TAF15 in tumor-associated macrophages enhances protumorigenic polarization and promotes cholangiocarcinoma progression

Abstract

Background & Aims

Intrahepatic cholangiocarcinoma (ICC) is a highly malignant and aggressive cancer. Tumor-associated macrophages (TAMs) are integral to the tumor microenvironment (TME), where they facilitate the malignant progression of ICC and reshape the TME. TATA-binding protein-associated factor 15 (TAF15), a protein that binds both DNA and RNA, plays a pivotal role in inflammatory signaling pathways and is abnormally expressed in TAMs in ICC. However, the specific function of TAF15 in ICC-associated macrophages remains to be elucidated. This study aimed to investigate the regulatory effect of TAF15 in ICC-associated macrophages on ICC progression.

Methods

The expression pattern of TAF15 in macrophages was assessed using multicolor fluorescence in ICC mouse tissues and patient samples (n = 5 per group). TAF15 expression in THP-1 cells was manipulated using CRISPR-Cas9 technology. The polarization index of TAMs, as well as the impact of TAMs on ICC proliferation, was evaluated through in vitro coculture. CUT&Tag and dual-luciferase reporter gene assay were used to identify potential regulatory elements of TAF15. M2pepLNP-siTAF15 was designed to target macrophages in a mouse ICC model for in vivo experiments, thereby confirming the role of TAF15 in TAMs on ICC progression (n = 5 per group).

Results

TAF15 is highly expressed in TAMs (p <0.05) and promotes the polarization of macrophages towards the M2 phenotype, thereby furthering the progression of ICC (p <0.01). Mechanistically, TAF15 transcriptionally activates SOCS1 (p <0.001), inhibits the JAK2/STAT1 pathway, and suppresses macrophage polarization towards the M1 phenotype. M2pepLNP-siTAF15 can effectively target TAMs in the treatment of ICC.

Conclusions

TAF15 has a pivotal role in ICC progression by affecting the phenotype of macrophages. Targeting TAF15 in TAMs emerges as a promising therapeutic strategy for the treatment of ICC.

Impact and implications

Our study provides the first evidence that TATA-binding protein-associated factor 15 regulates tumor-associated macrophages polarization through the SOCS1/JAK2/STAT1 axis, unveiling a novel immunotherapeutic target for cholangiocarcinoma. The developed M2pep-LNP-siTAF15 nanodelivery system not only overcomes the challenge of targeted delivery, but its remarkable antitumor efficacy highlights strong potential for clinical translation. This work fundamentally advances our understanding of stromal-immune crosstalk in cholangiocarcinoma while offering a clinically actionable therapeutic strategy.

Graphical abstract

Highlights

• •TAF15 expression is upregulated in cholangiocarcinoma tissues and TAMs.
• •TAF15 drives M2 polarization of TAMs and promotes ICC progression.
• •TAF15 promotes M2 polarization via transcriptional activation of SOCS1 and subsequent suppression of the JAK2/STAT1 pathway.
• •TAF15 drives M2 polarization-dependent TGFBI secretion and ECM remodeling to promote ICC progression.
• •Targeting TAF15 in TAMs using M2pepLNP-siRNA may be a promising therapeutic strategy for the treatment of ICC.

Introduction

Intrahepatic cholangiocarcinoma (ICC) is an aggressive and highly metastatic epithelial malignancy affecting the liver and biliary tract, and the incidence of ICC is increasing annually.¹ ICC is characterized by extensive connective tissue proliferation and a rich tumor immune microenvironment (TIME).² The clinical management of ICC is challenging because of its advanced stage at presentation, resistance to chemotherapy, and high rate of postoperative recurrence, resulting in a 5-year survival rate that remains below 25%.³ Despite the development of various therapeutic strategies, patient outcomes remain poor, largely owing to the inhibitory effects of the tumor microenvironment (TME).⁴

The TIME is complex and is composed primarily of cancer-associated fibroblasts (CAFs), tumor-associated macrophages (TAMs), neutrophils and vascular endothelial cells.⁵ TAMs, a critical component of the TIME, are plastic immune cells that infiltrate and are activated at the tumor site, and they exhibit tumor-promoting and immunosuppressive properties.^6,7 Macrophages polarize into two distinct states in response to external stimuli, namely, classically activated (M1) and alternatively activated (M2) macrophages.⁸ Lipopolysaccharide (LPS) or interferon-gamma activates transcription factors, such as the signal transducer and activator of transcription (STAT) family, nuclear factor κB (NF-κB), and interferon regulatory factor (IRF), initiating an inflammatory response that drives macrophage differentiation into the M1 phenotype, characterized by high expression of IL-1β, IL-6, tumor necrosis factor-alpha (TNF-α), CD86, and CD80.^9,10 Conversely, M2 differentiation is induced by IL-4 and IL-13 cytokines, which activate anti-inflammatory genes such as CD163, CD206, and transforming growth factor-beta (TGF-β).¹¹ The presence of M1 macrophages in solid tumors is generally associated with a better prognosis, whereas the presence of M2 macrophages is associated with increased angiogenesis, metastasis, and poor prognosis in various cancer types.¹² Macrophage polarization is a dynamic process, and inhibiting the differentiation of monocytes into M2-TAMs may offer an effective strategy for targeting TAMs in tumor therapy.

TATA-binding protein-associated factor 15 (TAF15), a member of the FET family that includes FUS (fused in sarcoma) and EWS (Ewing Sarcoma), is ubiquitously expressed across tissues and cells, and it functions as both a DNA- and RNA-binding protein.¹³ TAF15 associates with TBP to form the TFIID multisubunit complex, which initiates gene transcription in conjunction with RNA polymerase II.¹⁴ The N-terminal domain of TAF15 possesses transactivation and oncogenic capabilities.¹⁵ RNA sequencing has revealed that TAF15 influences a multitude of genes integral to the cell cycle and apoptosis, and TAF15 overexpression increases the proliferation of sarcoma, neuroblastoma, and leukemia cells, suggesting a significant role for TAF15 in human cancer progression.¹⁶ TAF15 is also known to be pivotal in several key inflammatory signaling pathways.¹⁷ Our recent study revealed that TAF15 is markedly upregulated in the livers of non-alcoholic steatohepatitis (NASH) model mice, where it interacts with p65, activates the NF-κB signaling pathway, increases the secretion of proinflammatory cytokines, and induces the polarization of M1 macrophages, thereby exacerbating NASH progression.¹⁸ Given that TAF15 is highly expressed in macrophages and that there is a close association between the inflammatory signaling pathway and the macrophage activation phenotype, we hypothesized that TAF15 may significantly contribute to macrophage polarization. Thus, the specific role of TAF15 in TAMs was investigated in the present study.

Materials and methods

Cell lines and cell culture

HuCCT1 cells were purchased from JCRB (Osaka, Japan). The HCCC9810, RBE, and THP-1 cell lines were purchased from the Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences (Shanghai, China). THP-1 cells were differentiated to a macrophage-like state via incubation with 200 nM phorbol 12-myristate 13-acetate (PMA) (GLPBIO, Shanghai, China), and they were then cocultured with tumor cells in a cell chamber (Corning, Shanghai , China).

To isolate liver macrophages, liver perfusion was performed using a preheated Hanks’ Balanced Salt solution containing calcium and magnesium at 37 °C (Beyotime), followed by 0.3% type IV collagenase (C6885-1G, Sigma). After digestion, the cells were removed and filtered with a 70-μm cell filter (Corning). The cell suspension was centrifuged at 4 °C and 1,500 rpm for 5 min, and the cell pellet was resuspended in Hanks’ Balanced Salt solution. After centrifugation at 50 × g for 5 min, the supernatant was retained. After an additional centrifugation at 50 × g for 5 min, the supernatant was retained. Finally, the cell pellet was collected after centrifugation at 350 × g for 10 min. The proportion of macrophages was measured by CD11b and F4/80.

All cells were cultured in RPMI 1640 or DMEM (Invitrogen, Waltham, MA, USA) supplemented with 10% FBS (Biological Industries, Cromwell, CT, USA), 100 U/ml penicillin (Invitrogen) and 100 μg/ml streptomycin (Invitrogen) in a humidified incubator with 5% CO₂ at 37 °C.

Animal studies

Male C57BL/6 mice (5–6 weeks old) and BALB/c nude mice (5 weeks old) were purchased from GemPharmatech Co., Ltd. (Nanjing, Jiangsu, China) and the mice were housed in a specific pathogen-free environment.

To generate the ICC model, 20 μg of the pT3-EF1α-HA-myr-AKT plasmid and 30 μg of the pT3-EF1α-YapS127A plasmid were injected into C57BL/6 mice via hydrodynamic injection. All the plasmids, including Sleeping Beauty transposase, pT3-EF1α-HA-myr-AKT, and pT3-EF1α-YapS127A, were kind gifts from Dr Zhang (Department of Pathology, Eastern Hepatobiliary Surgery Hospital, Second Military Medical University, Shanghai, China).

To construct a subcutaneous xenotransplantation model, HuCCT1 cells (2 × 10⁸) were cocultured with 1 × 10⁸ PMA-KO TAF15 THP-1 or PMA-THP-1 as a control, and then the cell mixtures were subcutaneously injected into the groin of 5-week-old male BALB/c nude mice (n = 5). The tumor size was measured with Vernier calipers, and the tumor volume was calculated using the following formula: tumor volume = length × (width)² × 0.5.

siNc (5′-ACGUGACACGUUCGGAGAATT-3′) and siTaf15 (5′-CAGUCAAGGCUAUGGACAATT-3′) with 2′-O-methylation (2′-OMe) modifications were synthesized by Corues Biotechnology. M2pepLNPs loaded with siRNA were prepared by Xi’an Ruixi Technology Co., Ltd. M2pepLNP-siRNA was injected through the tail vein at a dose of 2 mg/kg every 2 days for 13 days.

All experiments utilizing animals were reviewed and approved by the Ethical Committee and Animal Welfare Committee of Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School, Jiangsu, Nanjing, China.

Patient specimens

Pathologically proven ICC tissues were collected from five patients who underwent radical surgery at Drum Tower Hospital, Medical School of Nanjing University (Jiangsu, Nanjing, China). All experiments involving human samples were approved by the Medical Research Ethics Committee of Drum Tower Hospital Affiliated with the School of Medicine of Nanjing University. The patient information is shown in Table S1.

Colony formation assay

After counting, the cells were seeded into a six-well plate (1,000 cells/well) containing complete medium and cultured for 14 days according to different treatments. Representative images of the colonies were acquired, and the colonies were counted. A cell mass containing more than 50 cells was considered a colony. The individual wells of the six-well plate were scanned at 4 × magnification using an M7000 fluorescence microscope, and the number of colonies in the fields of view was counted to create a bar chart for the corresponding treatment groups.

Immunohistochemistry and immunofluorescence

The tumor tissue was embedded in paraffin wax, and immunohistochemistry (IHC) was performed according to standard procedures. The tumor slides were dewaxed, blocked, incubated with primary antibodies (overnight, 4 °C), and incubated with secondary antibodies (1 h, room temperature). Biotin-labeled secondary antibodies were used for IHC, and fluorescently labeled secondary antibodies were used for immunofluorescence. The following antibodies were used at various concentrations: anti-TAF15 (1:200, Affinity, DF2424), anti-F4/80 (1:50, Abcam, ab300421), and anti-CK19 (1:200, Proteintech, 10712-1-AP).

Flow cytometry

After treatment for the indicated times, the samples were detected by flow cytometry. The following flow cytometry antibodies were used: anti-human CD206 (551135, BD), anti-human CD206 (561763, BD), anti-mouse CD11b (101227, Biolegend), anti-mouse F4/80 (123109, Biolegend), anti-mouse CD86 (105013, Biolegend), anti-mouse CD206 (141703, Biolegend), and Alexa Fluor 488-labeled goat anti-rabbit IgG (H+L) (A0423, Beyotime).

Apoptosis was detected with an Annexin V-FITC Apoptosis Detection Kit (Beyotime).

To account for batch-to-batch variations in staining intensity and inherent differences in cell size, thresholds were dynamically adjusted based on experimental conditions. Compensation was calibrated using single-stain controls across all fluorescence channels. Our optimized gating strategy precisely delineated cell populations by adapting to distinct clustering patterns.

The samples were analyzed with a BD FACS Canto II flow cytometer (BD Biosciences, Milpitas, CA, USA) and FlowJo software (FlowJo_v10.8.1). FlowJo was provided by FlowJo, LLC (Ashland, OR, USA); Com Win was supplied by Com Win Biotechnology Co., Ltd. (Beijing, China).

RNA extraction and real-time quantitative PCR

Total RNA was extracted via TRIzol reagent (Invitrogen) according to the manufacturer’s instructions and reverse transcribed. A LightCycler and a SYBR real-time PCR (RT-PCR) kit (Com Win Biotech) was used for quantitative RT-PCR (qRT-PCR) analysis. The PCR primers used are shown in Table S2. The data were normalized to the level of actin expression in each sample.

Western blot analysis

Samples, including tissues and cells, were lysed in RIPA lysis buffer (Biosharp) supplemented with protease and phosphatase inhibitor cocktail (Roche Diagnostics GmbH, Mannheim, Germany) and phenylmethylsulfonyl fluoride (Biosharp, Hefei, China) for 15 min on ice. The proteins were subjected to Western blot analysis according to standard protocols. The following antibodies were used: anti-β-actin (Sigma, A5441, St. Louis, Missouri, USA), anti-TAF15 (Affinity, DF3382, Cincinnati, OH, USA), anti-cleaved PARP1 (Cell Signaling Technology, #9541), anti-cleaved caspase3 (Cell Signaling Technology, #9661, Danvers, Massachusetts, USA), anti-BAX (Cell Signaling Technology, #2772, Danvers, MA, USA), anti-Suppressor of cytokine signaling1 (SOCS1) (Absin, abs118094, Shanghai, China), anti-CD206 (ABclonal, A21014, Wuhan, Hubei, China), anti-Janus Kinase2 (JAK2) (Selleck, F0231, Houston, Texas, USA), anti-phospho-JAK2 (ABclonal, Y1007/1008, Wuhan, Hubei, China), anti-STAT1 (Selleck, F3253, Houston, Texas, USA), and phospho-STAT1 (S727) (Selleck, F0263, Houston, Texas, USA).

Cleavage under targets and tagmentation (CUT&tag)

For cleavage under targets and tagmentation (CUT&Tag) analysis, a Hyperactive Universal CUT&Tag Assay Kit from Vazyme (Nanjing, Jiangsu, China) was utilized. In brief, target cells were collected and counted, and the cell concentration was adjusted (usually 50,000–500,000 cells). After washing the cells two to three times with cold PBS, they were collected by centrifugation (500 × g, 5 min, 4 °C). The cells were resuspended in permeabilization buffer (0.1% Tween-20) and incubated on ice for 10 min. The cells were then washed twice with washing buffer, centrifuged, and resuspended in antibody binding buffer. Specific primary antibodies for target proteins were added, and the samples were incubated overnight at 4 °C or at room temperature for 2 h. The cells were washed two to three times with washing buffer, centrifuged, and resuspended in antibody binding buffer. Protein A/G-coupled secondary antibodies were added, and the samples were incubated at room temperature for 1 h. The cells were washed two to three times with washing buffer, centrifuged, and resuspended in Tn5 reaction buffer. Tn5 transposition enzyme (preloaded sequencing connector) was added, and the mixture was incubated at room temperature for 1 h. The Tn5 transposition enzyme cuts the DNA and simultaneously adds a sequencing splice. DNA fragments were extracted using DNA purification reagents. The DNA fragments were eluted (usually with 10–20 μl of elution buffer), and PCR amplification was performed using primers with sequencing connectors (typically 12–15 cycles). The reaction system included a DNA template, PCR primers, dNTPs and high-fidelity DNA polymerase. PCR purification kits were used to purify the amplified products for further quantitative detection by qPCR or agarose gel electrophoresis. The following antibodies were used: anti-TAF15 (1:50, Affinity, DF3382), anti-IgG (1:50, Abcam, ab200699).

Luciferase reporter assays

293T cells were transfected with pGL3-basic-Luc, pGL3-promoter-Luc, PCDNA3.1+SOCS1-WT, TAF15+SOCS1-MUT1, or TAF15+SOCS1-MUT2 using Lipofectamine 2000 Transfection Reagent according to the manufacturer’s instructions. Reporter gene activity was determined by normalization of the firefly luciferase activity to Renilla luciferase activity.

RNA sequencing

Total RNA was extracted from the tissue using TRIzol® Reagent according the manufacturer’s instructions. Then RNA quality was determined by 5300 Bioanalyser (Agilent) and quantified using the ND-2000 (NanoDrop Technologies). Only high-quality RNA sample (OD260/280 = 1.8∼2.2, OD260/230 ≥2.0, RQN ≥6.5, 28S:18S ≥1.0, >1 g) was used to construct sequencing library.

RNA purification, reverse transcription, library construction and sequencing were performed at Shanghai Majorbio Bio-pharm Biotechnology Co., Ltd. (Shanghai, China) according to the manufacturer’s instructions (Illumina, San Diego, CA, USA). The RNA-seq transcriptome library was prepared following Illumina® Stranded mRNA Prep Ligation from Illumina using 1 μg of total RNA. In short, firstly, messenger RNA was isolated according to polyA selection method by oligo (dT) beads and then fragmented by fragmentation buffer. Secondly, double-stranded cDNA was synthesized using a SuperScript double-stranded cDNA synthesis kit (Invitrogen) with random hexamer primers (Illumina). Then the synthesized cDNA was subjected to end-repair, phosphorylation and ‘A’ base addition according to Illumina’s library construction protocol. Libraries were size selected for cDNA target fragments of 300 bp on 2% Low Range Ultra Agarose followed by PCR amplified using Phusion DNA polymerase (NEB) for 15 PCR cycles. After quantified by Qubit 4.0, paired-end RNA-seq sequencing library was sequenced with the NovaSeq X plus sequencer (2 × 150-bp read length).

The raw paired end reads were trimmed and quality controlled by fastp with default parameters. Then clean reads were separately aligned to reference genome with orientation mode using HISAT2 software. The mapped reads of each sample were assembled by StringTie in a reference-based approach.

Enzyme-linked immunosorbent assay

Macrophages were stimulated with tumor conditional media (CM). After 48 h, the CM was collected, and the sample was diluted according to the recommended concentration for detection. The transforming growth factor-beta induced (TGFBI) ELISA test kit was purchased from CUSABIO (CSB-EL023450MO, CSB-E16665h, Wuhan, China).

Small interference transfected BMDM cells

Bone marrow-derived macrophages (BMDM) were isolated from tibias and femurs, and they were cultured with DMEM containing 20% FBS, glutamine, and 40 ng/ml M-CSF (ABclonal) for 7 days before harvesting adherent BMDM.

BMDM cells were inoculated into the culture plate to achieve a confluence of 70% cell density at transfection 1 day before transfection. For transfection, siRNA was first diluted to 100 nM in serum-free Opti-MEM medium with gentle mixing to avoid aggregation. Lipofectamine 2000 transfection reagent was separately prepared by diluting in Opti-MEM at a 1:50 ratio according to the manufacturer's protocol, followed by 5 min of incubation at room temperature. The diluted siRNA and transfection reagent were then combined at a 1:1 ratio, mixed gently by pipetting, and allowed to form complexes during a 20-min incubation at room temperature. Before transfection, the culture medium was aspirated and cells were washed once with PBS. The siRNA-lipid complexes were subsequently added dropwise to cells in serum-free medium, with gentle swirling to ensure even distribution. Cells were maintained under standard culture conditions (37 °C, 5% CO₂) for 4–6 h before replacing the transfection medium with complete medium. Knockdown efficiency was typically assessed after 24–72 h of additional culture.

Liposome generation

A lipid mixture composed of DOTAP, DOPE, cholesterol, and DSPE-PEG2000-M2pep was dissolved in 2 ml of anhydrous ethanol and transferred to a pear-shaped flask. Moreover, the siRNA was dissolved in a citrate buffer (50 mM citrate, pH = 4) containing 25% ethanol and then gradually added to the lipid mixture under gentle mixing. The resulting mixture was incubated for 20 min to facilitate complex formation. The preparation was subsequently sonicated and extruded through a liposome extruder equipped with a 100 nm polycarbonate membrane to achieve a uniform particle size. To remove unencapsulated siRNA, the solution was dialyzed using a nanodialysis device (polycarbonate membrane with a 30-nm pore size).

Statistical analysis

The results are reported as the means ± standard deviations of at least three independent experiments. Unpaired and paired bilateral Student’s t tests were used to assess statistical significance for two samples, and univariate ANOVA was used to assess statistical significance for two or more samples. The error bars on the graph represent the group mean ± standard error. Significance levels are indicated by asterisks (∗∗∗∗p <0.0001, ∗∗∗p <0.001, ∗∗p <0.01, and ∗p <0.05), and the values are provided in the graphs and/or legends.

To identify differentially expressed genes (DEGs) between two different samples, the expression level of each transcript was calculated according to the transcripts per million reads method. RSEM (Version 1.3.3, University of Wisconsin–Madison, Madison, WI, USA) was used to quantify gene abundances. Essentially, differential expression analysis was performed using the DESeq2 with |log₂fold change [FC]| ≧2 and p value ≤0.05 were considered to be significantly DEGs. In addition, functional-enrichment analysis including Reactome (version 86) was performed to identify which DEGs were significantly enriched and metabolic pathways at Bonferroni-corrected p value ≤0.05 compared with the whole-transcriptome background. Reactome enrichment analysis were carried out using clusterProfiler. Our RNA-seq data reference is: GRCh38 (Ensembl_111).

All data were analyzed using GraphPad Prism (version 6, GraphPad Software, San Diego, CA, USA).

Results

TAF15 is abnormally expressed in ICC and TAMs

To investigate the expression of TAF15 in ICC, we queried the TCGA public database and discovered that TAF15 was upregulated in ICC tissues (Fig. 1A). A YAP/AKT-driven ICC model was established via the Sleeping Beauty hydrodynamic system. Western blot analysis revealed greater TAF15 protein expression in tumor tissues than in adjacent nontumorous tissues (Fig. 1B). Furthermore, IHC revealed elevated CK19 and TAF15 expression in the model group compared with the normal group, which suggested high TAF15 expression in ICC (Fig. 1C). Multicolor immunofluorescence staining and flow cytometry of liver tissues from normal and model mice was performed to detect TAF15 expression in macrophages. Significantly higher TAF15 expression levels were present in macrophages from ICC tissues than in those from normal liver tissue (Fig. 1D, Fig. S1A and B). Together, these results confirmed that TAF15 expression in macrophages associated with cholangiocarcinoma is greater than that in macrophages from normal liver tissue.

Fig. 1.

The expression of TAF15 is upregulated in ICC tissues and TAMs. (A) Expression profile of TAF15 in cholangiocarcinoma (CHOL) based on the TCGA database. (B) Schematic diagram of in situ cholangiocarcinoma modeling in mice and Western blot analysis of TAF15 protein expression in mouse in situ cholangiocarcinoma. (C) H&E and immunohistochemistry analysis of CK19 and TAF15 for tissues of normal liver and tumor. Scale bar = 50 μm. (D) Representative images of F4/80 and TAF15 co-staining in liver sections. Scale bar = 200 μm. (E) QPCR analysis of TAF15 in macrophage from normal liver and YAP/AKT liver. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (F) Representative images of CD68 and TAF15 co-staining in liver sections from patients with cholangiocarcinoma, n = 5. Scale bar = 200 μm. (G) Schematic diagram of the coculture system involving tumor cells and macrophages. (H) QPCR analysis of CD86, HLA-DR, CD163, and CD206 in THP-1 and TAMs. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (I) Morphology of THP-1, PMA-THP-1, and TAMs. Scale bar = 50 μm. (J,K) QPCR and Western blot analysis of TAF15 in PMA-THP-1 and TAMs. Data are presented as mean ± SD, one-way ANOVA was performed for statistical analysis, n = 3. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. ICC, intrahepatic cholangiocarcinoma; P, para-tumor; T, tumor; PMA, phorbol 12-myristate 13-acetate; TAMs, tumor-associated macrophages; TAF15, TATA-binding protein-associated factor 15.

Given that TAMs play an immunosuppressive role in the ICC microenvironment and TAF15 is aberrantly expressed in macrophages in a fatty liver model, a recognized risk factor for cholangiocarcinoma, we investigated the role of TAF15 in TAMs.¹⁸ By isolating macrophages from the liver tissue of mice with cholangiocarcinoma, we confirmed that TAMs presented higher TAF15 expression than normal macrophages (Fig. 1E). We subsequently collected tumor and distal normal liver tissue samples from five patients with ICC for analysis, which revealed that TAF15 expression was significantly higher in human ICC-associated macrophages than in distal normal liver tissue (Fig. 1F, Fig. S1C). In addition, TCGA-CHOL data revealed that TAF15 expression was positively correlated with the expression of TAM inhibitory immune molecules, such as TREM2, ARG2, and PD-L1 (Fig. S1D–F). These results suggested that high TAF15 expression may help shape the inhibitory immune microenvironment.

Next, we cocultured macrophages differentiated from the THP-1 cell line with the HuCCT1, RBE, and HCCC9810 human bile duct cancer cell lines. The cells were treated with PMA to stimulate the formation of TAMs and simulate the TME of ICC, thereby constructing an ICC-TAMs model (Fig. 1G). Compared with THP-1 cells, qPCR analysis revealed increased levels of macrophage markers in TAMs (Fig. 1H), and morphological assessment indicated that TAMs underwent elongation and extension (Fig. 1I). Moreover, TAF15 expression in macrophages cocultured with tumor cells was markedly upregulated compared with that in PMA-stimulated THP-1 cells (Fig. 1J and K, Fig. S1G). Collectively, these results suggested that TAF15 expression is upregulated in TAMs in ICC and that these TAMs exhibit an immunosuppressive phenotype.

TAF15 facilitates the polarization of TAMs towards the M2 phenotype

After identifying the expression profile of TAF15 in TAMs in ICC, we examined the impact of TAF15 on macrophage polarization in vitro. Using the CRISPR-Cas9 genome editing system, we generated THP-1 cells with TAF15 knockout (KO) and overexpression (Fig. 2A). These cells were induced to differentiate into M0 macrophages using PMA, after which they were cocultured with tumor cells for 48 h to form TAMs, and the macrophage polarization index was then assessed. Western blot analysis revealed that CD206 protein expression was significantly reduced in KO TAF15 TAMs but was increased in TAF15-overexpressing (OE) TAMs (Fig. 2B). The qPCR results indicated that the M1-like macrophage polarization markers were upregulated in KO TAF15 TAMs, whereas the M2-like macrophage polarization markers were downregulated in KO TAF15 TAMs. Conversely, OE TAF15 TAMs presented the opposite pattern (Fig. 2C–F). Compared with NC TAMs, flow cytometry revealed that CD206 expression in KO TAF15 TAMs was downregulated, whereas CD206 expression in OE TAF15 TAMs was significantly upregulated (Fig. 2G–J). These findings suggested that TAF15 in TAMs promotes the polarization of TAMs towards the M2 phenotype.

Fig. 2.

TAF15 promotes the polarization of ICC-associated macrophages towards M2. (A,B) Expression profile of TAF15 and CD206 in TAMs following TAF15 knockdown and overexpression. (C–F) QPCR analysis of CD206, CD163, TGF-β, CD86, TNF-α, and IL-6 in TAMs. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 6. (G–J) Flow cytometry analysis of CD206 expression in TAMs and KO/OE TAF15 TAMs cocultured with ICC cells. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. ICC, intrahepatic cholangiocarcinoma; KO, knockout; OE, overexpressing; QPCR, quantitative PCR; TAMs, tumor-associated macrophages; TAF15, TATA-binding protein-associated factor 15; TGF-β, transforming growth factor-beta; TNF-α, tumor necrosis factor-alpha.

TAF15 in macrophages associated with ICC promotes the proliferation of cholangiocarcinoma cells both in vitro and in vivo

To further elucidate the role of TAF15 in TAMs in cholangiocarcinoma progression, we performed plate cloning experiments using tumor cells cocultured with TAMs and KO TAF15 TAMs (Fig. 3A and B). The proliferation of HuCCT1 and RBE cells was suppressed when these cells were cocultured with KO TAF15 TAMs. The expression of apoptotic proteins in HuCCT1 and RBE cells, as well as apoptosis analysis by flow cytometry, demonstrated an increase in the level of apoptosis in tumor cells following coculture with KO TAF15 TAMs, whereas the opposite effect was observed following coculture with OE TAF15 TAMs (Fig. 3C–E, Fig. S3A and B). HuCCT1 cells mixed with TAMs or KO TAF15 TAMs were subcutaneously injected into BALB/c mice to establish subcutaneous tumor models (Fig. 3F). In vivo assessments demonstrated that the HuCCT1+KO TAF15 TAMs group presented the smallest tumor volumes and weights, whereas the HuCCT1+TAMs group presented the largest tumor volumes and weights (Fig. 3G–J, Fig. S3C). IHC analysis further revealed that the HuCCT1 + KO TAF15 TAMs group had the lowest Ki67 labeling index (Fig. S3D). Collectively, these findings suggested that KO TAF15 in TAMs inhibited tumor progression both in vitro and in vivo.

Fig. 3.

Inhibition of TAF15 in TAMs can inhibit tumor growth. (A,B) Colony formation ability of ICC cells cocultured with TAMs or KO TAF15 TAMs, Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (C) Western blot analysis was conducted to measure the protein expression levels of cleaved PARP-1 (C-PARP1), cleaved CASPASE3 (C-CASPASE3), and BAX in ICC cells cocultured with TAMs or KO TAF15 TAMs, with ACTIN serving as a loading control. (D,E) Flow cytometry analysis the proportion of apoptosis. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (F) Schematic diagram of subcutaneous tumor experiment. (G,H,I) A comparison of subcutaneous tumor size and weight in nude mice following different treatment. Data are presented as mean ± SD, one-way ANOVA was performed for statistical analysis, n = 5. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001. ICC, intrahepatic cholangiocarcinoma; KO, knockout; TAMs, tumor-associated macrophages; TAF15, TATA-binding protein-associated factor 15.

TAF15 regulates macrophage polarization by transcriptionally activating SOCS1

To elucidate the specific mechanisms through which TAF15 influences the polarization of TAMs and the subsequent malignant progression of cholangiocarcinoma, we cocultured HuCCT1 cells with either TAMs or KO TAF15 TAMs and performed RNA sequencing, which revealed a significant number of differentially expressed genes (Fig. 4A). The sequencing data indicated that the mRNA expression levels of M1-related proinflammatory genes, including CDK1, CD80, IL-1β, and IL-6, were increased, whereas the mRNA expression levels of M2-related immunosuppressive genes, including MRC1, FABP5, CCL2, SPP1, TREM2, and CD163, were decreased, which aligns with the polarization study results (Fig. 4B). These findings suggest that TAF15 in TAMs drives macrophage polarization towards an M2-like phenotype.

Fig. 4.

TAF15 in TAMs activates SOCS1, inhibiting the JAK2/STAT1 signaling pathway. (A) Volcano plot of differentially and non-differentially expressed genes, as revealed by RNA sequencing analyses comparing KO TAF15 TAMs and control TAMs cocultured with HuCCT1, is presented. (B) A heatmap illustrates the gene expression patterns of M1 and M2 macrophage-related genes in KO TAF15 TAMs vs. control TAMs. (C) Downregulated differential genes between TAMs and KO TAF15 TAMs were identified with a threshold of |log₂FC| > 0.5 and p value adjusted (pval adj) <0.05. Additionally, target genes predicted to bind to the TAF15 protein from the ChIP-Atlas database were intersected with the differential genes from both groups. (D) A heatmap displays the top 10 differentially expressed genes between the two groups. (E,F) QPCR and Western blot analyses of SOCS1 in KO TAF15 and OE TAF15 TAMs were conducted. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (G,H) SOCS1 expression in the different groups was detected using the TAF15-targeted primer in the cut tag assay, with results displayed by QPCR and gel electrophoresis. (I) A schematic diagram depicts the luciferase reporter construct containing the SOCS1 promoter and the mutant construct (SOCS1-WT and Mut). (J) Luciferase activity driven by SOCS1 promoters containing wild-type or mutated putative TAF15 binding sites in response to wild-type TAF15 overexpression in 293T cells. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 3. (K) Western blot analysis depicts the expression patterns of the JAK2/STAT1 signaling pathway across different groups. Data are presented as mean ± SD, unpaired two-tailed Student’s t test. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. SOCS1, Suppressor of cytokine signaling1; JAK2, Janus Kinase2; STAT1, signal transducer and activator of transcription1; KO, knockout; QPCR, quantitative PCR; TAMs, tumor-associated macrophages; TAF15, TATA-binding protein-associated factor 15.

Given that TAF15 is a DNA-binding protein, we hypothesized that TAF15 functions as a transcription factor to regulate the transcriptional activation of key molecules involved in macrophage polarization. We intersected the downregulated differential genes identified by RNA sequencing with the potential downstream genes in the ChIP-Atlas database predicted to bind to the TAF15 protein, which yielded 221 overlapping genes (Fig. 4C). Based on RNA sequencing differential expression analysis, we re-ranked the 221 overlapping genes according to their log₂FC (from highest to lowest), with subsequent focus on the top 10 genes showing the most pronounced expression changes (Fig. 4D). Among these genes, the RIMBP3C gene, which had a low basal expression level, is not related to macrophage polarization according to literature searches. Thus, we focused on the second-ranked gene, namely, SOCS1. SOCS1, also known as a suppressor of cytokine signaling, is a negative feedback regulator of the JAK-STAT signaling pathway.¹⁹ SOCS1 negatively regulates this pathway by inhibiting STAT phosphorylation and dimer formation or by directly inhibiting JAK phosphorylation, thereby inhibiting the sustained proliferation and differentiation of cells.²⁰ Recent studies have indicated that SOCS1 inhibits M1 polarization and promotes M2 polarization in macrophages by suppressing the JAK2/STAT1 pathway.²¹ Detection of SOCS1 mRNA and protein expression in KO/OE TAF15 TAMs revealed that the SOCS1 protein and RNA expression levels were decreased in KO TAF15 TAMs but increased in OE TAF15 TAMs compared with control TAMs (Fig. 4E and F, Fig. S4A and B).

To determine whether TAF15 binds to the promoter region of SOCS1, we used a CUT&Tag assay utilizing a TAF15 antibody to precipitate the SOCS1 promoter DNA fragment. The resulting DNA fragments were amplified via PCR with specific primers and subsequently analyzed using qPCR and agarose gel electrophoresis (Fig. 4G and H). The CUT&Tag assay revealed that TAF15 was bound to the SOCS1 promoter. These results suggested that TAF15 directly promotes the transcriptional activation of SOCS1. Next, mutations were constructed with CUT&Tag-specific primers (Fig. 4I). Dual-luciferase assays revealed that TAF15 bound to both sites of SOCS1 DNA (1,375–1,394 and 1,444–1,458), indicating that TAF15 binds multiple sites in the SOCS1 promoter (Fig. 4J). Furthermore, we examined the expression levels of JAK2/STAT1 pathway-related proteins in KO/OE TAF15 TAMs. The phosphorylation levels of STAT1 and JAK2 were increased in KO TAF15 TAMs but suppressed in OE TAF15 TAMs (Fig. 4K).

These data suggested that TAF15 directly transcriptionally activated SOCS1, thereby inhibiting the JAK2/STAT1 pathway and influencing macrophage polarization.

TAF15 suppresses the JAK2/STAT1 signaling pathway through the transcriptional activation of SOCS1 to promote M2-like polarization and enhance the progression of ICC

To further validate the relationships among the TAF15-mediated transcriptional activation of SOCS1, subsequent inhibition of the JAK2/STAT1 pathway, and impact on macrophage polarization, we conducted a rescue experiment. Overexpression of SOCS1 in TAMs promoted the polarization of macrophages towards the M2 phenotype, inhibited the JAK2/STAT1 signaling pathway, and inhibited the apoptosis of tumor cells (Fig. S5A–E). We next overexpressed SOCS1 in KO TAF15 THP-1 cells and induced their differentiation into TAMs (Fig. 5A and B). SOCS1 overexpression upregulated CD163 and CD206 mRNA expression but downregulated TNF-α and IL-6 mRNA expression in KO TAF15 TAMs (Fig. 5C, Fig. S5F). Western blot and flow cytometry analyses revealed that SOCS1 overexpression counteracted the TAF15 KO-mediated reduction in CD206 expression in TAMs (Fig. 5D and E). Additionally, SOCS1 overexpression suppressed the activation of the JAK2/STAT1 pathway in TAMs following KO TAF15 (Fig. 5F). Compared with tumor cells stimulated by CM from KO TAF15 TAMs, tumor cells stimulated by CM from KO TAF15 TAMs OE SOCS1 increased cell proliferation and mitigated the apoptosis induced by CM from KO TAF15 TAMs (Fig. 5G, Fig. S5G). We then treated KO TAF15 TAMs with the ruxolitinib JAK2 inhibitor. Low-dose ruxolitinib inhibited the decrease in CD206 expression levels and the activation of STAT1 induced by KO TAF15 (Fig. 5I–K). The colony formation assay revealed that tumor cells treated with CM from KO TAF15 TAMs and low-dose ruxolitinib exhibited enhanced proliferation compared with those treated with CM from KO TAF15 TAMs alone, but there was no significant difference between tumor cells treated with CM from wild-type TAMs and those treated with CM from KO TAF15 TAMs (Fig. 5L, Fig. S5H).

Fig. 5.

The overexpression of SOCS1 in TAMs can rescue the polarized phenotype induced by TAF15 deletion. (A,B) QPCR and Western blot analyses were conducted to assess SOCS1 expression in TAMs with an empty vector and in TAMs overexpressing SOCS1. (C) QPCR analysis of macrophage markers in different treatments is presented. Data are presented as mean ± SD, unpaired two-tailed Student’s t test, n = 6. (D) Western blot analysis of CD206 in TAMs, KO TAF15 TAMs, and KO TAF15 OE SOCS1 TAMs cocultured with HuCCT1. (E) Flow cytometry analysis the expression of CD206 in different treatments. (F) Impact of JAK2/STAT1 pathway expression under various treatments was evaluated by Western blot. (G) Colony formation ability of HuCCT1 cells cocultured with TAMs, KO TAF15 TAMs, and KO TAF15 OE SOCS1 TAMs was examined. (H) Western blot analysis of BAX, cleaved CASPASE3 (C-CASPASE3), and cleaved PARP1 (C-PARP1) under different treatments is shown. (I,J) Western blot and flow cytometry analysis of CD206 in different groups. (K) Western blot analysis of STAT1 and p-STAT1 in different groups. Ruxo = 0.1 μm. (L) Colony formation ability of different groups. Ruxo = 0.1 μm. Data are presented as mean ± SD, one-way ANOVA was performed for statistical analysis, n = 3. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. KO, knockout; OE, overexpressing; QPCR, quantitative PCR; TAMs, tumor-associated macrophages.

Together, these findings demonstrated that TAF15 in TAMs downregulated the JAK2/STAT1 signaling pathway through the transcriptional activation of SOCS1, which inhibited M1 polarization but promoted M2 polarization, thereby impacting tumor progression.

TAF15 in TAMs facilitates M2-like polarization and enhances the secretion of TGFBI to promote tumor progression

To identify the intermediary molecules of TAF15 in TAMs that influence tumor cell survival, we analyzed the RNA sequencing data, which revealed significant enrichment of the extracellular matrix (ECM) remodeling pathway (Fig. 6A). The ECM formed by cancer cells regulates immune cells, leading to an immunosuppressive microenvironment that impacts tumor progression.²² On the basis of this gene set, we sorted the genes by log FC, which indicated that TGFBI presented the most pronounced downregulation following TAF15 KO (Fig. 6B). TGFBI, an ECM protein induced by TGF-β, is abnormally expressed in various cancers. A recent study has indicated that TGFBI is predominantly secreted by M2-like TAMs in glioblastoma multiforme.²³ Analysis of publicly available single-cell data from the GSE138709 dataset confirmed that TGFBI is also expressed primarily by macrophages in human cholangiocarcinoma samples (Fig. 6C). Further analysis revealed that TGFBI protein expression analysis was significantly reduced in KO TAF15 TAMs, and the opposite effect was observed in OE TAF15 TAMs (Fig. 6D). Additionally, recent studies have identified TGFBI as a potential biomarker for cholangiocarcinoma. The soluble form of the TGFBI protein specifically increases the malignancy of cholangiocarcinoma cells by activating the integrin β-1-dependent PPARγ signaling pathway.²⁴ Thus, we hypothesized that TAF15 in TAMs inhibits the JAK2/STAT1 pathway through the transcriptional activation of SOCS1, thereby promoting M2 polarization and TGFBI secretion, which in turn advances the malignant progression of bile duct carcinoma. To test this hypothesis, we treated bile duct cancer cells with recombinant TGFBI protein, which enhanced the proliferation of ICC cells (Fig. 6H). We next used ELISA to measure TGFBI expression in the supernatant of TAMs cocultured with tumor cells. TGFBI expression in the supernatant of KO TAF15 TAMs was significantly lower than that in the supernatant of TAMs. In addition, SOCS1 overexpression in TAMs and the addition of ruxolitinib rescued the downregulation of TGFBI caused by KO TAF15 (Fig. 6F, Fig. 6G–I).

Fig. 6.

TAF15 facilitates M2-like polarization of macrophages, enhancing TGFBI secretion and promoting the progression of ICC. (A) Reactome enrichment analysis of DEGs is depicted, with the top 10 enriched terms displayed. (B) Expression heatmaps of extracellular matrix organization. (C) Upper: a UMAP plot is colored by different clusters. Lower: a dot plot represents the mean expression of TGFBI across various clusters. (D) Western blot analysis examines the expression of TGFBI in different groups. (H) The colony formation ability of ICC cells treated with 0, 5, and 10 μg/ml recombinant TGFBI protein is assessed. (F–I) ELISA analysis measures the levels of TGFBI in the different CM of TAMs. Data are presented as mean ± SD, one-way ANOVA was performed for statistical analysis, n = 3. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. CM, conditional media; DEGs, differentially expressed genes; ICC, intrahepatic cholangiocarcinoma; TAF15, TATA-binding protein-associated factor 15; TGFBI, transforming growth factor-beta induced.

M2 peptide-modified liposomes targeting TAF15 in TAMs inhibit ICC progression

Given the absence of a TAF15 inhibitor and the challenges in targeting TAF15 within macrophages, we used M2 peptide-modified liposomes (M2pepLNP) formulations to deliver siRNAs against TAF15. We initially designed M2pepLNPs that encapsulated IR780, a near-infrared fluorescent dye. We subsequently established a YAP/AKT-driven cholangiocarcinoma mouse model and administered free IR780 or M2pepLNP/IR780 intravenously via the tail vein. Compared with free IR780, M2pepLNP/IR780 exhibited superior tumor targeting both in vivo and in vitro (Fig. 7A and B). Similarly, we coated M2pepLNPs with Dil, a red fluorescent dye with maximum excitation and emission wavelengths of 549 nm and 565 nm, respectively, and administered them intravenously via the tail vein in mice to assess cellular targeting. The self-luminescence of Dil and the F4/80 macrophage marker exhibited a high degree of colocalization, as determined by flow cytometry and immunofluorescence in frozen sections (Fig. 7C, Fig. S6A). These findings revealed precise targeting of M2pepLNPs to macrophages.

Fig. 7.

M2pepLNP-siTaf15 shows promise in targeting macrophages for the treatment of ICC. (A,B) In vivo and in vitro small animal imaging was conducted using free IR780 and M2pepLNP/IR780 in ICC mice via the tail vein. (C) Self-luminescence (Dil) with F4/80 multicolor immunofluorescence. (D,E) Western blot and QPCR analysis of Taf15 in BMDM after transfection with small interference for 48 h. (F) A schematic diagram illustrates the encapsulation of siTaf15 by M2pep-liposomes. (G) A schematic diagram depicts the treatment regimen of ICC mice with liposome materials. (H) Differential morphology of the liver is observed after treatment with M2pepLNP coated with siNc and siTaf15. (I) Weight of the liver is measured following treatment with M2LNP coated with siNc and siTaf15. (J) H&E and IHC analyses are conducted for CK19 and TAF15 in liver tissues treated with M2pepLNP-siNc and M2pepLNP-siTaf15. (K) H&E analysis of heart, lung, spleen, kidney in mice treated with M2pepLNP-siNc and M2pepLNP-siTaf15. (L) Flow cytometry analysis the expression of CD206 and CD86 of macrophage in mice liver. (M) ELISA analysis measures the levels of TGFBI in eyeball blood and liver tissue from mice. Data are presented as mean ± SD, unpaired two-tailed Student’s t test. ∗p <0.05, ∗∗p <0.01, ∗∗∗p <0.001, ∗∗∗∗p <0.0001. BMDM, bone marrow-derived macrophages; ICC, intrahepatic cholangiocarcinoma; IHC, immunohistochemistry; QPCR, quantitative PCR; TAF15, TATA-binding protein-associated factor 15; TGFBI, transforming growth factor-beta induced.

We then screened siRNAs targeting TAF15 to identify the optimal small interfering RNA with the lowest knockdown efficiency in BMDMs. The optimal siTaf15 achieved 70% knockdown efficiency (Fig. 7D and E). To increase the in vivo stability of the siTaf15, all pyrimidine bases in both strands of the siRNA were modified with 2′-OMe.²⁵ The voltage and particle size are as the picture shows (Fig. S6B and C). We next established a YAP/AKT in situ cholangiocarcinoma model via the Sleeping Beauty hydrodynamic system. After 10 days of intravenous plasmid injection, the mice were randomly treated with M2pepLNP-siNc or M2pepLNP-siTaf15 (Fig. 7F and G). M2pepLNP-siTaf15 significantly reduced the liver tumor load and liver weight (Fig. 7H and I). Immunohistochemical analysis revealed that the proportions of CK19 and Ki67 in the livers of the M2pepLNP-siTaf15 group were significantly lower than those in the M2pepLNP-siNc group (Fig. 7J, Fig. S6D and E). Moreover, the H&E results from the heart, spleen, lung, and kidney revealed that M2pepLNP-siNc/siTaf15 were well tolerated and significantly safe in mice (Fig. 7K). In addition, flow cytometry analysis revealed that the proportion of CD11b⁺F4/80⁺CD206⁺ cells was lower but that the proportion of CD11b⁺F4/80⁺CD86⁺ cells was greater in the M2pepLNP-siTaf15 group than in the control group (Fig. 7L, Fig. S6F). Detection of TGFBI expression in mouse serum and tissue revealed that TGFBI expression was downregulated after treatment with M2pepLNP-siTaf15 (Fig. 7M). These findings suggested that targeting TAF15 in TAMs with liposomes modified with the M2 peptide effectively reverses tumor polarization and inhibits the expression of TGFBI, thereby inhibiting the progression of ICC.

Discussion

TAF15, a member of the FET family (including FUS and EWS), is a DNA- and RNA-binding protein that plays a key role in multiple cancer types. TAF15 enhances the proliferation and migration of lung squamous cell carcinoma cells by stabilizing HMGB3 mRNA, thereby promoting disease progression.²⁶ TAF15 activates the GREM1-NF-kB pathway by promoting the stability of BRD4 mRNA, thus accelerating the development of osteoarthritis.²⁷ However, most of the recent studies on TAF15 have focused on parenchymal cells, but little is known about its role in interstitial cells. TAF15 promotes gene transcription and expression, and it plays a crucial role in several key inflammatory signaling pathways. In our previous studies, we reported that TAF15 was also highly expressed in macrophages in a mouse model of NASH,¹⁸ suggesting that TAF15 may play a role in the phenotypic shaping of macrophages.

The limited benefit of immunotherapy in cholangiocarcinoma immunotherapy is attributed mainly to the complex inhibitory TME.²⁸ TAMs are the most abundant immune cells in the TME and are a key component of the inflammatory circuits that drive tumor progression and metastasis.²⁹ At present, there are three treatment strategies for TAMs as follows: (1) cut off the source and eliminate the production of M2 TAMs, including inhibiting the transformation of monocytes into M2 TAMs and eliminating specific protumoral tissue-resident macrophages in liver cancer; (2) reprogramming M2-TAMs into M1-TAMs and CAR-Ms; and (3) block the communication between liver cancer M2 cells.³⁰ An inherent disadvantage of eliminating TAMs and blocking M2 cells communication is the loss of the native immunostimulative role of macrophages as primary phagocytes and antigen-presenting cells in solid tumors.³¹ Therefore, reprogramming or repolarization of immunosuppressive TAMs into immunostimulating TAMs is an attractive research direction.

In our previous study, we focused on the functions of TAF15 in hepatocytes. We found that TAF15 overexpression in hepatocytes promoted macrophage infiltration and proinflammatory cytokines secretion, thereby promoting the M1 polarization of macrophages.¹⁸ We found that TAF15 was also expressed in hepatic macrophages. However, the functions of TAF15 in macrophages remain unclear. In the current study, we shift our focus to the effect of TAF15 on TAMs. Our data reveal that TAF15 is upregulated in ICC-associated macrophages and promotes M2 polarization. Mechanistic studies indicate that in macrophages, TAF15 transcriptionally activates SOCS1, thereby suppressing the JAK2/STAT1 pathway. The role of TAF15 in NASH-related hepatocellular carcinoma (HCC) or its function in macrophages during NASH remains to be elucidated, and we plan to investigate these aspects in future studies. Our findings demonstrate a context-dependent duality of TAF15 function: hepatocyte-derived TAF15 promotes M1 macrophage polarization in inflammatory models, while macrophage-intrinsic TAF15 drives M2 polarization in TME. This cell type- and disease state-specific regulation suggests TAF15 may serve distinct immunomodulatory roles under different pathological conditions. Further mechanistic studies are warranted to elucidate these differential regulatory networks.

Starting with a mouse ICC in situ model and ICC patient specimens, we compared the expression of TAF15 in TAMs and relatively normal macrophages adjacent to cancer cells. TAF15 is highly expressed in TAMs and promotes the polarization of tumor macrophages towards the M2 phenotype, highlighting its carcinogenic potential. SOCS1 was identified as the downstream gene of TAF15 by transcriptome sequencing and intersection with ChIP-seq data in the ChIP-Atlas database. CUT&Tag and dual-luciferase reporter assays revealed that TAF15 bind to the promoter of SOCS1 and played a role in transcriptional activation.

SOCS1, a member of the SOCS protein family, is a classic negative feedback regulator of the JAK-STAT signaling pathway, and it plays a key role in regulating cytokine-triggered signaling pathways.³² According to a recent review, abnormal expression levels of SOCS1 disrupt macrophage-T cell interactions and disrupt the balance of macrophage subpopulation polarization.³³ SOCS1 inhibits STAT1 phosphorylation, and in the absence of SOCS1, the proinflammatory effects of M1 macrophages are enhanced by the induction of the JAK/STAT pathway.¹⁹ It has been reported that a SOCS1 peptide inhibits JAK2/STAT1 activation to inhibit M1 polarization of macrophages, resulting in a protective effect against pathological glomerular changes in mesangial proliferative glomerulonephritis.²¹ SOCS1 is a particularly potent inhibitor of JAK1 and JAK2. Owing to the inhibition of JAK through non-ATP competition mechanisms, SOCS1 achieves similar potency as ruxolitinib in the presence of intracellular ATP (∼5–10 mM).³⁴ Consistent with our study, ruxolitinib promotes the polarization of macrophages towards M2, which may explain its limited applications in solid tumor therapy.³⁵ The present study revealed that the loss of TAF15 in macrophages inhibited SOCS1 transcription and activated the JAK2/STAT1 pathway, thereby promoting the polarization of macrophages towards the M1 phenotype, inhibiting the expression of M2 polarization markers in macrophages, and inhibiting tumor proliferation. These manifestations are remedied by further overexpression of SOCS1 or administration of low doses of ruxolitinib. Our findings suggest that in ICC with OE TAF15 TAMs, targeting TAF15 or its downstream pathway, using SOCS1 inhibitors or JAK2 agonists, could reverse macrophage polarization and represent a promising therapeutic strategy. However, potential redundancy in these signaling pathways may limit their efficacy as monotherapies and warrants further investigation.

The present TAMs transcriptome sequencing results revealed that ECM remodeling pathway was highly enriched. The tumor cell ecosystem is nourished by its ECM.³⁶ Changes in the tumor ECM promote tumor cell growth and survival by providing key biomechanical and biochemical cues, and these changes actively promote the occurrence and development of cancer by regulating angiogenesis and immune function.³⁷ TGFBI is an extracellular matrix protein induced by transforming growth factor β, and it is abnormally expressed in various types of cancer. TGFBI promotes the progression of renal cell carcinoma by activating the PI3K/AKT/mTOR/HIF-1α pathway.³⁸ However, recent studies have shown that TGFBI is highly expressed in TAMs from human PACs and that the expression of TGFBI is positively correlated with the growth of human PACs.³⁹ In high-grade gliomas, TGFBI is specifically expressed in TAMs and preferentially secreted by M2 macrophages, but TGFBI is preferentially expressed in tumor stem cells under hypoxic conditions.^23,40 Research has indicated that TAMs secrete TGFBI, which contributes to immunosuppression in early-stage ovarian cancer and supports the maintenance of an immunosuppressive microenvironment in advanced high-grade serous ovarian cancer.⁴¹ Soluble TGFBI aggravates the malignancy of cholangiocarcinoma by activating the integrin β-1dependent PPARγ signaling pathway, and increased TGFBI expression may be associated with poor prognosis, suggesting that TGFBI may serve as a prognostic biomarker and therapeutic target for cholangiocarcinoma.²⁴ By analyzing single-cell data from the GSE138709 dataset, we found that TGFBI was also specifically expressed by macrophages in ICC and that exogenous TGFBI significantly promoted tumor growth. ELISA results also revealed that M2-like macrophages secreted more TGFBI. These results indicate that TAF15 in TAMs inhibits the activation of JAK2/STAT1 through the transcriptional activation of SOCS1, which promotes M2 polarization and TGFBI secretion, thereby promoting the malignant progression of bile duct carcinoma. TGFBI has been reported to bind αvβ3⁴¹ and αvβ5²³ integrins involved in the regulation of cell adhesion, metabolism, and the immune response.⁴² Future studies will be conducted to explore the mechanism by which exogenous TGFBI, especially TGFBI derived from TAMs, promotes cholangiocarcinoma.

Advances in nanotechnology have made it possible to study the efficacy of small molecule inhibitors.⁴³ Chen et al.⁴⁴ used M2pep-modified liposomes to target M2 macrophages, knockdown Yin Yang1, inhibit the lung metastasis of prostate cancer (PCa) cells, and produce synergistic antitumor effects with PD-1 blockade. Han et al.⁴⁵ used polylactic acid-glycolate nanoparticles (NPs) to support D-lactic acid (DL), and they further modified the DL-supported NPs with HCC membranes and M2 macrophage-binding peptides (M2peps) to form the DL@NP-M-M2pep nanoformula. The DL@NP-M-M2pep nanoformula transforms M2 TAMs into M1 TAMs, thereby reshaping the immunosuppressive TIME in HCC mice.⁴⁵ Chen et al.⁴⁶ constructed an engineered milk exosome (mExo) system modified with an M2pep (M2 macrophage-binding peptide) and 7D12 (an anti-EGFR nanobody) (7D12-MEXo-M2pep-siPDL1) to specifically deliver siPDL1 to M2 TAMs. Chen et al.⁴⁶ reported that the dual-targeted engineered mExos efficiently deliver siPDL1 into M2 macrophages and repolarize them into M1 macrophages, thus restoring the immune activity of CD8⁺ T cells and reshaping the TIME. These studies confirm the definitive macrophage-targeting effect of the present M2pep-LNP–siRNA system. Primary liver cancer is characterized by three subtypes, namely, HCC, ICC, and the combination of HCC and ICC; these subtypes present distinct epidemiological, morphological, and genomic characteristics, and they exhibit differential responses to therapy.⁴⁷ We believe that it is inevitable for nanomaterials to enter hepatocyte metabolism; however, we focused on ICC, and measured the proportion of cholangiocarcinoma epithelial marker CK19 to judge the progression of ICC. Via animal immunofluorescence and flow cytometry, we conclusively demonstrate that M2pep-LNP-siRNA delivery shows predominant tropism for TAMs rather than neoplastic cholangiocytes. We designed M2pepLNPs to encase siTaf15, effectively targeting TAF15 in TAMs in vivo, which significantly reduces the proportion of CD206⁺ macrophages, increases the proportion of CD86⁺ macrophages, and significantly reduces TGFBI serological levels. TAF15 is highly expressed in TAMs and promotes M2 macrophage polarization, thereby coordinating immunosuppression in ICC by establishing the TME. Mechanistic evaluation revealed that TAF15 further mediated the JAK2/STAT1/TGFBI axis through transcriptional activation of SOCS1, which affected ICC progression. Targeting TAMs with M2pepLNP-siTaf15 is a promising strategy to overcome ICC immunoresistance.

The present study reveals for the first time that TAF15 promotes the polarization of TAMs towards the M2 phenotype and that macrophage-specific KO of TAF15 inhibits tumor proliferation both in vivo and in vitro. TAF15 activates SOCS1 by transcription, which mediates the JAK2/STAT1/TGFBI axis, thereby affecting ICC progression. Targeting TAF15 in TAMs with M2pepLNPs-siRNA may be a promising therapeutic approach for the treatment of ICC.

Abbreviations

2′-Ome, 2′-O-methylation; BMDM, bone marrow-derived macrophages; CM, conditional media; CUT&Tag, cleavage under targets and tagmentation; DEGs, differentially expressed genes; ECM, extracellular matrix; ICC, Intrahepatic cholangiocarcinoma; IHC, immunohistochemistry; JAK, Janus Kinase; KO, knockout; M2pepLNP, M2 peptide-modified liposomes; NASH, non-alcoholic steatohepatitis; NF-κB, nuclear factor κB; NPs, nanoparticles; OE, overexpressing; PMA, phorbol 12-myristate 13-acetate; qRT-PCR, real-time quantitative PCR; SOCS1, Suppressor of cytokine signaling1; STAT, signal transducer and activator of transcription; TAF15, TATA-binding protein-associated factor 15; TAMs, tumor-associated macrophages; TGFBI, transforming growth factor-beta induced; TIME, tumor immune microenvironment; TME, tumor microenvironment; TNF-α, tumor necrosis factor-alpha

Financial support

This work was supported in part by the project funded by 1) National Natural Science Foundation of China (grant numbers 82400692 to S.Z.Y. and 82000548 to B.X.), 2) Jiangsu Provincial Key Research and Development Program, China(grant number BE2023655 to L.W.), 3) the key project of medical science and technology development of Nanjing Municipal Health Commission (grant number ZKX21032 to L.W.).

Authors’ contributions

Conceptualization; writing – original draft: YL, BX.

Investigation; formal analysis: TZ, HL, YP, SZ, QC, LiShan Wang, GB, NZ, Yue Zhou, JW, XF, Yaru Zhou, ZW, LX, Yun Zhu, SS.

Resources, supervision: SY, LZ, Lei Wang.

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Conflicts of interest

The authors declare that they have no competing interests.

Appendix A. Supplementary data

The following are the Supplementary data to this article.