Genetic alterations of ALKBH1 in STAD revealed by comprehensive analysis

We initially determined the frequency and type of ALKBH1 alterations in STAD, based on the STAD dataset from TCGA, and found that the ALKBH1 gene was mutated in 2.1% of all cancers by cBioPortal dataset (Fig. 1A). We further examined the genetic alterations of ALKBH1 in different tumor types in the TCGA dataset and observed that STAD tumors had the fourth highest frequency of genetic alterations, mainly mutations, Amplification, and Deep deletion (Fig. 1B). Additionally, we found that missense mutations from 389 amino acids were the primary type of genetic alteration in all TCGA tumor samples (Fig. 1C). Furthermore, our analysis of concurrent gene alteration frequencies of ALKBH1 alterations using the cBioPortal database revealed that there were 4028 samples with enriched gene alterations in both ALKBH1 altered and unaltered groups (Fig. 1D, E). Finally, we observed significant variations in the gain and loss of ALKBH1 in the copy number variation ratio distribution and boxplot (Fig. 1F).

Fig. 1

Frequency and functional enrichment analysis of ALKBH1 alterations in gastric cancer. A Analysis of various mutations in the ALKBH1 gene in human cancer data. B cBioPortal cancer genomics analysis of the frequency of ALKBH1 gene alterations in various types of cancer. C ALKBH1 protein domain map showing specific mutation sites. D, E Volcano and scatter plots showing frequency changes of genes associated with mutations in ALKBH1. F A combination of scatter and box plots to show a more detailed distribution and correlation view of copy number variation in gastric adenocarcinoma

We subsequently extracted data on ALKBH1 expression in STAD patients from the TCGA database and presented it using a waterfall plot that described the top 25 affected genes. Analysis of genetic variations across different ALKBH1 expression levels unveiled connections between these levels and frequently mutated genes (such as TTN, TP53, etc.) in STAD (Fig. 2A). To delve further into the impact of ALKBH1 mutations on STAD, we explored their interactions with other common genes involved in cancer progression, including TTN, PBRM1, and SETD2. In alignment with the findings in Fig. 2A, the mutated ALKBH1 gene demonstrated a more pronounced connection with TTN (Fig. 2B).

Fig. 2

The gene landscape and expression of ALKBH1 in STAD. A Fisher ‘s exact test was used to compare mutation frequencies between ALKBH1-high and -low groups, with the right panel showing mutation types, driving mutation types, and groups. B The study investigated the relationship between ALKBH1 and the nine highly mutated genes in STAD, with red lines indicating mutation sites. C Univariate and multivariate Cox regression models were used to calculate the risk ratio of ALKBH1 at different stages of STAD. D ALKBH1 expression was compared between STAD tumor and non-tumor tissues, with box plots showing expression levels at different stages of STAD E, metastatic status F and tumor grade G. ** P < 0.01, *** P < 0.001 versus the normal group

We then conducted univariate and multivariate Cox regression analyses to evaluate the relationship between ALKBH1 expression and overall survival (OS). Our findings revealed that stage, and ALKBH1 expression (high vs. low) were all significantly associated with OS. Furthermore, we performed multivariate Cox regression analysis on the same variables and discovered that risk score could act as an independent prognostic factor (P < 0.05) (Fig. 2C). To gain a more comprehensive understanding of ALKBH1 expression in various contexts, we examined its expression levels in different sample types, cancer stages, metastatic stages, and Helicobacter pylori infection in The Cancer Genome Atlas (TCGA) dataset (Fig. 2D–G). Our results indicate that ALKBH1 is significantly overexpressed in late stage, highly metastatic status, and highly malignant tumor grades.

Clinicopathological significance of ALKBH1 and its prognostic value in patients

We validated the clinical significance of ALKBH1 by analyzing immunohistochemical staining data from the Human Protein Atlas, which confirmed that ALKBH1 protein was upregulated in STAD tumor tissues compared to normal tissues (Fig. 3A). To further understand the prognostic value of ALKBH1, we performed Kaplan–Meier analysis and log-rank tests to examine the correlation between ALKBH1 expression and clinical follow-up data. We found that patients with high ALKBH1 expression had shorter overall survival (OS) time, post-progression survival (PPS), and progression-free survival (PFS) than those with low ALKBH1 expression (Fig. 3B–D). ALKBH1 was identified as an potential prognostic factor in intestinal, diffuse, and mixed types of STAD, but not in basal type (Fig. 3E–G). Additionally, ALKBH1 was found to be an independent factor in HER-positive and HER-negative patients (Fig. 3H and I).

Fig. 3

Analysis of the pathological changes of ALKBH1 in clinical gastric cancer. A Immunohistochemical analysis of ALKBH1 protein expression levels in STAD tissue samples from different patients based on the Human Protein Atlas. B–I Kaplan–Meier survival analysis indicating overall survival (OS) based on low/high expression of ALKBH1 C post-progression survival (PPS), D progression-free survival (PFS), E intestinal, F diffuse, G mixed, H HER + , I HER−. All survival analyses were based on 6 GEO datasets, including: GSE62254, GSE29272, GSE15459, GSE14210, GSE51105, GSE22377

In an effort to map transcriptomic signatures onto H&E-stained histological sections of a human reference tumor (PMID: 35931863) (Fig. 4A), we utilized ST technology. This approach involves sequencing spatially localized barcoding, which maps transcriptomic signatures directly onto histological images. Our analysis resulted in a total of 2384 counts measured, each with its own expression signature, and superimposed on the histological image of the tumor based on localized barcodes. An unsupervised clustering approach was used to group the spots according to the gene expression of each spot, with each cluster representing a specific cell type based on known marker genes and underlying histology. Space Ranger generated 10 unsupervised clusters (Fig. 4B) that were consistent with known marker genes for each tumor microenvironment cell, overlaid on histological features. We compared the biomarkers of STAD, AQP1, and PECAM1 with ALKBH1, which showed high levels in the lower right. Our analysis revealed that the spatial expression position of ALKBH1 was similar to that of the two biomarkers. Additionally, we examined the three gene expression levels in spots in high magnification regions to highlight histological features. Dot plots display the normalized, logarithmically transformed, and variance-scaled expressions of various cell clusters (y-axis) and signature genes (x-axis) in STAD snRNA-seq data (Fig. 4C). The data revealed that the expression of ALKBH1 was consistently higher than the other two gastric cancer biomarkers, AQP1 and PECAM1, across all 10 clusters. Moreover, the frequency of ALKBH1 gene alterations was significantly higher compared to AQP1 and PECAM1, as revealed by gene variation analysis (Fig. 4D).

Fig. 4

Gene expression of spatially transcriptomic-defined clusters in STAD. A The tissue sections were analyzed by spatial transcriptomics to identify clusters, and proper alignment with morphology was demonstrated through hematoxylin and eosin staining and cluster maps. B The spatial location and gene changes of ALKBH1, AQP1, and PECAM1 were visualized in different tissue sections using 10X Visium spatial gene expression. The variation of each gene in boxes i and ii was illustrated in an enlarged images. C A dot plot was used to show the levels of gene expression in different clusters. D A box plot was used to display alterations in the expression of three genes

Analyzing single-cell RNA sequencing database to evaluate the ALKBH1 transcriptomics

To evaluate the transcriptome of ALKBH1 in STAD at the single cell level and investigate the heterogeneity of various cell types in the STAD microenvironment, we conducted an analysis of the publicly available STAD single-cell RNA sequencing database (GSE162115). This database originated from two STAD patient samples and comprised of 11 cell types and 26 clusters. After removing batch effects and quality control, we analyzed a total of 35,308 cells and identified 26 significant cell populations using UMAP plots. Cell type-specific markers were identified based on top differentially expressed genes for each cluster, which were used for cell type classification. The expression and distribution of ALKBH1 in these single-cell sequencing databases were analyzed, and we found that ALKBH1 was highly expressed in the same region as the inflammatory response model. Additionally, TNFA signaling via NF-kB and xenobiotic metabolism were also generally increased, as shown in Fig. 5G, H.

Fig. 5

Single-cell RNA sequencing analysis enables the identification of immune cell populations. A and B Displays the relative proportions of each cell type found in the public dataset and the proportion of integrated immune cells present in the database. C and D The STAD cells from the GSE162115 dataset are visually represented using the unified flow approximation and projection (UMAP) technique and color-coded based on clusters. E The expression clusters of ALKBH1 are visualized using UMAP plots and subjected to gene set enrichment analysis (GSEA) F inflammatory response, G TNFA signaling via NFkB and H xenobiotic metabolism. I Visual representations of clusters of placental tissue from the GSE1599929 dataset identified using UMAP plots and bar graphs are shown through single-cell RNA sequencing. J The expression of the ALKBH1 gene and cell type biomarkers in different single-cell type clusters of tissues is shown in the heatmap

The strong association between ALKBH1 overexpression and the development of STAD was demonstrated in Fig. 2E, while Fig. 5F, G suggested that ALKBH1 was linked to immune cells. Consequently, we conducted further analysis on the impact of ALKBH1 on immune cell infiltration in STAD using another publicly available database. Through a single-cell RNA sequencing dataset (GSE159929) (Fig. 5I), we examined transcriptomic data, characterized heterogeneous cell populations, and investigated changes in immune cells in the tumor microenvironment. We obtained UMAP maps of 5,318 single cells and analyzed the specificity and distribution of ALKBH1 in different cell populations of STAD to determine the gene expression variation in each cell type. The heatmap revealed that ALKBH1 was highly expressed in gastric tissues, where the biomarkers of five macrophages also exhibited high expression levels (Fig. 5J).

Association of macrophage infiltration with overexpression of ALKBH1

In this study, we investigated the potential correlation between ALKBH1 and STAD by performing gene set enrichment analysis (GSEA) using differentially expressed genes (DEGs). We identified clusters highly associated with macrophages from the single-cell RNA-sequencing database shown in Fig. 5A–H. Since macrophages exhibit tissue-resident properties and possess pro- or anti-inflammatory functions, we further analyzed the relationship between ALKBH1 and tumor immune interactions (Fig. 6A). Given the critical role of the tumor immune microenvironment (TIME) in tumor progression, metastasis, and immune evasion, we conducted Spearman correlation analysis by TISIDB, which revealed a positive correlation between ALKBH1 expression and the majority of tumor-infiltrating macrophages in STAD (Fig. 6B). Moreover, we observed complex interactions between various T cell subpopulations and macrophages, which were the strongest intercellular communication detected in STAD, as retrieved from the scTIME portal (Fig. 6C).

Fig. 6

Investigation of the impact of ALKBH1 on the tumor immune microenvironment (TIME). A Immunological analysis of the immune infiltrate was conducted using the GSE162115 database, as illustrated. Moreover, a heat map was generated to demonstrate the association between ALKBH1 expression and immune infiltration in human gastric cancer cells. B A heat map is presented to exhibit the correlation between ALKBH1 expression and lymphocyte infiltration in human cancers. C A circular plot was constructed to visualize the potential ligand-receptor interactions, where the edge's width represents the strength of the interaction

We aimed to investigate the potential association between ALKBH1 expression and the tumor microenvironment of STAD, particularly in macrophages. We assessed the correlation between ALKBH1 and M0, M1, and M2 macrophage subtypes, and found a positive correlation (Fig. 7A). Furthermore, we examined macrophage biomarkers and observed a positive correlation between ALKBH1 and CD163, CD68, MARCO, MRC1, and MSR1 (Fig. 7B). To validate the association between ALKBH1 and immune cells, we compared low and high ALKBH1 expression groups in terms of immune-related functions. The difference between the low and high ALKBH1 groups was significant for immune effector process, regulation of immune response, production of immune response, innate immune response, T cell proliferation involved, but not for "Th1 cells" (Fig. 7C). Patients with high ALKBH1 expression and high macrophage infiltration exhibited shorter survival times than those with high gene expression only (Fig. 7D). We investigated whether ALKBH1 mutations also affect macrophages. Using the mutation module in Pan-cancer, we analyzed the effect of ALKBH1 mutations on immune cell infiltration in various cancer types. Our results showed that ALKBH1 mutations were the fourth most frequent in pan-cancer (Fig. 7E). We also examined the correlation between ALKBH1 mutations and macrophage biomarkers mentioned earlier. We found that the expression of CD163, MARCO, and MRC1 was significantly elevated in mutated ALKBH1, while CD68 and MSR1 showed no significant difference (Fig. 7F).

Fig. 7

Exploring the link between ALKBH1 and immune infiltration in STAD. A The relationship between ALKBH1 levels and various macrophage subtypes was explored. B The correlation between ALKBH1 and genes related to macrophages was investigated. C Effect of different ALKBH1 expression on immune cells. D The Kaplan–Meier plots of macrophage and ALKBH1 mRNA expression to visualize the survival differences in STAD. E The Gene_Mutation module assesses the incidence of ALKBH1 gene mutations across multiple cancer types. Examples of mutational profiles and immunogenicity analyses are presented. F The analysis of the association between ALKBH1 mutations and macrophages

In order to validate the relationship between ALKBH1 and macrophages in STAD, we conducted a triple-labeling immunofluorescence study of ALKBH1, CD163, and DAPI on the entire STAD in TMA. The STAD TMA samples were divided into early and late stages, and their fluorescence intensities were analyzed using panoramic tissue scanning. The high intensity of green fluorescence ALKBH1 and red fluorescence CD163 in the late-stage group can be observed in Fig. 8A, with both red and green fluorescence consistently increased. The magnified image clearly shows a large amount of ALKBH1 in gastric cancer tissues, with advanced tumors showing a large amount of red color, resulting in some tissues appearing yellow. After quantifying the fluorescence signals, we confirmed a positive correlation between ALKBH1 and CD163 (Fig. 8B). We also analyzed the signals of ALKBH1 and CD163 in each of the 60 biopsies and confirmed their positive correlation (Fig. 8C). Additionally, we analyzed the correlation between ALKBH1 and CD163 using the TMIER database, and the results were similar to those presented in Fig. 8D.

Fig. 8

Investigation of the correlation between ALKBH1 and CD163 expression in tumor biopsies during STAD progression. A Whole TMA was triple-labeled with immunofluorescence of ALKBH1, CD163 and DAPI, followed by panoramic tissue scanning. B The Pearson's correlation coefficient was used to display the overlapping values of ALKBH1 and CD163 fluorescence signals. C A scatterplot was used to illustrate the normalized Spearman correlation between ALKBH1 and CD163 for each patient reading. D The correlation between ALKBH1 and CD163 expressions in STAD from the CIBERSORT dataset was shown in a scatter plot. ** p < 0.01

Examination of the clinical and pathological importance of ALKBH1 in STAD from a pan-cancer perspective

To further investigate the potential clinical significance of ALKBH1, we examined its expression levels in various types of cancer and evaluated its prognostic value. They found that several types of cancer, including breast cancer (BRCA), cholangiocarcinoma (CHOL), esophageal carcinoma (ESCA), head-neck squamous cell carcinoma (HNSC), liver hepatocellular carcinoma (LIHC), lung adenocarcinoma (LUAD), lung squamous cell carcinoma (LUSC), prostate adenocarcinoma (PRAD), thyroid carcinoma (THCA), and uterine corpus endometrial carcinoma (UCEC), exhibited similar expression patterns to those observed in STAD (as shown in Fig. 9A). Next, the we performed Kaplan–Meier analysis to assess the association between ALKBH1 expression and overall survival (OS) in each of these cancer types. Their analysis revealed that high expression of ALKBH1 was significantly associated with shorter OS time in BRCA (as shown in Fig. 9B), ESCA (Fig. 9C), HNSC (Fig. 9D), LIHC (Fig. 9E), LUAD (Fig. 9F), and UCEC (Fig. 9G). These findings suggest that ALKBH1 may have a broad clinical value in predicting prognosis across multiple cancer types.

Fig. 9

Bioinformatic validation of the clinical significance of ALKBH1 across multiple cancer types. The clinical significance of ALKBH1 from a pan-cancer perspective. Expression levels of ALKBH1 in various cancers and comparison between tumor and normal tissues with statistical significance denoted by asterisks (A). Kaplan–Meier survival analysis of low/high ALKBH1 expression groups in BRCA (B), ESCA (C), HNSC (D), LIHC (E), LUAD (F and UCEC (G). ***p < 0.0001, **p < 0.01, *p < 0.05 between tumor and normal

Identification of potential ALKBH1 inhibitors for STAD through pharmacogenetic screening

The study aimed to identify potential drugs with efficacy against STAD by examining the GDSC database for drugs that displayed increased potency in the presence of high ALKBH1 expression. We conducted cross-association analyses between drug response and the CRSPR knockout of ALKBH1 using single-guide RNA (sgRNA) in different STAD cells. Our performed cross-correlation analyses to investigate the effects of 430 drugs on sgRNA-mediated ALKBH1 in various STAD cells (Fig. 10A). The analysis identified six small molecule drugs, namely Elesclomol, Vinorelbine, Alisertib, ZM447439, RU-SKI 43, and FTY-720, that displayed altered potency (Fig. 10B–G). The study found that STAD cell lines with high sgALKBH1 efficiency were more sensitive to these drugs. Overall, the results suggest that these drugs have the potential to be used as anticancer agents targeting ALKBH1 to regulate gastric cancer cell growth.

Fig. 10

Analysis of drug sensitivity and cytotoxicity in gastric cancer cells. A The scatter plot illustrates the cross-association scores of predictivity and descriptivity that were used to identify potent drugs with efficacy against STAD cells. To identify gene signatures and potential drugs, we queried the pharmacogenetics database for the ALKBH1 gene. We then evaluated the drug sensitivity of the sgALKBH1 gene to various chemical drugs in STAD cell lines. The boxplots B–G depict the log of the half maximal inhibitory concentration (IC50) values for six drugs, namely Elesclomol, Vinorelbine, Alisertib, ZM447439, RU-SKI 43, and FTY-720, that showed altered potency