CD9-positive tumor-associated macrophages promote renal cell carcinoma progression by activating the orphan nuclear receptor Nor1/Nr4a3

Abstract

Tumor growth, metastasis, and treatment response in renal cell carcinoma (RCC) are believed to be regulated by the tumor microenvironment (TME). However, the mechanisms underlying the genomic, transcriptomic, and epigenetic alterations in clear cell RCC progression remain incompletely understood. This study analyzed single-cell RNA sequencing data from public databases, including tissue samples from three patients with RCC and three matched pairs of immune cells from tumor tissue and peripheral blood to screen macrophages with differential expression in the TME. We found that a CD9⁺ macrophage population was identified in the RCC tumor tissue, suggesting a role for CD9-expressing macrophages in RCC progression. The function of these macrophages was further investigated through in vitro and in vivo experiments involving CD9 knockdown or overexpression. CD9 expression was elevated in RCC tissues compared to adjacent non-cancerous tissues. CD9 knockdown considerably inhibited the TME-induced macrophages (M2), while CD9-overexpressing macrophages markedly promoted RCC cell growth in vitro and in vivo. Moreover, this phenomenon may be associated with the orphan nuclear receptor NOR1//Nr4a3 activation by CD9. These findings indicated that macrophages in the TME exhibited distinct expression patterns during RCC progression and that CD9⁺ macrophages promoted RCC development, suggesting that CD9 may be a potential therapeutic target for inhibiting RCC progression.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-025-04060-x.

Introduction

Renal cell carcinoma (RCC) is a primary malignancy and the leading cause of kidney cancer-related deaths. While radical nephrectomy remains the standard approach for localized RCC, many individuals present with advanced or metastatic disease, complicating management [1]. Although targeted agents and immune checkpoint inhibitors have extended survival in advanced RCC, patient outcomes remain poor compared to those with early-stage disease [2]. Moreover, selecting the most effective treatment options remains a challenge for clinicians. Therefore, investigating the underlying mechanisms of RCC is crucial for identifying potential therapeutic targets and improving patient outcomes.

Notably, RCC is characterized by substantial immune cell infiltration [1]. The effectiveness of harnessing adaptive immunity through immune checkpoint blockade therapy in improving patient survival underscores the importance of the immune microenvironment in RCC [3]. However, bulk sequencing, which is limited in resolving distinct immune cell populations, cannot accurately characterize this microenvironment. A single-cell immune map of RCC generated by mass cytometry reveals diverse immune populations within tumors. Additionally, advances in single-cell RNA sequencing (scRNA-seq) have transformed our knowledge of the complex and dynamic interactions between tumor and immune cells [4].

Macrophages are an essential leukocyte component that infiltrate the tumor microenvironment (TME) [5, 6]. They are broadly classified into two phenotypes: M1 and M2. M1 macrophages, typically activated by Lipopolysaccharide (LPS) [7], promote antitumor responses by producing cytokines, recruiting immune cells, and directly phagocytosing tumor cells [8]. Conversely, M2 macrophages, typically induced by interleukin-4 (IL-4), support tumor growth and metastasis by facilitating angiogenesis, remodeling the extracellular matrix, recruiting leukocytes, and suppressing immune responses [9]. Importantly, macrophages in TME display a spectrum of phenotypes beyond the classic M1/M2 classification, reflecting their functional plasticity. They may lie between or outside this spectrum. Regardless of the polarization state, macrophages retain the capacity for plasticity and can switch between phenotypes based on the microenvironment. Consequently, altering macrophage phenotypes within the TME from immunosuppressive to immunoprotective is being explored for therapeutic applications [9].

CD9 (TSPAN29) is a tetraspanin protein with four transmembrane domains [10, 11] and is implicated in various diseases, including cancer, although its precise role remains unclear [12]. CD9 is crucial for cell adhesion and functions in leukocytes, endothelial cells, hematopoiesis, and blood clotting [13]. It also contributes to sperm–egg fusion, neurite extension, and muscle cell formation [14]. Bioinformatic assessment of CD9 regulatory regions highlighted E2F, NFkB, SP1, and STAT3 as key transcription factors linked to cancer development and prognosis [15]. The tetraspanin CD9 exhibits tumor-promoting and tumor-suppressing effects in cancer [16], based on the context. For instance, in aggressive triple-negative breast cancer cells, CD9 knockdown alters membrane protrusions and reduces tumor growth and metastasis in mouse models [17]. Elevated CD9 levels are linked to an epithelial phenotype and improved recurrence-free survival [4]. In breast cancer, type IV collagen triggers CD9 expression and cell migration via a DDR1-dependent pathway, an effect not observed in non-tumorigenic cells [18]. To date, the functional relevance of CD9 in RCC remains largely unexplored, and the biology of CD9⁺ macrophages within the RCC micro-environment is essentially uncharted territory. Research indicates that CD9-rich microdomains inhibit LPS-induced receptor assembly by blocking CD14 clustering, whereas reduced CD9 in macrophages increases inflammation in mice [19]. According to recent studies, histone deacetylase (HDAC) inhibitors block M2 polarization but not M1 polarization [20]. Another experiment demonstrated that HDAC inhibited the downregulation of CD9 expression in macrophages [21]. Epigenetic modifications can alter DNA and histone structure, playing a crucial role in gene regulation. In normal tissues, epigenetic mechanisms can provide genomic balance and chromosomal regulation. Conversely, enhanced metabolism and several other factors can induce abnormal epigenetic changes, possibly leading to malignant transformation [22]. Notably, HDAC can catalyze the removal of acetyl groups from the amino-terminal lysine residues of histones, a notable feature of epigenetic regulation that alters cancer growth [23]. Epigenetic modifications drive phenotypic changes in cancer and immune cells [24]. This evidence suggests a potential mechanism by which tumor cells promote CD9 expression in TAMs, with metabolic changes in tumors likely affecting CD9 expression in TAMs by inducing epigenetic changes, particularly HDAC-mediated epigenetic regulation. However, HDAC inhibitors vary in their activities depending on the specific HDAC targeted, the macrophage source, the in vivo assays used, and even the HDAC concentrations applied, all of which can influence macrophage polarization [25]. This suggests that different HDACs may have distinct mechanisms of action on CD9.

Neuron-derived orphan receptor 1 (NOR1, NR4A3) is an immediate-early transcription factor of the NR4A nuclear-receptor family. Lacking a known endogenous ligand, it is considered “orphan.” NOR1 is widely expressed-neurons, vascular smooth muscle, T cells, dendritic cells, cardiomyocytes, hepatocytes, pancreatic islets and multiple tumor types-and was originally cloned from apoptotic neurons, where it governs synaptic growth and cell migration. Up-regulated NOR1 participates in diverse pathophysiological events: it modulates apoptosis and tissue regeneration, orchestrates neuronal development and contextual fear memory, fuels inflammation, drives vascular smooth-muscle proliferation, fine-tunes insulin secretion, and promotes tumorigenesis [26, 27]. Despite mounting data, an in-depth portrait of NOR1 biology has yet to be assembled in the online literature. Studies show that HDAC inhibitors robustly up-regulate NR4A3 in leukaemia cells; the ensuing surge of NOR1 protein drives transcription of pro-apoptotic genes (PUMA, TRAIL, Bid, Bik) via promoter-specific histone hyper-acetylation, tipping the balance toward programmed cell death [28]. In acinar cell carcinoma, specific recombination phenotype [t (4; 9) (q13; q31)] may increase the expression level of NOR1, thereby promoting the progression of salivary gland acinar cell carcinoma by upregulating the NOR1 regulatory gene [29]. Some other findings establish NR4A3 factors as key drivers of CAR-T hyporesponsiveness in solid tumours and point to NR4A inhibition (genetic deletion, siRNA, or small-molecule antagonists) as a promising strategy to enhance CAR-T therapy for solid cancers [30].

This study conducted single-cell RNA sequencing of immune cells in RCC tumor tissue and compared the differences between RCC and peripheral blood immune cells. Subsequently, we validated the immune environment’s heterogeneity in RCC and its impact on RCC development through human tissue pathology, cell culture, and animal experiments. Based on our findings, CD9⁺ macrophages are vital in RCC progression through regulating the NOR1//Nr4a3 activation and may influence the response to immune or targeted therapy in clinical cohorts.

Materials and methods

Analysis of scRNA-seq data

We analyzed scRNA-seq data (GSE121636) from three patients with RCC using Seurat (version 4.0.5). Seurat objects were created for tumor-infiltrating and peripheral blood immune cells, and cells were filtered out based on gene count and mitochondrial or hemoglobin gene content. The datasets were then merged, normalized, and scaled, with 2,000 highly variable genes identified for further analysis. Principal component analysis was performed via RunPCA. Batch effects across samples were corrected using RunHarmony (version 0.1.0), with sample identity as the batch variable and default settings. Clustering was conducted at a 1 or 0.5 resolution using FindNeighbors and FindClusters, with UMAP employed to visualize cell populations. We identified distinct genes in cell clusters using the Wilcoxon rank-sum test with the FindAllMarkers function. Each cell cluster was then identified as a specific cell type based on classical marker genes in Table 1.

Table 1.

The classical marker genes of each cell cluster

Cell cluster	Classical marker genes
T Cells	CD3D, CD3E, CD8A, IL7R
B cells	CD19, CD79A, MS4A1, CD20
Plasma cells	IGHG1, MZB1, SDC1, CD79A
Monocytes and macrophages	CD68, CD163, CD14
NK Cells	FGFBP2, NKG7, CX3CR1

Clinical characteristics of patients and surgical specimens

This study included 20 RCC tumor specimens with pathologically confirmed clear cell RCC and their respective adjacent non-cancerous tissues from the Fourth Affiliated Hospital of China Medical University, with informed consent between December 2023 and December 2024. Patients with active inflammatory diseases and receiving immunosuppressive therapy were excluded. Table 2 summarizes clinical characteristics. The research strictly followed the guidelines stipulated by the Declaration of Helsinki and was sanctioned by the ethics committee of the Shengjing Hospital of China Medical University (Ethical Lot Number: 2024PS1753K).

Table 2.

Demographic and clinical characteristics of control and case group patients

Gender	Control group		Case group		P-value*
Sex, n (%)
Male	9	(45)	12	(60)	0.527
Female	11	(55)	8	(40)
Age, n (%)
≤ 65	20	(100)	13	(65)	0.008
> 65	0	(0)	7	(35)
Tumor size (cm), n (%)
≤ 4	8	(40)	0	(0)	0.000
4–7	8	(40)	0	(0)
> 7	4	(20)	0	(0)
T stage, n (%)
T 1–2	0		16	(80)	0.000
T 3–4	0		4	(20)
Lymph node metastasis, n (%)
Negative	20	(100)	17	(85)	0.231
Positive	0	(0)	3	(15)
Distant metastasis, n (%)
Negative	20	(100)	19	(95)	1.000
Positive	0	(0)	1	(5)

*p < 0.05, **p < 0.01, ***p < 0.001

Cell culture

Human monocytic leukemia (THP-1) cells and Mouse renal cancer cells (Renca) were obtained from the Chinese Academy of Sciences Cell Bank and cultured in RPMI 1640 medium with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (Biological Industries, Israel) at 37 °C in 5% CO₂. For experiments, cells were seeded in 6-well plates and differentiated into macrophages by treatment with 100 nM Phorbol-12-myristate-13-acetate (PMA) for 24 h. LPS, PMA, Poly-L-lysine (Sigma-Aldrich, USA), and Nigericin (InvivoGen, USA) were used as reagents. Plasmids (pcDNA 3.1+) and CD9 siRNA were obtained from GeneChem (China), and transfections were performed using Lipofectamine 3000 (Invitrogen) following the manufacturer’s instructions.

Bone marrow derived macrophage (BMDM)

Under aseptic conditions in a laminar-flow hood, tibiae and femora were dissected out and cleaned of surrounding muscle with fine scissors and forceps. The bones were transferred to a Petri dish containing sterile PBS, rinsed once, and moved to a second dish of fresh PBS. Both epiphyses were snipped off with scissors; a 1-mL syringe fitted with a 25-G needle was filled with PBS, inserted into each marrow cavity, and the marrow was flushed repeatedly until the bones turned white. The resulting marrow suspension was passed through a 70-µm cell strainer to remove debris and muscle fragments. Cells were pelleted at 500 × g for 5 min, resuspended, and subjected to ACK lysis buffer to remove erythrocytes. After a second spin (200 × g, 5 min), the pellet was resuspended in complete DMEM (10% FBS, 1% penicillin–streptomycin) supplemented with 10 ng/mL M-CSF and plated at 5 × 10⁵ cells/mL for culture.

Preparing conditioned medium (CM) from tumor cells

ACHN cells (ATCC) were maintained in RPMI 1640 medium with 10% FBS and 1% penicillin–streptomycin (50 mg/L penicillin, 100 µg/mL streptomycin) at 37 °C in a humidified incubator with 5% CO₂. Cells were used at approximately 70% confluence. For tumor-CM preparation, culture supernatant was collected after 24 h, centrifuged at 3,000 rpm for 10 min at 4 °C, filtered through a 0.22 μm membrane (Millipore), and stored at − 20 °C.

Preparation of macrophage CM and co-culture treatment

Cell culture media were harvested after a 12-h incubation period and centrifuged at 3,000 rpm for 10 min at 4 °C to prepare macrophage CM. To assess the impact of various macrophage conditions on ACHN cells, the ACHN cells were first washed and cultured with fresh medium or macrophage-derived CM for 24 h.

Western blot

Proteins were extracted following the manufacturer’s protocol and quantified using a bicinchoninic acid assay. Each sample (20 µg) was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes. After blocking with 5% skim milk for 1 h at 37 °C, membranes were incubated with primary antibodies overnight at 4 °C. After washing, secondary antibodies were applied for 1 h at 37 °C. Antibodies used included glyceraldehyde 3-phosphate dehydrogenase (1:500, CST), and CD9, CD68, CXCR4, NOR1, CD206, Arg1, N-Cadherin, E-Cadherin, and Vimentin (all 1:1000, Abcam).All Western blot images are representative of three biologically independent experiments (n = 3) performed on different days with independently prepared cell lysates. Densitometry data are expressed as mean ± SD of these three replicates.

Quantitative real-time PCR analysis (qRT-PCR)

Total RNA was isolated from cultured cells using TRIzol and reverse transcribed into cDNA. qRT-PCR was performed using TB Green on a Roche Light Cycler 96, with β-actin as the internal control. Cycling conditions included initial denaturation, 45 amplification cycles, and melting curve analysis to verify specificity. Relative gene expression was calculated using the 2^–ΔΔCT method, with primer sequences in Table 3.

Table 3.

Primer sequences used in real-time PCR analysis

Target gene (Mus)	Forward primer (5′−3′)	Reverse primer (5′−3′)
CD206	CTCTGTTCAGCTATTGGACGC	TGGCACTCCCAAACATAATTTGA
CD9	CCACCTGCAAGACCATCGAC	CTGGCGAGCCTTAGTTTGGAC
CD68	GCTTTGCAATCTCCCTGTTG	TTGATCCGGGTTCTTACCTG
Arg1	AACGGGAGGGTAACCATAAGC	TGATGCCCCAGATGGTTTTC
NOR1	AAGGATCCATGGATCAGGCCAA	TACTCGAGCCTATAACTCATCCA
CD80	GGAATACAACCAAGCAAGAGCA	TCTTGGGGCAAAGCAGTAGG
iNOS	CGCTACAACATCCTGGAG	ACATTCTGCTTCTGGAAACTA
MHC-II	GGACAAAGCCAACCTGGAAA	AGGACGTTGGGCTCTCTCAG
β-actin	GTGCTATGTTGCTCTAGACTTCG	ATGCCACAGGATTCCATACC

Migration and invasion assays

Migration and invasion assays were performed using Transwell inserts (Corning, New York, USA) with uncoated and matrix-coated membranes following the manufacturer’s guidelines. Following a 24-h incubation period, migrated and invaded cells were stained using a 0.5% crystal violet solution and visualized under a microscope.

Scratch assay

ACHN cells (6.0 × 10⁵) were plated 24 h before the assay. A scratch was made in the cell monolayer, and detached cells were removed by washing twice with a serum-free medium. Cell-free areas were then measured using ImageJ.

Tumor models

Four-week-old BALB/c athymic mice or immunocompetent mice were utilized based on institutional and national animal welfare guidelines. Each mouse received a subcutaneous injection of 5 × 10⁵ ACHN/Renca cells in 100 µL PBS. Three days after tumor establishment, transfected macrophages/BMDMs (5 × 10⁶ in 100 µL PBS) were administered via the tail vein. Tumor size and mouse weight were measured thrice weekly using calipers, and tumor volume was calculated as ½ × (length × width²). At study completion, tumors were fixed in 10% paraformaldehyde for histological analysis. Euthanasia was performed under pentobarbital sodium anesthesia by cervical dislocation.

Histopathological analysis

Tumor tissues were fixed in 10% paraformaldehyde for 24 h, dehydrated, and embedded in paraffin. Section (5 μm thick) were cut and dried at 37 °C, then stained with hematoxylin and eosin (H&E) to assess histology. Subsequently, xylene dewaxing was performed on the slides, followed by H&E staining based on the manufacturer’s guidelines.

Immunohistochemical and Immunofluorescence staining

Tissue sections were mounted, dried at 37 °C, dewaxed, and rehydrated through graded ethanol. After antigen retrieval, the sections were incubated with primary antibodies against CD9, CD68, and CXCR4, followed by biotin-conjugated secondary antibodies. Hematoxylin staining and standard dehydration were performed before sealing. For immunofluorescence, sections were labeled with fluorescent antibodies (CD206, F4/80), incubated in the dark, and visualized using a Nikon Eclipse 90i microscope.

TUNEL assay

The TUNEL assay was used to assess DNA fragmentation. After dewaxing and rehydration, the samples were fixed in 4% formaldehyde/PBS for 20 min. The Click-iT Plus TUNEL kit with Alexa Fluor 488 (Thermo Fisher) was used for detection. Nuclei were counterstained and mounted with Vectashield medium (Vector Laboratories). Ki-67 was employed as a proliferation marker (red fluorescence), while TUNEL-positive cells appeared green. Images were obtained with a fluorescence microscope.

Data analysis

All statistical analyses were conducted using R (version 4.1.2) and GraphPad Prism (version 8.0) software. Data are presented as mean ± standard deviation. Group comparisons were performed using a one-way analysis of variance with Tukey’s post hoc, Mann–Whitney U, or Wilcoxon rank-sum tests, as appropriate. Survival was analyzed using the log-rank test. Statistical significance was defined as *p < 0.05, with *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. Values above this threshold were considered non-significant.

Results

Construction of RCC immune cell atlas

We constructed an RCC immune cell atlas by performing an integrated analysis of scRNA-seq data from six samples. First, quality control measures were applied to the cell data, removing cells with mitochondrial genes and cells lacking gene copies (Fig. 1A). Next, the cells were scored based on their cell cycle (Fig. 1B). Cell clustering analysis was conducted to group cells, with the effects of different clustering resolutions on cell clustering evaluated (Fig. 1C). Using a cluster resolution of 1, we reduced the data dimensionality and labeled it into 27 cell clusters (Fig. 1D). Gene expression was compared between different cell clusters (Fig. 1E). Subsequently, the expression (Fig. 1F) and distribution status (Fig. 1G) of classical marker genes in each cell cluster were analyzed. Each cell cluster was identified as a specific cell type based on classical marker genes (Fig. 1H). Comparing cell clusters between different samples revealed that tumor immune cells contained heterogeneous cells compared to peripheral blood immune cells, particularly in the Mono/Macrophage cells (Fig. 1I).

Fig. 1.

Construction of the immune cell atlas using scRNA-seq data. (A) Seurat-based clustering analysis with resolutions ranging from 0 to 1. (B–D) UMAP representation of single cells grouped into distinct clusters. (E) Differential gene expression analysis across various clusters. (F) Marker gene expression for each identified cell type in RCC; dot size and color indicate the percentage (Pct. exp) and average proportional expression (Avg. exp. scale) of marker gene expression, respectively. (G) UMAP visualization of classical marker expression. (H) UMAP-based representation of immune cell populations. (I) UMAP visualization of immune cell populations within different samples. T denotes immune cells from the RCC tumor tissue, and P represents paired peripheral blood mononuclear cells

scRNA-seq analysis reveals differences in Mono/Macrophage composition between tumor and normal tissue immune cells

We extracted Mono/Macrophage cells from the RCC immune cell atlas and reclustered them to construct the RCC Mono/Macrophage atlas. At a clustering resolution of 0.3 (Fig. 2A), Mono/Macrophage cells were grouped into 12 subclusters (Fig. 2B). In this map, the T group was clustered on the lower left side of the UMAP map, including clusters 0, 2, 5, 6, 8, and 10. We analyzed heterogeneity genes among different clusters (Figs. 2C, D) and found that genes such as CD9, TNFAIP1, CXCR4, and NR4A3 (NOR1) were specifically expressed in tumors (Fig. 2E–G). Protein interaction networks revealed associations between these genes (except NR4A3) (Fig. 2H). To explore NR4A3’s relationship with the aforementioned genes, we analyzed the gene correlation between NR4A3 and other genes using the Cancer Genome Atlas database. We discovered a positive correlation between CD9 and NR4A3 (NOR1) (Fig. 2I). These findings suggest that CD9-positive macrophages may play a key regulatory role in RCC.

Fig. 2.

scRNA-seq analysis reveals differences in Mono/macrophage composition between tumor and normal tissue immune cells. (A) Seurat clustering results with resolution from 0–1. (B) UMAP representation of individual cells of different clusters. (C) Heatmap showing top-ranked differential genes (Wilcoxon test) in each macrophage subcluster. (D) Differential gene expression of daughter cell clusters in macrophages, where dot size and color represent percentage (Pct. exp) and average proportional expression (Avg. exp. scale) values of marker gene expression, respectively. (E) Expression of differential genes in different samples. (F) Expression of differential genes in subclusters of different clusters. (G) UMAP showed the expression of differential genes. (H) Protein interaction network. (I) Gene correlation analysis

Elevated CD9 expression in RCC tissue possibly associated with macrophage infiltration

Subsequently, we analyzed RCC tumor tissue, utilizing healthy tissue as a control (Fig. 3A). Immunohistochemical experiments revealed a higher expression of CD9 in RCC than in normal tissues. Concurrently, increased CD68 expression was observed in RCC, suggesting enhanced macrophage infiltration within the tumor tissue. This heightened macrophage infiltration might be attributed to the upregulation of the chemokine receptor CXCR4, which is associated with M2 macrophages (Figs. 3B and C). We also used double staining of CD9⁺CD68⁺ to present CD9⁺ macrophages, revealing the CD9⁺ macrophages were highly expressed in the tumor tissues (Fig. 3D). Based on these findings, both the protein expression and transcript levels of CD9, CD68, and CXCR4 were significantly elevated in RCC tissues (Figs. 3E–F).

Fig. 3.

High expression of CD9 in RCC tissue may be associated with macrophage infiltration. (A) H&E staining of RCC tumor tissue and tumor-adjacent tissue controls (400×). (B) Immunohistochemistry of CD9, CD68, and CXCR4 in RCC tumor tissues and the tumor-adjacent tissue controls (400×). (C) Semiquantitative immunohistochemistry. (D) Immunofluorescence staining for CD68 (green fluorescence) and CD9 (red fluorescence) (400×). (E) Protein expression of CD9, CD68, and CXCR4 in RCC tumor tissues and tumor-adjacent tissue controls. The data shown were the representative of three experiments.(F) Relative mRNA expression of CD9, CD68, and CXCR4 in RCC tumor tissues compared with the tumor-adjacent tissue controls. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

CD9 enhances macrophage M2 polarization by augmenting NOR1 expression and stimulating tumor growth

We first performed cell experiments to investigate the relationship between CD9 and tumor-associated macrophages (TAMs). Initially, we observed that the protein expression and transcriptional activity of CD9 and NOR1 were elevated in macrophages exposed to RCC cell CM compared to controls. The relative expression of CD206 and Arg1 proteins and mRNAs, which are markers of M2 polarization, was also increased in macrophages exposed to RCC cell-CM (Figs. 4A–C). These findings suggest that the tumor-associated environment stimulates M2 macrophage polarization and enhances CD9 and NOR1 expression.

Fig. 4.

CD9 Promotes M2 Polarization of Macrophages. (A, B) Relative protein expression levels of CD9, NOR1, CD206, and Arg1 in the macrophages. (C) Relative mRNA expression levels of CD9, NOR1, CD206, and Arg1 in the macrophages. (D, E) Relative protein expression levels of CD9, NOR1, CD206, and Arg1 in the macrophages with overexpressed or knocked down CD9. (F) Relative mRNA expression levels of CD9, NOR1, CD206, and Arg1 in macrophages with overexpressing or knocked down CD9. (G) Relative mRNA expression levels of CD80, iNOS, and MHC-II in the macrophages with overexpressed or knocked down CD9. (H, I) Relative protein expression levels of CD9, NOR1, CD206, and Arg1 in the macrophages under different treatment conditions. (J) Relative mRNA expression levels of CD9, NOR1, CD206, and Arg1 in the macrophages with the LPS disposed. (K) Relative mRNA expression levels of CD80, iNOS, and MHC-II in the macrophages with the LPS disposed. The data shown were the representative of three experiments. M denoted wild-type macrophages cultured alone; M + T(CM) represented the macrophages cultured in the tumor cell-conditioned medium; M + T(CM) + CD9 indicated the CD9-overexpressed macrophages cultured in tumor cell-conditioned medium; M + T(CM) + anti-CD9 signified the CD9-knockdown macrophages cultured in tumor cell-conditioned medium; M + LPS represented the LPS-stimulated macrophages (M1); M + LPS + CD9 indicated the LPS-stimulated macrophages overexpressing CD9. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

To investigate the impact of CD9 on TAMs, we generated macrophages with CD9 either overexpressed or knocked down. Exposure to tumor cell CM enhanced NOR1 protein expression and transcriptional activity in CD9-overexpressing macrophages, further promoting their M2 polarization. However, CD9-deficient macrophages displayed reduced NOR1 expression levels and were resistant to M2 polarization induced by the tumor environment (Figs. 4D–F). Additionally, we used CD80, iNOS, and MHC-II as M1 polarization markers and found that CD9 knockdown increased M1 differentiation by qRT-PCR (Fig. 4G). Moreover, CD9 overexpression facilitated LPS-induced M1 macrophages’ polarization to M2 (Figs. 4H–J) and reduced M1 polarization (Fig. 4K). Collectively, these experiments demonstrate that CD9 promotes M2 polarization of TAMs.

We subsequently examined the influence of macrophages on tumor growth in vitro. Initially, the proliferation and migration-related indicators N-Cadherin and Vimentin of the ACHN co-cultured with macrophages increased, while E-Cadherin decreased (Figs. 5A, B). Transwell and scratch experiments confirmed that TAMs enhanced tumor cell proliferation, invasion, and migration (Figs. 5C, D). Co-culturing CD9-overexpressing macrophages with tumor cells further enhanced tumor cell proliferation, invasion, and migration. Conversely, CD9 knockdown weakened the promotional effect of macrophages on tumor cell proliferation, invasion, and migration (Figs. 5E–H).

Fig. 5.

CD9 overexpression in macrophages promotes tumor growth. (A, B) Relative expression levels of V, E, and N proteins in tumor cells co-cultured with macrophages. (C) The TRANSWELL assay assessed the invasion and migration abilities of tumor cells. (D) Scratch assay for cells. (E, F) Relative expression levels of V, E, and N proteins in the tumor cells co-cultured with macrophages overexpressed or knocked down CD9. (G) The TRANSWELL assay evaluated the invasion and migration abilities of tumor cells. (H) Scratch assay for tumor cells. (I-J) Relative expression levels of V, E, and N proteins in the tumor cells co-cultured with CD9-overexpressed and NOR1-knockdown macrophages. (K) The TRANSWELL assay assessed the invasion and migration abilities of tumor cells. (L) Scratch assay for tumor cells. The data shown were the representative of three experiments. T denoted ACHN cells cultured alone; T + M(CM) represented ACHN cells co-cultured with macrophages; T + M(CM) + CD9 indicated the ACHN cells co-cultured with CD9-overexpressing macrophages; T + M(CM) + anti-CD9 signified the ACHN cells co-cultured with CD9-knockdown macrophages; T + M(CM) + CD9 + anti-NOR1 represented the ACHN cells co-cultured with both CD9-overexpressing and NOR1-knockdown macrophages. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

To demonstrate that CD9’s ability to promote tumor proliferation, invasion, and migration through macrophages is mediated by NOR1, we knocked down NOR1 in CD9-overexpressing macrophages. The results indicated that compared to CD9-overexpressing macrophages, macrophages with CD9 overexpression and NOR1 knockdown exhibited a reduced capacity for tumor proliferation, invasion, and migration (Figs. 5I–L). Overall, our in vitro findings confirm that CD9 enhanced the M2 polarization of TAMs by increasing NOR1 expression, thereby promoting tumor growth.

CD9⁺ macrophages promote tumor growth in mice

Previous findings have indicated that CD9-expressing macrophages promote tumor progression. To evaluate their impact on RCC cells, we employed a mouse xenograft model. Nude and immunocompetent mice were randomly divided into three groups, and a xenograft tumor transplantation mouse model was established by subcutaneously implanting ACHN/Renca cells. Following tumor transplantation, macrophages/BMDMs overexpressing or knocking down CD9 were injected the tail vein of mice, while wild-type macrophages were used in the control group. The results demonstrated that tumors in mice with CD9-overexpressing macrophages grew faster than those with wild-type macrophages, while CD9-knockdown macrophages delayed tumor progression (Fig. 6–7A).

Fig. 6.

CD9⁺ macrophages promote tumor growth in the nude Xenograft mouse model. (A)Tumor volume measurement in the mouse model. (B) Ki67 (red fluorescence) and TUNEL (green fluorescence) staining as indicators of cell viability and apoptosis, respectively (400×). (C) Immunofluorescence staining for CD206 (green fluorescence), F4/80 (red fluorescence), and merge (yellow fluorescence)(400×). (D) Immunofluorescence analysis of CD206 and F4/80 expression. (E) Immunohistochemistry of CD9 and NOR1 in the tumor tissues (400×). (F) Relative mRNA expression levels of CD9, CXCR4, CD206, and NOR1 in the tumor tissues from the different experimental groups. The data shown were the representative of three experiments. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Fig. 7.

CD9⁺ macrophages promote tumor growth in the immunocompetent Xenograft mouse model. (A) Tumor volume measurement in the mouse model. (B) Ki67 (red fluorescence) and TUNEL (green fluorescence) staining as indicators of cell viability and apoptosis, respectively (400×). (C) Immunofluorescence staining for CD206 (white fluorescence), F4/80 (green fluorescence), Arg1 (red fluorescence) and merge (yellow fluorescence)(400×). (D) Immunofluorescence analysis of CD206 and F4/80 expression. The data shown were the representative of three experiments. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001

Proliferation and apoptosis fluorescent double staining revealed that the tumor tissue of mice treated with CD9-overexpressing macrophages exhibited enhanced cell proliferation capacity, whereas CD9-knockdown macrophages promoted apoptosis in some tumor cells (Figs. 6–7B). Using CD206/Arg1 and F4/80 fluorescent staining to localize M2 macrophages in tumor tissues, we observed increased M2 macrophages in tumor tissues of mice treated with CD9-overexpressing macrophages (Fig. 6–7C, D). Moreover, the proteins and mRNA levels of CD9, NOR1, CXCR4, and CD206 were elevated in the CD9-overexpressing group (Figs. 6E, F). These results suggested that CD9-overexpressing macrophages promoted tumor growth in the xenografted mice.

Discussion

The tumor immune microenvironment helps facilitate tumor proliferation, metastasis, and immune evasion [31]. Previous research documented the heightened prevalence of macrophages in diverse individual tumor types [32]. Furthermore, studies have explored the molecular attributes of TAMs, particularly in the context of RCC [33]. However, the involvement of CD9⁺ macrophages in RCC has not been reported. Our study constructed an RCC cancer immune atlas and characterized the heterogeneity of macrophages in different RCC tissues. Analysis of single-cell RNA-seq data revealed CD9⁺ macrophages in RCC tumors, suggesting their potential as diagnostic and therapeutic targets. We also found high CD9 expression in RCC specimens, and its relationship with poor prognosis in RCC requires further investigation [34].

We analyzed heterogeneity genes among different clusters and found that genes such as CD9, TNFAIP1, CXCR4, and NR4A3 (NOR1) were specifically expressed in tumors by the bioinformatics analysis and we also found NOR1 had no direct interaction with CD9 but had a positive correlation (Figs. 2E–I). However, currently, there are no reports that have elucidated the interaction between CD9 and NOR1 in cancer. So we investigated whether CD9 can act on NOR1 and regulate the progression of RCC. The results indicated that compared to CD9-overexpressing macrophages, macrophages with CD9 overexpression and NOR1 knockdown exhibited a reduced capacity for tumor proliferation, invasion, and migration. CD9-overexpressing macrophages further enhanced tumor cell proliferation, invasion, and migration in vivo and in vitro. Earlier work has identified junctional adhesion molecule-A (JAM-A) as an epithelial adhesion receptor that concentrates at tight junctions. CD9 recruits monomeric JAM-A into a ternary complex with αvβ3 integrin; once engaged by (bFGF) released from JAM-A, this complex triggers MAPK signalling, thereby promoting endothelial migration and neovascularization [35, 36]. A separate study demonstrated that the GLP-1 receptor agonist exendin-4 blunts ERK–MAPK phosphorylation elicited by serum, leading to down-regulation of NOR1 expression [37]. In primary mouse embryonic fibroblasts, anisomycin-triggered MAPK signalling phosphorylates CREB, which then recruits the co-activator CBP to cAMP-response elements on the NOR1 promoter, thereby driving rapid transcription of NOR1 mRNA [26]. In vivo studies report that NOR1 is strongly expressed in areas containing CD68⁺/MR⁺M2 macrophages in human atherosclerotic plaques [38]. We therefore hypothesize that CD9 may drive pathophysiological processes by tuning the MAPK signaling axis, thereby modulating downstream NOR1 expression in the macrophages.

This study identified CD9⁺ macrophages in RCC tumor tissues and explored their molecular characteristics, which displayed a high degree of phenotypic consistency. Our findings indicated that CD9⁺ macrophages were an alternatively activated M2 subset that contributes to tumorigenesis and helps shape the TME. Furthermore, our results suggested that CD9 overexpression may promote the M2-type activation of TAMs through the NOR1 pathway, thereby facilitating tumor progression, bridging the knowledge gap on CD9⁺ macrophages in RCC. Consequently, CD9 emerges as a potential novel target for immunotherapeutic interventions in RCC.

However, our study has some limitations. We analyzed scRNA-seq data from only six samples. The current scRNA-seq findings are hypothesis-generating rather than definitive because only three patients were profiled. Expanding the sample size would undoubtedly enhance the accuracy and robustness of our findings. Another, we have not explored the specific mechanism by which CD9⁺ macrophages affected renal cancer cells. Next, we will further investigate the interaction between the two types of cells.

In conclusion, our scRNA-seq data analysis highlighted the macrophage differences between RCC tumors and adjacent tissues. We identified CD9 as the primary differentially expressed gene in macrophages between these two tissue types. The CD9⁺ macrophages represent an alternatively activated M2 subset that fosters tumorigenesis and impacts the tumor immune microenvironment. Moreover, our findings demonstrate that the knockdown of CD9 expression in macrophages impedes RCC cell growth in vivo, highlighting its potential as a valuable immunotherapeutic target.

Author contributions

Q. Z. and S. L. designed and led the experiments, performed data analysis, and wrote the paper. J. H., J. S. and X. S. prepared protein samples and performed proteomic analysis. P. P. and P. W. performed cell culture, cell biology, and molecular biology experiments. X.H., T. M., Z. B., and W.Y. performed bacterial culture and animal experiments. Q. C. advised on the data analysis, designed and supervised the experiments, and wrote the manuscript. All authors have read and approved the article.

Funding

Joint Research Project of Liaoning Provincial Technology Plan (Applied Basic Research Program, No. 2023JH2/101700312) and the Research Start-up Funds for High-level Talents in the Affiliated Hospital of Guangdong Medical University (GCC2022044 and GCC2022046).

Data availability

The datasets generated and/or analyzed during the current study are available in the GEO repository (GSE121636).

Declarations

Ethics approval and consent to participate

The study was approved by the ethics committee of the Shengjing Hospital of China Medical University and implemented after approval. All animal research was conducted in accordance with the Guiding Opinions on the Treatment of Laboratory Animals issued and the Laboratory Animal Guideline for Ethical Review of Animal Welfare issued by the National Standard GB/T35892-2018 of the People’s Republic of China, and all operations performed on mice have approved the China Medical University Standards for the Laboratory Animals Welfare and Ethical Review. All methods are reported in accordance with ARRIVE guidelines for the reporting of animal experiments (Permit Number: 2024PS1672K).

Consent for publication

All participants provided written informed consent prior to assessment of eligibility, and they 410 were informed that the data would be publishedAll participants provided written informed consent prior to assessment of eligibility, and they were informed that the data would be published.

Competing interests

The authors declare no competing interests