AKR1B10 as a novel prognostic biomarker linking methylation and immune escape in hepatocellular carcinoma.

Introduction
Globally, hepatocellular carcinoma (HCC) is the sixth most prevalent cancer and a major cause of cancer-related death. Risk factors for the development of HCC encompass hepatitis virus infection, alcohol intake, obesity, tobacco use, and diabetes mellitus [1]. Consequently, significant research endeavors have been dedicated on discovering new biomarkers to enhance the development of more effective therapies. Recent evidence indicates that the combination of ICIs and gene-targeted therapies presents considerable potential for cancer treatment [2]. Timely diagnosis markedly enhances patient outcomes; nevertheless, the prognosis for advanced-stage HCC remains unfavorable. Although medical technology has advanced, there are still few effective treatments for HCC [3]. As a result, it is absolutely essential to identify individual genes and outlining the molecular mechanisms responsible in HCC.
Aldo-keto reductases (AKRs) comprise a superfamily of monomeric, soluble, and NADPH-dependent oxidoreductases. These enzymes act on carbonyl groups within aldehydes and ketones, facilitating their reduction to alcohols during elimination reactions. AKR1 family, play significant roles in various biological processes, including cancer development and drug metabolism. Among these, AKR1B10, with its restricted expression in the intestines and adrenal glands, is noteworthy for its potential implications in cancer and metabolic diseases. AKRs, including AKR1B10, are often aberrantly expressed in tumors, contributing to oncogenic signaling and chemoresistance [4]. AKR1B subfamily members, including AKR1B1 and AKR1C2, have been linked to cancer invasion and migration, highlighting their importance in cancer progression [5]. AKR1B10 is involved in the metabolism of signaling metabolites, influencing immune responses and inflammatory diseases [6, 7]. AKRs are overexpressed in a variety of inflammation-associated pathological conditions, including cancer, atherosclerosis, asthma, gout, osteoporosis, uveitis, sepsis, and alcoholic liver cirrhosis [8]. The enzyme’s activity in reducing carbonyl compounds is crucial for detoxifying xenobiotics and regulating steroid hormones [9]. Conversely, AKR1B10 appears to be downregulated in gastrointestinal cancers [10]. This paradoxical expression pattern underscores the context-dependent nature of AKR1B10 in tumorigenesis, characterized by both overexpression in certain cancers and downregulation in others. Inhibitors targeting AKR isoforms, including AKR1B10, are being explored as potential anticancer agents, with some showing promise in clinical trials [11]. Despite the growing body of evidence regarding AKR1B10 in cancer, its precise function in HCC remains unclear. Gaining insight into the relationship between AKR structure and function can help develop more potent therapies [12].
Given the well-established role of AKR1B10 in cellular metabolism and its potential association with carcinogenesis, this study aimed to comprehensively evaluate its expression and protein levels in LIHC utilizing multiple publicly available online databases. Additionally, these resources were leveraged to analyze the prognostic and diagnostic potential of AKR1B10. Our collected data suggests that AKR1B10 may serve as a valuable biomarker for predicting outcomes in LIHC., potentially linked to both methylation and immune escape mechanisms. These findings indicate potential for the advancement of innovative immunotherapies and targeted treatments for LIHC in the future.

Methods

Clinical samples and TCGA datasets acquisition
The data regarding AKR1B10 mRNA expression were collected from TCGA datasets (http://cancergenome.nih.gov/) [13]. Moreover, we examined the upregulated expression data of AKR1B10 in 33 types of human cancer, and in 374 LIHC tissues with 50 normal liver tissues. Clinical characteristics of LIHC patients and related genes were also acquired. For validation purposes, tissue samples of normal liver and liver tumor were obtained from The First Affiliated Hospital of Chongqing Medical University, under ethical approval. Each group comprised a sample size of three (n = 2). To elucidate the regulatory mechanisms of AKR1B10 in HCC, we examined the relationship between AKR1B10 expression and 25 genes that are significantly correlated with it. Additionally, an analysis of differential gene expression was conducted to identify genes with varying expression levels between high and low AKR1B10 expression groups in LIHC samples (p < 0.05).

Protein expression analysis
The CPTAC database (https://proteomics.cancer.gov/programs/cptac) is a collection of data (genomics, proteomics, imaging), assays, and reagents, which are made available to the public to speed up cancer research and advance patient care [14]. UALCAN (http://ualcan.path.uab.edu/) is an accessible and an online platform for analyzing cancer proteomic data from the CPTAC dataset, was utilized to assess AKR1B10 protein expression in HCC.
The HPA database (https://www.proteinatlas.org/) is a publicly accessible resource that comprehensively maps the human proteome [15]. Here, the IHC staining results of AKR1B10 in normal liver tissue and HCC tissue were obtained from the HPA.

Survival and prognostic analysis
The Kaplan-Meier plotter (http://kmplot.com/) is a comprehensive public online database that assesses the influence of diverse gene expressions (miRNA, mRNA, protein) on survival across over 30,000 samples from 21 cancer types, including liver cancer (n = 364). The overall survival (OS), progression-free survival (PFS), disease-specific survival (DSS), and relapse-free survival (RFS) of AKR1B10 in liver cancer were assessed using patient samples categorized into two groups according to median expression levels. The hazard ratio (HR), along with a 95% confidence interval (CI) and log-rank P value, was also incorporated.

Gene and protein interaction analysis
GeneMANIA (http://www.genemania.org/) is a versatile and user-friendly platform for formulating hypotheses regarding gene function, analysis, and prioritizing genes for functional assays [16]. STRING (https://string-preview.org/) is an online database for functional enrichment analysis. This study employed GeneMANIA to construct gene-gene interaction networks and STRING to analyze PPI networks centered on AKR1B10, leveraging the online database for functional enrichment analysis of PPIs.

GO term and KEGG pathway enrichment analysis
Functional enrichment analysis, encompassing biological processes (BPs), KEGG Pathways (KPs), and molecular functions (MFs), was performed on the top 300 genes exhibiting the strongest positive correlation with AKR1B10, as identified within the TCGA database. Additionally, to elucidate the regulatory framework of AKR1B10 in LIHC, we investigated the correlation between AKR1B10 expression and 25 significantly associated genes (P < 0.05).

Genetic alteration analysis
The cBioPortal database for Cancer Genomics (https://www.cbioportal.org/) provides an open-access web platform for the interactive analysis of multidimensional cancer genomics datasets [17]. The cBioPortal was employed to identify various genetic alterations of AKR1B10 across multiple cancers, with a subsequent focus on the alterations in LIHC. A log-rank test was performed on KM plots for survival outcomes, including OS, DSS, PFS, and RFS, that were associated with AKR1B10 alterations.

Methylation analysis
MethSurv (https://biit.cs.ut.ee/methsurv/), a web portal offering univariable and multivariable survival analysis utilizing DNA methylation biomarkers derived from TCGA data, was developed to examine the DNA methylation sites of AKR1B10 and assess the prognostic significance of corresponding cytosine-guanine (CpG) methylation. DNA methylation status of the AKR1B10 regulator in various cancers were analyzed using the UALCAN database (http://ualcan.path.uab.edu/) to identify differential methylation patterns between tumor and normal tissues. The SMART (http://www.bioinfo-zs.com/smartapp/) facilitates the examination of the location of methylation tags across chromosomes [18]. The HR with a 95% CI for OS was calculated, and a P < 0.05 was deemed statistically significant.

Immune cells and drug sensitivity analysis
An extensive online platform for the systematic analysis of immune infiltrates across various cancer types, TIMER (http://timer.cistrome.org/) is used to input the gene expression profile data of tumor samples [19, 20]. The TIMER method was used in this study to better understand the connection between AKR1B10 levels in LIHC and six types of immune cells that enter tumors. These are B cells, CD4 + T cells, CD8 + T cells, neutrophils, macrophages, and DCs. TISIDB (http://cis.hku.hk/TISIDB/) is a digital platform that consolidates diverse data types related to tumor and immune system interactions [21]. The present study employed TISIDB to determine the expression of AKR1B10 and 28 distinct types of tumor-infiltrating lymphocytes (TILs) in various human malignancies. GSCALite (http://bioinfo.life.hust.edu.cn/web/GSCALite/) is an extensive online platform for the analysis of gene expression and drug sensitivity analysis [22].

Total RNA isolation and quantitative rt‒qpcr
Total RNA was extracted from liver tissue utilizing TRIzol reagent (Invitrogen, USA). Subsequently, cDNA was synthesized via reverse transcription employing a commercially available kit (RNAsimple Total RNA Kit (DP419) (TIANGEN Biotech Co., Ltd., China), strictly adhering to the manufacturer’s recommended protocol. RT-qPCR was then conducted using SYBR Green Master Mix on a RT-PCR system (Applied Biosystems). The relative expression levels of the target mRNA were normalized against an appropriate endogenous control (β-actin) and analyzed employing the 2 − ΔΔCt method. The primer sequences utilized in this study are available in supplementary file.

Statistical analysis
The online database utilized in this study employs established statistical methods, as previously described. The optimal cutoff value for AKR1B10 expression was determined by generating a receiver operating characteristic (ROC) curve using the SRplot webtools (https://www.bioinformatics.com.cn/srplot, accessed on 24 November 2024) [23]. A threshold of p < 0.05 or log-rank p < 0.05 was used to determine statistical significance.

Results

AKR1B10 expression analysis in hepatocellular carcinoma (HCC)
The mRNA expression pattern of AKR1B10 was evaluated across 33 cancer types using the TCGA database, providing a pan-cancer perspective. As shown in Fig. 1a, illustrates that AKR1B10 was markedly upregulated in 11 cancer types, including LIHC, in comparison to normal tissues. The observed mRNA expression patterns of AKR1B10 across diverse cancer types exhibited significant variability.
Subsequently, mRNA and protein expression levels of AKR1B10 in LIHC were analyzed using data from the TCGA and HPA databases. Figure 1b illustrates that paired data analysis demonstrated a significantly elevated mRNA expression of AKR1B10 compared to adjacent normal tissues (n = 50), with a median (IQR) of 110.794 (14.383-585.812) versus 4.376 (0.836–11.36, P < 0.001). AKR1B10 was evaluated at the protein level using the National Cancer Institute’s CPTAC dataset (Fig. 1c). To further validate the observed trends, RT-qPCR was performed in vitro (Fig. 2d). Comparative analysis with corresponding normal liver tissue samples revealed a statistically significant upregulation of AKR1B10 expression in HCC tissues (P < 0.001). Collectively, these finding indicated that the expression of AKR1B10 was markedly elevated in LIHC tumor tissues relative to normal tissues.
Moreover, immunohistochemical staining from the HPA database indicated that normal liver tissues exhibited negative or moderate AKR1B10 IHC staining (Fig. 1f), whereas LIHC tumor tissues displayed moderate to strong staining. Immunofluorescence analysis was performed in HepG2 cells to determine the subcellular localization of AKR1B10 relative to nuclear, microtubular, and endoplasmic reticulum markers. Figure 2 illustrates that AKR1B10 is situated within microtubules, the nucleus, and ER.

Assessment of the diagnostic and prognostic value of AKR1B10
To evaluate the diagnostic utility of AKR1B10 in differentiating LIHC from normal tissues, ROC curve analysis was performed. The resulting AUC was 1.000 (95% CI: 0.020-1.000) (Fig. 3a), suggesting that AKR1B10 may serve as a highly accurate biomarker for distinguishing between LIHC and normal tissues.
Next, to determine the prognostic significance of AKR1B10 expression in hepatocellular carcinoma (LIHC), patients were stratified into high and low expression groups based on AKR1B10 expression levels. Subsequently, Kaplan-Meier survival analysis was conducted to evaluate the association between AKR1B10 upregulation and patient outcomes. As depicted in (Fig. 3b and e), higher AKR1B10 expression was significantly correlated with improved OS, DSS, PFS, and RFS (all P < 0.001).
Further investigation was conducted to elucidate the connection between AKR1B10 mRNA regulation and prognosis, thereby enhancing the assessment of its prognostic significance. As depicted in (Fig. 3f), elevated AKR1B10 expression exhibited a significant association with increased risk across a range of clinical and demographic factors in LIHC, including T stage, tumor grade, AJCC stage, microvascular complications, alcohol consumption, viral infection, and demonstrating no significant association with race or gender.

Enrichment analysis of AKR1B10-interacting genes in HCC
To examine the essential fundamental biological process and molecular pathways associated with AKR1B10, we performed enrichment analyses using GeneMANIA and STRING databases. According to Figure 4a, AKR1B1, AKR1B15, LRAT, and SRXN1 are among the 20 most commonly changed genes that are closely linked to AKR1B10. Meanwhile, STRING analysis of the AKR1B10 protein interaction network identified 10 interacting proteins, including AKR1C3, CBR1, ALDH3A2, and ALDH7A1 (Fig. 4b).
Next, to investigate the co-expression genes of AKR1B10 and understand the gene sequence interactions, we utilized the TCGA database. Results identified the 25 genes that exhibited a positive relation with AKR1B10 in LIHC (Fig. 4c).
Furthermore, to elucidate the biological functions and pathways related with AKR1B10, KEGG pathway and gene ontology were conducted using the top 300 genes positively correlated with AKR1B10 expression. The top ten important terms from the KEGG, MF, and BP enrichment analyses are shown in (Fig. 4d). In terms of BP, AKR1B10 was upregulated in metabolic process, cellular aldehyde process, and alcohol catabolic process. In terms of KEGG pathways, AKR1B10 was enriched in metabolic process e.g., Glycerolipid metabolism, Histidine metabolism, Arginine metabolism. In terms of MF, AKR1B10 was enriched in oxidoreductase activity, NAD + activity, Aldehyde dehydrogenase activity.

Genetic alteration of AKR1B10 in HCC
Generally, the development of human cancers is attributed to the accumulation of genetic variations. As a result, the AKR1B10 genetic changes were examined in cancer samples and subsequently analyzed using the cBioPortal database to focus on LIHC. According to the analysis, distinct AKR1B10 genetic alterations were found in various cancers at varying frequencies (Fig. 5a). Furthermore, the predominant genetic alteration of AKR1B10 in LIHC was mutation and amplification, occurring at a rate of 9% (Fig. 5a and b). Additionally, Fig. 4c displays two pie charts that show the different types of mutations for better understanding. The percentage of samples with missense substitutions was about 38.61%, synonymous substitutions in approximately 20.14%, frameshift deletions in approximately 1.68%, and other types in approximately 6.71% (Fig. 5c). The substitution mutations occurred primarily at G > A (31.02%) and C > T (25.64%). Furthermore, a systematic investigation was conducted to determine the relationship between the clinical survival outcomes of HCC patients and genetic alterations in AKR1B10. Patients with genetic alterations of AKR1B10 in liver hepatocellular carcinoma (LIHC) exhibited improved DFS (P = 0.0363), but did not demonstrate significant differences in OS (P = 0.835), DSS (P = 0.957), or PFS (P = 0.0539) compared to patients without genetic changes in AKR1B10 (Fig. 5d).

DNA methylation of AKR1B10 and its gene correlation in LIHC
The methylation status of AKR1B10 was examined in LIHC by analyzing DNA methylation sites within the gene. The prognostic impact of CpG site methylation within the AKR1B10 gene was subsequently assessed using the MethSurv database. A total of 6 methylation CpG sites of AKR1B10 were identified, with cg01783195 exhibiting the DNA methylation level (Fig. 6a). Among the investigated CpG sites, methylation at cg01783195 demonstrated a significant association with the prognosis of LIHC patients. Hypermethylation at this site was associated with improved OS in these patients. (Table 1).

Methylation analysis revealed a significant inverse correlation between CpG site methylation within the AKR1B10 gene region and the expression of the AKR1B10 isoform, suggesting potential epigenetic regulation of AKR1B10. For example, At the cg12001930 site, a significant inverse correlation was observed between DNA methylation levels and isoform expression (R = -0.39, p < 7.9e-16), as illustrated in Fig. 6b. Similar negative correlations were observed at other CpG sites, suggesting that DNA methylation may play a crucial role in regulating the expression of this specific AKR1B10 isoform in hepatocellular carcinoma.
To further understand the regulatory network surrounding AKR1B10, (Fig. 6c-d) identified 25 genes that exhibited significant positive and negative correlations with AKR1B10 expression levels in LIHC using the TCGA database. This analysis offers significant insights into the potential functional interactions and regulatory pathways associated with AKR1B10 in hepatocellular carcinoma.

Correlation of AKR1B10 expression and immune cell infiltration in LIHC
To assess the correlation between AKR1B10 expression and immune cell infiltration in LIHC, we examined the TIMER database to analyze the relation between AKR1B10 and six distinct types of TILs. The results indicate an upward trend between AKR1B10 degree of expression and the infiltration of various immune cell subsets in the HCC tumor microenvironment. Specifically, Substantial positive correlations were observed between AKR1B10 expression and abundance of CD4 + T cells (Rho = 0.246, P = 3.94e-06), B cells (Rho = 0.121, P = 2.43e-02), monocytes (Rho = 0.176, P = 9.97e-04), macrophages (Rho = 0.311, P = 3.58e-09), and neutrophils (Rho = 0.11, P = 4.06e-02 (Fig. 7a). In addition, to link into the relationship between AKR1B10 and the TME, we analyzed data from the TISIDB, examining AKR1B10 expression across 28 different types of TILs in human cancers (Fig. 7b). These findings suggest that AKR1B10 may significantly influence the TME in HCC, potentially impacting immune surveillance and therapeutic response.

Drug sensitivity
We examined the drug sensitivity associated with AKR1B10 expression in tumors using GSCALite (Fig. 8). The expression of AKR1B10 exhibited a negative correlation with the 24% inhibitory concentration (IC50) values of five drugs: 17-AAG, Afatinib, Cetuximab, erlotinib, and Lapatinib. In addition, there was a positive relation between AKR1B10 and IC50 values of AR-42, Belinostat, BX − 912, CAY10603, CUDC − 101, CX − 5461, FK866, I − BET − 762, IPA − 3, OSI − 027, NPK76 − II − 72 − 1, PI − 103, PHA − 793887, PIK − 93, QL − X−138, SNX − 2112, TPCA − 1, THZ − 2−102 − 1, and Tubastatin.

Discussion
To develop potential therapeutic strategies, elucidating the molecular mechanisms of HCC is paramount for advancing our understanding of this disease. This study aimed to clarify the correlation of AKR1B10 with expression, prognosis, pathway analysis, genetic alterations, methylation, immune evasion, and drug sensitivity in LIHC. Our analysis demonstrated a significant upregulation of AKR1B10 in LIHC compared to adjacent non-tumor tissues validated by qRT-PCR. Moreover, elevated AKR1B10 expression correlated with advanced tumor stages. These findings, coupled with observed correlations between AKR1B10 expression, immune cell infiltration, and DNA methylation patterns, strongly suggest a critical role for AKR1B10 in HCC tumorigenesis and progression.
Aldo-keto reductases (AKRs) comprise a superfamily of enzymes that catalyze the NAD(P)H-dependent reduction of a diverse array of compounds containing carbon, including aldehydes, ketones, and quinones [24]. AKR1B10 demonstrates catalytic activity in the reduction of aliphatic and aromatic aldehydes, exemplified by the transformation of retinal to retinol. This enzymatic activity leads to retinoic acid depletion, thereby disrupting a critical signaling pathway that regulates cellular proliferation and differentiation [24, 25]. Furthermore, AKR1B10 plays a critical role in lipogenesis and fatty acid biosynthesis, while also detoxifying cytotoxic carbonyl compounds, and facilitates the emergence of resistance to multiple chemotherapeutic agents utilized in HCC therapy, including anthracyclines and cisplatin [26]. The overexpression of AKR1B10 is often associated with a less aggressive clinical course in lung cancers [27], AKR1B1 displays more widespread overexpression across various cancers and is frequently linked to poor patient outcomes [28]. Notably, AKR1B10 expression is elevated in several cancer cell lines, including medulloblastoma, pancreatic adenocarcinoma, and colon cancer, whereas AKR1B1 expression is prominent in bladder and renal carcinomas [29]. Interestingly, AKR1B1 expression is decreased in invasive breast and colorectal cancers but elevated in basal-like breast cancer [30]. Both AKR1B1 and AKR1B10 are upregulated in smokers and implicated in lung cancer development and progression [31]. Unlike AKR1B1, AKR1B10 expression is significantly reduced in gastric cancer and inversely correlates with both tumor stage and metastatic progression, suggesting its potential prognostic significance [10]. Importantly, overexpression of both AKR1B10 and AKR1C2 has been observed in Barrett’s esophagus, indicating a potential role in its pathogenesis and subsequent progression to esophageal cancer [32]. Finally, high AKR1B10 expression has been observed in the early stages of HCC, with subsequent downregulation as metastasis occurs [33]. Notably, high AKR1B10 expression in HCC tissues is correlates with favorable prognosis, longer RFS, and improved DSS [34]. Despite the considerable clinical importance of AKR1B10 in diagnosing and prognosticating HCC, its precise function and regulatory mechanisms in HCC progression remain inadequately understood, encompassing its associated AKR1B10 genes, pathway analysis, genetic modifications, methylation, immune evasion, and drug sensitivity assessments.
Utilizing STRING and GEPIA2 databases, we uncovered a set of genes co-expressed with AKR1B10 across diverse tissues and tumor types. Gene enrichment analysis of these co-expressed genes identified overrepresentation of biological processes associated with metabolic process, cellular aldehyde process, and alcohol catabolic process. Notably, our analysis highlighted strong correlations between AKR1B10 expression and the expression of several genes with well-established roles in these pathways, including AKR1B1, AKR1B15, LRAT, SRXN1 and their functions, e.g., Glycerolipid metabolism, Histidine metabolism, Arginine metabolism. These observations correspond with the findings of our gene enrichment analysis and offer significant insights into the potential molecular functions and regulatory networks associated with AKR1B10.
Genetic alterations, including mutations and deletions, are well-established drivers of cancer progression by disrupting critical cellular processes such as DNA replication, DNA damage repair, and cell cycle regulation (42). To investigate the potential role of genetic alterations in AKR1B10-mediated tumorigenesis, we analyzed the mutational landscape of AKR1B10 in TCGA database. Our analysis revealed a prevalence of AKR1B10 genetic alterations in LIHC at 9%%, with the majority represented by missense substitutions and amplification. Notably, these missense substitutions predominantly involved G > A and C > T transitions, consistent with known mutational signatures. Furthermore, the systematic study demonstrates the link between AKR1B10 genetic changes and clinical survival result in HCC patients. Analysis revealed a statistically significant connection between AKR1B10 genetic modifications and improved DFS (P = 0.0363) in HCC patients. However, no significant differences were found in OS (P = 0.835), DSS (P = 0.957), or PFS (P = 0.0539). In conclusion, although infrequent, genetic alterations in AKR1B10 may exert a subtle yet potentially significant influence on HCC development or progression.
DNA methylation levels changes were observed in cancer. In general, DNA hypermethylation may alter the DNA structure and subsequently suppress the transcription of the corresponding gene, thereby silencing it [35]. Nevertheless, the Methylation within CpG islands served as an epigenetic mechanism to modulate the expression of downregulated genes [36]. Consequently, the analyses revealed the presence of six methylation CpG sites of AKR1B10, some of which were correlated with prognosis in LIHC. Our analysis revealed that AKR1B10 undergoes DNA methylation in LIHC, with cg01783195 exhibiting the highest methylation level. Hypermethylation at cg01783195 was identified as a significant prognostic indicator for improved OS in HCC patients. Furthermore, A significant inverse correlation was observed between DNA methylation levels at the cg12001930 site within the AKR1B10 gene and AKR1B10 isoform expression, suggesting that DNA methylation may serve as a critical epigenetic regulator of AKR1B10 expression in HCC. Furthermore, to elucidate the regulatory network surrounding AKR1B10, we identified 25 genes that exhibited significant correlations with AKR1B10 expression levels in LIHC. These results provide valuable insights into the epigenetic and regulatory networks modulating AKR1B10 expression, with potential implications for HCC pathogenesis.
TME, critically influences tumor survival and progression, primarily through immune evasion mechanisms [46]. Tumors manipulate immune responses, enabling them to evade immune surveillance and destruction [37]. Consequently, extensive research on liver cancer has revealed that β-Catenin and PD-L1 facilitate immune evasion [38]. Additionally, PD-L1 expression can be induced both directly and indirectly by the enhanced activity of the STAT3 and NF-κB signals [1]. Meanwhile, activated hepatic stellate cells (HSCs) significantly enhance liver fibrosis through their immunomodulatory activity, characterized by the expression of proteins like PD-L1, which promotes the proliferation of immune cells suppression, including Tregs and MDSC [39, 40]. Liver metastasis is the most common pattern of distant disease progression after lymph node involvement, prompting numerous endeavors to elucidate that cancer immunity is a crucial component of metastasis in HCC [41, 42]. Furthermore, suppressive immune cells, such as MRC1 + CCL18 + M2-like macrophages, underwent significant spatial reprogramming in the metastatic microenvironment, despite the liver’s immunosuppressive polarization [42, 43]. Remarkably, liver metastases have the ability to drastically modify the immune environment of the liver through recruiting and spreading monocyte-derived macrophages and causing a systemic reduction in antigen-specific T cells [44]. Consequently, we examined the link between AKR1B10 expression and the TME in HCC. The TIMER database indicated that AKR1B10 expression correlated with tumor purity, CD8 + T cells, CD4 + T cells, macrophages, neutrophils, and dendritic cells. By synthesizing the aforementioned results and the expression of AKR1B10 in 28 types of TILs, it was hypothesized that AKR1B10 may influence immune infiltration in LIHC and subsequently impact its progression.
The investigation of immune evasion as a target for liver cancer therapy has resulted in notable progress in immunotherapeutic approaches. The identification of immune escape mechanisms has shifted the paradigm of immunotherapy, establishing it as a critical target for cancer treatment [45]. Nonspecific immune stimulation, immune checkpoint inhibitors, adaptive cell transfer, and dendritic cell-based vaccination have emerged as promising immunotherapeutic strategies for treating liver cancer patients [2]. The combination of bevacizumab, an anti-angiogenic agent, and atezolizumab, an immune-checkpoint inhibitor, has recently been shown to improve overall survival in LIHC when compared to sorafenib [3]. The GDSC database showed that the IC50 value of eight different drug types—including 17 − AAG, BMS − 509744, RDEA119, CGP − 60474, FH535, PD − 0325901, PD − 173074, and Trametinib—was negatively connected with the expression of AKR1B10. These medications may be able to stop the spread of cancer. This discovery has the potential to inform clinical drug selection and improve patient outcomes.

Limitations
Despite using large public datasets like TCGA and GTEx, this study has a number of inherent limitations that should be carefully taken into account. First off, these dataset’s retrospective design raises the possibility of biases in sample selection and treatment history, which could skew the associations that are found. Second, the current study lacks complementary in vivo experimental validation and primarily relies on bioinformatics analyses. The biological roles and underlying mechanisms of AKR1B10 in tumor progression must be confirmed by functional experiments, despite the strong correlations found between AKR1B10 expression, genetic changes, and survival outcomes.
Variations in immune cell population estimations can affect the immune infiltration analysis, which is based on several computational approaches. The accuracy of immune infiltration data may be impacted by the shortcomings of existing deconvolution algorithms due to the intrinsic heterogeneity of the TME. To confirm the connections between AKR1B10 and immune cells in the TME, more research using single-cell sequencing or spatial transcriptomics is required. Numerous correlational relationships are also examined in this study, such as those between immune infiltration, tumor mutational burden (TMB), and AKR1B10 expression and survival. In order to determine whether AKR1B10 directly affects these factors or if they are connected indirectly through other pathways, more research should focus on mechanistic studies.
Numerous essential pathways, including Wnt signaling and sphingolipid signaling, were found by pathway enrichment analyses; however, it is still unclear what role these pathways play biologically in relation to AKR1B10 function. To clarify the ways in which AKR1B10 alters these pathways in particular tumor contexts, more functional research is necessary. This study highlights the need for additional validation and thorough research to firmly establish the role of AKR1B10 in cancer and its potential as a therapeutic target by clearly recognizing these limitations.

Conclusion
This study showed that AKR1B10 was significantly upregulated and linked to immune evasion and poor prognostic methylation in LIHC. The study’s results indicate that increased AKR1B10 expression could function as a new targeted biomarker for more sophisticated and precise targeted therapies in LIHC.

Electronic supplementary material
Below is the link to the electronic supplementary material.