Metabolic reprogramming and immune evasion are two pivotal hallmarks of cancer, facilitating tumor development and progression [33]. Research has shown that alterations in tumor cell metabolism are partially driven by the recruitment of immune cells [34]. Our study identified 15 mediating relationships (12 with strong evidence, 3 with potential evidence), including 11 immune cells, 14 plasma metabolites, and one HCC trait. Mediation analysis supports the hypothesis that plasma metabolites play a mediating role in the influence of immune cells on HCC pathogenesis. This finding could offer new perspectives on the mechanisms underlying the initiation and progression of HCC, potentially revealing innovative targets for therapeutic intervention.
Our MR results revealed that Linolenate [alpha or gamma; (18:3n3 or 6)] positively mediate the causal relationship between CD19 on PB/PC and HCC (ME = 0.0389, MP = 19.3%), while negatively mediating the causal relationship between CD14 on Mo MDSC and HCC (ME=-0.0128, MP = 10.4%). Gamma linolenic acid (GLA) and alpha-linolenic acid (ALA) are members of the ω-6 family of polyunsaturated fatty acids which are converted into arachidonic acid via a series of elongation and desaturation reactions [35, 36]. According to Cui et al. ,GLA exhibits a chemo-protective effect against DEN induced HCC, and Feng et al. demonstrate that ALA suppresses HCC progression via the FXR/Wnt/β-catenin signaling pathway [37, 38]. Additionally, one study [39] showed that Farnesoid X receptor (FXR) ablation in DCs enhanced Treg cell generation. Moreover, Chimaeric Antigen Receptors (CARs) are a class of synthetic receptors that reprogram lymphocyte specificity and function and CARs targeting CD19 have demonstrated remarkable efficacy in B cell malignancies [40]. We observed a positive correlation between CD19 on PB/PC and HCC (OR = 1.224, 95%CI = 1.073–1.396). Tu et al. [41] report that S100A9 + CD14 + monocytes contribute to resistance to anti-PD-1 immunotherapy in advanced HCC by dampening T cell-mediated antitumor activity., while our findings imply that CD14 on Mo MDSC may reduce the risk of HCC. Collectively, these findings propose that Linolenate [alpha or gamma; (18:3n3 or 6)] could serve as a mediator, potentially regulating the FXR/Wnt/β-catenin signaling pathway, thereby modulating the immune cell effects on HCC.
CD4 on TD CD4 + cells is commonly acknowledged as CD4 on Terminally Differentiated CD4 + T cells, recognized as a biologically significant subset of T cells, playing crucial roles in anti-viral, anti-tumor, and autoimmune responses [42]. Research findings suggest that Studies indicate that CD4 T cell assistance is instrumental in sustaining and enhancing the functionality of effector CD8 T cells within tumor microenvironments, and recent studies have highlighted the pivotal role of CD4 T cell epitopes in augmenting CD8 T cell response in human cancers [43, 44]. A strong and effective CD4 T cell-mediated cytotoxicity was correlated with improved survival outcomes and reduced recurrence rates in HCC [45]. Our research aligns with these findings, suggesting that CD4 on TD CD4 + is negatively correlated with HCC risk. Further mediation analysis revealed that 9,10-DiHOME exerted negative mediating effect (ME=-0.00596, MP = 4.73%). 9,10-DiHOME is a metabolite of linolenic acid (LA). LA is initially converted to 9,10-EpOME and then to 9,10-DiHOME by cytochrome P450 epoxygenases and epoxide hydrolase, respectively [46, 47]. In a murine metabolite study, 9,10-DiHOME was identified as gut microbiota-derived lipophilic metabolites with Treg-inducing activity [48]. These insights propose that 9,10-DiHOME may act as a mediator, increasing Treg levels within the tumor microenvironment, thereby modulating CD4 on TD CD4 + effects on HCC. Furthermore, the pathway from CD4 on TD CD4 + to HCC is also mediated by the metabolites X-24306 and X-24307, accounting for 6.71% and 14.9% of the mediating effect, respectively.
Our research validated the existence of a causal relationship between 20 immune cell phenotypes and HCC. We discovered that CCR2 on monocytes is positively correlated with the development of HCC, which was corroborated by Li et al., who demonstrated that blockade of CCL2/CCR2 signaling with knockout of CCR2 or with a CCR2 antagonist inhibited the malignant growth and metastasis of HCC, reduced postsurgical recurrence, and enhanced survival [49], which is consistent with our MR results. The kynurenine derivative 3-hydroxyanthranilic acid has been shown to significantly suppress tumor growth in the HCC model by decreasing the percentage of F4/80lo CD64 PD-L1 + lo macrophages [50]. Furthermore, Tang et al. identified a potential increase in HCC risk associated with CD14 on CD14 + CD16- monocytes, aligning with our findings that CD64 on CD14- CD16 + monocyte is associated with an increased risk of HCC. Additionally, Shi et al. found that margin-infiltrating B lymphocytes featured an atypical memory phenotype (IgD − IgG + CD27 − CD38−), expressed surface markers characteristic of antigen-presenting cells, possessed tumor-killing potential on HCC [51], but our findings indicate that an increase in IgD-CD27-AC elevates the risk of HCC. Given that genetic variations are randomly distributed, MR analysis can offer more precise insights than experimental research. Previous outcomes might be triggered by some common confounders shared by HCC and IgD-CD27-AC. Jin et al. explored that long intergenic nonprotein coding RNA 942 promotes HCC cell proliferation and converts naive CD4 T cells into inducible Treg (iTreg) cells by regulating solute carrier family 7 member 11, suggesting a protective role for naive CD4 + T cells in HCC pathogenesis [52]. Moreover, it has been found that the knockdown of Tubulin γ-1 (TUBG1) has been shown to inhibit the proliferation, invasion, and migration of HCC cells and TUBG1 expression was negatively correlated with CD4 + regulatory T-cell [53], which is in line with our MR results and supports a causal relationship.
Metabolites play a crucial role in the early identification of individuals at high risk and in the prevention of HCC. Previous studies on the metabolomics of liver cancer patients have shown elevated levels of metabolites (including aromatic amino acids, and fatty acids) and depletion of metabolites (such as glutamine, and secondary bile acid) in the peripheral circulation of these patients [54–57]. Our findings find a causal relationship between the 36 plasma metabolites and HCC. Caffeic acid could inhibit glucose-regulated protein 75 attenuating the anti-apoptosis abilities and multi-drug resistance of HCC and contribute to inhibiting the proliferation and migration of liver cancer cells through the modulation of the AKT/CyclinD1/p21/p27 pathways [58, 59], our MR results further corroborate the conclusions drawn from prior studies. Furthermore, dysregulated bile acids (BAs) are closely tied to liver diseases and taurochenodeoxycholate (TCDCA) is a kind of hydrophobic BAs, has been identified by Yan et al. as a promoter of liver carcinogenesis [60]. Additionally, Perfluorooctanoic acid (PFOA) is a synthetic alkyl chain fluorinated compound that activates the PI3K/AKT/mTOR/4E-BP1 signaling pathway, thereby playing a pro-cancer effect on HCC cells [61], while our findings find PFOA is associated with a reduced risk of HCC, suggesting that this discrepancy may stem from residual confounding rather than a substantiated causal relationship. Glycerophospholipid metabolism encompasses the synthesis, degradation, and remodeling of glycerophospholipids. Glycerophospholipids are the most abundant phospholipids in mammalian cell membranes and can be divided into subcategories, such as phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylserine [62]. PCs serve as potential plasma biomarkers and help identify patients with hepatitis B virus-related liver disease who are susceptible to HCC [63]. Solute carrier family 27 member 4 (SLC27A4) is significantly overexpressed in HCC, overexpression of SLC27A4 significantly promoted the selective uptake of mono-unsaturated fatty acids, leading to elevated PE levels in HCC cells, consequently resulting in resistance to lipid peroxidation and ferroptosis [64].
This study pioneered the application of a comprehensive MR framework to dissect the causal relationships between immune cells, plasma metabolites and HCC. Furthermore, a pathway from the immune cell to HCC was constructed by employing a two-step MR and mediation analysis via plasma metabolites. MR offers distinct advantages over randomized controlled trials, including the mitigation of confounding factors and the circumvention of reverse causality biases. To ensure the robustness of our MR findings, a battery of sensitivity analyses was conducted. Nonetheless, this study has certain limitations. First, the absence of clinical data, such as age and sex, in the primary studies, precluded more granular subgroup analyses for HCC. Second, the predominant inclusion of East Asian individuals in GWAS constrains the generalizability of our findings to other ethnic groups. In addition, our conclusions remain theoretical and require further empirical validation through additional experimental and clinical research.