Graphical abstract graphic file with name fx1.jpg [65]Open in a new tab Highlights * • Downregulated LTA4H promotes HCC occurrence and progression * • LTA4H ablation fosters CD206^+ macrophage via LTBP1-mediated TGF-β activation * • LTA4H prevents LTBP1 mRNA maturation and processing in an HNRNPA1-dependent manner * • TGF-β blockage sensitizes LTA4H-deficient HCC to immunotherapy __________________________________________________________________ Yang et al. report that LTA4H deficiency promotes CD206^+ macrophage polarization through upregulating LTBP1 and downstream TGF-β secretion and activation in HCC. Mechanistically, LTA4H induces HNRNPA1 phosphorylation, enhancing their interaction and leading to functional inhibition of HNRNPA1 in regulating LTBP1 mRNA processing. Introduction Hepatocellular carcinoma (HCC) is the most common primary liver cancer, ranking as the sixth most prevalent cancer and the third leading cause of cancer-related death globally.[66]^1 Many patients are diagnosed at advanced stages, limiting curative treatment options and contributing to poor prognosis,[67]^2 characterized by low 5-year survival rates and frequent recurrence after surgery. The inherent heterogeneity of HCC further complicates treatment strategies.[68]^3 Chronic liver injury and inflammation, driven by factors such as viral hepatitis, alcohol-related liver disease, and nonalcoholic fatty liver disease, underlie HCC pathogenesis.[69]^1 Inflammatory mediators critically influence the tumor microenvironment (TME) and drive HCC progression. Understanding immune evasion mechanisms within the TME is key to improving immunotherapy outcomes in HCC. Tumor-associated macrophages (TAMs), which often adopt an M2-like phenotype, are abundant in HCC and serve as potential therapeutic targets.[70]^4 Chemokines such as transforming growth factor β (TGF-β) and CCL2 are central to TAM recruitment and functional remodeling.[71]^5^,[72]^6 Reprogramming TAMs via chemokine signaling may synergize with immune checkpoint inhibitors (ICIs), though further research is needed to refine combinations and identify predictive biomarkers for patient stratification. Effective predictive biomarkers are essential to address the heterogeneous response of patients with HCC. Promising evidence suggests that RNA sequencing data can identify predictive signatures of therapeutic response from the TME.[73]^7 However, translating these features into parameters measurable by immunohistochemistry (IHC) remains challenging but necessary for advancing clinical applications. Leukotriene A4 hydrolase (LTA4H) is a multifunctional zinc metalloenzyme with epoxide hydrolase and aminopeptidase activities. It catalyzes the conversion of leukotriene A4 (LTA4) to leukotriene B4 (LTB4), a potent inflammatory mediator acting through leukotriene B4 receptor 1 (BLT1) and leukotriene B4 receptor 2 (BLT2) receptors.[74]^8 LTA4H is highly expressed in several malignancies, including esophageal, skin, ovarian, and colorectal cancer,[75]^9^,[76]^10^,[77]^11 where it promotes tumor progression by enhancing cell proliferation.[78]^9^,[79]^10^,[80]^11^,[81]^12 However, its role in liver cancer, particularly in modulating the TME, remains poorly understood. In this study, we found that LTA4H deficiency promoted HCC occurrence and progression, correlating with poor prognosis. LTA4H-deficient models exhibited increased hepatocyte damage and M2-like macrophage infiltration as key drivers of HCC progression. Mechanistically, we showed that LTA4H regulates hepatocyte damage through c-Jun N-terminal kinases (JNK) activation and remodels the TME via the heterogeneous nuclear ribonucleoprotein A1 (HNRNPA1)-latent-transforming growth factor beta-binding protein 1 (LTBP1)-TGF-β signaling axis. We uncovered a previously undescribed role of nuclear LTA4H in mRNA processing and maturation, reprogramming macrophages to evade immune surveillance in HCC. Importantly, our findings support TGF-β blockade as an adjunct therapy to enhance ICI efficacy in patients with HCC with low LTA4H expression. Results Downregulation of LTA4H was associated with poor prognosis in patients with HCC To evaluate the clinical significance of LTA4H in HCC, we first analyzed the expression of LTA4H in human HCC tissue and paired adjacent tissues. IHC analysis of HCC sections and the tissue microarray (TMA) HCC cohort demonstrated that LTA4H expression was significantly downregulated in HCC ([82]Figures 1A–1C). Similar to the IHC results, the protein and mRNA levels of LTA4H were lower in HCC tissues compared to matched adjacent tissues ([83]Figures 1D–1F). We further examined the expression of LTA4H in a mouse HCC model induced by diethylnitrosamine (DEN). Consistent with our findings in clinical HCC samples, the transcript and protein levels of LTA4H were reduced in mouse HCC ([84]Figures 1G–1J). LTB4, a key mediator of LTA4H function, was significantly downregulated in both mouse and human HCC ([85]Figures 1K and 1L). Moreover, LTB4 levels showed a positive correlation with mRNA and protein level of LTA4H in clinical HCC samples ([86]Figure 1M). We further assessed the relationship between LTA4H expression and survival. Patient characteristics are listed in [87]Table S1. Both overall survival and recurrence-free survival were significantly shorter in patients with low LTA4H expression ([88]Figures 1N–1P). Univariate and multivariate Cox regression analyses identified LTA4H expression as an independent risk factor for predicting overall survival and recurrence-free survival ([89]Tables S1 and [90]S2). To validate these findings, we performed the same analysis of HCC samples from the published Chinese HCC patients with HBV infection (CHCC-HBV) cohort.[91]^13 Linear regression analysis revealed that LTA4H protein and transcript levels were decreased and positively correlated in the CHCC-HBV cohort ([92]Figures 1Q–1S). Low LTA4H protein expression was also associated with poor prognosis in the CHCC-HBV cohort ([93]Figure 1T). Together, these results indicated a repressive role for LTA4H in human HCC development. Figure 1. [94]Figure 1 [95]Open in a new tab Downregulation of LTA4H was associated with poor prognosis in patients with HCC (A) IHC detection of LTA4H in human HCC tissues. (B and C) TMA analysis of LTA4H in HCC and para-tumor tissues. (D and E) Western blot (WB) and quantification of LTA4H in clinical HCC samples. (F) qPCR detection of LTA4H in clinical HCC samples. (G–J) WB, qPCR, and IHC detection of LTA4H in the DEN-induced mouse HCC model. (K) ELISA detection of LTB4 in DEN-induced mouse HCC models. (L) ELISA detection of LTB4 in clinical HCC samples. (M) Pearson correlation of LTB4 with LTA4H mRNA and protein levels in clinical HCC samples. (N) Representative images of tumors with low and high LTA4H expression in the TMA HCC cohort. (O and P) Overall survival (OS) and recurrence-free survival (RFS) according to LTA4H level of patients with HCC in the TMA HCC cohort (n = 133). (Q and R) Comparison of LTA4H mRNA and protein levels in tumoral and adjacent tissues in the CHCC-HBV cohort. (S) Pearson correlation of LTA4H mRNA and protein levels in the CHCC-HBV cohort. (T) OS analysis according to the LTA4H protein level in the CHCC-HBV cohort. (A, J, and N) Scale bar: 100 μm. (B) Scale bar: 300 μm. (C) p value was calculated by Wilcoxon rank-sum test (n = 133), data represent mean ± SD. (E, F, and L) p value was calculated by paired Wilcoxon signed-rank test (n = 10). (O, P, and T) p value was calculated by log rank test. (H, I, and K) p value was calculated by Student’s t test (n = 3), data represent mean ± SD. (Q and R) p value was calculated by Wilcoxon rank-sum test (n = 159), the line and box represent median and upper and lower quartiles, respectively. All the replicates represent biological replicates. See also [96]Tables S1 and [97]S2. Hepatocyte-specific LTA4H ablation promotes hepatocarcinogenesis We constructed hepatocyte-specific LTA4H knockout (LTA4H^Δhep) and LTA4H knockout (LTA4H^KO) mice to study the role of LTA4H in a DEN-induced HCC model of LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice. LTA4H^Δhep and LTA4H^KO mice were generated by deleting the second exon of LTA4H ([98]Figure S1A). 40 weeks after DEN injection, the absence of LTA4H expression correlated with increased liver-to-body weight ratios and tumor numbers in LTA4H^Δhep and LTA4H^KO mice. Unexpectedly, LTA4H^KO mice developed significantly fewer tumors than LTA4H^Δhep mice ([99]Figures 2A–2D). Hepatocytic damage can be directly measured by alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels in the serum.[100]^14 Compared with LTA4H^f/f mice, LTA4H^Δhep and LTA4H^KO mice have higher plasma ALT and AST levels. However, no significant differences in plasma ALT or AST levels were observed between LTA4H^Δhep and LTA4H^KO mice ([101]Figures 2E and 2F). When mice were administered DEN followed by weekly injections of carbon tetrachloride (CCL[4]), similar results were observed, consistent with the DEN-induced HCC model ([102]Figures S1B–S1G). Based on these findings, we hypothesized that hepatocyte-specific LTA4H deficiency contributes to liver injury induced by DEN, while LTA4H derived from some stromal cells may influence tumor cell proliferation. To test this hypothesis, IHC of TUNEL and proliferating cell nuclear antigen (PCNA) was performed to quantify damaged and proliferative cells in the liver ([103]Figure 2G). The percentage of TUNEL^+ hepatocytes and PCNA^+ tumor cells was higher in LTA4H^Δhep and LTA4H^KO mice than in LTA4H^f/f mice. However, LTA4H^KO mice displayed more TUNEL^+ hepatocytes but fewer PCNA^+ tumor cells compared to LTA4H^Δhep mice ([104]Figures 2H and 2I). These findings suggest that LTA4H deficiency exacerbates DEN-induced hepatocytic damage, while LTA4H in some non-parenchymal cells may enhance tumor progression in HCC. Figure 2. [105]Figure 2 [106]Open in a new tab Hepatocyte-specific LTA4H ablation promotes hepatocarcinogenesis (A) Scheme of the DEN-induced HCC mouse model. (B) Gross liver tumor images from LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice post DEN treatment. (C and D) Liver weight/body weight ratios and tumor counts among LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice. (E and F) Serum ALT and AST levels for liver injury evaluation. (G) H&E, LTA4H, TUNEL, and PCNA staining of mouse liver sections. (H and I) Comparison of the proportions of TUNEL^+ and PCNA^+ cells among LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice. (G) H&E image scale bar: 500 μm, IHC image scale bar: 200 μm. (C–F, H, and I) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis (n = 6). Data represent mean ± SD. All the replicates represent biological replicates. See also [107]Figure S1. LTA4H ablation mediates hepatocytic damage via reducing JNK signal activation To investigate the possible role of LTA4H in hepatocytic damage, we isolated primary hepatocytes from LTA4H^f/f and LTA4H^KO mice without treatment. Bulk RNA was extracted from hepatocytes after 12 h in culture and subjected to RNA sequencing ([108]Figure 3A; [109]Table S3). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis showed that genes downregulated in LTA4H^KO hepatocytes were enriched in the mitogen-activated protein kinase (MAPK) signaling pathway and erb-b2 receptor tyrosine kinase (ErbB) signaling pathway compared with those in LTA4H^f/f hepatocytes ([110]Figure 3B). Cross-activation and feedback mechanism between the MAPK and ErbB signaling pathways have been reported.[111]^15 We first assessed the alterations in the MAPK signaling pathway (JNK, extracellular signal-regulated kinases (ERK), P38) and found that LTA4H knockout (KO) decreased JNK phosphorylation levels but not ERK and P38 phosphorylation levels ([112]Figure 3D). Differentially expressed genes in the ErbB signaling pathway ([113]Figure 3C) were validated by qPCR. epidermal growth factor (Egf), epidermal growth factor receptor (Egfr), and Nras were reduced in LTA4H^KO hepatocytes ([114]Figures 3E–3G). These findings underscore basal differences in JNK phosphorylation and gene expression between LTA4H^f/f and LTA4H^KO hepatocytes. Furthermore, Egf and Egfr expression levels were also downregulated in the adjacent liver tissues of LTA4H^Δhep and LTA4H^KO mice in the DEN-induced HCC model ([115]Figures 3H–3J). Figure 3. [116]Figure 3 [117]Open in a new tab LTA4H ablation mediates hepatocytic damage via reducing JNK signal activation (A) RNA sequencing scheme for LTA4H^f/f and LTA4H^KO primary hepatocytes. (B) KEGG enrichment of downregulated genes in LTA4H^KO hepatocytes. (C) Enriched genes in the ErbB signaling pathway in LTA4H^KO hepatocytes. (D) WB detection of key MAPK signaling molecules (total/phosphorylated p38 MAPK, JNK, and ERK) in LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO adjacent liver tissues from mice HCC models. (E–G) qPCR analysis of Egf, Egfr, and Nras in LTA4H^f/f and LTA4H^KO hepatocytes (n = 3). (H–J) Validation of Egf, Egfr, and Nras expression in adjacent liver tissues from LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice HCC models (n = 6). (K) Immunoblot of total/phosphorylated JNK and c-Jun in LTA4H^f/f and LTA4H^KO hepatocytes treated with DEN, ethanol (vehicle of LTB4), LTB4, JNK inhibitor, BLT1 inhibitor, and BLT2 inhibitor as indicated. (L) Comparison of the viability of LTA4H^f/f and LTA4H^KO hepatocytes subjected to different treatments (n = 3). (M) qPCR analysis of Il6 expression in LTA4H^f/f and LTA4H^KO hepatocytes under various treatments (n = 3). (E–G) p values were calculated by Student’s t test, data represent mean ± SD. (H–J, L, and M) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis, data represent mean ± SD. All the replicates represent biological replicates. See also [118]Figure S2; [119]Table S3. LTB4, a known promoter of JNK phosphorylation via BLT1 or BLT2 receptors,[120]^10^,[121]^16 was significantly reduced in both serum and liver tissues of LTA4H^Δhep and LTA4H^KO mice compared to LTA4H^f/f mice in the DEN-induced HCC model ([122]Figures S2E and S2F). Interestingly, no substantial differences in LTB4 levels were observed between LTA4H^KO and LTA4H^Δhep mice, suggesting that hepatocytes are the primary source of LTB4 in our model. To elucidate the underlying signal mechanism, hepatocytes were treated with DEN, LTB4, JNK inhibitor (JNK-IN-8), BLT1 inhibitor (U-75302), and BLT2 inhibitor ([123]LY255283). Western blot analyses revealed that LTB4 induced JNK and c-Jun phosphorylation in both LTA4H^f/f and LTA4H^KO hepatocytes. Inhibition of JNK, BLT1, or BLT2 individually significantly reduced JNK and c-Jun phosphorylation ([124]Figure 3K). Next, hepatocytic damage was evaluated using cell viability (CCK8) assays and the expression of damage-associated molecular pattern molecules (Il6, Il1b, and Tnfa). LTA4H^KO hepatocytes exhibited more severe damage than LTA4H^f/f hepatocytes. LTB4 treatment reversed the DEN-induced hepatocytic damage, which was partially mitigated by the JNK, BLT1, or BLT2 inhibitors individually ([125]Figures 3L, 3M, [126]S2A, and S2B). We further analyzed Egf and Egfr levels and found that LTB4 induced their expression in both LTA4H^f/f and LTA4H^KO hepatocytes, which was prevented by the inhibition of JNK, BLT1, and BLT2 ([127]Figures S2C and S2D). In a DEN-induced HCC model in BLT1^KO mice, results were consistent with those observed in LTA4H^Δhep and LTA4H^KO mice ([128]Figures S2G–S2I). BLT1^KO hepatocytes also exhibited greater severe hepatocytic damage than wild-type hepatocytes ([129]Figures S2J and S2K). In previous studies, we found that EGF/EGFR could promote JNK phosphorylation,[130]^17 which suggested a potential positive feedback pathway between JNK activation and the EGF/EGFR signaling pathway. In summary, our findings demonstrate a protective role of LTA4H in mitigating hepatocytic damage via JNK activation. LTA4H deficiency shifts the TME toward high infiltration of M2-like macrophages The role of LTA4H, a gene that regulates inflammation, in shaping the TME of HCC remains unclear. Tumor samples from LTA4H^f/f and LTA4H^KO DEN-induced mouse HCC models were analyzed using mass cytometry (CyTOF) to assess the percentage and functional status of immune cells ([131]Figures 4A, 4B, and [132]S3A–S3D; [133]Table S4). Analysis of 29 markers revealed 20 distinct cell clusters, showing increased infiltration of CD206^+ macrophages (C14, C16, and C20) and neutrophils (C6), alongside decreased infiltration of CD40^+ macrophages (positive CD86 expression, C2 and C11) and programmed death ligand 1 (PD-L1^+) dendritic cells (C9) in LTA4H-deficient tumors. The proportions of other immune cell clusters were comparable between LTA4H^f/f and LTA4H^KO HCC tumors ([134]Figures 4C–4E). The frequency of CD86^+ macrophages lacking CD40 expression (C10 and C17) remained unchanged between the groups. CD40 has been reported to stimulate antitumorigenic macrophage functions by regulating fatty acid and glutamine metabolism.[135]^18 CD40^+ macrophages exhibited opposing differentiation trajectories compared to CD206^+ macrophages, as revealed by a diffusion map analysis[136]^19 ([137]Figure 4F). Differential expression of CD206 and PD-L1 suggested that CD206^+ macrophages were immunosuppressive M2-like macrophages, whereas CD40^+ macrophages (C2 and C11) were M1-like macrophages ([138]Figures 4G, 4H, and [139]S3B). The ratio of CD206^+ macrophages to CD40^+ macrophages significantly increased in LTA4H^KO tumors ([140]Figure 4I). PD-L1^+ dendritic cells exhibited increased expression of CD11b and PD-L1 with decreased CD24 levels, suggesting an immunosuppressive phenotype ([141]Figure S3C). Figure 4. [142]Figure 4 [143]Open in a new tab LTA4H deficiency shifts the tumor microenvironment of HCC toward high infiltration of M2-like macrophage (A) Heatmap of 29 markers expression in each cell cluster. (B) t-distributed stochastic neighbor embedding (t-SNE) plot of CyTOF data from CD45^+ cells in LTA4H^f/f and LTA4H^KO HCC tumors (n = 5), annotated into specific cell types. (C and D) t-SNE plots showing cell subpopulations with annotations based on differentially expressed markers in CD45^+ cells from LTA4H^f/f and LTA4H^KO HCC tumors. (E) Proportion of differential cell clusters among CD45^+ cells in LTA4H^f/f and LTA4H^KO HCC tumors. (F) Diffusion map showing macrophage clusters as a phenotypic continuum. (G and H) Mean expression of CD206 and CD274 along diffusion component 1. (I) Ratio comparison of CD206^+ to CD40^+ macrophages in LTA4H^f/f and LTA4H^KO HCC tumors. (J) mIHC staining of F4/80, CD40, and CD206 verifying macrophage infiltration in LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO HCC tissues. Red arrow: CD206^+ macrophages; green arrow: CD40^+ macrophages. Scale bar: 20 μm. (K) Quantification of CD206^+ macrophages and the CD206^+/CD40^+ macrophage ratio in LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO HCC tumors. (E) The p value was calculated by Wilcoxon rank-sum test with Benjamini-Hochberg adjustment (n = 5), the line and box represent median and upper and lower quartiles, respectively. (I) The p value was calculated by Wilcoxon rank-sum test (n = 5), the line and box represent median and upper and lower quartiles, respectively. (K) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis (n = 5), data represent mean ± SD. All the replicates represent biological replicates. See also [144]Figures S3 and [145]S4; [146]Table S4. We confirmed these findings in tumor sections from LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice using multiplex IHC (mIHC). While there were no significant differences in PD-L1^+ dendritic cells among the groups, neutrophil infiltration was elevated only in LTA4H^KO tumors, indicating a contribution from non-parenchymal LTA4H deficiency to neutrophil infiltration ([147]Figure S3E). CD206^+ macrophages and the ratio of CD206^+ macrophages to CD40^+ macrophages were higher in LTA4H^Δhep and LTA4H^KO mice compared to LTA4H^f/f mice. Consistent with the tumor burden variations, LTA4H^KO mice showed lower levels of CD206^+ macrophages and a reduced ratio of CD206^+ macrophages to CD40^+ macrophages compared to LTA4H^Δhep mice ([148]Figures 4J, 4K, and [149]S3F), suggesting that macrophage phenotypic changes play a critical role in LTA4H-ablation-induced hepatocarcinogenesis. To elucidate LTA4H expression in human HCC, the analysis of published single-cell RNA sequencing data[150]^20^,[151]^21 revealed that myeloid cells were the primary non-parenchymal cells expressing LTA4H ([152]Figures S4A–S4D). Bone marrow-derived macrophages (BMDMs) from LTA4H^f/f and LTA4H^KO mice were polarized to M1 and M2 phenotypes using lipopolysaccharide + interferon-gamma or interleukin-4, respectively. LTA4H deficiency impaired both M1 (Cxcl9 and Cxcl10) and M2 polarization (Arg1 and Mrc1) ([153]Figures S4E–S4H), indicating that LTA4H expression in macrophages is essential for their polarization. Overall, hepatocyte-specific LTA4H ablation increases the frequency of M2-like macrophages in a mouse HCC model, highlighting its critical role in promoting HCC progression and shaping an immunosuppressive TME. Hepatocyte-specific LTA4H deficiency promotes M2-like macrophage polarization by upregulating LTBP1 expression and TGF-β secretion To explore how LTA4H deficiency promotes M2-like macrophage polarization, we examined differentially expressed secretory genes between LTA4H^f/f and LTA4H^KO hepatocytes ([154]Figure 3A). Nine secreted genes were significantly upregulated in LTA4H^KO hepatocytes ([155]Figure 5A), including Apob, Apoe, Mmp8, Lcn2, and Ltbp1, which have been implicated in M2 macrophage polarization.[156]^22^,[157]^23^,[158]^24^,[159]^25^,[160]^26 Validation experiments in primary hepatocytes confirmed the upregulation of Ltbp1 and Lcn2 in LTA4H^KO hepatocytes ([161]Figures 5B and [162]S5A–S5D). Further, Ltbp1 was upregulated in LTA4H KO liver tissues and Hepa1-6 cells but downregulated in LTA4H-overexpressing Hepa1-6 cells ([163]Figures 5C–5E and [164]S5E–S5G). Protein and mRNA analysis confirmed that LTBP1 was negatively regulated by LTA4H ([165]Figures 5F and 5G). High LTBP1 expression correlated with poor prognosis in patients from the TGCA-LIHC cohort ([166]Figure S5H). Figure 5. [167]Figure 5 [168]Open in a new tab Hepatocyte-specific LTA4H deficiency promotes M2-like macrophage polarization by upregulating LTBP1 expression and TGF-β secretion (A) Volcano plot of differentially expressed and secreted genes in LTA4H^KO vs. LTA4H^f/f hepatocytes ([169]Figure 3A). (B) Transcript level of Ltbp1 in LTA4H^f/f and LTA4H^KO hepatocytes (n = 3). (C and D) mRNA and protein levels of LTBP1 in liver tissues of LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO mice (n = 5). (E–G) mRNA and protein levels of LTBP1 in Hepa1-6 cells with LTA4H overexpression or knockout (n = 3). (H and I) ELISA for secreted TGF-β in supernatants of Hepa1-6 cells with LTA4H overexpression or knockout (n = 3). (J and K) ELISA for active and total TGF-β serum levels in LTA4H^f/f, LTA4H^Δhep, and LTA4H^KO DEN-induced HCC models (n = 4). (L) The coculture model of BMDMs and CM from scramble or LTA4H KO Hepa1-6 cells, CM pretreated with HCl/NaOH. (M and N) Transcript levels of Arg1 and Mrc1 in BMDMs under different treatments (n = 3). (O) CD206 immunofluorescence in BMDMs from coculture system under various treatments. Scale bar: 200 μm. (P) Quantification of CD206^+ macrophage in BMDMs from coculture system under various treatments (n = 3). (Q) mIHC staining for LTA4H, LTBP1, CD206, and CD68 in HCC tissues with low or high LTA4H expression from the TMA HCC cohort. Scale bar: 200 μm, enlarged images scale bar: 20 μm. (R) Quantification and Pearson correlation among LTA4H, LTBP1, and CD206^+ macrophages in the TMA HCC cohort. (S) Liver images and H&E staining of orthotopic HCC tumors (Hepa1-6 scramble and LTA4H KO) with or without macrophage depletion. Scale bar: 4 mm. (T) Tumor burden comparison across groups, determined by tumor-to-liver area ratio in sections (n = 5). (B, E, and H) p value was calculated by Student’s t test, data represent mean ± SD. (C, F, I–K, M, N, P, and T) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis, data represent mean ± SD. All the replicates represent biological replicates. See also [170]Figures S5 and [171]S6. LTBP1, a member of the latent TGF-β binding protein family, contributes to latent TGF-β secretion and activation. TGF-β plays a pivotal role in HCC progression by promoting M2 macrophage differentiation, which suppresses CD8^+ T cells, natural killer cells, and dendritic cell activity and enhances CD4^+ regulatory T cell activity.[172]^5 We detected total TGF-β secretion levels in Hepa1-6 cells and found that TGF-β secretion significantly decreased with overexpressing LTA4H and increased with LTA4H KO ([173]Figures 5H and 5I). Both total TGF-β and active TGF-β were elevated in the serum of the LTA4H^Δhep and LTA4H^KO DEN-induced HCC model ([174]Figures 5J and 5K). Notably, Tgfb expression remained unchanged between LTA4H-manipulated and control Hepa1-6 cells ([175]Figures S5I and S5J), indicating that LTA4H primarily modulates the secretion and activation of TGF-β. To test whether LTA4H affects M2-like macrophages via TGF-β, we cocultured BMDMs with conditioned media (CM) from Hepa1-6 cells with or without LTA4H KO. Latent TGF-β in the CM was activated using hydrochloric acid and neutralized with sodium hydroxide ([176]Figure 5L). LTA4H^KO CM promoted M2 polarization, which was abolished by TGF-β blockade ([177]Figures 5M–5P). Using mIHC in the TMA HCC cohort, we observed a significant negative correlation between LTA4H and both LTBP1 and CD206^+CD68^+ macrophages ([178]Figures 5Q and 5R). LTBP1 expression was elevated in HCC tumors and correlated with poor prognosis ([179]Figures S5K–S5M). Macrophage depletion using clodronate-containing liposomes inhibited tumor progression in the scramble group and abolished tumor enhancement in LTA4H KO tumors ([180]Figures 5S and 5T). Additionally, mIHC revealed that macrophage depletion reversed CD8^+ T cell exhaustion and increased the proportion of GZMB^+CD8^+ T cells ([181]Figures S6A–S6D). These findings also supported the direct role of M2-like macrophages in mediating the pro-tumorigenic effects of LTA4H deficiency. Collectively, LTA4H ablation promotes M2-like macrophage polarization through enhancing TGF-β secretion and activation, mediated by LTBP1 upregulation. LTA4H negatively regulates LTBP1 expression by inhibiting Ltbp1 mRNA maturation and processing mediated by HNRNPA1 We investigated the molecular mechanism by which LTA4H regulates LTBP1 expression. First, we explored the role of the LTA4H/LTB4/LTB4R pathway in hepatocytes. LTB4 treatment reduced Ltbp1 levels in both LTA4H^f/f and LTA4H^KO hepatocytes, which were restored by inhibition of BLT1 or BLT2. Notably, inhibiting BLT1 in LTA4H^KO hepatocytes caused a sharper Ltbp1 increase, suggesting an intrinsic inhibitory mechanism of LTA4H ([182]Figure S7A). Next, we assessed Ltbp1 mRNA stability and pre-mRNA levels in LTA4H^f/f and LTA4H^KO hepatocytes, as well as in Hepa1-6 cells with LTA4H overexpression or KO. Results showed no significant differences, indicating that LTA4H might regulate Ltbp1 expression via mRNA processing rather than transcription and stability ([183]Figures S7B–S7E). Interestingly, nuclear enrichment of LTA4H was observed in human and mouse HCC-adjacent tissues ([184]Figures S7F and S7G). We hypothesized that nuclear enrichment of LTA4H might directly participate in the regulation of mRNA processing and maturation of LTBP1. Anti-Flag-LTA4H co-immunoprecipitation (coIP) with liquid chromatography-tandem mass spectrometry (LC-MS/MS) identified 46 specific proteins ([185]Figure 6A; [186]Table S5). Enrichment analysis revealed that most of the proteins mainly involved in pathways related to protein-RNA binding and RNA processing and splicing ([187]Figure 6B). Furthermore, Tubulin Alpha 4a (TUBA4A) and HNRNPA1 were identified as potential interactors of LTA4H ([188]Figure 6C). HNRNPA1, also known as heterogeneous nuclear ribonucleoprotein A1, is mainly found in the cell nucleus and is involved in multiple cellular processes, including RNA processing, transport, and localization. Interestingly, LTBP1 was identified in a dataset of HNRNPA1 splicing targets.[189]^27 Intriguingly, knockdown of HNRNPA1 decreased LTBP1 protein and mRNA levels in Hepa1-6 cells ([190]Figures 6D and 6E). Next, coIP assays confirmed the interaction between LTA4H and HNRNPA1, which was RNA independent ([191]Figures 6F–6K). Figure 6. [192]Figure 6 [193]Open in a new tab LTA4H negatively regulates LTBP1 expression by inhibiting Ltbp1 mRNA maturation and processing mediated by HNRNPA1 (A) Venn diagram of proteins identified via anti-IgG and anti-Flag immunoprecipitation-mass spectrometry (IP-MS). (B) Gene Ontology (GO) and Reactome pathway analyses of 46 potential LTA4H-binding proteins. (C) Venn diagram of top LTA4H-binding candidates. (D and E) WB and qPCR for LTBP1 and HNRNPA1 expression in Hepa1-6 with Hnrnpa1 knockdown. (F and G) Reciprocal IP validated LTA4H-HNRNPA1 interaction in Hepa1-6 cells. (H and I) Reciprocal IP validated LTA4H-HNRNPA1 interaction in primary hepatocytes. (J and K) coIP tested LTA4H-HNRNPA1 interaction in LTA4H-overexpressing Hepa1-6 cells with or without RNaseA/T1 treatment. (L) Endogenous HNRNPA1 phosphorylation assessed via WB after coIP in LTA4H KO Hepa1-6 cells treated with LTB4 for various durations. (M) RIP-PCR validated HNRNPA1 and Ltbp1 mRNA interaction after LTB4 treatment in Hepa1-6 cells with or without LTA4H knockout. (N) coIP tested LTA4H-HNRNPA1 interaction in Hepa1-6 cells over time under LTB4 treatment. (O) RIP-PCR verified HNRNPA1-Ltbp1 mRNA interaction in cells described in (N). (P) LTA4H distribution was examined in nuclear and cytoplasmic extracts from Hepa1-6 cell with different treatments. (Q) WB examined the effects of LTB4, BLT1 inhibitor, and HNRNPA1 knockdown on LTBP1 expression in Hepa1-6 scramble and LTA4H KO cells. (R) Liver images and H&E staining of Hepa1-6 scramble and LTA4H KO orthotopic HCC tumors with Hnrnpa1 or Ltbp1 knockdown (n = 5). Scale bar: 4 mm. (E, M, and Q) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis (n = 3), data represent mean ± SD. All the replicates represent biological replicates. See also [194]Figures S7 and [195]S8; [196]Table S5. RNA immunoprecipitation (RIP)-qPCR revealed that LTA4H overexpression reduced HNRNPA1 binding to Ltbp1 mRNA ([197]Figure S7H). Because decreased Ltbp1 levels after LTB4 treatment were observed in LTA4H^KO hepatocytes ([198]Figure S7A), we conjectured that LTB4-induced HNRNPA1 phosphorylation might also inhibit HNRNPA1-mediated regulation of Ltbp1 mRNA processing and maturation. We examined the phosphorylation level of HNRNPA1 in LTA4H KO Hepa1-6 cells treated with LTB4 for various durations. The phosphorylation of HNRNPA1 increased significantly with time in LTA4H KO Hepa1-6 cells ([199]Figure 6L). Moreover, LTB4 treatment prevented the binding of HNRNPA1 and Ltbp1 mRNA in LTA4H KO Hepa1-6 cells, which suggested that phosphorylation of HNRNPA1 affected its interaction with Ltbp1 mRNA ([200]Figure 6L). In addition, LTA4H KO Hepa1-6 cells showed greater interaction between HNRNPA1 and Ltbp1 mRNA than did scramble Hepa1-6 cells in the presence or absence of LTB4, which suggested that the absence of an interaction between LTA4H and HNRNPA1 promoted the binding of HNRNPA1 and Ltbp1 mRNA ([201]Figure 6L). Furthermore, a decrease in the interaction between HNRNPA1 and Ltbp1 mRNA also exhibited a time-dependent gradient response to LTB4 treatment in normal Hepa1-6 cells ([202]Figures 6N and 6O). Collectively, these data verified that LTA4H inhibited LTBP1 expression via phosphorylation of HNRNPA1 through LTB4 or direct interaction. Unexpectedly, LTB4 treatment increased HNRNPA1-LTA4H binding, implying that phosphorylation of HNRNPA1 enhanced their interaction ([203]Figure 6N). Given the lack of a nuclear localization sequence in LTA4H, these findings suggested that upregulated LTA4H might promote HNRNPA1 binding through phosphorylation by LTB4, thereby facilitating its nuclear localization. Indeed, silencing HNRNPA1 prevented LTB4-induced nuclear translocation of LTA4H ([204]Figures S7I and S7J). BLT1 inhibition, but not BLT2, prevented LTB4-induced nuclear localization of LTA4H ([205]Figures 6P, [206]S7K, and S7L). Additionally, Hnrnpa1 knockdown reversed LTA4H ablation- or BLT1 inhibition-induced LTBP1 upregulation ([207]Figures S7M and [208]6Q). In the orthotopic HCC mouse model, Hnrnpa1 or Ltbp1 knockdown significantly hindered tumor progression induced by LTA4H deficiency ([209]Figures 6R and [210]S7N–S7P). The mIHC showed that Hnrnpa1 or Ltbp1 knockdown inhibited CD206^+F4/80^+ macrophage polarization, promoted CD8^+ T cell infiltration, decreased the proportion of PD-1^+CD8^+ T cells (PD-1: programmed cell death protein 1), and increased the proportion of GZMB^+CD8^+ T cells in LTA4H KO tumors ([211]Figures S8A–S8E). In summary, LTA4H inhibits HNRNPA1-mediated LTBP1 by facilitating HNRNPA1 phosphorylation through LTB4 or direct interaction. Moreover, LTA4H promotes its own nuclear localization through HNRNPA1. TGF-β blockade potentiates the efficacy of anti-PD-1 therapy in LTA4H KO tumor-bearing mice The aforementioned findings drove us to investigate whether disrupting the immunosuppressive crosstalk between tumor cells and macrophages by blocking TGF-β improves anti-PD-1 therapy efficacy in HCC. We constructed an orthotopic HCC mouse model using Hepa1-6 scramble or LTA4H KO cells and treated the mice with anti-PD-1, anti-TGF-β, or a combination of both on day 5 post transplantation. Significant tumor reduction was observed in Hepa1-6 scramble tumors treated with anti-PD-1 therapy or combination therapy, but LTA4H KO tumors showed resistance to anti-PD-1 monotherapy. Combination therapy significantly restricted tumor growth in LTA4H KO tumors compared to IgG or monotherapies ([212]Figures 7A–7F). The mIHC analysis showed that anti-PD-1 monotherapy failed to induce GZMB^+CD8^+ T cells and CD8^+ T cell infiltration in LTA4H KO tumors, likely due to abundance of CD206^+ macrophages. Anti-TGF-β reduced CD206^+ macrophages, but only combination therapy simultaneously and effectively decreased the proportion of PD-1^+CD8^+ T cells, increased the proportion of GZMB^+CD8^+ T cells, and promoted CD8^+ T cell infiltration in LTA4H KO tumors ([213]Figures 7G–7K. This explains the limited efficacy of anti-TGF-β monotherapy in LTA4H KO models, likely due to the pre-existing immunosuppressive microenvironment driving CD8^+ T cell exhaustion prior to treatments. Figure 7. [214]Figure 7 [215]Open in a new tab TGF-β blockade potentiates the efficacy of anti-PD-1 therapy in LTA4H knockout tumor-bearing mice (A) Tumor progression was monitored at 5, 9, and 13 days post inoculation using live bioluminescent imaging in mice injected with scramble or LTA4H KO Hepa1-6 cells and treated with IgG or blocking antibody as indicated. (B and C) Quantification of the liver tumor burden from live bioluminescent imaging studies. (D) Images of orthotopic HCC tumors from the indicated treatment groups. (E) H&E staining of Hepa1-6 orthotopic HCC tumors from treatment groups. Scale bar: 4 mm. (F) Comparison of tumor burden (tumor area/whole liver area) in orthotopic HCC tumors. (G) mIHC staining of CD206^+ macrophages, PD-1^+CD8^+ T cells, and GZMB^+CD8^+ T cells in Hepa1-6 scramble and LTA4H KO orthotopic HCC tumors with the indicated treatments. Scale bar: 20 μm. (H–K) Comparison of CD206^+ macrophages, CD8^+ T cells, PD-1^+CD8^+ T cells, and GZMB^+CD8^+ T cells among Hepa1-6 scramble and LTA4H KO orthotopic HCC tumors with the indicated treatments. (L) mIHC staining of LTA4H, CD68, and CD206 in human HCC samples with varying responses to ICIs. Scale bar: 100 μm. (M and N) Comparison of LTA4H expression in HCC tumor cells and the percentage of CD206^+ macrophages between responders (n = 6) and non-responders (n = 7). (B, C, F, and H–K) p values were calculated by one-way ANOVA with Tukey’s multiple comparison analysis (n = 5), data represent mean ± SD. (M and N) p value was calculated by Wilcoxon rank-sum test, data represent mean ± SD. All the replicates represent biological replicates. See also [216]Table S6. In addition, to investigate whether this finding can be extrapolated to clinical HCC, we analyzed the correlation between the LTA4H level in tumor cells and treatment response in patients with HCC receiving ICIs (ICI cohort, [217]Table S6). Responders exhibited significantly higher LTA4H levels and fewer CD206^+ macrophages than non-responders ([218]Figures 7L–7N). These findings suggest LTA4H as a potential biomarker for guiding HCC ICI treatment. Collectively, LTA4H deficiency exacerbated DEN-induced liver damage, while M2-like macrophages driven by TGF-β accumulation from LTA4H-deficient hepatocytes or tumor cells promoted HCC progression. Our study highlights the regulatory role of LTA4H in the TME of HCC through the LTA4H/HNRNPA1/LTBP1/TGF-β axis. Anti-TGF-β may serve as an effective strategy to sensitize LTA4H-deficient patients with HCC to ICIs. Decreased acetyl-CoA and increased HDAC1 were the main causes of reduced LTA4H expression in HCC To explore the molecular mechanism of LTA4H downregulation in HCC, we analyzed the CHCC-HBV cohort[219]^13 to identify proteins correlated with LTA4H expression and conducted pathway enrichment analyses ([220]Table S7). Proteins positively correlated with LTA4H were notably enriched in histone deacetylases (HDACs) and acetyltransferase-related pathways ([221]Figure S9A). Histone deacetylation, which increases chromatin compaction, typically suppresses gene transcription. Given the reported decrease in acetyl-CoA synthesis in HCC,[222]^28 we speculated that LTA4H downregulation is linked to histone deacetylation dysfunction. Trichostatin-A (TSA) and nicotinamide treatment enhanced the expression of Lta4h mRNA, with acetyl-CoA replenishing reagent (AAR) further augmenting this effect ([223]Figure S9B). Notably, TSA treatment alone significantly induced LTA4H expression, implicating HDACs as key regulators. The combination of AAR and TSA further boosted LTA4H expression compared to TSA alone ([224]Figures S9C and S9D). Analysis of the The Cancer Genome Atlas-Liver Hepatocellular Carcinoma (TCGA-LIHC) database and the CHCC-HBV cohort revealed elevated HDAC1 and HDAC2 expression in HCC, correlating with poor prognosis ([225]Figures S9E–S9H and [226]S10A–S10D). LTA4H negatively correlated with HDAC1 and HDAC2 in the CHCC-HBV cohort ([227]Figures S10H and S10I). Knockdown experiments in Hepa1-6 cells showed that silencing HDAC1, but not HDAC2, significantly increased LTA4H transcript and protein levels ([228]Figures S9I–S9K and [229]S10E–S10G). Analysis of our clinical HCC samples and TMA HCC cohort corroborated these findings ([230]Figures S9P–S9T, [231]S10J, and S10K). Using regulatory potential score from CistromeDB, we identified a strong association between H3K27ac modification and LTA4H expression ([232]Figures S10L and S10M). Given HDAC1’s role in deacetylating H3K27ac,[233]^29 chromatin immunoprecipitation sequencing (ChIP-seq) analysis confirmed that HDAC1 silencing promoted H3K27ac accumulation at the LTA4H promoter ([234]Figures S9U and S9V), a result further validated by ChIP-qPCR ([235]Figures S9W and S9X). These findings suggest that reduced acetyl-CoA levels and elevated HDAC1 expression suppress LTA4H expression by inhibiting H3K27ac modification at its promoter in HCC. Discussion LTA4H has been implicated in promoting tumor progression in cancers like skin, colorectal, and ovarian cancers by enhancing cell proliferation and metastasis. Additionally, it contributes to inflammation in lung diseases such as emphysema through Proline-Glycine-Proline (PGP) degradation and LTA4-to-LTB4 conversion, critical for neutrophil and macrophage recruitment.[236]^8^,[237]^30 However, its role in reshaping the TME remains underexplored. Recent findings in ovarian cancer suggest that LTA4H may foster an immunosuppressive TME via CCL5.[238]^11 In this study, we found that LTA4H ablation correlates with HCC occurrence and immunotherapy response. LTA4H ablation exacerbated liver damage and fostered M2-like macrophage infiltration in HCC. We identified TGF-β-dependent M2-like polarization as a mechanism by which LTA4H deficiency drives HCC progression. Metabolic enzymes have been shown to regulate nuclear gene expression, as seen with pyruvate dehydrogenase’s role in histone acetylation.[239]^31 Here, we revealed that LTA4H regulated LTBP1 expression and TGF-β secretion and activation through nuclear mechanisms. LTA4H interacts with HNRNPA1, a process enhanced by LTB4-induced HNRNPA1 phosphorylation, which disrupts HNRNPA1 binding to Ltbp1 mRNA, inhibiting Ltbp1 mRNA processing. Given LTBP1’s role in TGF-β signaling,[240]^32^,[241]^33^,[242]^34^,[243]^35 highly activated TGF-β signaling contributes to an exhausted TME and poor PD-1 blockade response in HCC. Functional studies showed that TGF-β blockade reversed immune suppression in LTA4H-deficient tumors, highlighting TGF-β inhibition as a promising strategy to enhance ICI efficacy.[244]^5^,[245]^36^,[246]^37 Therapies combining macrophage-targeting drugs with ICIs are emerging in HCC treatment. Clinical trials demonstrate the potent antitumor effects of blockade of TGF-β or its receptor across various cancers, including HCC, with an acceptable safety profile.[247]^38 Recent advancements involved the combination of anti-TGF-β antibodies or receptor inhibitors (galunisertib) with ICIs, which triggered robust antitumor immunity and tumor regression in a mouse model.[248]^39 M7824, a bifunctional fusion protein targeting PD-L1 and TGF-β, was developed to activate innate and adaptive immune responses and induce tumor regression in mouse models.[249]^40 Our findings supported TGF-β inhibition to improve ICI outcomes in LTA4H-deficient HCC. Additionally, our ICI cohort revealed that non-responders to ICIs exhibited significantly lower LTA4H expression, suggesting LTA4H as a predictive biomarker for immunotherapy response. We also identified key factors underlying LTA4H dysregulation in HCC, including reduced acetyl-CoA levels and elevated HDAC1 expression, which impair histone acetylation at the LTA4H promoter. Low levels of acetyl-CoA and high expression of HDAC1, which are the main causes of the dysregulation of histone acetylation, have been reported in HCC.[250]^28^,[251]^41 Given the absence of a nuclear localization sequence in LTA4H, we found that the nuclear translocation of LTA4H depended on HNRNPA1. Given the enrichment of LTA4H-binding proteins associated with mRNA splicing, more functions and mechanisms of nuclear LTA4H need to be further studied. In summary, LTA4H deficiency in HCC cells promotes TGF-β secretion and activation, and M2-like macrophage programming and immune escape. Our study underscores the potential of LTA4H as a biomarker for ICI response and highlights TGF-β inhibition to enhance therapeutic outcomes in LTA4H-deficient HCC. Limitations of the study This study systematically elucidates the role and mechanism of LTA4H in HCC progression via targeting the HNRNPA1/LTBP1/TGF-β axis. However, there are some limitations. In the macrophage depletion experiment, clodronate-containing liposomes were used to deplete all macrophages. While widely applied in studies of TAMs, this approach does not directly address the CD206^+ macrophage dependency of LTA4H in tumor progression. Additionally, we identified a strong association between LTA4H and mRNA splicing through its interaction with HNRNPA1, but further investigation is needed to fully understand the broader functions and mechanisms of nuclear LTA4H. We focused on investigating the role of LTA4H in hepatocytes and tumor cells during HCC progression. However, given its expression in immune cells, particularly macrophages, the function of LTA4H in these cells warrants further exploration in future studies. Resource availability Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Lei Chen (chenlei@smmu.edu.cn). Materials availability All materials used in this study are available from the [252]lead contact without restriction. Data and code availability Raw transcriptome sequencing data (in fastq format) are available on NGDC_SRA with accession number CRA022579 (NGDC: CRA022579). Raw data (.fcs) of mass cytometry were deposited in FlowRepository (FlowRepository: FR-FCM-Z7W7, [253]http://flowrepository.org/id/RvFrwdQZoBHMmZfhB4OY9X5wuLKDYIOF2ILVO SyjSoFd2XrTNexN54buIKjFlvwm). Raw LC-MS/MS data are available via ProteomeXchange with identifier PXD060326 (ProteomeXchange: PXD060326). Raw ChIP-seq data are available on NGDC (NGDC: CRA021482). This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the [254]lead contact upon request. Acknowledgments