Custom-developed CRISPR-Cas9 Huh7 control (YC-D001, Ubigene), Huh7 TREM1 KO (YKO-H838, Ubigene), HepG2 control (YC-C001, Ubigene), HepG2 TREM1 KO (YKO-H838, Ubigene) and Hep3B (HB-8064, ATCC) cell lines were used. The cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, MT10013CV, Corning) supplemented with 10% fetal bovine serum (FBS, SH3039602, Cytvia) and 100 u/mL penicillin/streptomycin (15-140-163, Gibco) in a 37°C incubator under humidified 5% CO2. For primary culture of human liver tumor cells, tissues were processed under sterile conditions by mincing into small fragments (~1–2 mm³) using sterile scalpels. The fragments were enzymatically digested in a solution of collagenase IV (17-104-019, Gibco™) and DNase I (AAJ62229MB, Thermo Scientific) at 37°C for 60 minutes with gentle agitation. Following digestion, the cell suspension was passed through a 70 µm cell strainer to remove undigested debris. The filtrate was centrifuged at 400 × g for 5 minutes, and the resulting cell pellet was washed with PBS. Cells were resuspended in DMEM/F12 media (SH3002301, Cytiva HyClone™) supplemented with 10% FBS, 20 ng/mL epidermal growth factor (354010, Corning), 5 µg/mL insulin (12-585-014, Gibco™), 1 µg/mL hydrocortisone (AC352450010, Thermo Scientific), and 100 U/mL penicillin/streptomycin, and then plated onto collagen-coated tissue culture dishes. Cultures were maintained at 37°C in a humidified incubator with 5% CO₂. Medium was replaced every 2–3 days, and cells were passaged when they reached 70–80% confluence.

TREM1 knockout (TREM1-KO) Huh7 and HepG2 cell lines were generated and validated by Ubigene (Guangzhou, China) using their proprietary Red Cotton™ CRISPR-Cas9 gene editing platform. Exon 2 of the human TREM1 gene was targeted using the following sgRNAs: for Huh7, sgRNA1: GTAGTCACATTTCACATCCAGGG and sgRNA2: AGCATGTGAGGCTCCTTGGGAGG; for HepG2, sgRNA3: ACTGGATGGGAATTCTTTGAAGG and sgRNA4: GAAAGCTTGGCAGATAATAAGGG. Electroporation parameters were optimized for each cell line (Huh7: 1600 V, 10 ms, 3 pulses; HepG2: 1230 V, 10 ms, 2 pulses).Single-cell clones were isolated by limiting dilution and screened for deletions in exon 2 by PCR and Sanger sequencing. RT-PCR was performed to confirm the loss of TREM1 transcript expression. All gene editing and validation steps were conducted by Ubigene, and fully characterized knockout clones were provided for downstream experiments. Representative validation data, including PCR and sequencing results for both HepG2 and Huh7 clones are shown in Supplementary Figures 1 and 2 .

NOD scid γ (NOD.Cg-Prkdc^scidIl2rgtm1Wjl/SzJ, JAX 005557) mice aged 4–6 weeks were purchased from Jackson Laboratory. Age- and sex-matched animals were included in all experiments. The mice were housed in a climate-controlled specific pathogen-free environment within the Augusta University animal facilities. The study protocol (2008-0051) was approved by Augusta University Institutional Animal Care and Use Committee (IACUC).

HCC patient samples were collected from Georgia Cancer Center Biorepository Bank following protocols approved by Institutional Review Board Committee. Informed consent was obtained from all subjects providing fresh tissues. The samples were snap frozen in liquid nitrogen. All experiments were carried out in accordance with relevant guidelines and regulations.

HCC and normal tissue expression profiles were obtained from The Cancer Genome Atlas (TCGA) database, in which mRNA expression level of TREM1 was analyzed. The expression data was also cross-referenced with TCGA clinical data. The Kaplan–Meier estimator was used to produce survival curves and assess the prognostic significance of TREM1 in HCC. Patients with aggressive HCC were categorized into high and low TREM1 expression groups based on 50% quantile as cut off for TREM1 expression.

Cells were cultured in 6-well plates until they reached 80% confluency. Cells were lysed in 40µl lysis buffer (FNN0021, Invitrogen) supplemented with 1mM PMSF and protease inhibitor cocktail (A32955, Thermo Fisher Scientific). 80µg of protein were resolved and separated by 4-20% Mini-PROTEAN^® TGX™ Precast gels (45461094, BIO-RAD), transferred onto PVDF membranes (IPFL85R, Millipore). Followed by blocking with SuperBlock blocking buffer (PI37580, Thermo Fisher Scientific) for 1h, membranes were incubated with primary antibodies - TREM1 monoclonal (FERMA548755, Invitrogen,1:1000), CD133 Rabbit polyclonal (A0219, Abclonal, 1:1000), Epcam Rabbit polyclonal (A23654, Abclonal, 1:1000), OCT4 Rabbit polyclonal (A7920, Abclonal, 1:1000), Nanog Rabbit polyclonal (A3232, Abclonal, 1:1000), Sox2 Rabbit polyclonal (A19118, Abclonal, 1:1000), β Actin (MAB8929SP, R&D Systems,1:1000), α tubulin (sc-53029, Santa Cruz Biotechnology, 1:1000) at 4°C overnight. After washing in TBST, the membranes were incubated with secondary antibodies - goat anti-mouse HRP-conjugated secondary antibody (AP127PMI, Millipore Sigma, 1:10,000), HRP-conjugated Anti-Mouse IgG (MAB8929-SP, R&D Systems, 1:10,000) and HRP-conjugated anti-Rabbit (AS014, Abclonal, 1:10,000) for 1h at room temperature (RT). After washing, the protein signals were visualized using ECL detection (45-002-401, Cytvia Amersham).

Total RNA was isolated using RNeasy Mini kit (74104, Qiagen). The cDNA was synthesized using RT² First Strand Kit (330404, Qiagen). qRT-PCR was performed using RT² SYBR Green qPCR Mastermix (330502, Qiagen) with an ABI Step OnePlus Real-Time PCR system (Applied Biosystems) according to the operator’s manual. The gene expression values were normalized to GAPDH. Each experiment was carried out in triplicate. Primers are provided in Supplementary Table .

HCC tumor and normal tissues (NCT046) were obtained from US Biomax. The sections were deparaffinized for multiplexed staining using the Opal protocol. According to the manufacturer’s instructions, paraffin-embedded tumor tissue samples were heated at 57°C overnight. Residual paraffin was removed using xylene, and the tissues were rehydrated through a series of graded alcohols to distilled water. Antigen retrieval was performed by microwave heating in AR6 solution (2620012, PerkinElmer). After cooling and washing with TBST, the tissues were blocked with antibody diluent (ARD1001EA, PerkinElmer) for 30 minutes at RT. The tissue sections were then incubated with primary antibody for 1 hour. After washing with TBST, the slides were incubated with HRP-conjugated secondary antibody (ARH1001EA, PerkinElmer) for 30 minutes at RT, followed by staining with Opal 650 (FP1496A, PerkinElmer) diluted at a 1:100 ratio in 1X amplification diluent (FP1498, PerkinElmer) for 10 minutes at RT. This process was repeated for staining with both primary antibodies: TREM1 Rabbit polyclonal antibody (11791-1-AP, Proteintech, 1:200) and alpha-Fetoprotein monoclonal antibody (13-949-982, Invitrogen, clone: 1E8, 20 µg/mL). Afterward, the sections were washed with TBST, and the nuclei were stained with ProLong™ Gold antifade reagent with DAPI (Invitrogen, P36931). Multiplexed imaging was conducted using the Opal 7-Color Manual IHC Kit, and the multiplexed sections were imaged at x20 magnification on the Vectra Automated Quantitative Pathology Imaging System.

Huh7 and HepG2 control and CRISPR-Cas9–mediated TREM1 KO cell lines were seeded in 6-well plates and cultured under standard culture conditions (see Cell Lines and Cell Culture). Once the cultures reached approximately 80% confluency, 10⁶cells were harvested and suspended in ice-cold FACS buffer (00422226,eBioscience) and incubated with following antibodies (all from BioLegend) at 4°C for 45 minutes in dark: TruStain FcX (clone: 93, 101319, 1:50 dilution), FITC anti-human CD133 (397907, clone: W6B3C1, 5 µL in 100 µL staining volume) and APC anti-human EpCAM (369809, clone: CO17-1A, 5 µL in 100 µL staining volume). Cells were acquired on the Attune NxT Acoustic Focusing flow cytometry platform (Thermo Fisher Scientific) and data were analyzed on FlowJo v10.0. Forward versus side scatter (FSC versus SSC) gating was used to exclude dead cells. Patient samples, vehicle and VJDT treated tumors: Tumors were processed into a single-cell suspension using gentle MACS Octo Dissociater with Heaters (Miltenyi Biotech) in combination with the tumor dissociation kit (Miltenyi Biotech). 1 million cells were collected, incubated with antibodies and acquired as described above.

CD133⁺ cells were isolated using the CD133 Microbead Kit (130-097-049, Miltenyi Biotec). Briefly, cells were incubated with CD133 microbeads and then passed through an LS column (130-042-401, Miltenyi Biotec) placed in the magnetic field of a MACS separator (130-042-302, Miltenyi Biotec) attached to a MACS multistand (130-042-303, Miltenyi Biotec). The magnetically labeled CD133⁺ cells were retained in the column, while the unlabeled cells flowed through and were collected as the CD133^- fraction. After removing the column from magnetic field, the retained CD133⁺ cells were eluted and subsequently incubated with EpCAM microbeads (130-061-101, Miltenyi Biotec). These cells were loaded onto a magnetic column to collect CD133⁺EpCAM⁺ cells. The CD133^-EpCAM^- cells were collected separately. Both fractions were subsequently cultured to obtain sufficient cells for further experiments.

Cells were harvested using trypsin-EDTA to ensure a single-cell suspension. Cells were washed by centrifugation at 1500 rpm for 5 minutes at 4°C and resuspended in ice-cold FACS buffer. Cell viability was confirmed to be above 90%. Fc block was performed by adding 100 µL of Fc block diluted in FACS buffer (1:50 ratio) to each sample, followed by incubation on ice for 20 minutes. Cells were stained with FITC anti-CD133 and APC anti-EpCAM antibodies diluted in FACS buffer at 4°C for 45 minutes in dark. Cells were washed three times by centrifugation at 1500 rpm for 5 minutes and resuspension in ice-cold FACS buffer. Cells were sorted using Invitrogen™ Bigfoot™ Cell Sorter (Thermo Fisher Scientific). Cells were sorted based on CD133 and EpCAM expression. Sorted cells were collected into appropriate collection tubes, and the purity of sorted populations was analyzed, aiming for at least 90% purity. Unstained, single-stained, and isotype controls were included for accurate gating.

Cell Counting Kit-8 (CCK-8, GK10001, Glpbio) was used to examine cell proliferation. 1 ×10⁴ cells were cultured in 96 well plate for 0,24,48 and 72h. after incubating with 10µl CCK-8 solution at 37°C for 1hr, absorbance at 440nm was measured using microplate reader (biotek synergy htx reader). The experiment was conducted in triplicate.

Migration assay was conducted with transwell plate (07-200-174, Corning). Briefly, 1ml medium containing FBS was added in the lower chamber and 600µl media containing 6 ×10⁴ cells were added in the upper chamber. After incubating for 24hr, the transwell inserts were fixed in 4% paraformaldehyde for 10mins. Subsequently, the cells were stained with 0.05% crystal violet dissolved in PBS for 30 minutes. The transwell plates were then washed with PBS and the cells in the upper chamber were removed using cotton swabs. The cells that had migrated to the lower chamber were observed and counted under a microscope. The experiment was conducted in triplicate.

Cells were collected, washed twice with cell staining buffer (420201), and resuspended in Annexin V Binding Buffer (422201) at a concentration of 1x10⁶ cells/mL. 100 µL of cell suspension was transferred into 5 mL test tube. 5µl each of FITC Annexin V (640905) and propidium iodide (421301) were then added to the cell suspension, followed by gentle vortexing. The cells were incubated at RT in the dark for 15 minutes. Subsequently, an additional Annexin V Binding Buffer was added, and the cells were immediately acquired on the Attune NxT Acoustic Focusing flow cytometry platform (Thermo Fisher Scientific) and data were analyzed on FlowJo v10.0. All reagents used were obtained from BioLegend.

Total RNA was isolated using RNeasy Mini kit (74104, Qiagen). The cDNA was synthesized using RT2 First Strand Kit (330404, Qiagen). The first strand cDNA was mixed with RT2 SYBR Green qPCR Mastermix (330502, Qiagen), and aliquoted into the 96-Well RT2 Profiler PCR Array for apoptosis (PAHS-012A, SA Biosciences). This assayed 84 key genes associated with apoptosis. RT-PCR was performed using ABI Step OnePlus Real-Time PCR system (Applied Biosystems) according to the operator’s manual. GAPDH was used for data normalization.

Cells were cultured in 6-well plates until they reached 80% confluency. Cells were collected and pelleted. 5ml cold 70% ethanol was added dropwise to the pellet and incubated at -20°C for 2 hours. Cells were washed twice in cell staining buffer and resuspended. 100µl cell suspension containing 10⁶ cells were transferred into sample tubes and incubated with 20µl Alexa Fluor^® 488 anti-human Ki-67 Antibody (350507, BioLegend, clone: Ki-67) at RT for 30 minutes in the dark. After washing with cell staining buffer, a 0.5ml staining buffer and 10 µl PI solution were added to the sample tubes. Following this the cells were acquired on the Attune NxT Acoustic Focusing flow cytometry platform (Thermo Fisher Scientific) and data were analyzed on FlowJo v10.0.

Cells were seeded in triplicate at a density of 500 cells/well in 6-well plates. After 14 days, the medium was aspirated, and cell colonies were washed twice with PBS. They were then fixed in 4% paraformaldehyde for 5 mins followed by staining with 0.05% crystal violet dissolved in PBS for 30mins. Excess dye was removed by rinsing twice with PBS. After drying, the colonies were counted under a microscope. The experiment was conducted in triplicate.

5000 CD133⁺EpCAM⁺ control and TREM1 KO cells per well were seeded on low attachment 6-well plates in serum free DMEM/F12 medium supplemented with 2% B27(12687-010, Gibco), 1% penicillin-streptomycin, 5 ng/mL epidermal growth factor (354010, Corning) and 10 ng/mL basic fibroblast growth factor (13256-029, Invitrogen). After 7 days of culture, spheres were counted and photographed under Keyence microscope. For pharmacological inhibition of TREM1, spheroids were treated with 10µM VJDT on day 7 and were counted on day 10. Both experiments were conducted in triplicate.

For CDX model, 200µl cell suspension containing 5000 freshly sorted CD133⁺EpCAM⁺ control and TREM1 KO cells mixed with Matrigel (356230, Corning) in a 1:1 ratio and were subcutaneously injected into the left flank of 5 weeks old NSG mice (n=6 per group). The mice were observed every other day for tumor growth for 60 days. After this, mice were sacrificed, and all tumor tissues were harvested for morphological assessment. For experiments involving pharmacological inhibition of TREM1, 1x10⁶ Huh7 cells were subcutaneously implanted onto 10 NSG mice. Mice were treated with either 20 mg/kg VJDT or vehicle (DMSO) intraperitoneally on day 10 after tumor cell injection and continued alternate days until day 22 (n=5 per group). VJDT is a novel TREM1 small molecule inhibitor that effectively blocks TREM1 signaling. VJDT design, development, dose optimization and toxicity studies are described in our prior work and is covered under the international patent application WO2022061226A1 (9). Tumor growth was measured alternate days using a digital caliper; tumor volumes were calculated using the formula V=L × W2 × (π/6), where L and W denote length and width of the tumor.

CD133⁺EpCAM⁺ control and TREM1 KO cells were sorted using FACS. Total RNA was isolated using Trizol reagent (5–596-026, Thermo Fisher Scientific) and purified using RNEasy mini kit (74104, Qiagen). The samples were submitted to Integrated Genomics Core Shared Resources at Augusta University (RRID: SCR_026483). Briefly RNA sequencing was carried out using an Illumina NovaSeq 6000, and filtered clean reads were mapped to a reference genome using hierarchical indexing for spliced alignment of transcripts 2 (HISAT2). Visualization of alignment results were verified using Integrative Genomics Viewer. RNA-seq data were analyzed using iDEP (integrated Differential Expression and Pathway analysis) (26). Differential expression analysis was performed using DESeq with Fold Change ≥1 and FDR < 0.05 set as screening criteria. Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis was used to identify the significantly affected pathways during TREM1 KO. Gene Ontology enrichment analysis for molecular function, biological process and cellular component was performed using IDEP.

Statistical analysis was conducted using GraphPad Prism 9 (version 9.0.0). Data are generally presented as mean ± SD, unless specified otherwise. Graphs display either group mean values ± SD for in vitro experiments or ± SEM for in vivo experiments, or individual values. If the data sets followed a normal distribution and comparisons were made between two experimental groups, an unpaired, two-tailed Student’s t-test was utilized. For in vitro studies, unpaired t-tests were used for comparing two groups, while for in vivo studies, a two-way ANOVA was employed for multiple comparisons of longitudinal tumor growth across various groups. A p-value of less than 0.05 was considered statistically significant.

To gain a comprehensive understanding of the role of TREM1 in liver hepatocellular carcinoma (LIHC), we analyzed the LIHC cohort from the TCGA database (27). The cohort was subdivided into three iClusters based on well-established parameters, including specific mutations (TP53, CTNNB1) and multi-omics data (27, 28). iCluster1 and iCluster3 represent patients with the more aggressive, rapidly proliferating LIHC and who have a worse prognosis, while iCluster2 patients exhibit more differentiated tumors and have better overall survival (27, 29). Interestingly TREM1 expression was consistently higher in iCluster1 and iCluster3 compared to iCluster2 and matched normal tissue ( Figure 1A ). Higher TREM1 expression also significantly correlated with worse overall survival in the combined iCluster1 and iCluster3 cohorts ( Figure 1B , p<0.05). These data suggest that high TREM1 expression is associated with more aggressive tumors and poor prognosis for HCC patients. Additionally, TREM1 expression increased with tumor stage, with stage 4 showing the highest expression ( Figure 1C ). Pathway analysis of the top 200 TREM1-correlated genes (Spearman correlation > 0.5) revealed that the inflammatory response was the most enriched pathway ( Figure 1D ). Other pathways of interest included regulation of cell activation, response to external stimuli, and cell activation, which differ from the conventional inflammatory pathways associated with TREM1. Genes associated with cell division, such as TUBA1C(Tubulin Alpha 1c), CCNJL(Cyclin J-Like), PLK3(Polo-Like Kinase 3), and HK3 (Hexokinase 3) (30–32), exhibited positive correlation with TREM1 expression and are associated with aggressive LIHC tumors ( Figure 1E ). To confirm our observations from the TCGA database, we performed immunofluorescence staining of LIHC tumor sections. This staining exhibited marked overlap between TREM1 and α-fetoprotein, a classical marker for HCC cells (33–35) ( Figure 1F ), indicating that the tumor cells themselves are positive for TREM1. The data also showed that TREM1 expression is significantly higher in LIHC sections compared to matched normal tissues (p<0.001). Additionally, analysis of Human Protein Atlas database revealed a moderate to high TREM1 protein expression in the cytoplasm/membrane of HCC sections ( Supplementary Figure 3 ). To further validate our findings and explore the role of TREM1 in liver cancer at the cellular level, we assessed TREM1 expression in human HCC cell lines (Huh7, HepG2, Hep3b) and one primary HCC patient (P1). RT-PCR analysis revealed elevated expression of TREM1 across all groups compared to resting THP-1 cells, which have negligible TREM1 expression (36) ( Figure 1G ). Western blot analysis detected TREM1 protein levels across all HCC cell lines tested at varying levels ( Figure 1H ). Our data indicates that TREM1 is expressed in liver cancer cells and tissues. It also suggests that TREM1 has a more direct role within these cancer cells, separate from its traditionally recognized involvement in inflammation.

To determine the specific role of TREM1 in HCC, we employed the CRISPR-Cas9 system to knock out TREM1 expression in HepG2 and Huh7 cell lines. Western blot analysis confirmed a marked reduction in TREM1 expression in the KO cells, although faint TREM1 bands were still observed, which may represent residual truncated protein isoforms with negligible functional capacity ( Figure 2A ). To further validate the knockout efficiency, we performed RT-PCR ( Figure 2B ) and flow cytometry analysis ( Supplementary Figure 4 ).Functional assays revealed that TREM1 silencing significantly inhibited cell proliferation (p<0.01), as demonstrated by CCK-8 assays, and reduced migration capacity in transwell assays in both Huh7 (p<0.05) and HepG2 (p<0.01) cell lines ( Figures 2C, D ). Additionally, Ki-67-based cell cycle analysis indicated a substantial decrease in the proportion of S-phase cycling cells (p<0.05) and a G1 cell cycle (p<0.05) arrest in Huh7 and HepG2 TREM1 KO lines, suggesting impaired cell cycle progression ( Figure 2E ). These results highlight the critical role of TREM1 in promoting cancer cell proliferation. To assess the impact of TREM1 knockout on apoptosis, we analyzed the proportions of apoptotic cell populations via Annexin V - PI flow cytometry. We observed a significant increase in late apoptotic cells in Huh7 (p<0.01) and HepG2 TREM1 KO (p<0.05) lines compared to control cells ( Figure 2F ). This finding suggests that TREM1 expression exerts an anti-apoptotic effect to support tumor cell survival, and that its silencing shifts the balance toward pro-apoptosis. To further characterize this phenomenon, we used the RT2 Apoptosis Profiler Array to screen 84 key genes involved in apoptosis and their perturbations during TREM1 silencing. We identified significant changes in the expression of apoptosis-related genes in Huh7 TREM1 KO cells in comparison to control ( Figure 2G ). TREM1 silencing led to significant (p<0.0001) upregulation of pro-apoptotic genes such as BAX, CASP3, and CASP14, indicating enhanced activation of apoptotic pathways (37, 38). Conversely, anti-apoptotic genes like BAG1 and BCL2, which promote cell survival (39, 40), were significantly downregulated in TREM1 KO cells ( Figure 2H ). A similar trend was observed in HepG2 cells, where TREM1 silencing led to a significant (p < 0.001) upregulation of pro-apoptotic genes such as BAX and CASP3, along with downregulation of anti-apoptotic genes including BIRC3 and BAG1 ( Supplementary Figures 5A, B ).These findings suggest that TREM1 functions as a vital regulator of cell survival in HCC by suppressing apoptosis and promoting proliferation, migration, and cell cycle progression. TREM1 abrogation disrupts these processes, thereby impairing overall tumor cell viability.

Our data demonstrated the integral role of TREM1 in HCC, as its silencing inhibited cell proliferation and migration ( Figures 2C, D ), disrupted cell cycle progression, and enhanced apoptosis ( Figures 2E, F ). Given this profound impact, we analyzed TREM1 expression in LCSLCs, which are integral drivers of HCC tumors (41, 42). Flow cytometry analysis revealed that a significant proportion (80 ± 6%) of CD133⁺EpCAM⁺ LCSLCs in Huh7 and HepG2 cell lines and the primary HCC P1 patient sample actively expressed TREM1 ( Figure 3A ). To further validate this observation, we performed RT-PCR on FACS-sorted CD133⁺EpCAM⁺ LCSLCs from Huh7, HepG2, and HCC P1. TREM1 expression was significantly elevated in the LCSLC population compared to the corresponding CD133^-EpCAM^- non-stem cell populations ( Supplementary Figure 6 ). To confirm the stemness of these cells, we MACS purified Huh7 CD133⁺EpCAM⁺ and CD133^-EpCAM^- cells. RT-PCR analysis showed significantly (p<0.01) higher expression of prominent stem cell factors—SOX2, OCT4, and NANOG—in CD133⁺EpCAM⁺ cells (43–45) compared to their negative fractions, corroborating their characterization as cancer stem-like cells ( Figure 3B ). Preliminary flow cytometry analysis revealed a significant reduction in the overall frequency of CD133⁺EpCAM⁺ LCSLCs in Huh7 (p<0.0001) and HepG2 TREM1 KO (p<0.01) groups compared to their respective controls ( Figure 3C ). RT-PCR analysis of MACS-purified LCSLCs showed significant downregulation of SOX2, OCT4, and NANOG during TREM1 silencing in both Huh7 and HepG2 cells compared to controls ( Figure 3D ). This was further validated at the protein level using Western blot analysis ( Figure 3E ). Spheroid formation assays using purified LCSLCs demonstrated that TREM1 silencing significantly inhibited spheroid formation capacity in both Huh7 (p<0.05) and HepG2 cells (p<0.01) ( Figure 3F ). Additionally, colony formation assays revealed that TREM1 ablation significantly inhibited clonogenicity and self-renewal capacity of LCSLCs in both cell lines ( Figure 3G ). We also evaluated the tumorigenic potential of TREM1-deficient and TREM1-sufficient Huh7 LCSLCs in an in vivo setting. We implanted 5000 FACS-purified Huh7 CD133⁺EpCAM⁺ LCSLCs from TREM1 KO or control groups into immunodeficient NSG mice to establish a CDX model ( Figure 3H ). Over a 60-day period, only 1 out of 6 mice implanted with TREM1 KO LCSLCs developed tumors, whereas in the TREM1-positive implants, tumor formation was significantly higher, with 4 out of 6 developing tumors. This stark contrast in tumor initiation underscores the crucial role of TREM1 in driving the tumorigenicity of LCSLCs.

Our previous data demonstrated the integral role of TREM1 in maintaining the tumorigenicity of CD133⁺EpCAM⁺ LCSLCs in Huh7 and HepG2 cells. To determine the molecular function of TREM1, we performed bulk RNA sequencing of FACS-purified CD133⁺EpCAM⁺ LCSLCs from Huh7 and HepG2, TREM1 KO, and control tumors. K-means clustering of the top 10,000 differentially expressed genes between TREM1-control and TREM1 KO cells resolved four different clusters. In Huh7 LCSLCs, TREM1 silencing led to the downregulation of genes involved in cell cycle regulation (CDK4, CDK2, CHK1) (46), DNA replication (MCM3, MCM2, MCM4, POLA1) (47, 48), G2/M transition (CDC25B, CDC25A, CDC25C, NDC80) (49, 50), and chromosome condensation (NCAPD2, NCAPH, SMC4, SMC2) (51), while genes associated with the double-strand break repair pathway (FANCM, RAD51C, RAD51, RAD54L) (52, 53) were upregulated ( Figure 4A ). Gene Set Enrichment Analysis (GSEA) incorporating the Gene Ontology (GO) biological process pathway identified significant downregulation in pathways related to chromosome separation, organization, and protein-DNA complex assembly ( Figure 4B ). GO cellular component analysis showed that a significant number of genes downregulated during TREM1 silencing were localized in the chromosome, nucleus, and nucleosome regions ( Figure 4C ). These observations indicate that in Huh7 LCSLCs, TREM1 silencing primarily affects nuclear structures, cell division processes, and chromosome complexes. Volcano plots revealed that TREM1 genetic ablation led to a significant decrease in stemness-related genes such as SOX1, SOX2, NANOG, WNT1, and SOX8 (54) in Huh7 LCSLCs ( Figure 4D ). Conversely, in HepG2 LCSLCs, genes involved in molecular function regulation (WNT16, SOX10, IL10) (55), signal receptor activity (CCL22, IL4, TYRP1), and G protein-coupled receptor (GPCR) activity (ADAM7, BMP7, IL5, PIK3CG, BMX, HGF) (56) were downregulated, whereas pro-apoptotic genes (BAX, BLK) were upregulated during TREM1 ablation ( Figure 5A ). GO biological process pathway enrichment analysis revealed a significant reduction in cytokine activity, receptor-ligand interactions, and signaling receptor activity, suggesting that TREM1 silencing in HepG2 LCSLCs predominantly affects extracellular signaling pathways ( Figures 5B, C ). Volcano plots revealed that along with TREM1, several cancer stemness-related genes including NANOG, SOX2, SOX8, and WNT16 (57–59) were significantly downregulated ( Figure 5D ). These findings suggest that TREM1 silencing alters key signaling pathways in a cell-type-specific manner, disrupting nuclear functions in Huh7 cells and extracellular signaling in HepG2 cells, while also impairing cancer stem cell properties in both cell lines.

We previously demonstrated that TREM1 KO LCSLCs exhibit significantly reduced tumorigenic and spheroid formation capabilities. To further validate these findings, we utilized VJDT, a small molecular inhibitor of TREM1 (7, 9, 60, 61). We subcutaneously injected 1x10⁶ Huh7 cells into 10 NSG mice. On day 10, five of these mice were treated with 20 mg/kg VJDT every alternative day until day 22. VJDT treatment significantly (p<0.05) reduced overall tumor volumes in comparison to the vehicle group ( Figure 6A ). Tumors were digested into single-cell suspensions for flow cytometry analysis, which revealed a significant reduction in the percentage of CD133⁺EpCAM⁺ LCSLCs in the VJDT-treated tumors ( Figure 6B ). Western blot analysis of Huh7 LCSLC tumors revealed decreased expression of stem cell-related proteins such as Sox2, Nanog, and Oct4 in the VJDT-treated group ( Figure 6C ). Additionally, we isolated 5000 Huh7 and HepG2 CD133⁺EpCAM⁺ LCSLCs using MACS and performed spheroid formation assays. Spheroids were formed by day 5. The spheroids were treated with 10 µM VJDT or vehicle from day 7. VJDT treatment significantly (p<0.01) inhibited spheroid formation by day 10 in Huh7 ( Figure 6D ) and HepG2 ( Supplementary Figure 7A ). Furthermore, RT PCR analysis of VJDT treated spheroids exhibited significant downregulation in key genes involved in molecular function (SOX10, IL10), signal receptor activity (CCL22, IL4) and GPCR activity (ADAM7, IL5) ( Supplementary Figure 7B ), corroborating previous results shown in bulk RNA sequencing. These findings confirm that TREM1 inhibition via VJDT effectively depletes the number of LCSLCs, reduces tumor size, and decreases spheroid formation. This also validates the results we obtained from using Huh7 CRISPR-Cas9 TREM1 KO cells and highlights the potential of VJDT as a therapeutic agent in targeting liver cancer stem-like cells.