE186344 (B), GSE158803 (C), GSE131907 (D), E-MTAB-8230 (E) datasets, respectively. UMAP, uniform manifold approximation and projection.

Single-cell analysis of macrophage subsets

Several studies have confirmed the close association between the infiltration of TAMs and the mechanism of brain metastasis in lung cancer. To explore changes in macrophage subsets within the tumor microenvironment caused by brain metastasis in lung cancer, understand the heterogeneity of tumor-related macrophages in brain metastasis, and clarify the actions and roles of macrophages in lung cancer brain metastasis, we re-analyzed macrophages in the aforementioned datasets. We identified and extracted 30,922 macrophages from the five datasets. The standard Seurat integration algorithm was applied to integrate macrophages from different sources, followed by dimensionality reduction and clustering analysis. Ultimately, we obtained 16 macrophage subsets. The UMAP diagram illustrates the spatial cluster distribution of all macrophage subsets (Figure 2A).

Figure 2 Analysis of macrophage subsets. (A) UMAP cluster plot shows the clustered distribution of all macrophage subsets. (B) Violin plot shows mitochondrial and ribosomal proportions of all clusters. (C) UMAP plot shows the cluster distribution of all patient sample source cells. (D) UMAP plot shows the cluster distribution of all source cells of the dataset. (E) Histogram compares the proportions of all clusters in all datasets. (F) Heatmap shows differential gene expression across all macrophage clusters. UMAP, uniform manifold approximation and projection.

Then, we calculated mitochondrial and ribosomal gene ratios for these 16 macrophage subsets to compare differences in cell states between the subsets. The violin plot illustrates the results for all subsets (Figure 2B). As depicted in the figure, the proportion of mitochondrial genes in all subsets remained below 20%, suggesting a low degree of apoptosis. Notably, subset 14 exhibited a higher proportion of mitochondrial genes, indicating a potentially elevated degree of apoptosis or mitochondrial activity compared to other subsets. Subset 15 demonstrated a higher proportion of ribosomal genes, suggesting it might be in a period of vigorous protein synthesis with more active ribosomal function. To assess distribution differences of tumor tissue cells in each patient’s sample and evaluate the data integration effect, we assigned patient labels to all cells. The UMAP plot displays the clustering distribution of cells from all patients (Figure 2C). Results indicated a uniform distribution of most patient samples on the UMAP figure, signifying that inter-group and technical differences between patient samples were appropriately eliminated. However, a few patients exhibited unique subclusters, suggesting individual differences among patients with brain metastasis from lung cancer in this dataset. For comparing integration effects across the five datasets, all cells were labeled by dataset source. The UMAP plot displays the spatial clustering of TAMs across datasets (Figure 2D). Notably, the distribution of TAMs appeared generally uniform across different datasets. Simultaneously, to visually compare the types and proportions of all subsets in each dataset and explore differences in the cellular composition of the tumor microenvironment, we extracted the number and proportion of all cell types in each dataset. The histogram illustrates the cell proportions of all cell subsets in the five datasets (Figure 2E). It was evident that the E-MTAB-8230 dataset only contained subset 13, and its subset components were substantially different from other datasets. E-MTAB-8230 data originated from normal brain tissue, whereas the other four datasets derived from normal lung tissue, lung cancer tissue, and brain metastasis tissue of lung cancer. This indicated significant differences in molecular expression between macrophages in normal brain tissue and TAMs. To analyze the differences in transcriptome expression among the subsets, we examined their differentially expressed genes. The top upregulated genes in 16 subsets are presented in a heatmap (Figure 2F).

Macrophage subsets specific for brain metastasis of lung cancer

To further elucidate subset differences among lung cancer brain metastases, lung cancer primary lesions, and normal brain tissues, the correlation analysis was performed on subsets from different groups. Initially, based on the classification of normal lung tissue, lung cancer tissue, normal brain tissue, and brain metastasis tissue of lung cancer, we illustrated the cluster distribution of all cells in the dataset (Figure 3A). Subsequently, we visually compared the proportion of total macrophages in the four groups using a pie chart (Figure 3B). The results indicated that macrophages from normal brain tissue were the least, potentially due to challenges in collecting normal brain tissue samples.

Figure 3 Macrophage subtype analysis of different tissue samples. (A) UMAP shows cells from normal lung tissue, normal brain tissue, primary tumor of lung cancer, and brain metastasis of lung cancer. (B) Pie chart compares the number and proportion of cells in different groups. (C) UMAP cluster plot shows the cluster distribution of all macrophage clusters in different tissues. (D) Histogram counts the proportion of all clusters in each group. (E) UMAP plot shows the expression of SPP1 in all cells. (F) UMAP shows TFF3 expression in all cells. BM, brain metastasis; NB, normal brain; NL, normal lung; PT, primary tumor; UMAP, uniform manifold approximation and projection.

To clearly compare the type and distribution of each subset within each group, we presented the clustering of all cell subsets in the four groups, separately (Figure 3C). The results revealed that many macrophage subsets exhibited tissue specificity, with subsets 11 and 13 almost exclusively present in lung cancer brain metastases or normal brain tissues, respectively. To visually compare the composition differences of macrophage subsets across all groups, the histogram was employed to depict the proportions of subsets in the four groups (Figure 3D). Subset distribution analysis showed that subsets 2 and 11 were predominantly concentrated in brain metastasis tissues of lung cancer; subsets 0, 5, 6, 8, and 10 were concentrated in normal lung tissues; subsets 1 and 12 were concentrated in lung cancer tissue cells, and subset 13 was present in normal brain tissues. Additionally, macrophage subsets 2 and 11 were marked with red boxes in the figure, indicating a significant increase in their proportion in brain metastasis tissues of lung cancer. This observation may be closely related to the pathological mechanism of brain metastasis from lung cancer and warrants further investigation. Based on the differential gene analysis of subsets, the most up-regulated gene was found to be secreted phosphoprotein 1 (SPP1) in subset 2 and trefoil factor family 3 (TFF3) in subset 11. To analyze the specificity of these two genes for subsets 2 and 11, we illustrated the expression distribution of SPP1 and TFF3 in all cells using UMAP (Figure 3E,3F). Clearly, SPP1 was predominantly expressed in subset 2, while TFF3 was mainly expressed in subset 11. Given this characteristic, subset 2 and subset 11 were referred to SPP1⁺ macrophages and TFF3⁺ macrophages, respectively.

Thus, we have identified two macrophage subsets associated with brain metastasis from lung cancer. Further analysis of these two subsets may enhance our understanding of the role of macrophages in the brain metastasis of lung cancer and unveil potential targets for combating brain metastasis from lung cancer.

Differential gene analysis of tumor-associated macrophage subsets in brain metastases from lung cancer

Through the integration of multiple single-cell datasets of macrophages, we have identified two macrophage subsets associated with brain metastasis of lung cancer, namely SPP1⁺ macrophages and TFF3⁺ macrophages. To delve into the underlying mechanisms and molecular principles of these two subsets in the process of brain metastasis from lung cancer, we employed a volcano plot to visualize the upregulated and downregulated genes in these subsets. Upregulated genes are depicted by red circles, while downregulated genes are represented by blue circles. The higher the value on the Y-axis, the smaller the P value; the larger the absolute value on the X-axis, the greater the difference in fold change. According to the differential gene analysis of the SPP1⁺ macrophage subset (Figure 4A), numerous cancer-promoting genes were significantly up-regulated, such as SPP1, C-C motif chemokine ligand 2 (CCL2), macrophage migration inhibitory factor (MIF), amphiregulin (AREG), ribonuclease A family member 1 (RNASE1), metallothionein 1X (MT1X), tissue inhibitor of metalloproteinases 1 (TIMP1), heme oxygenase 1 (HMOX1), metallothionein 1G (MT1G), and more. Notably, SPP1 is a secreted phosphoglycoprotein highly expressed in various tumors (18). Its non-RGD binding to CD44 at the C-terminal promotes the activation of signaling pathways (19), potentially leading to immune escape (20). Macrophage MIF, a homotrimeric protein, can promote the migration and angiogenesis of tumor cells as an autocrine factor (21). Its functional activities include down-regulation of tumor-inhibiting factors, up-regulation of cyclooxygenase-2 (COX-2) and prostaglandin E2 (PGE2), effective induction of angiogenesis, and promotion of tumor growth, proliferation, and invasion. AREG, as a membrane-anchored precursor protein, participates in near-secretory signaling of adjacent cells. It belongs to the epidermal growth factor (EGF) family of ligands, mediating cell survival, proliferation, and operation by binding to epidermal growth factor teceptor (EGFR) in both epithelial and mesenchymal cells, thus triggering various signal cascades. AREG serum levels and primary tumor expression are positively correlated with cancer progression, including invasion and metastasis (22). It has been demonstrated that AREG overexpression promotes tumor cell proliferation, invasion, and migration, induces DNA synthesis, prevents apoptotic cell death, and facilitates cell cycle progression (23).

Figure 4 Characterization of macrophage clusters 2 and 11 associated with brain metastasis from lung cancer. (A) Volcano plot shows upregulated and downregulated genes in cluster 2 (SPP1⁺ macrophage subset). (B) Volcano plot shows upregulated and downregulated genes in cluster 11 (TFF3⁺ macrophage subset). (C) Dot plot shows enrichment of signaling pathways in cluster 2 (SPP1⁺ macrophage subset). (D) Dot plot shows enrichment of signaling pathways in cluster 11 (TFF3⁺ macrophage subset). FC, fold change.

The TFF3⁺ macrophage subset also exhibited significant upregulation of several oncogenes, including TFF3, trefoil factor 1 (TFF1), C-C motif chemokine ligand 4 (CCL4), C-C motif chemokine ligand 3 (CCL3), anterior gradient 2 (AGR2), BPI fold-containing family A member 1 (BPIFA1), BPI fold-containing family B member 2 (BPIFB2), C-C motif chemokine ligand 3 like 3 (CCL3L3), C-C motif chemokine ligand 4 Like 2 (CCL4L2), and secretoglobin family 3A member 1 (SCGB3A1) (Figure 4B). TFF3 is a secreted protein with regulatory functions, known to promote epithelial-mesenchymal transformation by activating the MAPK/ERK pathway. The process of epithelial-mesenchymal transformation is a key pathway in tumor occurrence, development, and metastasis. It is speculated that TFF3 promotes the metastasis and proliferation of tumor cells due to its biological behavior (24). TFF1, a member of the trefoil factor family, is recognized as an important oncogene. A recent study has reported that TFF1 can promote tumor cell invasion, inhibit apoptosis, and stimulate tumor angiogenesis (25).

CCL2, CCL3, and CCL4, as the CC family members, were markedly overexpressed in the aforementioned two macrophage subsets. Some studies have observed the overexpression of CCL2 in NSCLC tissues, and its high levels were associated with strong infiltration in the lung tissues of NSCLC patients. The enhanced expression of CCL2 has been linked to increased metastasis of NSCLC cells in nude mice. CCL3, also known as macrophage inflammatory protein-1a (MIP-1a), serves as a chemokine for monocytes and lymphocytes. It plays a crucial role as a key regulator in the immune microenvironment, mediating the transport of immune cells in inflammation and cancer (26). CCL4 has the ability to recruit regulatory T cells and tumorigenic macrophages, impacting other resident cells in the tumor microenvironment, such as fibroblasts and endothelial cells. This recruitment promotes their tumorigenic potential, thereby facilitating the occurrence and progression of tumors.

Subsequently, to analyze the distinctions in signaling pathways between macrophage subsets associated with brain metastasis from lung cancer and macrophages in other normal tissues or lung cancer tissue, we analyzed the signaling enrichment pathways of differential genes in the SPP1⁺ macrophage and TFF3⁺ macrophage subsets. The findings revealed that differential genes in SPP1⁺ macrophages and TFF3⁺ macrophages were similarly enriched in the NOD-like receptor signaling pathway, Chemokine signaling pathway, NF-κB signaling pathway, and TNF signaling pathways (Figure 4C,4D). These pathways were significantly activated in both brain metastasis-associated macrophages. Upon activation, NOD-like receptors induced the release of various inflammatory factors, fostering the production of cytokines and chemokines, and contributing to cell proliferation and angiogenesis through a series of signaling pathways. Chemokines within the chemokine pathway can regulate tumor immune response, tumor cell proliferation, and invasion. The NF-κB pathway, consisting of classical and non-classical pathways, is activated by various stimuli and regulates the expression of various pro-inflammatory genes. Dysregulation of NF-κB activity could lead to inflammation-related diseases and cancer. TNF, as a key regulator of the TNF signaling pathway, promotes tumor growth by enhancing tumor cell proliferation, migration, invasion, and angiogenesis.

Discussion

This study analyzed and integrated single-cell transcriptome data related to brain metastasis from lung cancer to compare the differences in cell components and alterations in the tumor microenvironment among normal lung tissue, lung cancer tissue, normal brain tissue, and brain metastasis from lung cancer. Our focus was on the distinctions in tumor-infiltrating macrophages in brain metastasis of lung cancer. We observed a significant increase in the cell proportion of SPP1⁺ macrophages and TFF3⁺ macrophages in brain metastasis of lung cancer. Upon analyzing the differentially expressed genes of these two subsets relative to other subsets, we discovered that they not only markedly increased the cell proportion in brain metastasis tissues but also upregulated a series of factors that promoted tumor growth, angiogenesis, and metastasis. Specifically, SPP1⁺ macrophages exhibited high expressions of SPP1, CCL2, MIF, AREG, while TFF3⁺ macrophages exhibited high expressions of TFF3, TFF1, CCL4, CCL3, AGR2, CCL3L3, CCL4L2. Further pathway enrichment analysis revealed that the NOD-like receptor signaling pathway, Chemokine signaling pathway, NF-κB signaling pathway, and TNF signaling pathway were highly activated in these subsets. These signaling pathways are widely reported to contribute to tumor progression and metastasis.

TAMs are highly heterogeneous, with complex and diverse functions, including the promotion of tumor proliferation and metastasis, neovascularization, and the induction of an inhibitory immune microenvironment (27). TAMs are generally categorized into M1-type macrophages and M2-type macrophages. In the tumor microenvironment, M1 macrophages mediate pathogen defense, secreting inducible nitric oxide synthase (ions) and pro-inflammatory cytokines. On the other hand, M2 macrophages are involved in tissue repair, promoting angiogenesis and facilitating tumor progression (28). However, the classification of M1 and M2 is insufficient to capture the full heterogeneity of macrophages in the tumor microenvironment. In our study, SPP1⁺ macrophages and TFF3⁺ macrophages expressed high levels of cytokines typically secreted by M1 or M2 macrophages, such as CCL2, CCL3, and CCL4. CCL2 binds to its receptor CCR2, triggering intracellular signal transduction in tumors and other cell types. The CCL2/CCR2 signaling pathway supports the proliferation and survival of tumor cells, induces migration and invasion of cancer cells, stimulates inflammation and angiogenesis, and promotes epithelial-mesenchymal transition (EMT) and the expression of MMP2 and MMP9 to enhance cancer cell invasion. TAMs recruited by CCL2 and myeloid-derived suppressor cells (MDSCs) induce angiogenic switching and inhibit immune-mediated attacks on cancer cells (29). Our study reveals that SPP1⁺ macrophages express high levels of CCL2, suggesting their potential contribution to the CCL2 in the tumor metastasis microenvironment. This subset may promote brain metastasis from lung cancer by inducing epithelial-mesenchymal transformation of NSCLC cells. Additionally, CCL3 has been implicated in the development of various tumors and is closely associated with immune surveillance and tolerance. It serves as a prognostic biomarker in solid and hematological malignant tumors (30). High levels of CCL2 and CCL3 are linked to increased infiltration of regulatory T cells (Tregs), TAMs, and MDSCs within tumors, enabling tumor cells to evade immune surveillance (31). Chemokine (C-C motif) ligand 4 (CCL4), also known as macrophage inflammatory protein-1 (MIP-1b), plays a crucial role in immune response, inflammation, and tumor progression. A growing body of evidence suggests that CCL4 controls various tumor-related functions, including proliferation, invasion, metastasis, and angiogenesis (32,33). Elevated CCL4 concentrations predict adverse outcomes in patients with lung adenocarcinoma (33-35). These chemokines may serve as essential mediators for macrophages associated with brain metastasis from lung cancer to promote tumor cell migration. Investigating their sources is crucial for a deeper understanding of the mechanism underlying brain metastasis from lung cancer.

SPP1, highly expressed in the macrophage subset associated with brain metastasis from lung cancer, has captured our attention. SPP1 exhibits differential expression between tumor and normal tissues and is notably upregulated in various malignant tumors. High expression of SPP1 plays a pivotal role in tumor initiation and metastasis. During the invasion of tumor cells into the extracellular matrix, SPP1 promotes the high expression of matrix metalloproteinases (MMPs) in tumor cells through the NF-κB-dependent signaling pathway. This cascade leads to the degradation of the cell basement membrane and extracellular matrix, facilitating tumor infiltration and metastasis, ultimately contributing to poor prognosis (36). SPP1, expressed by bone marrow cells and tumor cells, acts as an immune checkpoint by binding to CD44 on the surface of T lymphocytes, inhibiting T lymphocyte proliferation, and conferring immune tolerance to host tumors (37). SPP1 is been closely associated with increased pleural effusion volume and intrapleural tumor dissemination in NSCLC (37,38). When secreted by immune cells, SPP1 promotes angiogenesis by binding to integrin avp3 and activating PI3K and ERK pathways on endothelial cells, thereby stimulating the production of vascular endothelial growth factors (39). This, in turn, promotes the occurrence, progression, and metastasis of lung tumors. It has been indicated that SPP1 expression on TAMs is linked to the poor prognosis and chemotherapy resistance of lung adenocarcinoma. SPP1 serves as a potentially useful marker for detecting monocyte-derived TAMs (40), providing a theoretical basis for therapeutic strategies targeting TAMs.

Another macrophage subset associated with brain metastasis from lung cancer exhibits high expression of TFF3 and TFF1. TFF3 plays a pivotal role in the occurrence, proliferation, differentiation, migration, invasion, and apoptosis of various human malignant tumors (41). Knocking out TFF3 not only inhibits the proliferation, migration, and invasion of tumor cells but also significantly enhances the sensitivity of tumor cells to radiotherapy and chemotherapy (42). TFF3 is highly expressed in the serum, lung tissues, and lung cancer cell lines in patients with lung squamous cell carcinoma, lung adenocarcinoma, and small cell lung cancer (43). TFF1, on the other hand, can promote tumor cell invasion, inhibit apoptosis, and contribute to tumor angiogenesis (25). Therefore, the roles of TFF3 and TFF1 in tumor progression and metastasis deserve increased attention. This paper describes the sources of macrophages, and the results also indicate their potential close associations with brain metastasis from lung cancer. Further investigation is worthwhile.

Our research is founded on the analysis of transcription data at the single-cell level, which has inherent limitations. While some new macrophage subsets implicated in brain metastasis from lung cancer have been identified, their specific correlations and molecular mechanisms at the protein level necessitate further investigation and validation with a larger dataset. Such analysis is crucial for refining treatment strategies for brain metastasis from lung cancer and can offer valuable insights for subsequent studies. Once the functions and mechanisms of macrophage subgroups identified in this study are clearly elucidated, our research will present new avenues and perspectives for clinical strategies targeting macrophages. This includes bispecific antibody drugs that target SPP1 or TFF3 alongside CD47, a widely developed drug target on the surface of macrophages across various tumor types. Furthermore, treatment strategies focusing on SPP1 or TFF3 can be integrated with currently approved drugs for brain metastasis from lung cancer, such as platinum-based chemotherapy or third-generation EGFR-TKI targeted therapies.

Conclusions

In conclusion, our study delved into the role and heterogeneity of macrophages in brain metastasis from lung cancer. We identified distinct macrophage subsets in normal lung tissue, lung cancer tissue, normal brain tissue, and tissue of brain metastasis from lung cancer. Two macrophage subsets were found to be closely associated with tumor promotion and brain metastasis from lung cancer, and their potential functional mechanisms were inferred through differential gene expression. This contributes to a better understanding of the mechanisms underlying brain metastasis from lung cancer and aids in the development of drug targets. Furthermore, it offers a novel perspective for clinical strategies targeting macrophages or the tumor microenvironment to treat brain metastasis from lung cancer.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2489/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2489/prf

Funding: This work was supported by Key Construction Disciplines of Provincial and Municipal Co construction of Zhejiang (No. 2023-SSGJ-002) Jiaxing Key Laboratory of Precision Treatment for Lung Cancer.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-2024-2489/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Martínez-Espinosa I, Serrato JA, Ortiz-Quintero B. MicroRNAs in Lung Cancer Brain Metastasis. Int J Mol Sci 2024;25:10325. [Crossref] [PubMed]
2. Kim SY, Park HS, Chiang AC. Small Cell Lung Cancer: A Review. JAMA 2025;333:1906-17. [Crossref] [PubMed]
3. Zhang J, Jin J, Ai Y, et al. Differentiating the pathological subtypes of primary lung cancer for patients with brain metastases based on radiomics features from brain CT images. Eur Radiol 2021;31:1022-8. [Crossref] [PubMed]
4. Brower JV, Robins HI. Erlotinib for the treatment of brain metastases in non-small cell lung cancer. Expert Opin Pharmacother 2016;17:1013-21. [Crossref] [PubMed]
5. Guo Z, Song J, Hao J, et al. M2 macrophages promote NSCLC metastasis by upregulating CRYAB. Cell Death Dis 2019;10:377. [Crossref] [PubMed]
6. He BP, Wang JJ, Zhang X, et al. Differential reactions of microglia to brain metastasis of lung cancer. Mol Med 2006;12:161-70. [Crossref] [PubMed]
7. Sanchez LR, Borriello L, Entenberg D, et al. The emerging roles of macrophages in cancer metastasis and response to chemotherapy. J Leukoc Biol 2019;106:259-74. [Crossref] [PubMed]
8. Gao J, Liang Y, Wang L. Shaping Polarization Of Tumor-Associated Macrophages In Cancer Immunotherapy. Front Immunol 2022;13:888713. [Crossref] [PubMed]
9. Lavin Y, Kobayashi S, Leader A, et al. Innate Immune Landscape in Early Lung Adenocarcinoma by Paired Single-Cell Analyses. Cell 2017;169:750-765.e17. [Crossref] [PubMed]
10. Li R, Tong R, Zhang Z, et al. Single-cell sequencing analysis and transcriptome analysis constructed the macrophage related gene-related signature in lung adenocarcinoma and verified by an independent cohort. Genomics 2022;114:110520. [Crossref] [PubMed]
11. Hafemeister C, Satija R. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome Biol 2019;20:296. [Crossref] [PubMed]
12. Hao Y, Hao S, Andersen-Nissen E, et al. Integrated analysis of multimodal single-cell data. Cell 2021;184:3573-3587.e29. [Crossref] [PubMed]
13. Laughney AM, Hu J, Campbell NR, et al. Regenerative lineages and immune-mediated pruning in lung cancer metastasis. Nat Med 2020;26:259-69. [Crossref] [PubMed]
14. Gomes-Duarte A, Venø MT, de Wit M, et al. Expression of Circ_Satb1 Is Decreased in Mesial Temporal Lobe Epilepsy and Regulates Dendritic Spine Morphology. Front Mol Neurosci 2022;15:832133. [Crossref] [PubMed]
15. Nieto P, Elosua-Bayes M, Trincado JL, et al. A single-cell tumor immune atlas for precision oncology. Genome Res 2021;31:1913-26. [Crossref] [PubMed]
16. Kim N, Kim HK, Lee K, et al. Single-cell RNA sequencing demonstrates the molecular and cellular reprogramming of metastatic lung adenocarcinoma. Nat Commun 2020;11:2285. [Crossref] [PubMed]
17. Cesur MF, Basile A, Patil KR, et al. A new metabolic model of Drosophila melanogaster and the integrative analysis of Parkinson's disease. Life Sci Alliance 2023;6:e202201695. [Crossref] [PubMed]
18. Lamort AS, Giopanou I, Psallidas I, et al. Osteopontin as a Link between Inflammation and Cancer: The Thorax in the Spotlight. Cells 2019;8:815. [Crossref] [PubMed]
19. Gao C, Mi Z, Guo H, et al. Osteopontin regulates ubiquitin-dependent degradation of Stat1 in murine mammary epithelial tumor cells. Neoplasia 2007;9:699-706. [Crossref] [PubMed]
20. Pang X, Gong K, Zhang X, et al. Osteopontin as a multifaceted driver of bone metastasis and drug resistance. Pharmacol Res 2019;144:235-44. [Crossref] [PubMed]
21. Sakellariou GT, Iliopoulos A, Konsta M, et al. Serum levels of Dkk-1, sclerostin and VEGF in patients with ankylosing spondylitis and their association with smoking, and clinical, inflammatory and radiographic parameters. Joint Bone Spine 2017;84:309-15. [Crossref] [PubMed]
22. Busser B, Sancey L, Brambilla E, et al. The multiple roles of amphiregulin in human cancer. Biochim Biophys Acta 2011;1816:119-31. [Crossref] [PubMed]
23. Jiang J, Zhao W, Tang Q, et al. Over expression of amphiregulin promoted malignant progression in gastric cancer. Pathol Res Pract 2019;215:152576. [Crossref] [PubMed]
24. Lin X, Zhang H, Dai J, et al. TFF3 Contributes to Epithelial-Mesenchymal Transition (EMT) in Papillary Thyroid Carcinoma Cells via the MAPK/ERK Signaling Pathway. J Cancer 2018;9:4430-9. [Crossref] [PubMed]
25. Chaiyarit P, Jaresitthikunchai J, Phaonakrop N, et al. Proteolytic effects of gingipains on trefoil factor family peptides. Clin Oral Investig 2018;22:1009-18. [Crossref] [PubMed]
26. Mitchell DA, Batich KA, Gunn MD, et al. Tetanus toxoid and CCL3 improve dendritic cell vaccines in mice and glioblastoma patients. Nature 2015;519:366-9. [Crossref] [PubMed]
27. Anderson NR, Minutolo NG, Gill S, et al. Macrophage-Based Approaches for Cancer Immunotherapy. Cancer Res 2021;81:1201-8. [Crossref] [PubMed]
28. Murray PJ, Wynn TA. Protective and pathogenic functions of macrophage subsets. Nat Rev Immunol 2011;11:723-37. [Crossref] [PubMed]
29. Yumimoto K, Sugiyama S, Mimori K, et al. Potentials of C-C motif chemokine 2-C-C chemokine receptor type 2 blockers including propagermanium as anticancer agents. Cancer Sci 2019;110:2090-9. [Crossref] [PubMed]
30. Burger JA, Quiroga MP, Hartmann E, et al. High-level expression of the T-cell chemokines CCL3 and CCL4 by chronic lymphocytic leukemia B cells in nurselike cell cocultures and after BCR stimulation. Blood 2009;113:3050-8. [Crossref] [PubMed]
31. Chanmee T, Ontong P, Konno K, et al. Tumor-associated macrophages as major players in the tumor microenvironment. Cancers (Basel) 2014;6:1670-90. [Crossref] [PubMed]
32. Lien MY, Tsai HC, Chang AC, et al. Chemokine CCL4 Induces Vascular Endothelial Growth Factor C Expression and Lymphangiogenesis by miR-195-3p in Oral Squamous Cell Carcinoma. Front Immunol 2018;9:412. [Crossref] [PubMed]
33. Mollica Poeta V, Massara M, Capucetti A, et al. Chemokines and Chemokine Receptors: New Targets for Cancer Immunotherapy. Front Immunol 2019;10:379. [Crossref] [PubMed]
34. Sivina M, Werner L, Rassenti L, et al. Dynamic changes in CCL3 and CCL4 plasma concentrations in patients with chronic lymphocytic leukaemia managed with observation. Br J Haematol 2018;180:597-600. [Crossref] [PubMed]
35. Wang Y, Liu T, Yang N, et al. Hypoxia and macrophages promote glioblastoma invasion by the CCL4-CCR5 axis. Oncol Rep 2016;36:3522-8. [Crossref] [PubMed]
36. Wai PY, Kuo PC. The role of Osteopontin in tumor metastasis. J Surg Res 2004;121:228-41. [Crossref] [PubMed]
37. Klement JD, Paschall AV, Redd PS, et al. An osteopontin/CD44 immune checkpoint controls CD8+ T cell activation and tumor immune evasion. J Clin Invest 2018;128:5549-60. [Crossref] [PubMed]
38. LaFleur MW, Muroyama Y, Drake CG, et al. Inhibitors of the PD-1 Pathway in Tumor Therapy. J Immunol 2018;200:375-83. [Crossref] [PubMed]
39. Dai J, Peng L, Fan K, et al. Osteopontin induces angiogenesis through activation of PI3K/AKT and ERK1/2 in endothelial cells. Oncogene 2009;28:3412-22. [Crossref] [PubMed]
40. Matsubara E, Yano H, Pan C, et al. The Significance of SPP1 in Lung Cancers and Its Impact as a Marker for Protumor Tumor-Associated Macrophages. Cancers (Basel) 2023;15:2250. [Crossref] [PubMed]
41. Kannan N, Kang J, Kong X, et al. Trefoil factor 3 is oncogenic and mediates anti-estrogen resistance in human mammary carcinoma. Neoplasia 2010;12:1041-53. [Crossref] [PubMed]
42. Perera O, Evans A, Pertziger M, et al. Trefoil factor 3 (TFF3) enhances the oncogenic characteristics of prostate carcinoma cells and reduces sensitivity to ionising radiation. Cancer Lett 2015;361:104-11. [Crossref] [PubMed]
43. Qu Y, Yang Y, Ma D, et al. Increased trefoil factor 3 levels in the serum of patients with three major histological subtypes of lung cancer. Oncol Rep 2012;27:1277-83. [Crossref] [PubMed]