FARSB Serves as a Novel Hypomethylated and Immune Cell In�ltration Related Prognostic Biomarker in Hepatocellular carcinoma

Purpose Hepatocellular carcinoma (HCC) is a prevalent tumor with high morbidity, and an unfavourable prognosis. FARSB is an aminoacyl tRNA synthase, and plays a key role in protein synthesis in cells. Furthermore, previous reports have indicated that FARSB is overexpressed in gastric tumor tissues and is associated with a poor prognosis and tumorigenesis. However, the function of FARSB in HCC has not been studied. Results The results showed that FARSB mRNA and protein levels were upregulated in HCC and were closely related to many clinicopathological characteristics. Besides, according to multivariate Cox analysis, high FARSB expression was linked with a shorter survival time in HCC and may be an independent prognostic factor. In addition, the methylation level was negatively associated with the expression of FARSB. Furthermore, enrichment analysis showed that FARSB was related to the cell cycle. And TIMER analysis revealed that the FARSB expression was closely linked to tumor purity and immune cell in�ltration. The TCGA and ICGC data analysis suggested that FARSB expression is greatly related to m6A modi�er related genes. Potential FARSB-related ceRNA regulatory networks were also constructed. What’s more, based on the FARSB-protein interaction network, molecular docking models of FARSB and RPLP1 were constructed. Finally, drug susceptibility testing revealed that FARSB was susceptible to 38 different drugs or small molecules.


Introduction
Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer and constitutes more than 90% of the primary tumor of the liver.HCC is now the world's fth most widespread cause of cancer and the second leading cause of cancer death (Asafo-Agyei and Samant, 2022).Nowadays, the morbidity of HCC worldwide is still rising, owing to the absence of obvious signs in the early days and rapid development of HCC (Yamamoto et al., 2018, Liu et al., 2019).Moreover, the ve-year rate of survival for HCC is only 18%, as more than 60% of HCC patients are identi ed in late stages and have a poor prognosis (Siegel et al., 2013).At present, the test of serum alpha-fetoprotein (AFP), abdominal ultrasound, and triple-phase helical CT are the main diagnostic methods for HCC (Bruix and Sherman, 2005, Zhang et al., 2004, Choi et al., 2001, Arguedas et al., 2003).Regrettably, the accurancy and speci city of these approaches is still unsatisfying, especially in the early stages of disease (Masuzaki et al., 2012).Meanwhile, numerous studies have demonstrated that biomarkers display promising diagnostic abilities in HCC, such as AFP-L3,GP73, DCP, GPC3, SCCA, and OPN (Wang andWei, 2020, Tsuchiya et al., 2015).So far, recognition of effective biomarkers in HCC monitoring and early diagnosis remains de cient.As a result, there is a pressing need to evaluate innovative treatment targets and early diagnostic markers to enhance the prognosis of HCC patients.FARSB (Phenylalanyl-TRNA Synthetase Subunit Beta) is an aminoacyl tRNA synthase (ARSs), it has been shown to be associated with brain calci cation, interstitial lung disease and liver cirrhosis (Zadjali et al., 2018).Nevertheless, liver cirrhosis is one of the leading causes of hepatocellular carcinoma.ARSs are essential enzymes for protein synthesis in cells (Havrylenko and Mirande, 2015).As molecular adapters to translate mRNA into proteins, they are already present in the most primitive prokaryotes (Kim et al., 2011).Recent studies suggest that ARSs play key roles in controlling transcription, translation, splicing, apoptosis, in ammation, immune response, tumorigenesis, and other important physiological and pathological processes (Park et al., 2005, Zou et al., 2021), implying that ARSs have potential as novel therapeutic targets and drugs in multiple pathways (Kwon et al., 2019).It also take a part in biological processes closely linked to the development of cancer.Exemplify, meththiamide-tRNA synthetase combines with CDK4 and promotes cell cycle, tyrosine-trna synthases can also act as secreted cytokines to shape the tumor microenvironment (Kwon et al., 2018, Son et al., 2014).And moreover, ARSs are participated in the pathway of tumorigenesis by combining with proteins that interact with ars.In HCC, AIMP3, one of the macromolecular protein complex cofactors of ARSs, is reduced in HCC (Park et al., 2006), and overexpression of GRS was considered for a biomarker of poor prognosis in patients (Thul and Lindskog, 2018).Recent report contained that FARSB could be a biomarker in gastric cancer (Gao et al., 2021).Nevertheless, the research about FARSB biological function and mechanism in hepatocellular carcinoma has not been put forward, and FARSB correlation with prognosis is unclear.
In our study, the possible mechanisms of FARSB in liver cancer were analyzed using bioinformatics methods based on data from multiple online open databases.To start with, the relation between differential expression of FARSB in hepatocellular carcinoma and patients' poor prognosis was explored, and included the mechanism of FARSB in liver cancer, such as tumor in ltrating immune cells, m6a methylation and drug sensitivity.Except the online open database, we also obtained tumor samples from patients from Second A liated Hospital of Nanchang University for experiments to support our view.Our study is the rst to nd out the important function of FARSB in HCC, and demonstrates a possible mechanism of FARSB in HCC.For an example, the recruitment of certain immune cells.In conclusion, our results suggest that FARSB is a novel hepatocellular carcinoma biomarker and may be used as a prognostic factor for human HCC.

Patients and Tumor Specimens
We gathered 40 pairs of para n-embedded HCC and corresponding normal tissues from 40 patient samples from the Second A liated Hospital of Nanchang University from January 2018 to January 2021.The above patients did not receive adjuvant chemotherapy.Entire patients endorsed a knowledgeable consent form, also, this research was accepted by the Ethics Committee of the Second A liated Hospital of Nanchang University.

Data Collection and Processing
Hepatocellular carcinoma (HCC) clinical and mRNA expression level had collected from the TCGA Database and ICGC Database.For TCGA, as far as the gene expression pro le in concerned, that research included 374 LIHC samples and 50 normal samples, and the data variety of mRNA expression pro le was HTSeq-FPKM.And our group obtained clinical information from 377 patients.Besides, RNA-seq data also collected from the ICGC website (https://dcc.icgc.org/projects/LIRI-JP). [LINC-JP] Liver Cancer -NCC, JP datasets included 202 normal samples and 243 tumor samples (Song et al., 2021).Furthermore, we also downloaded RNA-seq data and clinical information from GEO database (https://www.ncbi.nlm.nih.gov/geo/).The GSE76421 datasets which have 52 normal samples and 112 tumor samples.

TIMER Database Analysis
TIMER (https://cistrome.shinyapps.io/timer/) is a library of capabilities for systematic analysis of immune in ltrations in different cancer types (Liu et al., 2020, Li et al., 2017).Which contain 1000 samples of thirty-two different tumor types.In tumor tissue, the relationship between RNA-seq expression pro le data and the immune cell in ltration degree could be embodied.In other words, it can re ect the correlation hidden in cancer and immune cells.We analyse expression of FARSB in different tumor and interaction of FARSB expression and immune in ltration, along with conventional immune cells.After that we utilized SCAN model to nd out the correlation of immune cell in ltration and FARSB CNV.Additionally, our researcher utilized TIMER to study the interaction hidden in FARSB expression and various immune gene markers.Subsequently, our team visualized the correlation of FARSB expression level and immune checkpoint gene expression level by the correlation module.

TISCH Analysis
TISCH (http://tisch.comp-genomics.org/)give us a user-friendly tool to systematically visualize, search and download gene expression atlas of the tumor microenvironment from various cancer types (Sun et al., 2021).We utilize TISCH to expound FARSB expression level in different immune cells.

cBioPortal Analysis
The cBioPortal for Cancer Genomics (https://www.cbioportal.org/) is a website resource for exploring,visualizing, and analyzing multidimensional cancer genomics data (Gao et al., 2013).This website help us study association of FARSB promoter methylation level and expression in HCC with "Plot" module (Liu et al., 2021c).Additionally, the "Mutations" module was utilized to found out secondary structure of FARSB and its physical interaction protein in HCC.

HPA Analysis
HPA (https://www.Proteinatlas.org/)uses antibody methods for immunostaining of tissues and cell lines, along with contrasting expression analysis of proteins in normal and tumor tissues.For our analysis, the immunohistochemical outcomes of FARSB in tumor tissues and normal tissues were concluded by HPA database.

GEPIA Analysis
Gene expression pro ling transactional estimation GEPIA (http://gepia.cancer-pku.cn/) is an interactional website app consisting of tumor and normal tissue sample data, which can be used to visualize clinicopathological characteristics (Tang et al., 2017).The tumor data comes from the TCGA database.Furthermore, the m6A-related genes LRPPRC, RBM15B and HNRNPA2B1'overall survival in HCC was also displayed in this website.

HCCDB Analysis
HCCDB: Integrative Molecular Database of Hepatocellular Carcinoma (http://lifeome.net/database/hccdb) is a free to use database which include 15 open access for researchers online datasets (Lian et al., 2018).During our article, it accomplish the mission that demonstrate differential expression of FARSB in HCC successfully.

Protein-Protein Interaction Network (PPI) Analysis
The online website (STRING, https://string-db.org) for nding interacting genes were utilized for PPI network construction and pivot gene screening (Szklarczyk et al., 2019).Besides, with the purpose of exploring the correlation amongst the top 500, our team used STRING database for analysis, medium con dence = 0.4 for screening, and visualization using Cytoscape software.To search cluster subnetworks, the Cytoscape Molecular Complex Detection (MCODE) plug-in has been utilized (Shannon et al., 2003).The acquiescent parameters are listed below: degree cutoff = 5, node score cutoff = 0.2, k-core = 9, maximum depth = 100.

GSCALite analysis
GSCALite (http://bioinfo.life.hust.edu.cn/web/GSCALite/) is a multifunctional genomics online portal for analyzing various bioinformatics related information (Liu et al., 2018a).In our research, the GSCALite was used to analyze the pathway activities and drug sensitivity of FARSB, FARSA and USP8.

Kaplan-Meier Plotter Database Analysis
The Kaplan-Meier database (http://kmplot.com/analysis/)can be utilized to appraise effect of genes on the survival of tumor tissue samples (Yuan et al., 2019, Peng et al., 2017).We used more than three hundreds LIHC samples to evaluate the interaction hidden in FARSB expression and overall survival (OS), relapse-free survival (RFS), progression-free survival (PFS), and disease speci c survival (DSS).We also explored the difference in LIHC patients' survival under different immune cell subtypes.A 95% con dence interval and logrank p < 0.05 were studied analytically important.

UALCAN Analysis
UALCAN (http://ualcan.path.uab.edu/) is an interactive evaluation and excavation website of the Internet, which can be used to analyze the relationship between tumor and normal specimens and the relative expression of genes with diverse clinicopathological characteristics (Chandrashekar et al., 2017, Liao et al., 2020).UALCAN was used to investigated various clinicopathological characteristics of FARSB and the methylation of FARSB promoter.The research further utilized UALCAN to compared patients survival time between FARSB higher expression and FARSB less expression in different clinicopathological characteristics.

GeneMANIA Analysis
GeneMANIA (genemania.org)can locate genes whose connected with input genes (Nguyen andKim, 2022, Zhang et al., 2022).We apply its tool to identi ed the gene which have physical interaction with FARSB.

LinkedOmics Database Analysis
LinkedOmics database (http://www.linkedomics.org/login.php)contains multi-omics data and clinical data for 32 cancer types and a total of 11158 patients from The Cancer Genome Atlas (TCGA) project (Vasaikar et al., 2018).We utilized the LinkFinder module of LinkedOmics to research the differently expressed genes connected with FARSB in TCGA LIHC (n = 515).Perform statistical analysis on the conclusions and display them in volcano maps and heat maps.Gene set enrichment analysis (GSEA) has been utilized for LinkedOmics functional modules to conduct gene ontology (GO) analysis and KEGG pathway analysis.False discovery rate (FDR) less than 0.01 is a noteworthy expression, P-values less than 0.05 is a signi cant related gene.

Prediction and Construction of ceRNA Networks
The TargetScan (http://www.targetscan.org),DIANA-microT (http://diana.imis.athenainnovation.gr/DianaTools/index) and RNAinter (http://www.rnainter.org)online websites has been used to nd out target miRNAs of FARSB (The miRNA which appear in three database at the same time and have negative correlation with FARSB will be con rmed).At next step, miRNet2.0(www.mirnet.ca/miRNet/home.xhtml)and starBase3.0(www.starbase.sysu.edu.cn) were put to use to forecast target lncRNAs of the screened miRNAs (The lncRNA which appear in two database at the same time and have negative correlation with target miRNA will be con rmed).In the last stage, an lncRNA-miRNA-mRNA (FARSB) ceRNA network in HCC will be completed by researcher.

Statistic Analysis
R software (version 3.6.3/4.1.2) had utilized to complete all statistical analyses amongst our research.
The "limma", "ggplot2" and "bee swarm" packages of "R" and the rank-sum test were utilized to detect the difference of FARSB expressions between LIHC samples and normal samples.Logistic regression were utilized to scrutinize the relevancy between FARSB expression and clinicopathological features.Afterwards, we used the Kaplan-Meier method to generate survival curves.Besides, the "survival" and "survminer" packages of " R were used to construct Cox models, and the elements striking related to prognosis had ltered out by Univariate and Multivariate Cox regression analysis (P < 0.05).After that, the ROC curve created by "survival ROC" was utilized to estimate the predictive ability of FARSB expression level to one-year, three-year, or ve-year survival.

Protein Data Bank Database Analysis
Protein data bank (PDB) (https://www.rcsb.org/) is the single worldwide archive of structural data of biological macromolecules (Berman et al., 2000, Berman and Gierasch, 2021).We utilized PDB to nd out spatial structures of FARSB and it related proteins.The structure of FARSB comes from PDB ID: 3L4G, and RPLP1 comes from PDB ID: 2LBF.Then we through the ZDOCK (https://zdock.umassmed.edu/)to study the pattern of combined docking between FARSB and RPLP1, and visualization used PYMOL (version 1.5.0.3).

The Farsb Expression Is Striking Increased In Hcc Patients
With the aim of illuminating FARSB expression in HCC, we utilized the TCGA, TIMER, HCCDB, and ICGC databases rstly to plot different forms of visual representation.According to TIMER databases, we discovered that FARSB expression levels in HCC and many other tumors were signi cantly up-regulated (Fig. 1A).For the sake of reliability, we also utilized the HCCDB to validate our results.The experimental results were in line with our expectation with FARSB expression being higher in HCC than normal tissue (Fig. 1B).Next, we compared 50 adjacent normal tissue samples and 377 HCC samples, and the result indicated that FARSB expression was clearly increased (Fig. 1C).After that, contrast with adjacent normal samples, FARSB expression in 58 pairs of LIHC tumor samples was signi cantly increased (Fig. 1D).
Meanwhile, the result of Fig. 1C was also veri ed by data sets LIRI-JP (liver cancer-RIKEN, JP) downloaded from the ICGC and GSE76427 downloaded from GEO (Fig. 1E-F).Additionally, based on HPA data, our team con rmed that the Phenylalanine-tRNA ligase beta subunit expression level of tumor tissues in the liver of patients with HCC was remarkably higher than normal tissues.(Fig. 1G-H).Furthermore, bene ting from the experimental method of IHC staining, our team further studied the expression of FARSB protein in HCC tissues (Fig. 1I).In brief, FARSB expression was higher in HCC tissues than in normal tissues.

Clinicopathologic Features of HCC Patients Could be Affected by FARSB Expression.
Given of high FARSB expression in HCC patients, for the purpose of nding out the mechanism of FARSB in development of HCC, we utilized the UALCAN online tool to investigate the relationship between clinicopathological features and FARSB expression.The results showed that FARSB expressed signi cantly different in age, cancer stages, tumor grade, TP53 methylation and race.What's more, there were signi cant differences among subgroups (Fig. 2A-E).For instance, FARSB expression level would be higher in stage 3 and 2 than in stage 1 (Fig. 2B).Meanwhile, it was also higher in grade 3 than in grade 1 and 2 (Fig. 2C).Besides, to analyze the relatinship between FARSB expression and clinicopathological variables in greater depth, we utilized logistic regression analysis.The result showed that high FARSB expression was obviously related to age, grade, stage and tumor size (Table 1).It was obvious that FARSB expression was closely related to clinicopathological variables; addtionally, FARSB expression may play a part in tumor progression of HCC.Hence, FARSB had the potential to be a biology marker for screening highrisk HCC patients.

Prognostic Value Of Farsb In Hcc
For the purpose of investigating the link between FARSB expression and HCC prognosis, the Kaplan-Meier Plotter was utilized.And correlational analysis showed that patients who had high-expressed FARSB had short OS, DFS, PFS, and DSS compare with low-expressed FARSB.(Fig.3A-D, P < 0.05).Moreover, the UALCAN database results re ected that higher FARSB expression was obviously related with shorter survival of HCC patients in BMI, gender, and race subgroups (Fig. 3E-H, P < 0.05).These results discovered that high-expressed FARSB is closely related to poor prognosis of HCC.

Overexpression Of Farsb Is An Independent Prognostic Factor Of Hcc
Similar with the results above, the survival curves and ROC curves darwn by R software proved that the high-expressed FARSB was strongly linked with the low overall survival rate, and AUC of the prognosis model at 1, 3 and 5 years was 0.687, 0.670, and, 0.618, respectively, inferring that overexpression of FARSB led to poor prognosis of HCC patients is accurate (Fig. 4A-B).Then, the univariate and multivariate Cox models were then used to examine the link between clinical characteristics and prognosis in HCC patients.Univariate analysis perform that stage, T stage,and FARSB expression were closely related to worse OS.Multiple-factor analysis reveal, FARSB expression was found to be an independent prognostic factor (Table .2).These were directly re ected in the forest map.These results indicated that the overexpress FARSB was related to poor prognosis so FARSB could independently predict the prognosis of HCC possibly.

The Promoter Methylation Level Of Farsb In Hcc Patients
DNA methylation is one of the essential epigenetic mechanisms.Recent reports have revealed that DNA methylation participates in HCC metastasis and proliferation via epigenetic regulation of oncogenes and tumor suppressor genes.
Therefore, we detected the methylation level of the FARSB promoter in HCC tissues through the MethSurvwebsite.The correlation heat map showed that 12 sites were hypomethylated (Fig. 5A).The results of Fig. 5A were then veri ed by the UALCAN database, promoter methylation in HCC tissues were lower than in normal tissues (Fig. 5B).Then, we looked into the link between FARSB promoter methylation and FARSB expression levels by cBioporta.The results revealed that FARSB promoter methylation level was inversely linked to FARSB expression.In other words, when FARSB was highly expressed, its promoter was poorly methylated (Fig. 5C).Besides, the promoter methylation of the FARSB subgroup was assessed in relation to various clinical characteristics.We discovered that promoter methylation level was positively connected with patient age, gender, cancer stage, tumor grade, and nodal metastasis status (Fig. 5D-H).Furthermore, there were signi cant differences in the methylation of the FARSB promoter among different races (Fig. 5I).Moreover, in the heatmap, 3 of the 12 CpG sites linked with hypomethylation were connected with a worse prognosis (P < 0.05), including cg20634234, cg25138017, and ch.2.4457098R (Fig. 5J-L).These results testi ed that the hypomethylation of FARSB is bound up with a poor prognosis in HCC .
In Hcc, The Gene Most Associated With Farsb And Cellular Pathway That Farsb Participate Given that FARSB expression was markedly correlated with the HCC prognosis, we tried to nd out the pathway in which FARSB was involved in order to better understand the biological functions of FARSB in HCC.To start with, we utilized "LinkFinder" in LinkedOmics to explore FARSB co-expression genes in HCC.
According to the Spearman test, there were 9879 genes (dark red dots) had a positive correlation with FARSB, while 10042 genes (dark green dots) had a negative connection (false discovery rate, FDR < 0.01) (Fig. 6A).Next, a heatmap depicted the top 50 genes that were positively and negatively linked with FARSB (Fig. 6B-C).Following that, the GO term annotation discovered that the FARSB co-expressed genes were primarily involved in DNA-templated transcription, among other things (Fig. 6D).KEGG pathway analysis showed that it was rich in cell cycle, DNA replication, and other processes.(Fig. 6E).
In contemplation of investigating underlying regulatory pathway of FARSB in HCC, we chosen the higher and lower FARSB expression groups to display GSEA analysis.Some important pathways, like oxidative phosphorylation.Otherwise some metabolism pathways such as purine metabolism, pyrimidine metabolism and glutathione metabolism.Also, RNA-related pathway, rna degradation and biosynthesis of aminoacyl tRNA (Fig. 7A-F).In a recent study of HCC, the pathway oxidative phosphorylation was found to have up-regulated expression in NK cell.Besides, FARSB may be connected with angiogenesis and immune that are closely associated with tumor development [2,3].This result inferred that FARSB might be connected to the cell cycle.

Correlation Of Farsb Expression With Cell Cycle
In view of the results above, we speculated that FARSB was related to the cell cycle.So STRING database was performed on top 500 co-expressed genes to make the PPI network, and Cytoscape was applied for nding the critical module, which we marked in yellow (Fig. 8A).The MCODE score for the genes with the highest scores consisted of CDCA3, NCAPG, KIF2C, CCNB1, PTTG1, and CDC20 (Fig. 8B).And we investigated that there existed an obvious correlation coe cient between FARSB and these genes through GEPIA analysis (Fig. 8C, Fig. 8E).What's more, we have done prognosis analysis of these genes by the Kaplan-Meier Survival Method, which showed that all of these 6 genes were oncogenes that were related to poor prognosis (Fig. 8D, Fig. 8F).The results of pathway analysis proved that all 6 genes in the module were related to the cell cycle in HCC, and based on the above analysis, we inferred that FARSB might be connected to the cell cycle.

Associations Between Farsb And Immune In ltration In Hcc
To lucubrate the mechanism of the mechanism of FARSB in hepatocellular carcinoma, scRNA-seq was used to detect the expression of FARSB in different cell types of HCC.The result demonstrated that FARSB was highly expressed in immue cells group in all LIHC data sets(Fig.9A).Then, we investigated the link between FARSB expression and the in ltrating immune cells (hepatocellular carcinoma) by utilized TIMER "Gene" module.According to this graph that FARSB expression was positively related to the tumor purity and in ltration of some immune cells (Fig. 9B, P < 0.05).In order to nd which type of immune cells that FARSB was mainly expressed, researcher tested the expression of FARSB in various immune cells by scRNA-seq.This gure suggest that FARSB was highly expressed in T cells(Fig.9C).
Additionally, TISIDB was used to detect FARSB expression in various immune subgroups in HCC.Our team defect FARSB express highly in C1 and C2 clusters and low in C6 (Fig. 9D).We also looked into the connection between FARSB expression and T cell checkpoints (such as PD1, PD-L1, LAG3, CD80, B7-H3, PD-L2, IDO1, and TIM-3) through the GEPIA database(Table .3).The expression of FARSB has a signi cant relationship with PD-1, PD-L1, LAG3, CD80, B7-H3, and TIM-3 in HCC (Fig. 9E).Altogether, the data suggested that FARSB expression is intimately related to the degree of the in ltration of T cells.

Correlation Analysis Of Farsb Mrna Expression Levels And Markers Of Various Immune Cell Subsets
To further con rm the correlation among FARSB expression and different immune cells in ltration, we investigated the connection of FARSB expression and the levels with the expression of immune cell gene markers in HCC by using the TIMER database.(Table.4)These ndings suggested that FARSB was highly associated with gene markers of most immune cells (Fig. 10A-E).Moreover, we examined the link between FAESB and diverse T cell immune markers(Table .5).The results suggest that FARSB expression was remarkably related to 29 of the 42 T cell markers in HCC.The data suggested that FARSB expression may in uence the occurrence and development of hepatocellular carcinoma by regulating the degree of in ltration of T cells.
On the basis of previous reports, chemokines help to regulate immune cell migration and provide an antitumor immune response.Chemokines also help to generate and recruit immune cells, which forms a protumor microenvironment.
Thus, we assumed that T cells could be recruited by some speci c cytokines, resulting in a pro-tumor microenvironment and a poor prognosis for HCC.Using the R software, we found that three chemokines and eight receptors were positively correlated with FARSB expression in HCC.Following that, based on plenty of assays, we summarized that CCL26, CX3CL1, and CCR8 can generate and recruit Th2.The scatter diagram showed concrete correlations (Fig. 10F).In addition, using immune-related chemokine and receptor expression in HCCLM3 cells, we discovered that CCL26, CX3CL1, and CCR8 expression were all down-regulated in shFARSB cells (Fig. 10G).This suggested that FARSB may affect microenvironments primarily by in uencing Th2 immune in ltration.

Prognostic Analysis Of Farsb Expression In Hcc In The View Of Immune Cells
According to the results above, FARSB is attributed with immune in ltration in HCC.Moreover, we looked into whether FARSB expression affects the prognosis of HCC caused by immune invasion.The Kaplan-Meier Plotter had been utilized to conduct a prognostic evaluation for patients in different FARSB expression level across various immune cell subsets.We observed that FARSB expression didn't show any obvious difference in the prognosis of HCC under different in ltration levels of these immune cells (Fig. 11C-H).However, high-expressed FARSB was linked with a poor prognosis in enriched Th1 and Th2, while there was no strong association between its decreased subgroups (Fig. 11A-B).Therefore, our conjectural theory is FARSB may affect the prognosis of HCC by enriching Th1 and Th2.
These ndings indicated that FARSB may affect the prognosis who suffer from hepatocellular carcinoma by recruiting Th1 and Th2 to in uence immune in ltration.Overall, our study further con rmed that FARSB could make a difference in HCC's development.It also leads to poor prognosis by modulating immune in ltration..

Relationship Between Farsb Expression And M6a Modi cation In Hcc
N6-methyladenosine (m6A) is the commonest modi cation of eukaryotic RNAs which is required for a number of biological processes.Previous research has veri ed that abnormal regulation of m6A modi cation is linked to a wide range of cancer in different human organs, such as lung, breast and liver.
By analyzing TCGA and ICGC HCC data, we detected a link between FARSB expression and twenty-one m6A interrelated genes expression of HCC.In the TCGA database, FARSB expression was correlated positively with LRPPRC, RBM15B, HNRNPA2B1, YTHDF1 (Fig. 12A, P < 0.01).Furthermore, FARSB expression was correlated positively with RBM15B, LRPPRC, YTHDF1, HNRNPC, HNRNPA2B1 in ICGC data sets (Fig. 12B, P < 0.01).We divided TCGA samples into two groups according to the expression of FARSB.We tried to compare the expression of genes involved in m6A modi cation between the two groups.As shown in Fig. 12, the m6A modi cation was not the same in high and low groups with the FARSB expression in HCC (Fig. 12C).Compared to the group of low expression, the expression of 5 genes in the FARSB high expression group was increased (P < 0.05).Both expression correlation and differential expression of genes, were presented in Venn's diagram, including LRPPRC, RBM15B, HNRNPA2B1, and HNRNPC.(Fig. 12D).The scattering plot shows the relationship between FARSB and m6A related genes expression(Fig.12E) Then, we used the Kaplan-Meier curve to reveal that high expression of LRPPRC, RBM15B, and HNRNPA2B1 was intensely associated with a poor prognosis of HCC (P < 0.001) (Fig. 12F).
These results claim that, in HCC, FARSB may have a close relation with the m6A modi cation, speci cally via its interactions with LRPPRC, RBM15B, and HNRNPA2B1, all of which eventually in uence the progression and prognosis of HCC.

Farsb-related Cerna Regulatory Network
In recent year, many research articles highlighted the regulatory role of lncRNA-miRNA-mRNA ceRNA networks in cancers.Hence we established a FARSB ceRNA regulatory network in HCC.

Farsb Ppi Network (Physical) And Molecular Docking Demonstration
Generally speaking, protein interactions can perform biological functions and metabolic reactions.A FARSB protein interaction web had been established by GeneMANIA.(Fig. 14A) This result indicated FARSB had physical interaction with 11 different proteins.In these proteins, RPLP1, which has been validated that played a important role in HCC progression.In next step, cBioPortal database was utilized to research the secondary structure of FARSB and RPLP1 (Fig. 14B-C), which included various chemical modi cation sites, like Phosphorylation, acetylation, ubiquitination, and methylation.Furthermore, we analyzed tertiary structure of protein FARSB and RPLP1 obtained from the PDB database, and the results of molecular docking shown that they can interact to form dimers (Fig. 14D).

Cancer Pathway Activity And Drug Sensitivity
Based on the above results, we used the GSCALite tool to evaluate the possible role of the top ve GeneMANIA selected genes in the classical cancer pathways.As shown in our results, these genes, especially FARSB, could activate Apoptosis, Cell Cycle, DNA Damage Response, Hormone AR pathways, TSC/mTOR and inhibit EMT, Hormone ER, PI3K/AKT, RAS/MAPK, RTK pathways to play a regulatory role in the cancer process (Fig. 15A).In addition, the cell with high-expressed FARSB were sensitive to 38 drugs or small molecules (Fig. 15B).These ndings showed novel and selectable treatment options for HCC patients with high FARSB expression .

Discission
Hepatocellular carcinoma (HCC), the main pathological histological type of primary liver cancer, is one of the most common malignancies today.And it's the leading cause of cancer-related deaths worldwide.
Early detection and treatment is an effective means of improving survival in patients with hepatocellular carcinoma.However, the most commonly used biomarker for HCC detection, alpha-fetoprotein (AFP), still shows low sensitivity and heterogeneous speci city at various cut-off points.Identi cation of useful biomarkers for monitoring HCC remains inadequate (Piñero et al., 2020).It is therefore of interest to explore more biomarkers for the post-treatment prognosis of HCC, and they may have more clinical utility in the near future.In our article, we combined bioinformatic analysis and in vitro experiments to analyse the expression of FARSB in HCC.We also further surveyed its connection with immune in ltration, m6A modi cation and drug sensitivity.
In our research, we rst identi ed elevated mRNA levels of FARSB in hepatocellular carcinoma compared to normal tissues through the Timer online website.Immediately after we veri ed the results of signi cantly higher FARSB mRNA expression using TCGA, ICGC, and GEO databases.We next examined FARSB expression at the protein level in clinical samples.Our research exposed that FARSB was highly expressed in hepatocellular carcinoma.In conclusion, we noticed that FARSB was highly expressed in HCC at both mRNA and protein levels, using public databases and HCC clinical samples.
After discovering the close relationship between expression and clinicopathological features, we then went on to discover the prognostic value of FARSB expression in hepatocellular carcinoma.Kaplan-Meier Plotter survival analysis exhibited that high FARSB expression had correlation with four types of poor survival including OS, DFS, PFS, DSS.Next the overall survival OS of patients at high FARSB expression was validated using R and the results showed that patients with high FARSB expression had lower overall survival.ROC curve analysis presented that FARSB expression had predictive value for the evaluation of survival in HCC patients.We then further explored the prognostic value of FARSB expression in hepatocellular carcinoma using single multifactor Cox analysis, and the results showed that FARSB could be considered as an independent predictor.Therefore, to further investigate the possible mechanism of FARSB overexpression in liver cancer, we studied the methylation of FARSB promoter in liver cancer, and observed that in HCC patients, the degree of promoter methylation was negatively correlated with FARSB expression.In addition, our analysis exposed that the FARSB promoter was poorly methylated in HCC and correlated with clinical stage, histological grade, lymph node metastasis, age, and gender.Meanwhile, the FARSB promoter is hypomethylated at many CpG sites and is related to poor prognosis.These studies suggest that the high expression of FARSB in HCC may be due to hypomethylation of FARSB promoter, and both high expression of FARSB and hypomethylation predict poor prognosis in HCC patients.
In order to further explore the biological function and potential molecular mechanism of FARSB expression affecting the occurrence and development of HCC, we analyze the enrichment of genes signi cantly related to FARSB expression in HCC.FARSB may form a regulatory network with these genes to promote the occurrence and development of HCC.GO and KEGG results showed that the biological functions of FARSB were associated with cell cycle, DNA replication, acute in ammatory response and base excision repair.Some genes in the regulatory network have been con rmed to affect the occurrence and development of HCC through cell cycle.For example, WDR4 promotes HCC cell proliferation by inducing G2/M cell cycle conversion and inhibiting apoptosis (Xia et al., 2021), while DTYMK up-regulation enhances HCC growth and proliferation by promoting cell cycle (Zhou et al., 2021).GSEA enriched these genes co-expressed with FARSB and found that this regulatory network could promote oxidative phosphorylation, base resection and repair, pyrimidine and purine metabolism, and inhibit cell apoptosis.When FARSB is highly expressed, "oxidative phosphorylation", "base resection and repair", "pyrimidine metabolism, purine metabolism" and other pathways are signi cantly up-regulated, while "apoptosis" pathway is signi cantly down-regulated, which is consistent with the characteristics of liver cancer progression.Changes in liver metabolism are critical to the development of liver disease, and Currently, only 10-20% of the population bene tting from immunotherapy (Rudd, 2020).Therefore, it is of great interest to discover more new biomarkers to improve prognosis and individual tumor treatment.
Considering that FARSB is highly correlated with HCC immune cell in ltration, to continue to understand the contribution of FARSB in HCC immunotherapy, we revealed the relationship between FARSB and HCC immune checkpoints, and we explored that high expression of FARSB was closely associated with PD1, PD-L1, CTLA-4, CD80, B7-H3, TIM-3.These results revealed that patients with high FARSB expression may bene t from immunotherapy for HCC, and that rational application of CTLA-4 and PD-1 or its ligands CD80 and PD-L1 inhibitors will facilitate restore anti-tumor immune responses, which in turn will provide long-term bene ts to patients.To explore in depth the role of FARSB in tumor immunity, we investigated the correlation between FARSB and immune cell biomarkers in HCC.The results present an obvious positive relationship between FARSB expression and biomarkers of T cells which further demonstrates a positive correlation between FARSB and immune cell in ltration.
The continuous injury and regeneration of hepatocytes can be caused by the accumulation of immune cells to tumors which can also promote the development of HCC (Giannelli et  The interaction of protein is usually necessary to implement biological function and metabolic reactions, and protein -protein interaction by the regulation of modi cation after translation (Sadeghi et al., 2018).Therefore, we used the GeneMANIA database to establish an interaction network between FARSB and other tumor-related proteins.The interaction network showed that FARSB could directly interact with RPLP1.Ribosomal protein LP1 (RPLP1) is a member of the Ribosomal protein L12P family (Artero-Castro et al., 2009), which takes an oncogene part in hepatocellular carcinoma (Xie et al., 2021).Next, we continued to study the secondary structure of FARSB and RPLP1, and found that FARSB and RPLP1 have multiple modi cation sites, which can be modi ed after translation.These results suggest that targeted protein modi cations can alter the expression of FARSB and RPLP1, and FARSB may act by interacting with RPLP1.Finally, to further explore the mode of direct interaction, we forecasted encouraging binding sites by a molecular docking model.Our result could furnish a basis for future experimental studies.
In conclusion, our study found hypommethylation and overexpression of FARSB in HCC, which was substantially connected to clinicopathological features and poor prognosis of HCC patients.FARSB expression affected cell cycle and immune microenvironment of HCC.And up regulation of FARSB expression promoted tumor immune cell in ltration and checkpoint expression.This suggests a new direction for tumor immunotherapy in patients with high expression of FARSB, and is closely related to m6A modi cation and drug sensitivity.This suggests that FARSB can be used as a diagnostic marker and immune-related therapy target, with an opportunity to refresh diagnostic and treatment choices of HCC. Figures More and more studies have shown that epigenetic modi cation can affect the accumulation of genetic changes during the development of liver cancer (Sciacovelli and Frezza, 2017, Serrano-Gomez et al., 2016, Schulze et al., 2015, Fujimoto et al., 2012, Totoki et al., 2014), in which abnormal DNA methylation has been detected could gradually increases with the progression of cancer (Nishida et al., 2012).
mitochondrial oxidative phosphorylation contributes to the development or progression of hepatocellular carcinoma(Piccinin et al., 2019).Meanwhile, inhibition of apoptosis has also been proved to be an important factor in the progression of HCC(Zhang et al., 2021), for example, inhibition of apoptosis induced can promote the progression of HCC(Sui et al., 2015).In conclusion, the signaling pathways associated with high expression of FARSB can promote the progression of cancer.Whether FARSB can promote the development of HCC by regulating cell cycle, oxidative phosphorylation and apoptosis signaling pathways remains to be further studied.Recently, a large number of studies have proved that immune microenvironment (TME) takes a crucial part in tumor genesis and development(Binnewies et al., 2018, Quail andJoyce, 2013), and immune cell in ltration is closely associated with tumor tissue formation, survival and metastasis (Hinshaw and Shevde, 2019).TGF-β derived from TME in HCC can promote the expression of tim-3 in tumor-associated Macrophages (TAMs).Activated TAM promotes tumor growth and immune tolerance through NF-κB/IL-6 pathway(Yan et al., 2015).At the same time, tumor immune cell in ltration can also in uent the prognosis of cancer patients and the e cacy of immunotherapy(Razmkhah et al., 2019, Ren et al.,  2018).For example, in HCC, intertumor neutrophil in ltration indicates a poor prognosis associated with CXCL5 overexpression(Zhou et al., 2012).To further understand the role of FARSB in HCC, we investigated the association between FARSB gene expression and immune cells.First, we detected the expression of FARSB in a variety of immune cells.Next, we continued to explore the connection between the expression of FARSB and immune cell in ltration, and the ndings displayed that the expression level of FARSB was related to many kinds of immune cells in HCC.Our result showing that FARSB may take a certain part in the immune in ltration of HCC.With a better understanding of mechanisms of cancer progression, we have discovered more ways to treat tumors (Xin Yu et al., 2019), among which immune checkpoint blockade (ICB) has been a great success and immune checkpoint inhibitors (ICI) are used as rst-line therapy for advanced HCC, such as inhibition of CTLA-4 and PD-1 expression.Immunotherapies that inhibit the expression of CTLA-4 and PD-1 have been effective in the treatment of various tumors (Litak et al., 2019, Wessely et al., 2020, Xin Yu et al., 2020), and tumor immunotherapy requires su cient immune cells to in ltrate the tumor microenvironment and su cient immune checkpoint expression to achieve e cacy (Chae et al., 2018).

Figure 2 Box
Figure 2

Figure 3 Comparison
Figure 3

Figure 4 Construction
Figure 4

Figure 14 Physical
Figure 14

Table 1
Logistic analysis of the association between FARSB expression and clinical characteristics.

Table 2
Univariate and Multivariate COX regression analysis of factors associated with OS in Liver cancer patients.

Table 3
Correlation analysis between FARSB and immune checkpoint of T cells in TIMER

Table 4
Correlation analysis between FARSB and gene markers of different types of immune cells in TIMER

Table 5
Correlation analysis between FARSB and gene markers of different types of T cells in TIMER (Guo et al., 2020))2019Kryczek et al., 2011)l., 2016).Chemokines take a important part in the recruitment and activation of immune cells.Also participate in tumor progression, invasion and metastasis(Kryczek et al., 2016, Kryczek et al., 2011).Interestingly, We noticed that FARSB expression was positively related to chemokines CCL26, CX3CL1 and CCR8, a key surface molecule of TH2 cells, which played a role in chemotactic TH2 cells in HCC (Julia, 2012,Hamilton et al., 2014), suggesting that FARSB expression had closely connection with TH2 cell in ltration in HCC.At the same time, studies have shown that different molecular mechanisms affect the progression of liver cancer by triggering characteristic responses in speci c immune cell subsets.For example, TH2 cytokines can promote the spread and metastasis of cancer cells in various cancers(Berraondo et al., 2019).An imbalance in the TH1/TH2 cell ratio have relations with reduced survival rate in patients with breast, melanoma, esophageal, and colon cancers (Speiser et al., 2016).It comparing to the low expression group.Among them, studies have shown that LRPPRC can promote G1/S conversion and cell proliferation in hepatocellular carcinoma(Liu et al., 2021b).RBM15B has been identi ed as an independent prognostic indicator of melanoma(Liu et al., 2021a).HNRNPA2B1 can promote the progression of esophageal cancer by up-regulating fatty acid synthase as a carcinogen(Guo et al., 2020).In conclusion, our results suggest that the poor prognosis of patients with higher expression of FARSB may be connected to m6A modi cation.And FARSB may in uent the methylation level of HCC mRNA through its connection with LRPPRC, RBM15B, and HNRNPA2B1, ultimately leading to poor prognosis of HCC patients.Subsequently, we created a ceRNA regulatory network based on prediction.Since FARSB's ceRNA regulatory network is derived from bioinformatics analysis, we need more experiments to con rm this network in future studies.With the progress of molecular biology, more and more attention has been paid to the development of HCC treatment drugs and their clinical bene ts.It has been reported that advanced HCC patients bene t from tivantinib treatment (Santoro et al., 2013).However, the limitations of drug therapy such as moderate e cacy of targeted drugs for liver cancer, lack of therapeutic response biomarkers and susceptibility to drug resistance still exist (Luo et al., 2021).In our study, HCC cell lines were sensitive to 38 drugs or small molecules when FARB is highly expressed.In particular, HCC cell lines with high expression of FARSB are sensitive to tivantinib, which may provide a new treatment option for HCC patients with high expression of FARSB.
has not been reported.Our results show that there is a certain correlation between FARSB and m6A modi cation.Eight M6A-related genes were differentially expressed when FARSB was overexpressed and underexpressed.Four genes most related to FARSB expression were screened out by Venn diagram, including LRPPRC, RBM15B, HNRNPA2B1 and HNRNPC.In addition, by constructing kaplan-Meier curves, we found that the survival time of LRPPRC, RBM15B and HNRNPA2B1 with high expression of FARSB was shorter