Incisional biopsy samples (tumor and adjacent normal tissues) were collected from 20 clinically diagnosed patients with HCC between August 2014 and April 2016, and were sent for histological confirmation according to classical histopathological features. The distance of the adjacent tissue from the tumor was >5 cm. Patients with other systemic tumors, infections or incomplete clinical data were excluded. All samples were obtained from the Department of Hepatobiliary Surgery of The First Affiliated Hospital of Fujian Medical University (Fuzhou, China). The mean age of the patients was 53 years (age range, 41–65 years). Clinical and laboratory parameters of patients such as disease severity were evaluated using the Child-Pugh score (23). Tumor staging was based on the Barcelona-Clinic Liver Cancer system (24). All patients signed informed consent forms in accordance with the Declaration of Helsinki. The study protocol was approved by the Ethical Committee of The First Affiliated Hospital of Fujian Medical University [approval no. (2015) 132].

For immunohistochemistry, three areas of tumor specimens and adjacent normal tissues from 20 patients with HCC were fixed with 10% formalin for 24 h at 37°C and embedded in a paraffin block. Subsequently, the samples were de-paraffinized in two changes of xylene for 5 min each at 37°C and rehydrated in 100, 95, 70 and 50% alcohol for 5 min each at 37°C, and finally sectioned into 5-µm-thick slices. For antigen retrieval, samples were treated with 10 mM citrate buffer (pH 6.0) at 100°C for 10 min. After quenching endogenous peroxidase with 3% hydrogen peroxide for 30 min at 37°C, 0.5% bovine serum albumin (BSA; cat. no. ST025-5 g; Beyotime Institute of Biotechnology) was used for 1 h at 37°C to block non-specific binding. The sections were then incubated overnight at 4°C with polyclonal rabbit anti-human TREM1 antibody (1:200; cat. no. YT5133; ImmunoWay Biotechnology Company). Subsequently, the sections were incubated with the appropriate goat anti-rabbit IgG secondary antibody (1:1,000; cat. no. ab6721; Abcam) for 60 min at 37°C according to the manufacturer's instructions. Images of the labeled specimens were captured with an Olympus CX41 light microscope (Olympus Corporation; magnification, ×400) and analyzed using ImageJ software (v1.4; National Institutes of Health). All experiments were performed in triplicate.

Immunohistochemical double staining (CD206+Arg-1, CD16+iNOS and CD206+TREM1) was performed using the DouMaxVision• immunohistochemical double staining test kit (cat. no. KIT-9998; Fuzhou Maixin Biotech Co., Ltd.) according to the manufacturer's protocol, and the number of positively stained cells in five high-power fields was counted under an Olympus CX41 light microscope (Olympus Corporation; magnification, ×400). Samples were incubated with anti-CD16 (1:200; cat. no. ab46629), anti-iNOS (1:200; cat. no. ab53769), anti-CD206 (1:200; cat. no. ab64693) and anti-Arg-1 primary antibodies (1:200; cat. no. ab133543) for 2 h at 37°C. Biotinylated goat anti-rabbit/mouse IgG secondary antibodies (1:1,000; cat. no. KIT-9998; Fuzhou Maixin Biotech Co., Ltd) were added for 60 min at 37°C.

The human leukemia monocytic cell line (THP-1), the 293T cell line and the HepG2 and MHCC97H cell lines were purchased from The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences. THP-1 cells were cultured in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal calf serum (Gibco; Thermo Fisher Scientific, Inc.) and 0.05 mM β-mercaptoethanol (Sigma-Aldrich; Merck KGaA). 293T, HepG2 and MHCC97H cell lines were cultured in high-glucose DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 10% fetal calf serum. All cell lines were cultured with 5% CO₂ at 37°C.

The third generation of lentiviral vector (hU6-MCS-Ubiquitin-EGFP-IRES-puromycin) containing short hairpin (sh)RNAs against TREM1 (shTREM1) used in the present study were produced by Shanghai GeneChem Co., Ltd., and were used for TREM1 gene knockdown. The target sequences of shRNAs were as follows: shTREM1-1, 5′-AGCCAGAAAGCTTGGCAGATA-3′; shTREM1-2, 5′-GAGGATCATACTAGAAGACTA-3′; and shTREM1-3, 5′-GTGGAAGATTCTGGACTGTAT-3′. Three plasmids (Shanghai GeneChem Co., Ltd.), namely 20 µg GV248 carrying the target sequence, 15 µg pHelper1.0 (carrying gag, pol and rev genes) and 10 µg pHelper2.0 (carrying the VSV-G gene), were mixed with Lipofectamine^® 2000 (cat. no. 11668030; Invitrogen; Thermo Fisher Scientific, Inc.) for 15 min at room temperature. Subsequently, the mixture was transfected into 293T cells for 6 h at 37°C to produce lentivirus. Lentiviral particles were collected and the virus titer were detected to be 2×10⁹ TU/ml. Transfections of shRNA into THP-1 cells were performed using polybrene (cat. no. REVG0001; Shanghai GeneChem Co., Ltd.) according to the manufacturer's instructions, and lentivirus at a multiplicity of infection of 50 was chosen for further experiments. After 12 h at 37°C, the medium was changed into RPMI-1640 medium supplemented with 10% fetal calf serum and 0.05 mM β-mercaptoethanol containing no virus. The fluorescence intensity was observed under a fluorescence microscope at ×100 and ×400 magnification. Puromycin (5 µg/ml) was used to screen the shRNA-transfected THP-1 cells after 7 days at 37°C. The knockdown efficiency of shTREM1 was tested by quantitative PCR assay and western blot analysis. A scrambled non-specific shRNA (5′-TTCTCCGAACGTGTCACGT-3′) was used as a negative control (sh-ctrl).

M2 macrophages were derived from THP-1 cells. shTREM1-treated macrophages were derived from shTREM1 THP-1 cells stimulated with 100 ng/ml phorbol 12-myristate 13-acetate (PMA; Sigma-Aldrich; Merck KGaA) for 24 h at 37°C. After PMA treatment, 20 ng/ml IL-4 and 20 ng/ml IL-13 (PeproTech, Inc.) were added for 24 h at 37°C (25).

After incubating M2 and shTREM1 macrophages with fresh RPMI-1640 medium for an additional 24 h at 37°C, the cells' conditioned media (CM) was harvested, centrifuged at 3,000 × g for 15 min at 37°C, filter-sterilized through a 0.22-µm nylon filter (Nalgene; Thermo Fisher Scientific, Inc.) and finally stored at −80°C for further experiments.

For immunofluorescence double labeling, cells were fixed in 4% paraformaldehyde for 20 min at 37°C and permeabilized with 0.1% Triton X-100 for 10 min at 37°C. After blocking non-specific binding with 0.1% BSA for 30 min at 37°C, cells were incubated with primary antibodies against CD163 (cat. no. ab156769; Abcam) and TREM1 antibody (1:200; cat. no. YT5133; ImmunoWay Biotechnology Company) at a 1:500 dilution overnight at 4°C. Subsequently, the samples were incubated with Alexa Fluor^® 488-labeled donkey anti-rabbit IgG antibody (1:1,000; cat. no. A-21206; Thermo Fisher Scientific, Inc.) and Alexa Fluor^® 546-labeled donkey anti-mouse IgG antibody (1:1,000; cat. no. A-10036; Thermo Fisher Scientific, Inc.) for 30 min at room temperature, and 1 µg/ml DAPI (Invitrogen; Thermo Fisher Scientific, Inc.) was used for nuclear labeling for 5 min at room temperature. Images of the labeled specimens were captured with a Zeiss confocal laser scanning microscope (Carl Zeiss AG) at ×2,000 magnification and analyzed using ImageJ software. All treatments were performed in triplicate.

Total cellular RNA was extracted using TRIzol^® reagent (Invitrogen; Thermo Fisher Scientific, Inc.) and was reverse transcribed into cDNA using PrimeScript™ 1st strand cDNA synthesis kit (cat. no. 6110A; Takara Bio, Inc.) according to the manufacturer's protocol. Subsequently, qPCR was performed using a SYBR Green kit (Takara Bio, Inc.) with specific primers for TREM1, iNOS, CD16, IL-12, CD163, Arg-1, IL-10 and GAPDH (used as the reference gene). PCR settings were adjusted according to the manufacturer's instructions: Initial denaturation at 95°C for 30 sec and 40 cycles at 95°C for 5 sec and 60°C for 30 sec. All samples were assayed in triplicate and relative mRNA expression was quantified using the 2^−∆∆Cq method (26). The primer sequences used were as follows: TREM1 forward, 5′-GGCAGATAATAAGGGACGGAGAG-3′ and reverse, 5′-CATTCGGACGCGCAGTAAA-3′; iNOS forward, 5′-GAGCCAGGCCACCTCTATGT-3′ and reverse, 5′-GTCCTCGACCTGCTCCTCAT-3′; CD16 forward, 5′-CCAGTGTGGCATCATGTGG-3′ and reverse, 5′-ATTGAGGCTCCAGGAACACC-3′; IL-12 forward, 5′-CTGAAGAAGATGGTATCACCTGGAC-3′ and reverse, 5′-TTAGAACCTCGCCTCCTTTGTG-3′; CD163 forward, 5′-TTCCTGTTCTGGACGTGTGG-3′ and reverse, 5′-AGCTGGACCACAGCCAAGTT-3′; Arg-1 forward, 5′-GGTGTTGCCTGCTGCCTTCC-3′ and reverse, 5′-GTTCTGAAGAGGTGAGTGGCTGTC-3′; IL-10 forward, 5′-TCAGAGAGGGGGTTAGACCTG-3′ and reverse, 5′-GAGTTGGTCCTGCCAGACTT-3′; GAPDH forward, 5′-CCCTTCATTGACCTCAACTACATG-3′ and reverse, 5′-TGGGATTTCCATTGATGACAAGC-3′.

Cells were washed with cold PBS, and then lysed with radioimmunoprecipitation assay buffer with phenylmethylsulfonyl fluoride (Beyotime Institute of Biotechnology). Cell lysates were collected and then centrifuged at 12,000 × g for 15 min at 4°C. The bicinchoninic acid method was used to detect protein sample concentrations. A total of 20 µg protein/lane was separated via 10% SDS-PAGE, and the proteins were transferred to polyvinylidene fluoride membranes. The membranes were blocked against non-specific binding with 5% non-fat milk for 30 min at 37°C, followed by overnight incubation at 4°C with primary antibodies (all 1:1,000) against TREM1 (cat. no. YT5133; ImmunoWay Biotechnology Company), iNOS (cat. no. ab53769), CD16 (cat. no. ab46629), IL-12 (cat. no. ab106270), CD163 (cat. no. ab156769), Arg-1 (cat. no. ab133543), IL-10 (cat. no. ab34843), phosphorylated (p)-PI3K (cat. no. ab138364), p-AKT (ab183758), p-mTOR (cat. no. ab84400), PI3K (cat. no. ab151549), AKT (cat. no. ab179463), mTOR (cat. no. ab2732), snail (cat. no. ab229701), E-cadherin (cat. no. ab15148) and GAPDH (cat. no. ab181602) (all Abcam). After three washes with TBS with 0.3% Triton X-100, the membranes were incubated with secondary HRP-conjugated goat anti-rabbit IgG (cat. no. ab6721) or goat anti-mouse IgG (cat. no. ab6789) secondary antibodies (both 1:2,000; Abcam) for 1 h at room temperature. The membranes were then washed three times and exposed via enhanced chemiluminescence reagents (BeyoECL Moon kit; cat. no. P0018FS; Beyotime Institute of Biotechnology). The protein bands were captured using Invitrogen iBright FL1500 Imaging System (Thermo Fisher Scientific, Inc.) and analyzed using ImageJ software.

Transwell chambers (8-µm; EMD Millipore) were used to evaluate cancer cell migration and invasion as previously described (27). A total of 80 µl high-glucose DMEM (Gibco; Thermo Fisher Scientific, Inc.) supplemented with 2% fetal calf serum (Gibco; Thermo Fisher Scientific, Inc.) containing 2×10⁵ HepG2 and MHCC97H cells was added into the upper chamber inserts. Serum-free RPMI-1640 medium, M2 macrophage-CM, shTREM1 macrophage-CM and sh-ctrl macrophage-CM were added into the bottom chambers to evaluate their influence on cancer cell migration. Before seeding HepG2 and MHCC97H cells, Matrigel (100 µl; Becton, Dickinson and Company) was placed into the upper chamber for invasion assays for 4 h at 37°C, and the remaining steps were the same as for migration. After a 24-h incubation at 37°C, the upper inserts were washed three times in PBS, fixed with 4% formaldehyde for 20 min at 37°C and stained with crystal violet for 2 h at 37°C. Five fields of view were randomly photographed with an Olympus CX41 light microscope (Olympus Corporation; magnification, ×100) and the numbers of migrated and invaded cells per field were determined using ImageJ software.

A total of 2×10⁵ HepG2 and MHCC97H cells were seeded into culture inserts (Ibidi GmbH); the culture inserts were then removed, and a 500-µm wide gap was generated. Cells were incubated with serum-free RPMI-1640 medium, M2 macrophage-CM, shTREM1 macrophage-CM and sh-ctrl macrophage-CM for 24 h at 37°C. Five fields of view were randomly photographed with an Olympus CX41 light microscope (Olympus Corporation; magnification, ×100) and the gap closure rate was estimated by calculating the gap area with ImageJ software.

HepG2 and MHCC97H cells were seeded into 96-well plates at a density of 1×10⁴ cells/well. Cells were incubated with serum-free RPMI-1640 medium, M2 macrophage-CM, shTREM1 macrophage-CM and sh-ctrl macrophage-CM for 24 h at 37°C. According to the instructions of a BrdU Cell Proliferation Assay kit (cat. no. 6813; Cell Signaling Technology, Inc.), BrdU was added to wells and incubated for 12 h at 37°C. Subsequently, cells were fixed and denatured using 100 µl Fix/Denature solution (part of the aforementioned kit) for 30 min at 37°C. Cells were washed with cold PBS three times. Cells were incubated with 100 µl anti-BrdU antibody (part of the aforementioned kit) for 1 h at 37°C and were then washed three times with cold PBS. Subsequently, HRP-conjugated secondary antibody (part of the aforementioned kit) was added for 1 h at 37°C. A total of 100 µl Substrate Solution was then added and incubated for 5 min at 37°C. Finally, 100 µl Stop Solution was added to stop the enzyme reaction, and the levels of BrdU incorporation was determined using an ELISA reader at a wavelength of 450 nm (Bio-Rad Laboratories, Inc.).

All continuous variables were presented as the mean ± SD. The paired Student's t-test was used for comparative analysis of paired cancerous and adjacent normal tissues. The unpaired Student's t-test was applied for comparisons between 2 groups for the other assays, and one-way ANOVA followed by Tukey's multiple comparison test was used when >2 groups were evaluated. Two-sided P<0.05 was considered to indicate a statistically significant difference. All statistical analysis was performed using SPSS version 19.0 (IBM Corp.).

The clinical characteristics of the study population are described in Table SI. Immunohistochemistry was performed to determine the expression levels of TREM1 in HCC tissues and adjacent normal tissues. The percentage of TREM1⁺ cells per field in HCC tissues was significantly higher than that in adjacent normal tissues (11.23±0.8840 vs. 2.253±0.8662%; P<0.01; Fig. 1A-C). Immunohistochemical double staining was used to detect the tumor-associated macrophage phenotype in HCC by analyzing CD206 and Arg-1 expression. The results revealed that the infiltration of M2 macrophages in liver cancer tissues was significantly increased compared with that in adjacent normal tissues (3.50±0.5000 vs. 0.5100±0.1813%; P<0.01; Fig. 1D-F). No double-positive cells of CD16 and iNOS (M1-associated markers) were observed, indicating no M1 macrophage infiltration in liver cancer tissues and adjacent normal tissues (Fig. 1G-I). In addition, immunohistochemical double staining with the M2 macrophage surface markers CD206 and TREM1 revealed that TREM1 was highly expressed on the surface of M2 macrophages in HCC tissues (16.51±0.9327 vs. 6.644±1.221%; P<0.01; Fig. 1J-L). The current results suggested that M2 macrophages mainly infiltrate HCC tissues, while no infiltration of M1 macrophages was observed in both HCC and adjacent normal tissues. M2 macrophages in HCC tissues highly express TREM1. These results indicated that TREM1 expression on M2 macrophages may serve an important role in the progression of HCC.

To establish a model of human M2 macrophages, human leukemia monocytic cells (THP-1) were stimulated using PMA, IL-4 and IL-13. Double-immunofluorescence labeling was used to identify co-expression of CD163 and TREM1 in M2 macrophages (Fig. 2A). RT-qPCR and western blot analysis revealed that TREM1 expression in M2 macrophages was significantly higher than in THP-1 cells (Fig. 2B and C). The mRNA expression levels of iNOS, CD16 and IL-12 (M1 macrophage-associated markers) were not significantly different between THP-1 cells and M2 macrophages, while the expression levels of CD163, Arg-1 and IL-10 (M2 macrophage-associated markers) were significantly increased in M2 macrophages compared with in THP-1 cells (Fig. 2D and E). Similarly, protein expression levels of iNOS, CD16 and IL-12 exhibited no significant difference between THP-1 cells and M2 macrophages, while protein expression levels of CD163, Arg-1 and IL-10 were significantly upregulated in M2 macrophages compared with in THP-1 cells (Fig. 2F). Overall, the present data indicated that M2 macrophages were successfully induced from the THP-1 monocytic cell line.

To discover possible functions of TREM1 in macrophage polarization, shRNAs were used to knock down TREM1 expression. As shown in Fig. 3A, shTREM1-2 was the most effective at TREM1 silencing and stable regulation of TREM1 in THP-1 cells and was therefore used for subsequent experiments. Western blot analysis revealed that transfection with shTREM1-2 resulted in a significant downregulation of TREM1 expression compared with sh-ctrl-transfected cells (Fig. 3B). RT-qPCR and western blot analysis demonstrated that iNOS, CD16 and IL-12 expression was significantly upregulated in shTREM1 macrophages, while CD163, Arg-1 and IL-10 expression was significantly downregulated in shTREM1 macrophages compared with in sh-ctrl macrophages (Fig. 3C-E). Overall, the present results indicated that TREM1 downregulation inhibited M2 macrophage polarization and promoted a shift towards the M1 phenotype.

A co-culture system was used to investigate the effect of shTREM1 macrophages on the migration and invasion of cancer cells. The results revealed that M2 macrophage-CM (M2-CM) significantly increased cell gap closure (Fig. 4A), as well as migration and invasion (Fig. 4B), in HepG2 and MHCC97H cell lines compared with RPMI-1640 medium (1640); meanwhile, shTREM1 macrophage-CM (shTREM1-CM) significantly decreased this effect. The present results indicated that downregulation of TREM1 can inhibit the migration and invasion of tumor cells.

A co-culture system was used to investigate the effect of shTREM1 macrophages on EMT of cancer cells. The results revealed that M2-CM significantly increased the protein expression levels of snail and significantly decreased those of E-cadherin in HepG2 (Fig. 5A) and MHCC97H (Fig. 5B) cells compared with RPMI-1640 medium; meanwhile, shTREM1-CM reversed this effect. These results indicated that downregulation of TREM1 inhibited the EMT of tumor cells. A BrdU kit was used to investigate the effect of shTREM1 macrophages on the proliferation of cancer cells. The results demonstrated that M2-CM significantly increased proliferation of HepG2 and MHCC97H cells compared with RPMI-1640 medium; meanwhile, shTREM1-CM reversed this effect (Fig. 5C). The present results indicated that downregulation of TREM1 inhibited the proliferation of tumor cells.

To discover the molecular mechanisms of the effects of silencing TREM1 in macrophages and inhibition of M2 polarization of macrophages, western blotting was used to determine the protein expression levels of p-PI3K (p85), PI3K, p-AKT, AKT, p-mTOR and mTOR. As illustrated in Fig. 6, the ratios of p-PI3K/PI3K, p-AKT/AKT and p-mTOR/mTOR were significantly upregulated in M2-polarized macrophages, while downregulation of TREM1 in macrophages reversed this effect. These results indicated that the PI3K/AKT/mTOR signaling pathway may be involved in the regulation of macrophage polarization by TREM1.