Abstract

Background

   The polarization status and function of tumor-associated macrophages
   (TAMs) influence tumor progression and patients’ prognosis.
   CCAAT/enhancer-binding proteins (CEBPs) family are important
   transcriptional factors in macrophages differentiation physically and
   pathologically. This study aims to explore the mechanism of CEBPs in
   TAMs polarization in clear cell renal cell carcinoma (ccRCC) immune
   microenvironment and its impact on immune checkpoint blockers (ICBs)
   therapy.

Methods

   The expression of CEBPs in ccRCC was analyzed by single-cell
   transcriptome and western blot. Immunofluorescence and in-vitro
   polarization assay were used to evaluate the effect of CEBP delta
   (CEBPD) on TAMs. Chromatin immunoprecipitation sequencing was used to
   explore targets of CEBPD. Dual-luciferase reporter assay and
   electrophoretic mobility shift assay were performed to confirm the
   regulation of CEBPD to RGS2. Specimens of patients received ICB therapy
   were used to analyze the relationship between CEBPD and immunotherapy.

Results

   The study identified CEBPD as a key transcriptional factor in ccRCC TAM
   polarization. Upregulation of CEBPD correlates to decreased M1/M2 ratio
   of TAMs and poorer clinical outcomes. CEBPD inhibited M1-like
   polarization in vitro and in vivo via the RGS2/PAR2 axis. Furthermore,
   CEBPD also affected the therapeutic efficacy of ICB.

Conclusion

   This study revealed CEBPD regulated TAM polarization via the
   CEBPD/RGS2/PAR2 axis. Targeting CEBPD may be a potential approach and a
   complementary strategy to ICB therapies in ccRCC.

   Keywords: Kidney Cancer, Macrophage, Immunotherapy
     __________________________________________________________________

WHAT IS ALREADY KNOWN ON THIS TOPIC

     * Tumor-associated macrophages (TAMs) are highly plastic and their
       polarization status affects clinical outcomes of patients with
       clear cell renal cell carcinoma (ccRCC). Previous studies have
       shown that the CCAAT/enhancer-binding protein (CEBP) family is
       involved in the differentiation of macrophages but the specific
       roles of family members in TAM polarization are not thoroughly
       revealed.

WHAT THIS STUDY ADDS

     * Our study showed for the first time that a CEBP family member, CEBP
       delta (CEBPD) regulated TAM polarization via the CEBPD/RGS2/PAR2
       axis and promoted tumor progression. Furthermore, CEBPD was related
       to the efficacy of immune checkpoint blockers (ICBs) in ccRCC.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

     * Our findings revealed the role of CEBPD in regulating TAM
       polarization and targeting CEBPD may enhance the ICBs therapeutic
       efficacy in patients with ccRCC.

Introduction

   Clear cell renal cell carcinoma (ccRCC), is the most common histologic
   subtype of kidney malignancies.[55]^1 Nearly 25% of patients with ccRCC
   are diagnosed at the metastatic stage, and a third of the patients who
   receive resection of localized tumors will have a recurrence.[56]^2
   Tyrosine kinase inhibitors (TKIs) are applied to treat patients with
   advanced or metastatic renal cell carcinoma, but most patients develop
   resistance to TKI therapy.[57]^3 In recent years, treatment with immune
   checkpoint blockers (ICBs) in ccRCC patients have made remarkable
   success, which reveals enormous potential for immune therapy. However,
   low response rates and drug resistance still remain intractable
   problems. Thus, there is an urgent need to investigate the tumor immune
   microenvironment and explore new treatment targets.

   Due to the dysfunction of von Hippel-Lindau protein, ccRCC is
   characterized by a highly hypoxic tumor microenvironment (TME).
   Hypoxia-related signal pathways are hyperactivated, such as HIF, which
   lead to excessive infiltration of tumor-associated macrophages (TAMs)
   in the TME.[58]^4 5 TAMs are preferentially recruited in the hypoxic
   region and gradually polarize into immune-suppressive M2-like
   phenotype,[59]^6 7 whereas the proportion of proinflammatory M1-like
   TAMs decreases. This phenotype shift represented a risk factor for poor
   prognosis, shorter survival and worse therapeutic efficacy.[60]8,[61]10
   Enhancing the transformation of TAMs into antitumor M1 phenotype could
   provide a therapeutic strategy.[62]^11 However, the molecular mechanism
   of influencing the TAMs polarization in the hypoxic TME remains to be
   further specified.

   CCAAT/enhancer-binding proteins (CEBPs), are a group of transcription
   factors significant in the development and differentiation of
   macrophages.[63]^12 13 Previous studies reported that CEBPs can be
   regulated by hypoxic stimulation.[64]14,[65]16 However, whether CEBPs
   play a significant role in TAM polarization in the hypoxic TME, still
   remains to be clarified. The CEBP family is comprised of six members:
   CEBPα, CEBPβ, CEBPγ, CEBPδ, CEBPε, CEBPζ. Family members have a high
   degree of homology structurally and share many common targets,[66]^17
   which makes it complicated to explore their individual function.
   Therefore, it’s necessary to analyze the expression pattern of CEBPs in
   ccRCC transcriptional landscape by single-cell transcriptome and
   combined with large-scale specimen verification in order to clarify the
   role of CEBPs played in the TAMs polarization in hypoxia TME of ccRCC.

   In this study, we conducted a comprehensive analysis of the CEBP family
   in ccRCC TAMs and identified CEBP delta (CEBPD) as a pivotal
   transcriptional factor influencing TAM polarization. Combining patient
   specimens and in-vitro polarization assay, we confirmed that CEBPD
   reduces the M1/M2 ratio and leads to poorer survival outcomes.
   Furthermore, through DNA-protein interactive assay, we reported RGS2
   was the target of CEPBD and elucidated the involvement of the RGS2/PAR2
   axis in TAM polarization. Finally, we explored the impact of CEBPD on
   patients received ICB therapy.

Materials and methods

Processing of single cell RNA-seq data

   Process single cell RNA-seq (scRNA-seq) data using R (V.4.3.0). Use the
   Read10X function in Seurat (R package, V.4.3.0.1) to separately import
   unique molecular identifier (UMI) count matrices generated from 20
   samples into R. To remove the low-expressed genes and low-quality
   cells, cells that had fewer than 200 genes, over 6,000, UMI over
   10,000, or more than 25% mitochondrial reads were depleted, and the
   genes expressed in at least 3 cells were kept from the original data.
   Follow the DoubletFinder (R package, V.2.0.4) to computationally infer
   and remove doublets in each sample individually, with default
   parameters. All of this information was merged into one Seurat object
   using the merge function. UMI counts were normalized using the
   “NormalizeData” function. The 2000 highly variable genes were
   identified by the “FindVariableFeatures” function. Principal component
   analysis (PCA) was performed on the single-cell expression matrix using
   the ‘‘RunPCA’’ function and the top 30 principal components were used
   for clustering with the Louvain graph-clustering method. Then, we used
   the “Runharmony” to perform the batch-effect correction for each
   dataset. Subsequently, Uniform Manifold Approximation and Projection
   (UMAP) was established based on harmony dimensions for cell clustering
   analysis. As to subclusters, the same methods were used for recognizing
   the variable genes, reducing dimensions and clustering.

Cell type annotation and cluster marker identification

   “FindAllMarkers” function was used to find markers for each of the
   identified clusters and annotated on the basis of the expression of
   canonical markers of particular cell types based on the published
   single-cell transcriptome data of kidney cancer.

   The gene set variation analysis (GSVA, R package, V.1.48.2) was applied
   to estimate pathway activity scores for single cells and visualize it
   through the heatmap using gene sets of C2 and C5 collection obtained
   from the molecular signature database. The differential activities of
   pathways were calculated using the limma R package.

Gene set variation analysis

   Regulon activity analysis pySCENIC (V.0.12.1) was used to build the
   gene regulatory network in all cells combined with cis-Target human
   motif database (V.10). Raw expression data were extracted from the
   Seurat data. Enriched motifs were identified with default parameters of
   pySCENIC. AUCell function assigns specific values to each cell to
   represent their regulon activity based on Jensen-Shannon divergence.

Cell culture and transfection

   THP-1 cells were obtained from the cell bank of the typical culture
   preservation committee of the Chinese Academy of Sciences. THP-1 cells
   were cultured in RPMI 1640 (Gibco) supplemented with 10% fetal bovine
   serum (Gibco). Human peripheral mononuclear cells were obtained from
   donors and were used for isolating CD14+ monocytes. 50 ng/mL Phorbol
   12-myristate 13-acetate (PMA) (Selleck) were added for 48 hours to
   induce M0 macrophages. 100 ng/mL lipopolysaccharides (LPS) (Selleck)
   and interferon-gamma (IFN-γ) (20 ng/mL) (PeproTech) were added to
   induce M1-like macrophages and interleukin (IL)-4 (20 ng/mL)
   (PeproTech) was added to induce M2-like macrophages.

   THP-1 cells were transduced with short hairpin RNA lentivirus targeting
   CEBPD (CEBPD-kd1 and CEBPD-kd2) and RGS2 (RGS2-kd1 and RGS2-kd2),
   overexpressing lentiviral vector (RGS2-oe) or control lentivirus.
   Lentiviruses were purchased from OBIOSH. Cells infected with virus were
   selected by 2 µg/mL puromycin (Selleck). PAR1 inhibitor (HY-14993),
   PAR2 inhibitor (HY-P1260) and PAR2 agonist (HY-P4803) were purchased
   from MCE.

Patient specimens

   Written informed consent was also obtained from all the patients. 196
   patient specimens were derived from patients underwent radical or
   partial nephrectomy at Xinhua Hospital between 2016 and 2018. All tumor
   specimens were pathologically confirmed ccRCC. After resection, the
   samples were immediately frozen in liquid nitrogen and stored at −80℃
   for further studies. Specimen collection for research purposes was
   approved by the Clinical Research Ethics Board of Xinhua Hospital.

Western blotting assay

   Total protein was extracted from the cultured cells or specimens using
   lysis buffer according to manufacturer’s protocol, and resolved by
   sodium dodecyl sulfate (SDS) polyacrylamide gels, and then transferred
   to polyvinylidene fluoride (PVDF) membranes. The antibodies were
   provided in the [67]online supplemental materials. The immunoreactive
   bands were analyzed by chemiluminescence.

Flow cytometry

   Cells (1×10^6) were incubated with Fc receptors blocker, then GhostDye
   staining of dead cells was incubated at 4°C in the dark for 30 min and
   centrifuged to remove the supernatant. Then, the cells were resuspended
   in fluorescence-activated cell sorting (FACS), after which
   extracellular staining was performed for the following proteins: CD86
   and CD206. The cells were incubated in cell staining buffer at 4°C in
   the dark for 30 min. The cells were analyzed using a Beckman Analytical
   Flow Cytometer and analyzed using CytExpert software (Beckman Coulter).
   The following antibodies were used: anti-CD86, PE (Miltenyi Biotec,
   130-116-160); anti-CD206, PE (Invitrogen, MA5-23594).

Tissues and cells immunofluorescence assay

   Tissue sections were deparaffinized, rehydrated with graded ethanol,
   incubated with 0.3% hydrogen peroxide for 20 min, and blocked with 10%
   goat serum. THP-1 cells were fixed on slides with 4% paraformaldehyde,
   permeabilized with a buffer containing 0.1% Triton X-100 and 0.25%
   bovine serum albumin (BSA), and blocked with goat serum and BSA. The
   slides were incubated with primary antibody overnight at 4℃. Next,
   slides were labeled with secondary antibody for 1 hour at room
   temperature. After washing three times with phosphate-buffered saline,
   the slides were subsequently stained with 4',6-diamidino-2-phenylindole
   (DAPI) for visualization of the nucleus. Images were scanned and
   analyzed using a FLUOVIEW vs200 microscope (Olympus). In multiplex
   immunohistochemistry (mIHC), we used CD86 to label M1-like macrophages
   and CD163 to label M2-like macrophages. Tumors with a CEBPD positivity
   rate exceeding 50% in TAMs were classified into the CEBPD-high group,
   while those below 50% were low group. The antibodies were provided in
   the [68]online supplemental materials.

RNA extraction and real-time quantitative PCR

   Total RNAs were isolated using TRIzol reagent (Invitrogen), and 2 µg of
   total RNA was subjected to reverse transcription using Superscript III
   transcriptase (Invitrogen). Real-time quantitative PCR (RT-qPCR) was
   conducted using a Bio-Rad CFX96 system with SYBR green to determine the
   messenger RNA expression level of a gene of interest. Expression levels
   were normalized to the ACTB level using the 2−∆∆Ct methods. The
   sequences of the primers were provided in [69]online supplemental
   materials.

Chromatin immunoprecipitation sequencing (ChIP-seq)

   Cells were cross-linked with 1% formaldehyde solution, then lysed and
   sonicated. The sonicated chromatin was precleared with chromatin
   immunoprecipitation (ChIP) Grade protein A/G for 1 hour. Then the
   lysates were incubated at 4℃ overnight with the target-specific
   antibody CEBPD (Invitrogen, PA5-30262) or negative control anti-IgG
   antibody. DNA was eluted from immunoprecipitated complexes, reverse
   cross-linked, and purified. High-quality ChIP DNA was used for
   sequencing. The ChIP-seq results were visualized with Integrative
   Genomics Viewer. The number of CEBPD binding sites in the promoter
   region of interested genes was analyzed to find target genes that can
   be transcriptionally regulated by CEBPD.

Dual-luciferase reporter assay

   The full-length segments and truncated segments of the RGS2 promoter
   were amplified by PCR from genome DNA of ccRCC tissues and were
   purified by Quick Gel Extraction Kit (TransDetect, China). The purified
   DNA segments were inserted upstream of the firefly luciferase
   complementary DNA (cDNA) in the pGL3-basic vector to construct
   pGL3-RGS2-Full-length, pGL3-RGS2-Truncated#1 and pGL3-RGS2-Truncated#2.
   The coding sequence of CEBPD was amplified by PCR from cDNA of ccRCC
   tissues and was also purified and inserted downstream of the promoter
   in the pcDNA3.1 vector to construct pcDNA3.1-CEBPD. Cells were
   transfected using Lipo3.0 Efficient Transfection Reagent (SuperKine,
   China). Dual-luciferase reporter assay was conducted using the
   Double-Luciferase Reporter Assay Kit (TransDetect, China) as per
   manufacturer’s protocol. All experiments were repeated at least three
   times. The primers used to construct pcDNA3.1-CEBPD and
   pGL3-RGS2-promoter were provided in [70]online supplemental materials.

Electrophoretic mobility shift assay (EMSA)

   The full-length segments of RGS2 promoter were obtained as described
   above. Recombinant Human CEBPD protein was purchased from Cusabio
   (catalog#EP005182HU). The RGS2-promoter DNA and CEBPD proteins were
   mixed at indicated concentrations in binding buffer and incubated at
   room temperature for 30 min. A 1% agarose gel containing
   0.5×Tris-Borate-EDTA (TBE) buffer was prepared. The incubated mixture
   was then subjected to the gel and run in 0.5×TBE buffer at 10 V/cm for
   1 hour on ice. After electrophoresis, the gel was steeped in 0.5×TBE
   buffer containing 4×Ultra Gelred (Vazyme, China) to stain nucleic acid
   for ∼40 min and then scanned using an ultraviolet emitter. After that,
   the gel was steeped in FastBlue (Biosharp, China) to stain protein for
   2 hour and then photographed. Equal doses of BSA were used as a
   negative control.

Statistical analysis

   All data were analyzed by GraphPad Prism V.9.0 and presented as the
   mean±SD. Statistical tests for data analysis included Student’s t-test,
   Spearman correlation analysis. P values were determined by a two-tailed
   Student’s t-test for comparing two groups. P values were indicated with
   *p<0.05, **p<0.01, and ***p<0.001 on graphs. Graphs not labeled with an
   asterisk indicate that the differences between the test groups and the
   control groups were not statistically significant. All in vitro
   experiments were repeated with at least three replicates.

Results

CEBPD upregulates in ccRCC and correlates to suppressed immune response

   To provide a landscape of ccRCC TME, we performed scRNA-seq on 13 ccRCC
   samples and seven adjacent kidney samples. Following UMAP
   dimensionality reduction, these cells clustered into seven groups
   annotated with specific marker genes ([71]online supplemental figure
   S1A–C). Myeloid cells were further categorized into monocytes,
   macrophages (including TAM1, TAM2 and TAM3), and dendritic cells
   ([72]figure 1A,B, [73]online supplemental figure S2A,B and S3). Our
   clustering strategy of TAMs was adapted from the study of Mei et
   al.[74]^18 The TAM1 cluster was similar to tissue-resident macrophages.
   The TAM2 cluster was defined as an immune-suppressive subtype,
   transcriptionally expressing a high level of CD163. The TAM3 cluster
   was described as a proinflammatory subtype, with an increasing
   expression of M1-like markers like IL-1β. We noticed a significant
   decrease in TAM3 in ccRCC tissues compared with adjacent kidney
   tissues, indicating a decline in inflammation and antitumor immune
   response ([75]figure 1C). We hypothesized that CEBP-mediated macrophage
   differentiation contributes to this phenomenon. Analysis of CEBP family
   expression in myeloid cells from ccRCC and adjacent tissues revealed
   that the expression of CEBPB and CEBPD was the highest among all
   members ([76]figure 1D). Notably, CEBPD in TAM3 differed significantly
   between tumors and adjacent tissues. Further analysis of CEBP family
   transcriptional activity showed a marked increase in CEBPD activity in
   TAM3 in ccRCC compared with adjacent kidney tissues ([77]figure 1F),
   while CEBPB and other members showed no significant differences
   ([78]figure 1E and [79]online supplemental figure S2C,D). Moreover,
   CEBPD-high ccRCCs had a significantly lower proportion of TAM3 compared
   with CEBPD-low ccRCCs ([80]figure 1G). Pathway enrichment analysis
   showed CEBPD upregulation in ccRCCs was positively associated with
   negative regulation of IL-6 pathway, negative regulation of monocyte
   chemotaxis and macrophage fusion ([81]figure 1H), while it was
   negatively associated with positive regulation of T cell
   differentiation ([82]figure 1H). Similarly, CEBPD upregulation
   correlated to low expression of IL-1α, IL-1β, IL-6, suggesting impaired
   antitumor function ([83]figure 1I). These results indicated that the
   activation of CEBPD in ccRCC leads to a decrease and dysfunction of
   proinflammatory TAMs and suppression of immune response. To identify
   downstream of CEBPD, we analyzed differentially expressed genes
   ([84]figure 1J) and performed function enrichment ([85]online
   supplemental figure S4) in TAM3 between CEBPD-high and CEBPD-low
   groups, and RGS2 showed upregulation.

Figure 1. CEBPD is upregulated in ccRCCs and correlated with impaired immune
response. (A) The cell type clusters of myeloid cells. (B) The group clusters
of myeloid cells. (C) The proportion of myeloid cells in ccRCCs and adjacent
kidney tissues. (D) The expression of CEBPD family in cell type clusters, and
the yellow dot represents high level of average expression, while the purple
dot represents low level of average expression. Besides, the size of the dot
represents the percentage of cells expressing the gene. Transcription factor
activity of (E) CEBPB and (F) CEBPD in myeloid cells. (G) Proportion of 5
groups of myeloid cells in CEBPD-high and CEBPD-low ccRCCs. (H) Pathway
enrichment analysis of 13 ccRCC samples. (I) Heatmap of cytokines expression
of 13 ccRCC samples. (J) Volcano map of differentially expressed genes
between CEBPD-high and low TAM3. Data were presented as mean±SD. A
significant difference between the groups, **p<0.01, ***p<0.001, and
****p<0.0001. ccRCC, clear cell renal cell carcinoma; CEBP,
CCAAT/enhancer-binding protein; CEBPD, CEBP delta; DC, dendritic cell; GSVA:
gene set variation analysis; IL, interleukin; ns, not significant; RCC, renal
cell carcinoma; TAM, tumor-associated macrophage.

   [86]Figure 1
   [87]Open in a new tab

CEBPD upregulation in ccRCC tissues correlates to low M1/M2 proportion and
poorer clinical prognosis

   We analyzed the protein levels of CEBP family members in six matched
   patient samples. Compared with adjacent kidney tissues, ccRCC tissues
   showed elevated levels of CEBPD ([88]figure 2A), while other CEBP
   family members did not show significant differences. Elevated CEBPD
   expression in ccRCC was related to inferior overall survival
   probability in The Cancer Genome Atlas (TCGA) cohort ([89]figure 2B).
   To further investigate changes in the immune microenvironment in ccRCC
   tissues, we performed a multiplex immunofluorescence assay.[90]^19 We
   used CD86 to define M1-like antitumoral TAMs and CD163 to define
   M2-like protumorous TAMs. Results showed a higher infiltration of
   macrophages in ccRCC compared with adjacent kidney tissue ([91]figure
   2D, [92]online supplemental figure S5). Besides, tumors with higher
   expression of TAMs CEBPD showed increased CD163, which was considered
   as M2-like TAMs, while M1 marker CD86 was reduced, compared with tumor
   with lower expression of TAMs CEBPD ([93]figure 2D,E). Furthermore,
   higher TAM-CEBPD patients showed worse overall survival probability
   than the low group ([94]figure 2C, [95]online supplemental table S1).
   These results further indicated that CEBPD upregulation would disturb
   the proportion of macrophage subtypes, which may lead to poor clinical
   outcomes.

Figure 2. CEBPD upregulation influenced TAM proportion and was associated
with poorer prognosis. (A) Western blot results showed the protein expression
of CEBPA, CEBPB, CEBPD, CEBPG, CEBPZ in six patient specimens. (B)
Kaplan-Meier curves of TCGA-KIRC cohort. (C) Kaplan-Meier curves of survival
probability in Xinhua Hospital cohort. Patients were divided by IHC score.
(D) Representative immunofluorescence images of adjacent kidney tissue
(left), CEBPD-low tumor (middle) and CEBPD-high tumor (right), and (E)
quantitative analysis of M1-like and M2-like TAMs.Data were presented as
mean±SD. A significant difference between the groups, *p<0.05. ccRCC, clear
cell renal cell carcinoma; CEBP, CCAAT/enhancer-binding protein; CEBPD, CEBP
delta; IHC, immunohistochemistry; KIRC, kidney renal clear cell carcinoma;
TAM, tumor-associated macrophage; TCGA, The Cancer Genome Atlas.

   [96]Figure 2
   [97]Open in a new tab

CEBPD attenuates the M1 polarization of macrophages in vitro

   To explore the role of CEBPD in macrophage polarization, we conducted
   experiments using human acute monocytic leukemia THP-1 cells, which
   were induced by PMA into the M0 stage. M0 THP-1 exhibited low levels of
   both M1 and M2 markers, and the knockdown or overexpression of CEBPD
   did not affect the basal expression of these markers ([98]online
   supplemental figure S6). We then induced M1 polarization using LPS and
   IFN-γ, and M2 polarization using IL-4. Notably, results of flow
   cytometry showed CEBPD knockdown resulted in an increasing proportion
   of cells expressing CD86 ([99]figure 3A). In addition, western blot and
   RT-qPCR showed CEBPD knockdown increased the expression of IL-1β and
   IL-6, indicating enhanced M1 polarization ([100]figure 3B,C). However,
   M2 polarization was not affected by CEBPD knockdown ([101]figure 3D–F).
   Similar results were shown in monocytes that CEBPD knocked down could
   enhance M1 polarization while not affecting M2 polarization
   ([102]online supplemental figure S7). Immunofluorescence showed that
   CEBPD knockdown enhanced the translocation of p65 from cytoplasm to the
   nucleus ([103]figure 3G). These findings supported that CEBPD impeded
   M1 polarization via the NF-κB pathway. Overexpression of CEBPD was
   found to reduce the production of proinflammation factors, potentially
   impairing the antitumor effect of macrophages within the TME.

Figure 3. CEBPD knockdown enhanced M1 type polarization of THP-1 cells. (A)
CD86 positive cells in CEBPD-kd1, CEBPD-kd2 and control group. (B–C) CEBPD
knockdown enhanced the expression of IL-1β and IL-6. (D) CD206 positive cells
in CEBPD-kd1, CEBPD-kd2 and control group. (E–F) CEBPD knockdown did not
affect the expression of M2 markers. (G) CEBPD knockdown promoted the
translocation of p65 from cytoplasm to nucleus. A significant difference
between the groups, ***p<0.001. CEBPD, CCAAT/enhancer-binding protein delta;
DAPI: 4',6-diamidino-2-phenylindole; IFN-γ, interferon-gamma; IL,
interleukin; LPS, lipopolysaccharides; ns, not significant; UN, uninduced,

   [104]Figure 3
   [105]Open in a new tab

CEBPD binds to RGS2 promoter region and upregulates RGS2 expression

   To elucidate the underlying mechanism of CEBPD in TAM polarization,
   ChIP sequencing (ChIP-seq) was performed in THP-1 cells to identify
   downstream genes regulated by CEBPD. The analysis revealed that nearly
   1/4 binding regions of CEBPD were located in promoter regions
   ([106]figure 4A), with the binding peaks almost concentrated in
   transcription start sites ([107]figure 4B). Differential fold change
   and peak score analysis of downstream genes regulated by CEBPD
   highlighted RGS2 as having the highest fold change and peak score
   ([108]figure 4C,D). Functional enrichment analysis of these genes
   indicated involvement in the production of molecular mediators of
   immune response ([109]figure 4E). Visualization with IGV software
   confirmed that the CEBPD binding peak was positioned upstream of the
   transcription start site of RGS2 ([110]figure 4F). Additionally, a
   positive correlation between RGS2 and CEBPD was validated in TAM3 using
   scRNA-seq ([111]figure 4G), and a similar result was observed in the
   TCGA-KIRC (kidney renal clear cell carcinoma) database ([112]figure
   4H).

Figure 4. Chromatin immunoprecipitation sequencing analysis of CEBPD. (A) Pie
chart of binding region of CEBPD. (B) Genomic distribution of CEBPD binding
peaks. (C) Fold changes and (D) peak score of genes regulated by CEBPD. (E)
Function enrichment of genes regulated by CEBPD. (F) Visualization of binding
peak of CEBPD on promoter region of RGS2. Correlation analysis between CEBPD
and RGS2 in (G) TAM3 and (H) TCGA-KIRC database. ccRCC, clear cell renal cell
carcinoma; CEBPD, CCAAT/enhancer-binding protein delta; ChIP, chromatin
immunoprecipitation; FC, fold change; KIRC, kidney renal clear cell
carcinoma. TAM, tumor-associated macrophage; TCGA, The Cancer Genome Atlas.

   [113]Figure 4
   [114]Open in a new tab

   To confirm the direct transcriptional regulation of RGS2 by CEBPD, we
   performed dual-luciferase reporter assay first. We designed a 2000 bp
   promoter sequence containing CEBPD binding peak, and identified three
   potential binding sites by JASPR database ([115]figure 5A, top).
   Through sequential truncation of these binding sites, we determined
   that the first site was responsible for binding and transcriptional
   activity ([116]figure 5A, bottom). Next, the western blot assay showed
   RGS2 increased as CEBPD was overexpressed, while RGS2 decreased as
   CEBPD was knocked down ([117]figure 5B). To further confirm the direct
   binding between CEBPD protein and RGS2 promoter, we performed an
   electrophoretic mobility shift assay (EMSA). As shown in [118]figure
   5C, a shifted band of the RGS2 promoter appeared as the concentration
   of CEBPD protein increased, indicating the DNA-protein complex formed
   by direct binding between the RGS2 promoter and CEBPD protein
   ([119]figure 5D, top). In contrast, no shift bind was observed when BSA
   was used instead of CEBPD protein ([120]figure 5D, bottom). These
   results established that CEBPD directly binds to the promoter region of
   RGS2, thereby promoting its transcription and expression.

Figure 5. CEBPD upregulated RGS2 by binding to its promoter region. (A)
Dual-luciferase reporter assay results showed CEBPD bound to the first site
on RGS2 promoter region. Western blot assay showed RGS2 increased as CEBPD
was overexpressed while RGS2 decreased as CEBPD was knocked down in (B) THP-1
cells and (C) human monocytes. (D) EMSA results confirmed the combination
between CEBPD protein and RGS2 promoter region. Data were presented as
mean±SD. A significant difference between the groups, **p<0.01, ***p<0.001,
and ****p<0.0001. BS, binding site. BS, binding site; BSA, bovine serum
albumin; CEBPD, CCAAT/enhancer-binding protein delta; EMSA, electrophoretic
mobility shift assay.

   [121]Figure 5
   [122]Open in a new tab

RGS2 inhibits M1 polarization through deactivating the PAR2 pathway

   We further explored whether RGS2 was the target of CEBPD involved in
   regulating macrophage polarization. Western blot analysis showed that
   THP-1 cells and monocytes with CEBPD knockdown expressed more IL-1β and
   IL-6; however, the overexpression of RGS2 reversed this trend
   ([123]figure 6A, [124]online supplemental figure S8A). Additionally,
   CEBPD knockdown promoted the translocation of p65 to the nucleus, but
   this effect was diminished by RGS2 overexpression ([125]figure 6B).

Figure 6. RGS2 inhibited M1 polarization through deactivating PAR2. (A) CEBPD
knockdown increased IL-1β and IL-6, whereas RGS2 overexpression reversed this
effect. (B) CEBPD knockdown promoted the translocation of p65 from cytoplasm
to nucleus, whereas RGS2 overexpression reversed this effect. (C) PAR2 was
upregulated during M1 polarization. (D) RGS2 knockdown activated
PAR2-mediated MAPK signaling pathway. (E) PAR2 inhibitors reduced IL-1β and
IL-6. (F) PAR2 inhibitors impeded the nuclear translocation of p65. CEBPD,
CCAAT/enhancer-binding protein delta; DAPI: 4',6-diamidino-2-phenylindole;
IFN-γ, interferon-gamma; IL, interleukin; LPS, lipopolysaccharides; PAR,
proteinase-activated receptor; UN, uninduced.

   [126]Figure 6
   [127]Open in a new tab

   RGS2, a member of the regulator of G protein signaling family,
   functions as GTPase activating proteins. RGS family inhibits
   GPCR-mediated signaling by accelerating the GTPase activity of G
   protein alpha subunits, thus driving them into their inactive GDP-bound
   form.[128]^20 21 Proteinase-activated receptors (PARs) are a subfamily
   of G protein-coupled receptors with four members, PAR1, PAR2, PAR3,
   PAR4. PARs play an important role in inflammation[129]22,[130]24 and
   are reported to interact with RGS family members.[131]25,[132]27 We
   observed an increase in PAR2 expression during the M1 polarization of
   macrophages, indicating the involvement of the PAR2-associated pathway
   in this process ([133]figure 6C). Further investigation showed PAR2
   agonists increased phosphorylated ERK and NLRP3 levels, whereas PAR2
   inhibitors had the opposite effect ([134]figure 6D, [135]online
   supplemental figure S8B). RGS2 knockdown also elevated phosphorylated
   ERK and NLRP3, which could be reversed by PAR2 inhibitors ([136]figure
   6D, [137]online supplemental figure S8B). Additionally, PAR2 inhibitors
   reduced IL-1β and IL-6, effectively blocking M1 polarization
   ([138]figure 6E, [139]online supplemental figure S8C), while PAR1
   inhibitor did not affect this outcome. Furthermore, PAR2 inhibitors
   impeded the nuclear translocation of p65 ([140]figure 6F). These
   results demonstrated that RGS2 mitigated IL-1β and IL-6 production by
   deactivating PAR2 via the MAPK/NF-κB signaling pathway ([141]figure 7).

Figure 7. Schematic diagram of CEBPD affecting M1-type polarization of
macrophages through the RGS2/PAR2 axis. CEBPD, CCAAT/enhancer-binding protein
delta; IL, interleukin; PAR, proteinase-activated receptor.

   [142]Figure 7
   [143]Open in a new tab

CEBPD correlates with ICB therapeutic efficacy

   Given that ICBs are widely applied in patients with advanced and
   metastatic ccRCC, and that TAMs are also associated with the efficacy
   of ICBs,[144]^28 29 we further examined whether CEBPD influences the
   therapeutic effect of ICBs. We performed mIHC on ccRCC patients who
   received postoperative ICB therapy. The results showed that patients
   respond to ICB therapy had low expression of CEBPD, whereas patients
   who did not respond to ICB therapy had significantly elevated
   expression of CEBPD ([145]figure 8A,B). Meanwhile, the response group
   showed a higher level of CD86 and lower CD163 compared with
   non-response group ([146]figure 8A). Additionally, low-CEBPD patients
   had longer overall survival compared with high-CEBPD patients
   ([147]figure 8C). Furthermore, we investigated the association between
   CEBPD and the reported pathways through which TAMs influence ICB
   therapeutic efficacy by TCGA and scRNA-seq ([148]online supplemental
   figure S7). The result showed that CEBPD was not correlated with VEGF
   or TGF-β pathway, while it was negatively correlated with M1 macrophage
   gene set score. These results indicated that higher CEBPD expression in
   TAMs correlated with worse ICB efficacy, which may mainly mediate the
   polarization of TAMs. Additionally, in vivo experiments demonstrated
   enhanced efficacy of anti-PD1 therapy in CEBPD-knockout mice compared
   with wild type ([149]online supplemental figure S10), which indicated
   that targeting CEBPD may serve as a complement to existing
   immunotherapy.

Figure 8. CEBPD correlated with ICB efficacy in patients with RCC. (A)
Representative immunofluorescence images of patient respond to ICB therapy
(left) and patient non-respond to ICB therapy (right), and (B) quantitative
analysis of CEBPD intensity in two groups. (C) Kaplan-Meier curves of
survival probability in CEBPD-high group and CEBPD-low group. CEBPD,
CCAAT/enhancer-binding protein delta; ICB, immune checkpoint blocker; RCC,
renal cell carcinoma.

   [150]Figure 8
   [151]Open in a new tab

Discussion

   Macrophages are an important part of the immune system to eliminate
   tumor cells. In ccRCC, though macrophages are abundant in the TME, it
   seems like macrophages are “hijacked” by tumor cells and lose antitumor
   function. The ratio of M1/M2 TAMs is a dynamic indicator influenced by
   tumor progression and therapy. Low M1/M2 ratio represents an
   immunosuppressive TME and leads to poor clinical outcomes. It is well
   established that TAMs are susceptible to polarization to the M2
   phenotype and contribute to tumorigenesis, metastasis, tumor
   angiogenesis, immunosuppression and drug resistance.[152]30,[153]33 In
   recent years, more specific clusters of TAMs were identified, some of
   which expressed both traditional M1 and M2 markers. But we can still
   broadly categorize TAMs into proinflammatory type and immunosuppressive
   type. Therefore, future research should focus on reversing this trend
   to enhance the ratio of proinflammatory TAMs while reducing
   immunosuppressive types.[154]^34 35

   In this study, bioinformatics analysis indicated that CEBPD negatively
   influences antitumor effects. In vitro experiments showed that the
   knockdown of CEBPD promoted the M1 type polarization of macrophages
   without affecting M2 type polarization. However, multiple
   immune-fluorescence analyses on patient specimens showed that tumors
   with high CEBPD expression had fewer M1-like TAMs and more M2-like TAMs
   compared with tumors with low CEBPD expression. This revealed a
   competitive dynamic between M1 and M2 polarization in the TME.[155]^36

   Previous studies have demonstrated that TAMs impact the efficacy of
   ICBs.[156]37,[157]39 TAM-derived cytokines can influence the expression
   of PD-L1. For instance, in pancreatic ductal adenocarcinoma,
   TAM-derived TGF-β upregulates tumor PD-L1 expression via the AKT/NF-κB
   pathway.[158]^40 In bladder cancer, PEG2 can increase the expression of
   PD-L1 in tumor-infiltrated myeloid cells and exclude CD8^+ T
   cells.[159]^41 Besides, TAM-expressed ligands, such as V-domain
   Ig-containing suppressor of T-cell activation, can block
   anti-PD-1/PD-L1 immune efficacy.[160]42,[161]44 Moreover, TAMs can also
   express PD-1; Gordon et al found that PD-1^+ TAMs exhibited a reduced
   phagocytic capacity compared with PD-1^− TAMs,[162]^45 indicating
   impaired capacity to present tumor antigen and failure to activate T
   cells. In our study, patients with ccRCC who responded to ICB therapy
   showed lower TAM levels of CEBPD, alongside a higher proportion of M1
   TAMs and a lower proportion of M2 TAMs. Prior research has established
   a correlation between macrophage polarization status and the
   therapeutic outcomes of ICBs,[163]46,[164]49 which is consistent with
   our findings. Although the precise mechanism by which CEBPD influences
   anti-PD-1 efficacy remains to be elucidated, our findings suggest that
   CEBPD could be a potential target for immunotherapy in ccRCC.

   We reported for the first time that RGS2 is involved in TAM
   polarization. To confirm that RGS2 is directly regulated by CEBPD, we
   performed ChIP-seq, dual-luciferase assay and EMSA. RGS2, acting as
   GTPase activating protein, modulates G protein alpha subunits and
   drives G proteins into inactive GDP-bound form.[165]^20 21 Prior
   research has indicated that RGS2 plays a suppressive role in
   proinflammatory responses of macrophages. Macrophages derived from
   RGS2-KO mice showed increased response to TLR2 agonist and tended to
   produce more iNOS.[166]^50 Lee et al revealed that RGS2 could repress
   STAT3-mediated transcriptional activation of Nox1, thereby attenuating
   innate immunity.[167]^51 In this study, we discovered that RGS2
   deactivates PAR2 and inhibits the MAPK/NF-κB pathway, resulting in
   blocked M1 polarization and reduced production of IL-1β and IL-6.
   Targeting the RGS2/PAR2 axis holds promise for enhancing M1 TAM
   activation and promoting their proinflammatory functions against
   tumors.

Conclusion

   Our study demonstrated that upregulation of CEBPD is associated with
   impaired M1 polarization in ccRCC. We identified the CEBPD/RGS2/PAR2
   axis involved in this process. Furthermore, our findings suggested that
   targeting CEBPD could serve as a promising immunotherapy approach and a
   complementary strategy to ICB therapies.

Supplementary material

   online supplemental file 1
   [168]jitc-13-7-s001.docx^ (4.9MB, docx)
   DOI: 10.1136/jitc-2024-010898

Footnotes

   Funding: This work was sponsored by the National Natural Science
   Foundation of China (No. 82330094, 82072806, 82173265); Leading health
   talents of Shanghai Municipal Health Commission (2022LJ002); Shanghai
   Rising-Star Program (23QC1401400); the Natural Science Foundation of
   Shanghai (23ZR1441300); Hospital Funded Clinical Research, Xin Hua
   Hospital Affiliated to Shanghai Jiao Tong University School of Medicine
   (21XHDB06).

   Provenance and peer review: Not commissioned; externally peer reviewed.

   Patient consent for publication: Not applicable.

   Ethics approval: This study involves human participants and was
   approved by Ethics Committee of Xinhua Hospital Affiliated to Shanghai
   Jiao Tong University School of Medicine, Approval No. XHEC-D-2024-140.
   Participants gave informed consent to participate in the study before
   taking part.

   Correction notice: This article has been corrected since it was first
   published. The corresponding author information has been updated. 7th
   August 2025.

Data availability statement

   Data are available upon reasonable request.

References