Functional Analysis of Macrophages in the Progression of Liver Cirrhosis and Liver Cancer Based on Single-Cell Transcriptomic Sequencing

Ren Lingxuan, Lu Ziqi, Qi Wei*, Feng Zhijie*

Department of Gastroenterology, the Second Hospital of Hebei Medical University, Hebei Key Laboratory of Gastroenterology, Hebei Institute of Gastroenterology, Hebei Clinical Research Center for Digestive Diseases, Shijiazhuang 050000, China

*Corresponding authors: Feng Zhijie, Chief Physician/Professor; E-mail: fengzhijiefzj80@126.com; Qi Wei, Attending Physician; E-mail: chinaqiwei@yahoo.com

Note: Ren Lingxuan and Lu Ziqi are co-first authors.

Abstract

Background: Hepatic macrophages play a vital role in host defense mechanisms and maintaining internal homeostasis, and represent a major cellular component involved in liver injury and repair. Macrophages derived from monocytes differ from tissue-resident macrophages in terms of gene regulation and specific functions. Over 90% of primary liver cancers develop on the basis of cirrhosis, making the dynamic changes of macrophages during cirrhosis-to-cancer progression a subject worthy of investigation.

Objective: To dissect the transcriptomic differences between hepatic macrophages of different origins, analyze the dynamic patterns of macrophage changes during cirrhosis and hepatocellular carcinoma (HCC) progression, and explore potential strategies for preventing cirrhosis from progressing to HCC.

Methods: Single-cell transcriptomic data from healthy, cirrhotic, and HCC tissues were obtained from the GEO database. Healthy and cirrhotic data were derived from the GSE136103 dataset, comprising samples from 5 healthy livers and 5 cirrhotic livers. HCC data were obtained from the GSE149614 dataset, comprising samples from 10 HCC patients. The Seurat package was used to cluster cirrhotic and HCC samples and identify cell types. Three macrophage subclusters were extracted from cirrhotic samples, and the top 200 specifically expressed genes in each subcluster were analyzed using the Metascape online analysis tool. Cirrhosis-specific genes in macrophage subclusters were extracted and subjected to KEGG functional analysis to explore macrophage functions in cirrhosis. CellChat was employed to analyze intercellular interactions in cirrhotic and HCC single-cell transcriptome data, comparing differences in macrophage communication between the two conditions. Macrophages from healthy, cirrhotic, and HCC tissues were integrated using Harmony for batch effect correction, then imported into Monocle for pseudotime analysis to construct an evolutionary trajectory from healthy liver to cirrhotic liver to HCC macrophages. The limma package was used to identify genes continuously upregulated or downregulated during this trajectory, followed by functional enrichment analysis.

Results: Unsupervised clustering identified three macrophage subclusters (Mac1, Mac2, and Mac3) based on marker gene expression patterns. Mac1 originated from tissue-resident macrophages (Kupffer cells), while Mac2 and Mac3 originated from blood monocytes, with their numbers significantly increased in cirrhotic tissue. In cirrhotic tissue, Mac1 exhibited upregulation of adaptive immune system-related functions, whereas Mac2 and Mac3 showed downregulation of phagosome-related functions and antigen presentation capabilities.

Substantial differences existed in macrophage communication with other cell types between cirrhotic and HCC samples. Certain intercellular communications occurred exclusively in cirrhotic macrophages, including IFN-II and CD40 signaling pathways. After batch effect correction, pseudotime analysis of healthy, cirrhotic, and HCC macrophages revealed a specific temporal relationship among the three groups. We identified 81 genes continuously downregulated during this process, but no continuously upregulated genes. Functional analysis revealed enrichment for bacterial immune response among the downregulated genes.

Conclusion: Cirrhotic macrophages can be divided into three subclusters: Mac1 derived from liver-resident Kupffer cells, and Mac2 and Mac3 derived from blood monocytes. Many immune-related cellular communications in cirrhosis, such as IFN-II and CD40 pathways, disappear in HCC. The evolution from healthy liver to cirrhotic liver to HCC macrophages involves continuous downregulation of bacterial immune responses, which may exacerbate the harmful effects of gut microbiota translocation caused by portal hypertension. For cirrhotic patients, early treatment of portal hypertension-induced leaky gut may represent an important therapeutic strategy.

Keywords: Macrophages; Liver cirrhosis; Liver fibrosis; Liver neoplasms; Single-cell transcriptomic sequencing; Intercellular interactions

Funding: Hebei Natural Science Foundation Project (H2021206314)

DOI: 10.12114/j.issn.1007-9572.2023.0596

Introduction

Liver cirrhosis is a common digestive system disease characterized by extensive hepatic fibrosis. Recent studies indicate that 844 million people worldwide suffer from chronic liver disease, with 2 million deaths annually. Despite rising incidence, no effective anti-fibrotic therapies currently exist. Critically, over 90% of primary liver cancers develop on the basis of cirrhosis. The liver contains the highest proportion of macrophages among solid organs, and these cells play crucial roles in maintaining hepatic and systemic homeostasis. Beyond immune functions, macrophages regulate hematopoietic microenvironments, influence metabolism, mediate tissue repair, and modulate embryonic tissue maturation. In fibrotic livers, fibrosis-associated macrophages originate from recruited blood monocytes that differentiate within fibrotic niches, known as fibrotic niche-associated macrophages, which highly express genes including SPP1, LGALS3, CCL2, CXCL8, PDGFB, and VEGFA. Single-cell RNA sequencing (scRNA-seq) provides a novel approach for investigating disease pathogenesis with unprecedented resolution at the single-cell level. This study employs scRNA-seq to examine transcriptomic differences among macrophage subpopulations in liver fibrosis and explore dynamic changes in macrophage-related genes and pathways during cirrhosis-to-cancer progression.

Methods

Data Acquisition

We obtained single-cell transcriptomic data for healthy, cirrhotic, and HCC tissues from the Gene Expression Omnibus (GEO) database. Healthy and cirrhotic data were derived from the GSE136103 dataset, including samples from 5 healthy livers and 5 cirrhotic livers. HCC data were obtained from the GSE149614 dataset, comprising samples from 10 HCC patients.

Data Processing and Cell Type Identification

Downloaded single-cell transcriptome data for healthy, cirrhotic, and HCC tissues were imported into the Seurat package for cell type identification. First, quality control was performed using the PercentageFeatureSet function to calculate mitochondrial proportions, and cells with mitochondrial content exceeding 20% were filtered out. Data were then normalized using the NormalizeData function with the "LogNormalize" method, which normalizes gene expression for each cell by total expression, multiplies by a scaling factor (default 10,000), and log-transforms the results. Next, variable features were identified using FindVariableFeatures, returning 2,000 highly variable genes per dataset for downstream principal component analysis (PCA). The ScaleData function performed linear transformation as a standard preprocessing step before PCA. JackStraw and Elbow plot commands determined data dimensionality. KNN algorithm was applied for clustering followed by nonlinear dimensionality reduction (t-distributed stochastic neighbor embedding, tSNE). The FindMarkers command identified differentially expressed genes between cell types as biological markers, which were compared against the CellMarkers database for cell annotation. The Subset function extracted macrophages for subsequent analysis.

Differential Gene Functional Analysis

Genes specifically upregulated or downregulated in cirrhotic macrophages were converted to gene IDs using org.Hs.eg.db, followed by KEGG functional enrichment analysis using clusterProfiler. The limma package was used to analyze genes specifically upregulated or downregulated during cirrhosis-to-cancer progression.

Analysis of Intercellular Communication in Fibrotic and Tumor Macrophages

CellChat is an R package that quantitatively infers and analyzes cell-cell communication networks from scRNA-seq data. It requires gene expression data as input and integrates prior knowledge of ligand-receptor interactions and cofactors to establish probabilities of cell-cell communication, providing multiple visualization methods. We performed intercellular communication analysis between Mac1, Mac2, Mac3 from fibrotic tissues and other cell types, as well as between tumor tissue macrophages and other cells, to identify macrophage communication pathways in fibrotic versus tumor tissues.

Pseudotime Analysis of Healthy, Cirrhotic, and Tumor Macrophages

Macrophage data from healthy, fibrotic, and tumor tissues were integrated and batch effects were removed using the Harmony package. The integrated data were then imported into Monocle, and the DDRTree function was applied for dimensionality reduction to determine cell transition order. The differentialGeneTest function identified genes that changed with pseudotime.

Functional Enrichment Analysis Using Online Tools

Metascape is a powerful gene functional annotation tool for batch analysis of genes and proteins. We used Metascape to functionally analyze specifically enriched genes.

Results

Macrophage Subclustering and Functional Analysis

Extracted mononuclear macrophage data were divided into 10 clusters, including monocytes (Mono1, Mono2, Mono3), macrophages (Mac1, Mac2, Mac3), and dendritic cells (cDC1, cDC2, pDC). The dittoSeq function displayed the proportion of each subcluster in healthy and cirrhotic groups, showing that Mac1 was more abundant in healthy tissue, while Mac2 and Mac3 were more prevalent in cirrhotic tissue. Three macrophage clusters were extracted for further analysis. Mac1 highly expressed CD163 and MARCO, indicating tissue-resident macrophage (Kupffer cell) identity, while Mac2 and Mac3 highly expressed TREM2, CD9, and MNDA, suggesting monocyte-derived origins.

The top 200 marker genes from Mac1, Mac2, and Mac3 were imported into Metascape for functional enrichment analysis. Mac1 functions primarily involved "innate immune response," "regulation of immune response," and "inflammatory response." Mac2 functions centered on "ribosome, cytoplasm," "TRBP-containing complex (subunits of DICER, RPL7A, EIF6, MOV10, and 60S ribosomal particles)," and "ribosome assembly." Mac3 functions were enriched in "angiogenesis," "wound response," and "positive regulation of cell motility."

Analysis of cirrhosis-specific upregulated and downregulated genes in each subcluster revealed that Mac1 showed predominantly upregulated genes, while Mac2 and Mac3 showed mainly downregulated genes. KEGG functional analysis of upregulated genes in Mac1 and downregulated genes in Mac2 and Mac3 demonstrated that Mac1 exhibited upregulation of adaptive immune system-related functions in cirrhotic tissue, whereas Mac2 and Mac3 showed downregulation of phagosome-related functions and antigen presentation capabilities, suggesting potential immune functional decline.

Cell Communication Analysis in Cirrhosis and Hepatocellular Carcinoma

Significant differences existed in macrophage communication with other cell types between cirrhotic and HCC samples. Certain intercellular communications occurred exclusively in cirrhotic macrophages, including IFN-II and CD40 signaling pathways. In cirrhotic samples, IFN-II signals originated from T cells and NK cells and were received by Mac1, Mac2, and Mac3 through IFNGR1 and IFNGR2. CD40 signaling was restricted to monocyte-derived Mac2 and Mac3, with signals from T cells received by ITGA5, ITGB1, ITGAM, and ITGB2. These signaling communications were absent in HCC samples, reflecting dramatic changes in macrophage subcluster communication from cirrhosis to cancer and providing further evidence of immune functional decline.

Pseudotime Analysis of Macrophages in Cirrhosis-to-Cancer Progression

Cirrhosis-to-cancer progression typically spans several years. We leveraged pseudotime analysis to examine this dynamic process. After batch effect removal, integrated single-cell transcriptome data from healthy, cirrhotic, and HCC macrophages were jointly analyzed. The analysis constructed a pseudotime trajectory from healthy liver to cirrhotic liver to HCC macrophages. Overall, healthy and cirrhotic liver macrophages showed similar temporal patterns, while HCC macrophages were distinctly separated.

Functional Analysis of Macrophages in Cirrhosis-to-Cancer Progression

The evolution from healthy liver to cirrhotic liver to HCC macrophages can be considered a single pathophysiological process. Differential genes between healthy and cirrhotic liver macrophages and between cirrhotic liver and HCC macrophages were extracted and intersected. No continuously upregulated genes were identified, but 81 continuously downregulated genes were found. Functional enrichment analysis revealed these genes were highly associated with bacterial immune response.

Discussion

Early-stage cirrhosis is often accompanied by inflammatory responses, where macrophages primarily function to clear bacteria and cellular debris. Studies have shown that macrophages secrete various cytokines such as TNF-α and IL-6, which promote hepatocyte growth and repair. During mid-stage cirrhosis, as inflammatory responses intensify, macrophages gradually transform into pro-inflammatory cells, exacerbating chronic liver inflammation and cirrhosis progression. In late-stage cirrhosis, macrophages become a major source of fibrogenic cells, secreting TGF-β and other cytokines that promote fibroblast proliferation and transformation, leading to hepatic fibrosis.

Literature confirms that monocytes with angiogenic potential accumulate at liver regeneration sites, and studies have shown that during chronic liver disease, macrophages assist in forming complex vascular networks by secreting pro-angiogenic growth factors. Our findings support this, as Mac2 and Mac3, as monocyte-derived macrophages, showed functional enrichment for angiogenesis. Mac2 subcluster enriched for VEGFA-VEGFR2 signaling, while angiogenesis ranked first in Mac3 functional enrichment. Mac1, as liver-resident Kupffer cells, primarily performed innate immune functions.

Cirrhosis is associated with gut-liver axis dysfunction, where increased portal pressure causes intestinal edema, leading to gut dysbiosis, barrier damage, and increased bacterial translocation. As antigen-presenting cells, macrophages may undergo functional changes in response to increased bacterial translocation. We found that while cirrhotic macrophages showed enrichment for antigen presentation functions, monocyte-derived Mac2 and Mac3 subclusters exhibited downregulation of antigen presentation.

Pseudotime analysis and subsequent functional analysis of macrophages during cirrhosis-to-cancer progression suggest a synergistic effect between portal hypertension-induced leaky gut and macrophage functional changes. The 81 continuously downregulated genes were highly enriched for bacterial immune response functions. Research has indicated that altered gut microbiota in cirrhotic patients correlates with mortality, and that transjugular intrahepatic portosystemic shunt (TIPS) can reconstruct gut microbiota by reducing portal pressure. The gradual decline of macrophage immune function during cirrhosis-to-cancer progression, representing a failure of "gatekeeper" functions, may be a hallmark event in cirrhosis development and progression to HCC.

Given that cirrhosis-to-cancer progression typically requires several years, the numerous continuously downregulated genes in macrophages may indicate that the HCC macrophage phenotype already exists during the cirrhotic stage years earlier. This early-emerging dysfunctional macrophage phenotype, significantly enriched for bacterial immune response functions, may be related to abnormal gut-liver axis status under portal hypertension. TIPS represents a potential intervention to reduce portal pressure and alleviate leaky gut, though whether this can slow cirrhosis-to-cancer progression by intervening in gut microbiota translocation requires further investigation.

In summary, this study employed bioinformatics approaches to analyze public data from healthy, cirrhotic, and HCC samples, exploring macrophage functions during cirrhosis-to-cancer progression. Through cell communication and pseudotime analyses, we comprehensively characterized macrophage phenotypic changes during disease progression, providing insights for potential strategies targeting hepatic macrophages to prevent cirrhosis-to-cancer evolution. The continuous downregulation of bacterial immune responses during healthy-to-cirrhotic-to-HCC macrophage evolution may have synergistic effects with gut microbiota translocation through the portal vein, promoting chronic liver inflammation and cirrhosis progression. For cirrhotic patients, early treatment of portal hypertension-induced leaky gut to interrupt its synergy with macrophage immune decline may be an important therapeutic strategy.

Author Contributions: Ren Lingxuan conceived the study, designed the research, performed data analysis, generated figures, and wrote the initial draft. Lu Ziqi collected and screened data, participated in study conception, design, and manuscript writing. Qi Wei conceived the study, designed the research, and drafted the manuscript. Feng Zhijie conceived the study, designed the research, revised the final version, and takes responsibility for the manuscript.

Conflict of Interest: None declared.

References

[1] Chung BK, Øgaard J, Reims HM, et al. Spatial transcriptomics identifies enriched gene expression and cell types in human liver fibrosis[J]. Hepatol Commun, 2022, 6(9): 2538-2550. DOI: 10.1002/hep4.2001.

[2] Kim E, Viatour P. Hepatocellular carcinoma: old friends and new tricks[J]. Exp Mol Med, 2020, 52(12): 1898-1907. DOI: 10.1038/s12276-020-00527-1.

[3] Tacke F, Zimmermann HW. Macrophage heterogeneity in liver injury and fibrosis[J]. J Hepatol, 2014, 60(5): 1090-1096. DOI: 10.1016/j.jhep.2013.12.025.

[4] Ramachandran P, Dobie R, Wilson-Kanamori JR, et al. Resolving the fibrotic niche of human liver cirrhosis at single-cell level[J]. Nature, 2019, 575(7783): 512-518. DOI: 10.1038/s41586-019-1631-3.

[5] Marcellin P, Kutala BK. Liver diseases: a major, neglected global public health problem requiring urgent actions and large-scale screening[J]. Liver Int, 2018, 38(Suppl 1): 2-6. DOI: 10.1111/liv.13682.

[6] Angulo P, Kleiner DE, Dam-Larsen S, et al. Liver fibrosis, but No other histologic features, is associated with long-term outcomes of patients with nonalcoholic fatty liver disease[J]. Gastroenterology, 2015, 149(2): 389-397.e10. DOI: 10.1053/j.gastro.2015.04.043.

[7] Ramachandran P, Henderson NC. Antifibrotics in chronic liver disease: tractable targets and translational challenges[J]. Lancet Gastroenterol Hepatol, 2016, 1(4): 328-340. DOI: 10.1016/S2468-1253(16)30110-8.

[8] Liang WC, Feng ZJ, Rao ST, et al. Diarrhoea may be underestimated: a missing link in 2019 novel coronavirus[J]. Gut, 2020. DOI: 10.1136/gutjnl-2020-320832.

[9] Dick SA, Wong A, Hamidzada H, et al. Three tissue resident macrophage subsets coexist across organs with conserved origins and life cycles[J]. Sci Immunol, 2022, 7(67): eabf7777. DOI: 10.1126/sciimmunol.abf7777.

[10] Yan J, Zhang YM, Yu HR, et al. GPSM1 impairs metabolic homeostasis by controlling a pro-inflammatory pathway in macrophages[J]. Nat Commun, 2022, 13(1): 7260. DOI: 10.1038/s41467-022-34998-9.

[11] Guilliams M, Bonnardel J, Haest B, et al. Spatial proteogenomics reveals distinct and evolutionarily conserved hepatic macrophage niches[J]. Cell, 2022, 185(2): 379-396.e38. DOI: 10.1016/j.cell.2021.12.018.

[12] Aizarani N, Saviano A, Sagar, et al. A human liver cell atlas reveals heterogeneity and epithelial progenitors[J]. Nature, 2019, 572(7768): 199-204. DOI: 10.1038/s41586-019-1375-8.

[13] Ashburner M, Ball CA, Blake JA, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium[J]. Nat Genet, 2000, 25(1): 25-29. DOI: 10.1038/75556.

[14] Schauer D, Starlinger P, Zajc P, et al. Monocytes with angiogenic potential are selectively induced by liver resection and accumulate near the site of liver regeneration[J]. BMC Immunol, 2014, 15: 50. DOI: 10.1186/s12865-014-0050-3.

[15] Ramirez-Pedraza M, Fernández M. Interplay between macrophages and angiogenesis: a double-edged sword in liver disease[J]. Front Immunol, 2019, 10: 2882. DOI: 10.3389/fimmu.2019.02882.

[16] Albillos A, Martin-Mateos R, Van der Merwe S, et al. Cirrhosis-associated immune dysfunction[J]. Nat Rev Gastroenterol Hepatol, 2022, 19(2): 112-134. DOI: 10.1038/s41575-021-00520-7.

[17] Unanue ER. Antigen-presenting function of the macrophage[J]. Annu Rev Immunol, 1984, 2: 395-428. DOI: 10.1146/annurev.iy.02.040184.002143.

[18] Gedgaudas R, Bajaj JS, Skieceviciene J, et al. Circulating microbiome in patients with portal hypertension[J]. Gut Microbes, 2022, 14(1): 2029674. DOI: 10.1080/19490976.2022.2029674.

[19] Schierwagen R, Alvarez-Silva C, Madsen MSA, et al. Circulating microbiome in blood of different circulatory compartments[J]. Gut, 2019, 68(3): 578-580. DOI: 10.1136/gutjnl-2018-316227.

[20] Li MH, Li K, Tang SH, et al. Restoration of the gut microbiota is associated with a decreased risk of hepatic encephalopathy after TIPS[J]. JHEP Rep, 2022, 4(5): 100448. DOI: 10.1016/j.jhepr.2022.100448.

[21] Gitto S, Vizzutti F, Baldi S, et al. Transjugular intrahepatic Porto-systemic shunt positively influences the composition and metabolic functions of the gut microbiota in cirrhotic patients[J]. Dig Liver Dis, 2023, 55(5): 622-628. DOI: 10.1016/j.dld.2022.11.017.

(Received: October 13, 2023; Revised: January 3, 2024)
(Editor: Jia Mengmeng)