Mannose-binding lectin 2 secreted by hepatocellular carcinoma cells recruits and activates natural killer cells to reshape an immune-activated microenvironment

Figures Abstract Crosstalk between hepatocellular carcinoma (HCC) and the tumor microenvironment (TME) is pivotal for the initiation and management of HCC. The infiltration and function of natural killer (NK) cells in the TME are frequently hindered. However, it is unclear whether a crucial regulatory factor originating from HCC cells directly modulates NK cell activity to evade immune surveillance. In this study, we found that mannose-binding lectin 2 (MBL2) expression was markedly decreased in HCC and positively correlated with HCC prognosis. MBL2 inhibited the proliferation and migration of HCC cells intracellularly. Human and murine co-culture systems of HCC and NK cells were established to demonstrate that secreted MBL2 recruited and activated NK cells in the TME, particularly upregulating the infiltration of NKp46+ NK cells. Furthermore, secreted MBL2 promoted the production of IL-13 and IL-25 by NK cells, resulting in a decrease in exhausted cytotoxic T lymphocytes. Mechanistically, MBL2 interacts with the integrin β1 receptor, activating the FAK/AKT pathway and increasing PD-L1 expression on NK cells. Our discovery identifies MBL2 as an NK cell–activating cytokine, initiating the integrin β1/FAK/AKT pathway in NK cells and reshaping an immune-activated microenvironment of HCC. Strategies to up-regulate MBL2 may enhance the anti-PD-L1 immunotherapy efficacy and serve as a potential therapeutic approach for HCC. Citation: Liao H, Yang J, Cai L, Chi L, Wang C, Xu Y, et al. (2026) Mannose-binding lectin 2 secreted by hepatocellular carcinoma cells recruits and activates natural killer cells to reshape an immune-activated microenvironment. PLoS Biol 24(5): e3003793. https://doi.org/10.1371/journal.pbio.3003793 Academic Editor: Elena Rainero, The University of Sheffield, UNITED KINGDOM OF GREAT BRITAIN AND NORTHERN IRELAND Received: April 14, 2025; Accepted: April 24, 2026; Published: May 20, 2026 Copyright: © 2026 Liao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability: All data underlying the findings described in this manuscript are fully available without restriction. Quantitative data are provided in S1 Data, and raw immunoblot images are available in S1 Raw Images. Custom R scripts used for data analysis, along with the raw immunoblot images, are publicly accessible on Zenodo (DOI: https://doi.org/10.5281/zenodo.19657056). Furthermore, all raw flow cytometry FCS files and corresponding gating strategies have been deposited in a separate Zenodo repository (DOI: https://doi.org/10.5281/zenodo.19396187). Funding: This work was supported by the National Natural Science Foundation of China (grant numbers 82072627 and 82373159 to M.P.) and the Key-Area Research and Development Program of Guangdong Province (grant number 2023B1111020008 to M.P.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Funding websites:PRC-NSFC https://grants.nsfc.gov.cn; PRC-KARDPGP https://pro.gdstc.gd.gov.cn. Competing interests: The authors have declared that no competing interests exist. Abbreviations: anti-ASGM1, anti-asialo GM1; CM, culture medium; CTLs, cytotoxic T lymphocytes; DCN, decorin; DSS, disease-specific survival; ED, extracellular domain; FMO, fluorescence minus one; GSEA, gene set enrichment analysis; GZMB, granzyme B; HBV, hepatitis B virus; HBx, Hepatitis B virus X protein; HCC, hepatocellular carcinoma; IFN, interferons; LDH, lactate dehydrogenase; Lr-NK, liver-resident NK; MBL2, mannose-binding lectin 2; MHC-I, major histocompatibility complex class I; NK, natural killer; OPG, osteoprotegerin; OS, overall survival; p-AKT, phosphorylated AKT; PBMCs, peripheral blood mononuclear cells; PFS, progression-free survival; PVDF, Polyvinylidene fluoride membranes; rMBL2, recombinant MBL2; RFS, recurrence-free survival; RGD, Arg-Gly-Asp peptide; RT, room temperature; SP, signal peptide; TME, tumor micro environment; WB, western blot 1. Introduction Hepatocellular carcinoma (HCC) is one of the most common malignant tumors and the third leading cause of cancer-related death worldwide [1]. Interactions between HCC cells and the immune system play a crucial role in the development and treatment of HCC [2]. Natural killer (NK) cells in the liver play a pivotal role in innate immunity, serving as the first line of defense in host immune surveillance and a major checkpoint in the immune evasion process of HCC cells [3]. NK cells possess intrinsic anti-tumor functions, whereby activated NK cells kill tumor cells by releasing cytotoxic granules and inducing apoptosis, thereby inhibiting tumor growth [4]. Moreover, NK cells produce immunomodulatory molecules, such as interferons (IFN), which activate other immune cells, such as macrophages and T cells, to enhance immune responses [5,6]. The activating receptors on NK cells (NKG2D and NKp46), along with the cytotoxic proteins granzyme B (GZMB) and perforin, play critical roles in the recognition and elimination of tumors by NK cells [7]. In addition to directly killing HCC cells through cell-to-cell contact, activated NK cells produce IFN-γ, which positively influences anti-tumor immunity against HCC [8]. However, owing to tumor antigen alterations and an increase in immunosuppressive cells within the tumor microenvironment (TME), NK cells can fail to properly recognize and eliminate target cells. Myeloid-derived suppressor cells in patients with HCC directly inhibit the cytotoxic effects of NK cells in a contact-dependent manner, whereas regulatory T cells suppress IFN-γ production in tumor-infiltrating NK cells and diminish their cytotoxic capabilities [9]. Therefore, elucidating the alterations in NK cell functional status during tumor initiation and progression is crucial. Cytokines in the TME play a critical regulatory role in influencing the quantity and activity of NK cells [10,11]. IL-12 and IL-18 are crucial factors for NK cell activation and enhanced cytotoxicity, [12–14], whereas excessive IL-10 production in HCC is linked to impaired NK cell function [15]. However, whether cytokines derived from HCC cells directly regulate NK cell activity to evade immune surveillance remains unclear. Mannose-binding lectin 2 (MBL2) is primarily synthesized by hepatocytes and then secreted into the extracellular matrix and plasma [16]. MBL2 either activates the complement activation lectin pathway, exerting an indirect immunoregulatory effect, or binds to phagocytic cell lectin receptors, acting as a direct modulator [17]. Thus, deficiency or dysfunction of MBL2 increases the risk of infection and autoimmune diseases [18]. Previous research on liver diseases has shown that low MBL2 expression is associated with chronic hepatitis B virus (HBV) infection [19]. Moreover, polymorphisms in MBL2 are associated with susceptibility to HCC and can influence its progression and prognosis [20,21]. Although MBL2 is a crucial regulatory factor in HCC, its role in modulating innate immunity within HCC microenvironment, as well as the specific immune cell types involved in tumor immune regulation, has yet to be fully elucidated. In this study, we aimed to identify cytokines derived from HCC cells that are associated with HBV infection and that possess the potential to reshape an immunotherapy-activated TME. We investigated the intracellular and extracellular regulatory role of MBL2 in HCC proliferation and invasion. Further, we elucidated the molecular mechanisms by which MBL2 recruits and activates NK cells in the HCC microenvironment. Our results demonstrate that MBL2 can up-regulate the expression of PD-L1 in both HCC and NK cells. These findings suggest that strategies to up-regulate MBL2 can enhance the effectiveness of PD-L1 immunotherapy and serve as a therapeutic approach for HCC. 2. Results 2.1. Endogenous MBL2 inhibits the proliferation and migration of HCC cells Recent studies have identified CTCs in the bloodstream as a marker of HCC early metastasis. We performed noninvasive quantification of CTCs using liquid biopsy techniques. High-throughput transcriptome sequencing between HCC tissues with CTC-High level and CTC-Low levels (SRA data: PRJNA912860) revealed differential genes promoting metastasis and immune escape in hepatocellular carcinoma. The above sequencing results were combined with The Human Secretome and Membrane Proteome (HPA SPOCTOPUS), HBV-related HCC dataset (OEP000321), The Immunology Database and Analysis Portal (ImmPort) and GSE45436 dataset to identify the differential cytokines derived from HCC cells with potential for TME regulation. The intersection result showed MBL2 has the potential to modulate the HCC immune microenvironment, which is secreted into the extracellular matrix of HCC and correlates with tumor proliferation and metastasis (Fig 1A). MBL2 is predominantly synthesized by hepatocytes and secreted into the circulation in humans [16]. To determine the basal secretion level of MBL2 in human HCC cell lines, culture medium (CM) supernatants of Hep3B, HepG2, Huh7, MHCC97-L (97L), MHCC97-H (97H), and LM3 cells were collected. ELISA assays were utilized to assess the secretion of MBL2 in HCC cell lines. The results indicated that the MBL2 secretion levels of Huh7 and 97L cells were relatively high, whereas the MBL2 secretion of Hep3B, HepG2 and LM3 cells was relatively weak (Fig 1B). We used q-PCR and WB to clarify the expression of MBL2 in HCC cells at the transcriptional and translational levels, respectively. q-PCR highlighted that the MBL2 mRNA expression was relatively high in Huh7 and relatively low in Hep3B, 97H, and LM3 cells, after normalizing the mRNA of MBL2 in Hep3B (Fig 1C). The Human Protein Atlas database showed that the expression of MBL2 was highest in Huh7 cells, whereas the expression in Hep3B cells was relatively low (S1A Fig), which was consistent with our results. In the WB results, Huh7 and 97L cell lines showed higher MBL2 levels than other cell lines, whereas Hep3B and LM3 cells showed lower MBL2 expression (Fig 1D). Therefore, Huh7 cells were employed for MBL2 knockdown, and Hep3B and LM3 cells were adopted to construct cell lines stably overexpressing the MBL2 gene. Secretion of MBL2 to the extracellular compartment is blocked by the absence of signal peptides in HCC. To discern both the intracellular and extracellular impacts of endogenous MBL2 on HCC progression, full-length MBL2 molecules and truncated MBL2 proteins lacking the signal peptide (MBL2ΔSP) were generated following the human genome sequence (Fig 1E). After employing lentiviral transfection, qPCR (S1B Fig) and WB (Fig 1F) confirmed the stable construction of MBL2-knockdown cell lines (Huh7-MBL2-shRNA1&2) and the establishment of full-length MBL2 and MBL2ΔSP overexpression cell lines (Hep3B-MBL2, Hep3B-MBL2ΔSP, LM3-MBL2, and LM3-MBL2ΔSP) at the transcriptional and translational levels, respectively. (A) Venn diagram of the PRJNA912860 dataset, GSE45436 dataset, SPOCTOPUS, OEP000321 dataset and ImmPort. (B) ELISA was used to measure extracellular MBL2 in the culture supernatants of HCC cell lines. (C, D) qPCR and WB were used to detect MBL2 at the transcriptional and translational levels. (E) Schematic illustration of MBL2 constructs. (F) WB was used to assess the knockdown efficiency of MBL2 in human Huh7 cells and the MBL2 expression levels in human Hep3B and LM3 cells. (G, H) Impact of endogenous MBL2 on HCC proliferation was assessed using CCK-8 (G) and colony formation assays (H). (I) Transwell assays were employed to assess the impact of endogenous MBL2 on HCC cell migration. (J) Nude mouse subcutaneous HCC models of the vector group, MBL2 group, and MBL2ΔSP group. (K, L) Growth curve and tumor weight of the subcutaneous tumor model in nude mice. All values are shown as mean ± SD. *p Gels’ function was utilized to plot lane intensity histograms. To subtract background noise, profile peaks were enclosed with straight baselines. The Wand tool was then used to calculate the integrated density (area) of each peak. Final relative expression levels were determined by normalizing these values against their corresponding loading controls (β-Actin or β-Tubulin). Source data are available in S1 Raw Images. Flow cytometry Flow cytometry was employed to quantify the expression of cell surface markers (NKG2D, NKp46, PD-L1, PD-1, TIM-3) and intracellular cytokines (GZMB, perforin, IFN-γ) in the indicated cell populations. Briefly, cells were washed twice with flow cytometry wash buffer (PBS containing 1% FBS and 0.5 mM EDTA) and incubated with Fc Block (#553,142; BD Pharmingen) at room temperature (RT) for 10 min to prevent nonspecific antibody binding. Approximately 1 × 10⁶ cells were resuspended in 100 μL of staining buffer and stained with Fixable Viability Stain 700 (1:1000; #564997; BD Biosciences) to exclude dead cells. NK92 cell degranulation was assessed by adding anti-CD107a and BFA/Monensin Mixture (5 μL/mL; #CS1002, MultiSciences) at the onset of a 6-hour stimulation period, followed by surface staining. For surface antigen detection, cells were incubated with fluorochrome-conjugated monoclonal antibodies (listed in S3 Data) at 4 °C for 30 min in the dark. For intracellular cytokine detection, NK cell-derived cytokines were stimulated with PMA/Ionomycin Mixture (5 μL/mL; #CS1001, MultiSciences) and their intracellular accumulation was blocked using BFA/Monensin Mixture (5 μL/mL) for 6 hours at 37 °C. After surface staining, cells were fixed and permeabilized using BD Cytofix/Cytoperm Fixation/Permeabilization Kit (#554714, BD Pharmingen) according to the manufacturer’s protocol, followed by intracellular staining with relevant antibodies (listed in S3 Data) at RT for 30 min. All staining steps were performed in staining buffer under light-protected conditions. Data acquisition was conducted using a BD LSRFortessa flow cytometer (BD Biosciences), and data analysis was performed with FlowJo software (Tree star). Fluorescence minus one (FMO) controls and isotype controls were included to ensure gating accuracy and control for nonspecific staining. Immunohistochemistry (IHC) IHC was performed on 2.5 μm sections of formalin-fixed, paraffin-embedded HCC and adjacent noncancerous tissues. Antigen retrieval was conducted using citrate buffer (pH 6.0) or EDTA buffer (pH 8.0). After blocking endogenous peroxidase activity and nonspecific binding, sections were incubated overnight at 4 °C with primary antibodies against MBL2 (1:1000, DF4152; Affinity, Shanghai, China). Detection was performed using an HRP-conjugated secondary antibody and DAB staining kit (Zsbio), followed by nuclear counterstaining with Mayer’s hematoxylin. Slides were independently evaluated by three pathologists to assess the expression of specific markers in tumor-infiltrating lymphocytes within both the tumor core and the tumor margin. The tumor margin was defined as a region extending 200 μm inward from the histologically identifiable invasive edge of HCC tissues. Protein expression was scored by multiplying the percentage of positive cells (0–4: 0% = 0; 1%–25% = 1; 26%–50% = 2; 51%–75% = 3; > 75% = 4) by staining intensity (0 = none; 1 = weak; 2 = moderate; 3 = strong). A total score ≤5 was classified as low expression, while >5 indicated high expression. Discrepancies were resolved by joint review. Enzyme-linked immunosorbent assay (ELISA) A total of 5 × 10⁵ HCC cells or NK cells were cultured and subjected to the indicated treatments for 48 hours. After incubation, cell culture supernatants were collected and concentrated using Amicon Ultra centrifugal filters by centrifugation at 4,000 × g for 30 min at 4 °C. The concentrations of MBL2, GZMB, and perforin in the supernatants were quantified using ELISA kits from Cloud-Clone Corporation, following the manufacturer’s protocols. Briefly, samples and standards were added to 96-well plates pre-coated with specific capture antibodies and incubated. After washing to remove unbound substances, biotin-labeled detection antibodies were applied, followed by HRP-conjugated streptavidin. Substrate solution was added for color development, and the reaction was terminated with a stop solution. The optical density (OD) was measured at 450 nm. Sample concentrations were calculated by interpolation from a standard curve generated with known concentrations of the respective proteins. Immunoprecipitation (IP), co-IP and RNA-IP (RIP) Cells were lysed in RIPA buffer supplemented with protease inhibitors (1:100; Epizyme). Equal amounts of total cell lysates (500 μg of protein) were incubated with the indicated antibodies and Protein A/G agarose beads (sc-2003; Santa Cruz) overnight at 4 °C. Following incubation, the beads were washed three times with lysis buffer to remove non-specifically bound proteins. Co-IP was conducted to evaluate the interaction between MBL2 and integrin β1. Protein complexes were resolved on a 10% SDS-PAGE gel and transferred onto PVDF membranes. Membranes were blocked with 5% nonfat milk in Tris-buffered saline and then incubated with primary antibodies, followed by HRP-conjugated secondary antibodies. Protein signals were detected using an ECL chemiluminescence reagent (FD8030; FDbio) and visualized with a chemiluminescence detection system (Bio-Rad). To investigate the binding of HBx protein to MBL2 mRNA, RIP was performed using the EZ-Magna RIP Kit (#17-10086; Merck) in conjunction with qPCR. Briefly, 293T cells were transfected with the HBx plasmid for 48 hours, lysed in RIP lysis buffer, and incubated with magnetic beads conjugated with anti-HBx or control IgG antibodies. After three washes with the wash buffer, RNA bound to the magnetic beads was eluted and used as a template for qPCR analysis. The relative abundance of MBL2 mRNA in the eluates was quantified to assess the HBx-MBL2 mRNA interaction. Statistical analysis Statistical analyses were performed using SPSS 22.0 (IBM). Data are presented as mean ± SD unless otherwise specified. Normality and variance homogeneity were evaluated using Shapiro–Wilk and Levene’s tests, respectively. Student t test or one-way ANOVA followed by Tukey’s post-hoc test was applied for group comparisons in qPCR, flow cytometry, IHC, CCK-8, colony formation, Transwell analyses, LDH assays and ELISA assays. For all in vivo animal experiments, nonparametric counterparts were utilized, specifically the Mann–Whitney U test for two-group comparisons and the Kruskal–Wallis test followed by Dunn’s multiple comparisons test for multi-group analyses. Kaplan–Meier curves with the log-rank test were used for survival analysis. Statistical significance was set to p < 0.05, whereas “ns” indicated no significance. Supporting information S1 Fig. (A) The Human Protein Atlas database showed that the expression of MBL2 was highest in Huh7 cells, whereas the expression in Hep3B cells was relatively low. (B) qPCR was used to confirm the stable construction of MBL2-knockdown cell lines (Huh7-MBL2-shRNA1&2) and the establishment of full-length MBL2 and MBL2ΔSP overexpression cell lines. CCK-8 assays were performed to evaluate the proliferative effects of recombinant MBL2 at three concentrations (0.25 ng/mL, 0.50 ng/mL, and 0.75 ng/mL) on Hep3B (C) and LM3 cells (D), with the PBS-treated group serving as a control. All values are shown as mean ± SD. ***p < 0.001. ns indicates no significance. Source data are available in S1 Data. rMBL2, recombinant mannose-binding lectin 2; MBL2ΔSP, MBL2 proteins lacking the signal peptide. https://doi.org/10.1371/journal.pbio.3003793.s001 (TIF) S2 Fig. (A) Flow cytometry revealed that the proportion of CD3+ CD8+ T cells within the CD3+ T cell population was not directly altered by rMBL2 treatment. (B) A schematic representation illustrating the flow cytometry gating strategy and corresponding FMO controls (NKG2D, NKp46, PD-L1, GZMB, and IFN-γ) for peripheral CD3− CD56+ NK cells. (C) rMBL2 treatment significantly elevated NKG2D expression on peripheral NK cells, as assessed by flow cytometry. (D) Flow cytometry scatter plots demonstrated that rMBL2 treatment enhanced NKp46 expression on peripheral NK cells. All values are shown as mean ± SD. ***p < 0.001. Source data are available in S1 Data. PBMCs, peripheral blood mononuclear cells. https://doi.org/10.1371/journal.pbio.3003793.s002 (TIF) S3 Fig. (A and B) Sequential gating workflow and CD107a FMO control for CD107a degranulation assay following (A) stimulation with rMBL2 or (B) co-culture with Hep3B-MBL2 cells. (C) CD107a expression was significantly increased in rMBL2-treated NK cells compared to the PBS control. (D) Co-culture with Hep3B cells overexpressing MBL2 also upregulated CD107a expression on NK92 cells. (E) Flow cytometry scatter plots illustrating GZMB production in the rMBL2-treated peripheral NK cells. (F) Representative scatter plots showing CD3− CD56bright and CD3− CD56dim NK cell subsets in PBMCs treated with PBS or rMBL2 (250 ng/mL) for 48 hour. (G and H) Flow cytometry assays revealed that rMBL2 stimulation significantly increased NKG2D expression on both CD3− CD56bright and CD3− CD56dim NK cell subsets compared to the PBS control group. (I) Schematic diagram illustrating the workflow of flow cytometry for the detection of murine NK cell membrane receptors and intracellular proteins. (J) Schematic illustration of the co-culture model involving Hepa1–6 cells and murine liver NK cells. (K) Schematic representation of the co-culture model of NK92 cells and HCC cells. (L) Transwell migration assays were performed to evaluate the effect of MBL2 on NK cell recruitment. (M) LDH assays demonstrated increased LDH levels in the Hep3B-MBL2 groups after co-culture with NK92 cells. All values are shown as mean ± SD. **p < 0.01, ***p < 0.001. ns indicates no significance. Source data are available in S1 Data. FMO, Fluorescence Minus One; LDH, lactate dehydrogenase. https://doi.org/10.1371/journal.pbio.3003793.s003 (TIF) S4 Fig. (A, B) The Gene-concept Networks represented the membrane proteins that interact with MBL2 as identified through mass spectrometry analysis. (C) The heatmaps represented the correlation between integrin β1 and canonical pathways derived from Reactome gene sets within the transcriptional profile of infiltrating NK cells in HCC. (D) Flow cytometry indicated that MBL2 enhanced the proportion of the NKp46⁺ NK cells and the GZMB⁺ NK cells, an effect that was reversed upon RGD peptide treatment. Source data are available in S1 Data. RGD, Arg-Gly-Asp peptide. https://doi.org/10.1371/journal.pbio.3003793.s004 (TIF) S5 Fig. (A) The heatmap revealed differential genes in eukaryotic transcriptome sequencing to investigate the molecular mechanisms underlying MBL2 activation in NK cells. (B) The GSE183349 dataset was used to perform correlation analysis between integrin β1 and canonical pathways within infiltrating NK cells in HCC (R = 0.62). (C) qPCR demonstrated that MBL2 downregulated the expression of IFNGR2 in HCC transcriptionally. (D) Flow cytometry revealed that MK2206 could reverse MBL2-induced upregulation of NKG2D+ NK and NKp46+ NK cells. (E) Sequential gating strategy and corresponding FMO controls (NKp46, NKG2D, GZMB, and IFN-γ) in AKT pathway rescue experiments. Representative flow cytometry scatter plots showing a significant reduction in PD-1⁺ CTLs (F) and TIM-3⁺ CTLs (G) in the IL-13 and IL-25 treatment groups compared with the PBS control group. Neutralization of IL-13 and IL-25 reversed the suppression of PD-1⁺ CTL (H) and TIM-3⁺ CTL (I) populations mediated by MBL2-activated NK cells. All values are shown as mean ± SD. ***p < 0.001. Source data are available in S1 Data. MK2206, MK-2206 2HCl, a pan-AKT inhibitor. https://doi.org/10.1371/journal.pbio.3003793.s005 (TIF) S6 Fig. (A) IHC staining revealed a marked reduction in PD-1 expression in the Luc-MBL2 + anti -PD-L1 group compared with the Luc-NC + anti–PD-L1 group. (B–D) Kaplan-Meier survival analysis revealed discernible differences in (B) progression-free survival (PFS), (C) disease-specific survival (DSS) and (D) recurrence-free survival (RFS) associated with high MBL2 expression. https://doi.org/10.1371/journal.pbio.3003793.s006 (TIF) S1 Data. Source data for all figures. An Excel spreadsheet containing the underlying numerical data for Figs 1B–1D, 1F–1I, 1K, 1L, 2A–2K, 2M, 2N, 3F–3H, 3K–3M, 4A–4G, 4J–4O, 5A–5C, 5E–5I, 5M, 5O–5Q, 6A, 6B, 6D–6F, 6H, 7A, 7C–7F, S1A–S1C, S2A, S2C, S3C, S3D, S3G, S3H, S3L, S3M, and S5C. https://doi.org/10.1371/journal.pbio.3003793.s007 (XLSX) S3 Data. Supplementary Materials. (A) Primary antibodies in western blot assays; (B) Oligonucleotide information; (C) Fluorochrome-conjugated antibodies in flow cytometry. https://doi.org/10.1371/journal.pbio.3003793.s009 (DOCX) S4 Data. Western blot quantification and statistical analyses. https://doi.org/10.1371/journal.pbio.3003793.s010 (PDF) S1 Raw Images. Raw images of all blots and gels. https://doi.org/10.1371/journal.pbio.3003793.s011 (PDF) Acknowledgments We sincerely acknowledge the TCGA, ICGC, ANTE, and GEO database owners for providing their platforms and the contributors for uploading meaningful datasets. We gratefully acknowledge Prof. Zhizhang Wang and Prof. Qianbing Zhang (Southern Medical University) for their expert guidance and valuable advice regarding the flow cytometry analysis. Part of Fig 8 was drawn using pictures from Servier Medical Art (https://smart.servier.com/), licensed under a Creative Commons Attribution 4.0 International License (https://creativecommons.org/licenses/by/4.0/). References - 1. Sung H, Ferlay J, Siegel RL, Laversanne M, Soerjomataram I, Jemal A. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2021;71:209–49. - 2. Xue R, Zhang Q, Cao Q, Kong R, Xiang X, Liu H, et al. Liver tumour immune microenvironment subtypes and neutrophil heterogeneity. Nature. 2022;612(7938):141–7. pmid:36352227 - 3. Tang F, Li J, Qi L, Liu D, Bo Y, Qin S, et al. A pan-cancer single-cell panorama of human natural killer cells. Cell. 2023;186(19):4235-4251.e20. pmid:37607536 - 4. Crinier A, Narni-Mancinelli E, Ugolini S, Vivier E. SnapShot: natural killer cells. Cell. 2020;180(6):1280-1280.e1. pmid:32200803 - 5. Chiang J, Chen P-C, Pham J, Nguyen C-Q, Kaur K, Raman SS, et al. Characterizing hepatocellular carcinoma stem markers and their corresponding susceptibility to NK-cell based immunotherapy. Front Immunol. 2023;14:1284669. pmid:37954598 - 6. López-Soto A, Gonzalez S, Smyth MJ, Galluzzi L. Control of metastasis by NK cells. Cancer Cell. 2017;32(2):135–54. pmid:28810142 - 7. Rocca Y, Pouxvielh K, Marotel M, Benezech S, Jaeger B, Allatif O, et al. Combinatorial expression of NK cell receptors governs cell subset reactivity and effector functions but not tumor specificity. J Immunol. 2022;208(7):1802–12. pmid:35288470 - 8. Lin Q, Rong L, Jia X, Li R, Yu B, Hu J, et al. IFN-γ-dependent NK cell activation is essential to metastasis suppression by engineered Salmonella. Nat Commun. 2021;12(1):2537. pmid:33953170 - 9. Zhou Y, Cheng L, Liu L, Li X. NK cells are never alone: crosstalk and communication in tumour microenvironments. Mol Cancer. 2023;22(1):34. pmid:36797782 - 10. Liu Y, Xun Z, Ma K, Liang S, Li X, Zhou S, et al. Identification of a tumour immune barrier in the HCC microenvironment that determines the efficacy of immunotherapy. J Hepatol. 2023;78(4):770–82. pmid:36708811 - 11. Albini A, Noonan DM. Decidual-like NK cell polarization: from cancer killing to cancer nurturing. Cancer Discov. 2021;11(1):28–33. pmid:33277305 - 12. Zhou T, Damsky W, Weizman O-E, McGeary MK, Hartmann KP, Rosen CE, et al. IL-18BP is a secreted immune checkpoint and barrier to IL-18 immunotherapy. Nature. 2020;583(7817):609–14. pmid:32581358 - 13. Clark JT, Weizman O-E, Aldridge DL, Shallberg LA, Eberhard J, Lanzar Z, et al. IL-18BP mediates the balance between protective and pathological immune responses to Toxoplasma gondii. Cell Rep. 2023;42(3):112147. pmid:36827187 - 14. Heidemann SC, Chavez V, Landers CJ, Kucharzik T, Prehn JL, Targan SR. TL1A Selectively enhances IL-12/IL-18-induced NK cell cytotoxicity against NK-resistant tumor targets. J Clin Immunol. 2010;30: 531–8. - 15. Sun H, Huang Q, Huang M, Wen H, Lin R, Zheng M, et al. Human CD96 correlates to natural killer cell exhaustion and predicts the prognosis of human hepatocellular carcinoma. Hepatology. 2019;70(1):168–83. pmid:30411378 - 16. Choteau L, Parny M, François N, Bertin B, Fumery M, Dubuquoy L, et al. Role of mannose-binding lectin in intestinal homeostasis and fungal elimination. Mucosal Immunol. 2016;9(3):767–76. pmid:26442658 - 17. Kalia N, Singh J, Kaur M. The ambiguous role of mannose-binding lectin (MBL) in human immunity. Open Med (Wars). 2021;16(1):299–310. pmid:33681468 - 18. Stravalaci M, Pagani I, Paraboschi EM, Pedotti M, Doni A, Scavello F, et al. Recognition and inhibition of SARS-CoV-2 by humoral innate immunity pattern recognition molecules. Nat Immunol. 2022;23(2):275–86. pmid:35102342 - 19. Yuen MF, Lau CS, Lau YL, Wong WM, Cheng CC, Lai CL. Mannose binding lectin gene mutations are associated with progression of liver disease in chronic hepatitis B infection. Hepatology. 1999;29(4):1248–51. pmid:10094971 - 20. Eurich D, Boas-Knoop S, Morawietz L, Neuhaus R, Somasundaram R, Ruehl M, et al. Association of mannose-binding lectin-2 gene polymorphism with the development of hepatitis C-induced hepatocellular carcinoma. Liver Int. 2011;31(7):1006–12. pmid:21733090 - 21. Lin Y, Su C, Niu J, Guo Z, Cai L. Impact of mannose-binding lectin 2 polymorphism on the risk of hepatocellular carcinoma: a case-control study in Chinese Han population. J Epidemiol. 2015;25(5):387–91. pmid:25787238 - 22. Li T, Fu J, Zeng Z, Cohen D, Li J, Chen Q, et al. TIMER2.0 for analysis of tumor-infiltrating immune cells. Nucleic Acids Res. 2020;48(W1):W509–14. pmid:32442275 - 23. Llovet JM, Castet F, Heikenwalder M, Maini MK, Mazzaferro V, Pinato DJ, et al. Immunotherapies for hepatocellular carcinoma. Nat Rev Clin Oncol. 2022;19(3):151–72. pmid:34764464 - 24. Cheng J, Yan J, Liu Y, Shi J, Wang H, Zhou H, et al. Cancer-cell-derived fumarate suppresses the anti-tumor capacity of CD8+ T cells in the tumor microenvironment. Cell Metabolism. 2023;35: 961–78.e10. - 25. Hui L, Chen Y. Tumor microenvironment: sanctuary of the devil. Cancer Lett. 2015;368(1):7–13. pmid:26276713 - 26. Li J, Li H, Yu Y, Liu Y, Liu Y, Ma Q, et al. Mannan-binding lectin suppresses growth of hepatocellular carcinoma by regulating hepatic stellate cell activation via the ERK/COX-2/PGE2 pathway. Oncoimmunology. 2018;8(2):e1527650. pmid:30713782 - 27. Nakagawa T, Kawasaki N, Ma Y, Uemura K, Kawasaki T. Antitumor activity of mannan-binding protein. In: Methods in enzymology. Elsevier; 2003. pp. 26–33. - 28. Brown KS, Ryder SD, Irving WL, Sim RB, Hickling TP. Mannan binding lectin and viral hepatitis. Immunol Lett. 2007;108(1):34–44. pmid:17157924 - 29. Swierzko AS, Kilpatrick DC, Cedzynski M. Mannan-binding lectin in malignancy. Mol Immunol. 2013;55(1):16–21. pmid:23062612 - 30. Ma Y, Uemura K, Oka S, Kozutsumi Y, Kawasaki N, Kawasaki T. Antitumor activity of mannan-binding protein in vivo as revealed by a virus expression system: mannan-binding proteindependent cell-mediated cytotoxicity. Proc Natl Acad Sci U S A. 1999;96(2):371–5. pmid:9892640 - 31. Gu X, Ji Q, Wang H, Jiang M, Yang J, Fang M, et al. Genetic variants of mannose-binding lectin 2 gene influence progression and prognosis of patients with hepatitis B virus infection in China. Clin Res Hepatol Gastroenterol. 2016;40(5):614–21. pmid:26857650 - 32. Slagle BL, Bouchard MJ. Hepatitis B Virus X and regulation of viral gene expression. Cold Spring Harb Perspect Med. 2016;6(3):a021402. pmid:26747833 - 33. Kim G-W, Siddiqui A. Hepatitis B virus X protein recruits methyltransferases to affect cotranscriptional N6-methyladenosine modification of viral/host RNAs. Proc Natl Acad Sci U S A. 2021;118(3):e2019455118. pmid:33397803 - 34. Qin W-H, Yang Z-S, Li M, Chen Y, Zhao X-F, Qin Y-Y, et al. High serum levels of cholesterol increase antitumor functions of nature killer cells and reduce growth of liver tumors in mice. Gastroenterology. 2020;158(6):1713–27. pmid:31972238 - 35. Jaeger BN, Donadieu J, Cognet C, Bernat C, Ordoñez-Rueda D, Barlogis V, et al. Neutrophil depletion impairs natural killer cell maturation, function, and homeostasis. J Exp Med. 2012;209(3):565–80. pmid:22393124 - 36. Neglia L, Oggioni M, Mercurio D, De Simoni M-G, Fumagalli S. Specific contribution of mannose-binding lectin murine isoforms to brain ischemia/reperfusion injury. Cell Mol Immunol. 2020;17(3):218–26. pmid:30967639 - 37. He Y, Tian Z. NK cell education via nonclassical MHC and non-MHC ligands. Cell Mol Immunol. 2017;14(4):321–30. pmid:27264685 - 38. Chang Y-T, Prompsy P, Kimeswenger S, Tsai Y-C, Ignatova D, Pavlova O, et al. MHC-I upregulation safeguards neoplastic T cells in the skin against NK cell-mediated eradication in mycosis fungoides. Nat Commun. 2024;15(1):752. pmid:38272918 - 39. Sanborn KB, Rak GD, Maru SY, Demers K, Difeo A, Martignetti JA. Myosin IIA associates with NK cell lytic granules to enable their interaction with F-Actin and function at the immunological synapse. J Immunol. 2009;182:6969–84. - 40. Zhang Z, Yuan B, Lu N, Facchinetti V, Liu Y-J. DHX9 pairs with IPS-1 to sense double-stranded RNA in myeloid dendritic cells. J Immunol. 2011;187(9):4501–8. pmid:21957149 - 41. Baggio LL, Varin EM, Koehler JA, Cao X, Lokhnygina Y, Stevens SR, et al. Plasma levels of DPP4 activity and sDPP4 are dissociated from inflammation in mice and humans. Nat Commun. 2020;11(1):3766. pmid:32724076 - 42. Guo D, Zhang D, Ren M, Lu G, Zhang X, He S, et al. THBS4 promotes HCC progression by regulating ITGB1 via FAK/PI3K/AKT pathway. FASEB J. 2020;34(8):10668–81. pmid:32567740 - 43. Hu C, Ni Z, Li B-S, Yong X, Yang X, Zhang J-W, et al. hTERT promotes the invasion of gastric cancer cells by enhancing FOXO3a ubiquitination and subsequent ITGB1 upregulation. Gut. 2017;66(1):31–42. pmid:26370108 - 44. Tan S, Xu Y, Wang Z, Wang T, Du X, Song X, et al. Tim-3 hampers tumor surveillance of liver-resident and conventional NK cells by disrupting PI3K signaling. Cancer Res. 2020;80(5):1130–42. pmid:31848194 - 45. Zhang Y, Li Y, Zhang Z, Zheng X, Peng H, Tian Z, et al. CD49a targeting enhances NK cell function and antitumor immunity. Cancer Immunol Res. 2025;13(1):139–51. pmid:39570767 - 46. Zhang Y, Li Y, Zhang Z, Zheng X, Peng H, Tian Z, et al. CD49a targeting enhances NK cell function and antitumor immunity. Cancer Immunol Res. 2025;13(1):139–51. pmid:39570767 - 47. Green BL, Myojin Y, Ma C, Ruf B, Ma L, Zhang Q, et al. Immunosuppressive CD29+ Treg accumulation in the liver in mice on checkpoint inhibitor therapy. Gut. 2024;73(3):509–20. pmid:37770128 - 48. Kataoka S, Manandhar P, Lee J, Workman CJ, Banerjee H, Szymczak-Workman AL, et al. The costimulatory activity of Tim-3 requires Akt and MAPK signaling and its recruitment to the immune synapse. Sci Signal. 2021;14(687):eaba0717. pmid:34131021 - 49. Dong W, Wu X, Ma S, Wang Y, Nalin AP, Zhu Z, et al. The Mechanism of Anti-PD-L1 antibody efficacy against PD-L1-negative tumors identifies NK Cells expressing PD-L1 as a cytolytic effector. Cancer Discov. 2019;9(10):1422–37. pmid:31340937 - 50. Dong S, Guo X, Han F, He Z, Wang Y. Emerging role of natural products in cancer immunotherapy. Acta Pharm Sin B. 2022;12(3):1163–85. pmid:35530162 - 51. Mayall JR, Horvat JC, Mangan NE, Chevalier A, McCarthy H, Hampsey D, et al. Interferon-epsilon is a novel regulator of NK cell responses in the uterus. EMBO Mol Med. 2024;16(2):267–93. pmid:38263527 - 52. Kong R, Wang N, Han W, Bao W, Lu J. IFNγ-mediated repression of system xc- drives vulnerability to induced ferroptosis in hepatocellular carcinoma cells. J Leukoc Biol. 2021;110(2):301–14. pmid:34318944 - 53. Abiko K, Matsumura N, Hamanishi J, Horikawa N, Murakami R, Yamaguchi K, et al. IFN-γ from lymphocytes induces PD-L1 expression and promotes progression of ovarian cancer. Br J Cancer. 2015;112(9):1501–9. pmid:25867264 - 54. Garcia-Diaz A, Shin DS, Moreno BH, Saco J, Escuin-Ordinas H, Rodriguez GA, et al. Interferon receptor signaling pathways regulating PD-L1 and PD-L2 expression. Cell Rep. 2017;19(6):1189–201. pmid:28494868 - 55. Ma R, Li Z, Tang H, Tian L, Shah Z, Barr T, et al. The NK cell receptor NKp46 acts as an essential regulator of ILC1 proliferation, effector function and anti-aml activity. Blood. 2023;142(Supplement 1):3925–3925. - 56. Sen Santara S, Lee D-J, Crespo Â, Hu JJ, Walker C, Ma X, et al. The NK cell receptor NKp46 recognizes ecto-calreticulin on ER-stressed cells. Nature. 2023. - 57. Gauthier L, Morel A, Anceriz N, Rossi B, Blanchard-Alvarez A, Grondin G, et al. Multifunctional natural killer cell engagers targeting NKp46 trigger protective tumor immunity. Cell. 2019;177(7):1701-1713.e16. pmid:31155232 - 58. Yang C, Zhang H, Zhang L, Zhu AX, Bernards R, Qin W, et al. Evolving therapeutic landscape of advanced hepatocellular carcinoma. Nat Rev Gastroenterol Hepatol. 2023;20(4):203–22. pmid:36369487 - 59. Zhu AX, Abbas AR, de Galarreta MR, Guan Y, Lu S, Koeppen H, et al. Molecular correlates of clinical response and resistance to atezolizumab in combination with bevacizumab in advanced hepatocellular carcinoma. Nat Med. 2022;28(8):1599–611. pmid:35739268 - 60. Liao Q, Zhou X, Wu L, Yang Y, Zhu X, Liao H, et al. Gut microbial metabolite 4-hydroxybenzeneacetic acid drives colorectal cancer progression via accumulation of immunosuppressive PMN-MDSCs. J Clin Invest. 2025;135(11):e181243. pmid:40179015