Validation of CDC45 as a novel biomarker for diagnosis and prognosis of gastric cancer

Expression and transcription analysis

The TIMER2.0 (Li et al., 2017) was utilized to investigate the CDC45 expression among 33 different cancer types in The Cancer Genome Atlas (TCGA) project. Clinical and RNA-seq data of 375 gastric cancer patients and 32 normal samples were obtained from the TCGA database through the UCSC Xena website (https://xena.ucsc.edu/), and 210 normal samples were obtained from GTEX databases (https://www.gtexportal.org/home/). Additionally, four sets of gastric cancer chip datasets, namely GSE13911, GSE29272, GSE118916, and GSE66229, were downloaded from the GEO (https://www.ncbi.nlm.nih.gov/geo/) database and utilized as validation sets to investigate the differential expression of CDC45 in this study. The log2 [TPM (transcripts per million) +1] or log2 [FPKM (fragments per kilo base per million) +1] transformed expression data were employed for data analysis. The analysis process of this study is presented in Fig. 1. Detailed information about the data sources can be found in Table S1.

Diagnostic and survival analysis

The diagnostic value of CDC45 was evaluated employing the pROC package (Robin et al., 2011) in R version 3.6.3 (R Core Team, 2020). The resulting ROC curves were plotted using the ggplot (Wickham, 2009). To verify the diagnostic performance, the GSE66229 dataset, which included 300 gastric cancer samples and 100 paired normal gastric tissue samples, was employed. The correlation between CDC45 gene expression and overall survival was analyzed using the Kaplan-Meier Plotter in the R package. Furthermore, three additional GC chip datasets (GSE66229, GSE84433, and GSE84437) were obtained from the GEO database and used for prognostic validation.

Clinical specimen collection and RT-PCR analysis

Human GC tissue specimens and adjacent normal mucosa tissues were obtained from GC patients who underwent surgical treatment and were confirmed by pathology in Liutie Central Hospital (Liuzhou, China). Samples that had undergone radiotherapy or chemotherapy as well as those with unclear pathological diagnosis and incomplete clinical data were excluded. The clinicopathological characteristics are shown in Table S2. Tissues were stored in RNAlater solution (Qiagen, Valencia, CA, USA) at −80 °C. All procedures were approved by the Institutional Review Board of Liuzhou Liutie Central Hospital (KY2023-075-01). All patients provided written informed consent. The study methods conformed to the standards established by the Declaration of Helsinki.

In preparation for RNA extraction, frozen tissues (n = 12) were then crushed in liquid nitrogen. The total RNA from these tissues (100 mg per microdissection sample) was extracted using the SteadyPure Quick RNA Extraction Kit (AG21023; Accurate Biology, Shenzhen, China) following the instructions provided by the manufacturer. DNase I (AG12001; Accurate Biology, Shenzhen, China) was applied to digest DNA contaminants. The concentrations of total RNA were quantified using a NanoQ micro-volume Spectrophotometer (CapitalBio, Beijing, China). The reverse transcription of mRNA (1 μg RNA per samples) was carried out using the GoScript™ Reverse Transcription System (A5001; Promega, Madison, WI, USA). The mRNA expression levels were measured using the GoTaq® Master Mix kit (A6001; Promega) on a Roche LightCycler 480 II Real-time fluorescence quantitative PCR machine. The primers of RT-qPCR were designed in the website Primer Bank (https://pga.mgh.harvard.edu/primerbank/index.html). Primers were synthesized by Shangya Biotechnology (Fuzhou, China) and GAPDH was used as a reference following the 2^−ΔΔCT method. CDC45: forward: 5′-AGAGCATAAAGAACAGTTCCGCTA-3′; reverse, 5′-GACTGGCCTGTGAGTGTCAC-3′.

Immunohistochemistry (IHC)

Immunohistochemical (IHC) staining of gastric tissues was carried out on 3-µm-thick sections of formalin-fixed and paraffin-embedded (FFPE) tissue. The staining procedure followed standard protocols and utilized the EliVision™ Plus kit (Maixin Biotechnology, Fuzhou, China) as described previously (Chen et al., 2023). The sections were stained using antibodies against CDC45 (JJ091-04) (1:200) (Hangzhou HuaAn, Biotechnology, China) according to the manufacturers’ instructions. Hematoxylin was used for counterstaining in IHC. IHC staining were observed with a Leica light microscope (Leica DMI4000 B; Leica, Wetzlar, Germany).

Somatic mutations and drug sensitivity analysis

Somatic mutation data of the TCGA-STAD cohort (n = 396) was acquired from TCGA GDC Data Portal (https://portal.gdc.cancer.gov/). Analysis and visualization of the differences in markers of TCGA molecular classification between the high and low CDC45 expression groups were conducted by the “maftools” package (Yang et al., 2023). The cancer-related chemotherapeutic drug sensitivity was predicted via the Genomics of Drug Sensitivity Database following the previous study (Tang et al., 2020).

Immune infiltration analysis

For the analysis of the relative abundance of 22 immune cell subtypes among TCGA-STAD, the CIBERSORTx website (https://cibersortx.stanford.edu/) was used. Additionally, the relationship between the expression of CDC45 and immune infiltration in gastric cancer (GC) was investigated (Liu et al., 2018). Based on ESTIMATE, a ratio of immune-stromal components was calculated in each sample of cancer, and three scores were displayed: ImmuneScore, StromalScore, and ESTIMATEScore.

Co-expressed analysis

To identify the 255 co-expressed genes of CDC45 (Spearman’s correlation value > 0.3; p < 0.05), the “Co-expression” module of cBioPortal (https://www.cbioportal.org/) was utilized (Ding et al., 2022). From this set, the top six co-expressed genes were selected based on their cor-values. Detailed information about the data sources of co-expressed genes can be found in Table S3. Subsequently, the correlation analysis of these genes was visualized in cBioPortal. Additionally, the survival analysis of the top six co-expressed genes was conducted in GEPIA2 (http://gepia2.cancer-pku.cn/#index).

Pathway and enrichment analysis

To elucidate the potential functions of CDC45, DEGs (differentially expressed genes) between the CDC45 high and CDC45 low groups were identified using the Limma package, with the criteria of an adjusted p-value < 0.05 and a |logFC| > 1.5. Additionally, the “Cluster Profiler” R package was employed to conduct and visualize KEGG and GO (Gene Ontology) analysis, including molecular function (MF), biological process (BP), and cell composition (CC). Enrichment to a meaningful pathway was determined based on an adjusted p-value < 0.05 (Yu et al., 2012). Furthermore, Gene Set Enrichment Analysis (GSEA) (https://www.gsea-msigdb.org) was performed to identify functional and biological pathways associated with low and high expression of CDC45, using the DEGs sets between the CDC45 high and CDC45 low groups (Subramanian et al., 2005). The gene sets for pathway and enrichment analysis were listed in Table S4.

Statistical analysis

The statistical analysis was conducted using R software (version 4.0.3 & 3.6.3; R Core Team, 2020). The comparison of differential expression between cancer tissues and adjacent tissues was performed using a paired t-test. The two groups were compared using the Mann–Whitney U test. Correlation analyses were conducted using Spearman’s correlation. A significance level of p < 0.05 was used to determine statistical significance.

Upregulated expression of CDC45 in gastric cancer

Initially, the expression of CDC45 was investigated across various cancer types using RNA-seq data from TCGA. It was observed that CDC45 expression was upregulated in different malignancies (Fig. 2A), particularly in gastric cancer (GC) where mRNA levels were significantly higher compared to both normal tissues and adjacent normal tissues (Figs. 2B–2D). To further validate these findings, CDC45 expression was assessed in multiple independent datasets (GSE13911, GSE29272, GSE118916, and GSE66229), consistently demonstrating significantly elevated levels of CDC45 expression in GC tissues (Figs. 2E–2H). In addition to confirming the differential expression of CDC45via qRT-PCR, as depicted in Fig. 2J. We further investigated the protein expression levels of CDC45. Our analysis, utilizing data from the Human Protein Atlas (HPA) (Fig. 2I) and validated through immunohistochemistry (IHC) analysis as shown in Fig. 2K, revealed significant disparities between normal and GC tissues. Collectively, these findings provide compelling evidence that patients with GC exhibit heightened CDC45 expression at both the mRNA and protein levels, thereby implying its potential involvement in the pathogenesis of GC.

Potential significance of CDC45 in GC diagnosis and prognosis

Based on the aforementioned findings, it can be inferred that the overexpression of CDC45 may possess significant diagnostic and prognostic implications in gastric cancer. In order to substantiate this hypothesis, a comprehensive analysis and validation were conducted using multiple datasets. Notably, the diagnostic potential of CDC45 was conducted rigorously evaluated in the TCGA cohort, wherein it exhibited a strong diagnostic capability to discriminate gastric cancer (GC) cases. This was evidenced by the receiver operating characteristic (ROC) analysis, which yielded an impressive area under the curve (AUC) value of 0.911 (95% CI [0.873–0.948]) (Fig. 3A). Moreover, within the training set (TCGA+GTEx), encompassing both cancerous and non-cancerous samples, CDC45 demonstrated exceptional diagnostic accuracy. The ROC analysis further substantiated the reliability of CDC45 as a diagnostic marker for gastric cancer (GC), as evidenced by a robust area under the curve (AUC) exceeding 0.950 (Fig. 2B). To reinforce our findings, we conducted an independent validation in the GSE66229 cohort. Once again, CDC45 exhibited a substantial level of efficacy in distinguishing between GC and normal samples, with an AUC of 0.928 (95% CI [0.903–0.954]) (Fig. 3C). AFP and CEA are commonly used tumor markers in gastrointestinal tumors. Based on the TCGA-STAD data, we plotted the ROC curves of AFP and CEA for the diagnosis of gastric cancer (Fig. S1). The results showed that the AUC area of CEA and AFP in the diagnosis of gastric cancer was 0.669, 0.593, respectively, which was much lower than CDC45 (AUC = 0.911). Additionally, the effectiveness of CDC45 in differentiating between early tumor pathology (stage I and II) and normal controls was confirmed through ROC analysis (Fig. 3D). Collectively, our findings consistently indicate the up-regulation of CDC45 in GC and emphasize its potential as a valuable diagnostic biomarker.

To assess the prognostic value of CDC45 in patients with gastric cancer (GC), Kaplan-Meier (KM) survival curves were constructed using the TCGA, GSE66229, GSE84433, and GSE84437 datasets. In the TCGA-STAD dataset (Fig. 3E), patients with high expression of CDC45 exhibited a significantly lower risk of mortality (p = 0.014; HR = 0.66) compared to those with low expression. This trend was also observed in the GSE66229 cohort (Fig. 3F), where high CDC45 expression was associated with a better OS outcome (p = 0.028; HR = 0.70). To ensure the robustness of our findings, we further validated our results in two additional independent datasets. The validation of the GSE84433 dataset consistently supported our initial findings, demonstrating that GC patients in the CDC45 high-expression group had a favorable overall survival (p = 0.015; HR = 0.68) (Fig. 3G). Likewise, the GSE84437 dataset also confirmed the association between high CDC45 expression and improved overall survival in GC patients (p = 0.025; HR = 0.73) (Fig. 3H). Taken together, these findings highlight the potential clinical significance of abnormal CDC45 expression as a prognostic indicator for patients with GC.

Somatic mutations landscapes analyses

We further conducted a comprehensive examination of the genomic characteristics of CDC45-based GC subgroups (Fig. 4A). Our analysis revealed that the high-CDC45 group exhibited significant mutations in the following five genes: TTN, MUC16, SYNE1, ARID1A, and CSMD3. This finding suggests a potential association between elevated CDC45 mRNA expression and the genomic stability of gastric cancer (GC) patients. Previous research has indicated variations in the efficacy of chemotherapy among GC patients based on their molecular classifications (Sohn et al., 2017). To further explore this phenomenon, we assessed the sensitivity of patients in the CDC45 high and low expression groups tot chemotherapy. The results demonstrated that patients in the CDC45 high expression group exhibited higher sensitivity to chemotherapeutic drugs, including 5-Fluorouracil, docetaxel, cisplatin, and elesclomol (Figs. 4B–4E). This suggests that docetaxel, paclitaxel, and cisplatin may be appropriate chemotherapeutic regimens for GC patients with high CDC45 expression.

Correlations between CDC45 expression and immune infiltration in GC

In this study, we employed the CIBERSORT method to examine the association between CDC45 expression and tumor-infiltrating immune cells in gastric cancer (GC). The 375 GC samples were categorized into high and low expression groups based on the median value of CDC45 expression, and the relative abundance of 22 immune cell types was analyzed. The distribution and correlations of these immune cell types in GC are depicted in Figs. 5A and 5B. The findings of the study indicate that there is a significant association between high CDC45 expression and increased infiltration of various immune cell types, including M0 macrophages, M1 macrophages, NK resting cells, activated CD4+ memory T cells, and T follicular helper cells (Fig. 5B, all p < 0.05). Conversely, low CDC45 expression is associated with other immune cell types, such as naive B cells, memory B cells, resting CD4+ memory T cells, regulatory T cells, plasma cells, activated NK cells, monocytes, and resting mast cells (Fig. 5B, all p < 0.05). Additionally, the results demonstrate a positive correlation between CDC45 expression and the infiltration of antitumor cells, including M0 macrophages, M1 macrophages, and activated CD4+ memory T cells (Fig. 5C, all p < 0.05), while low CDC45 expression was associated with increased infiltration of immunosuppressive cells, particularly regulatory T cells (Fig. 5C, all p < 0.05). These findings suggest that GC patients with low CDC45 expression may have a stronger ability to evade immune responses.

The ESTIMATE method was employed to calculate immune scores, stromal scores, and ESTIMATE scores for each sample. Notably, significant negative correlations were observed between CDC45 expression and stromal score, immune score, and ESTIMATE score in GC (Figs. 5D–5F), indicating that CDC45 exerts a substantial influence on the infiltration levels of stromal and immune cells in GC.

Genes co-expressed with CDC45 in GC

To identify the top co-expressed genes with CDC45, the CBioPortal web server was utilized to analyze three distinct studies from TCGA: TCGA Firehose Legacy, TCGA Nature 2014, and TCGA Pan-Cancer Atlas. Through the examination of these studies, a total of 255 genes were identified as exhibiting positive co-expression with CDC45 representing the overlapping genes found within the three databases (Fig. 6A). The Spearman’s correlation values for this co-expression were presented in Fig. 6B. To ascertain the most highly co-expressed genes with CDC45, a ranking was established based on adjusted p values. Consequently, the top six co-expressed genes were determined through this analysis. It is plausible that these genes possess a robust association with CDC45 expression and potentially fulfill significant functions within shared biological processes or pathways. The results of the correlation analysis demonstrated a significant positive correlation between CDC45 and several genes, including protein disulfide-isomerase A5 (CDCA5), denticle less E3 ubiquitin protein ligase homolog (DTL), Non-SMC condensing I complex subunit H (NCAPH), RAN binding protein 1 (RANBP1), Origin recognition complex subunit 1 (ORC1), and lRAD54L (Fig. 6C). These findings suggest that the expression levels of CDC45 and these genes are likely to be regulated in a coordinated manner, indicating potential functional relationships or shared regulatory mechanisms. Survival analysis indicated higher expression of CDCA5, NCAPH and RAD54L were significantly correlated with more favorable prognosis (p = 0.022, 0.037 and 0.042, respectively).

Function enrichment analysis of potential CDC45 regulatory network in GC

A total of 56,494 genes were examined (Table S4), where in differential expression genes were visualized using the volcano plot (Fig. 7A). In order to gain further insights into the role of CDC45 and its potential interactions with other genes, we conducted the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses (Fig. 7B). The KEGG pathway enrichment analysis unveiled that the molecular mechanisms associated with CDC45, and its associated genes primarily revolved around signal pathways involved in the regulation of sarcous function (Fig. 7C). GO enrichment was various aspects The GO enrichment analysis was conducted to investigate the involvement of CDC45 in three aspects of biological processes (BP), cellular components (CC) and molecular functions (MF). The primary focus of enrichment was on contractile fiber, contractile fiber part and muscle system processes, as depicted in Figs. 7D–7F. Furthermore, Gene Set Enrichment Analysis (GSEA) revealed significant enrichment of gene sets related to G protein-coupled receptor ligand binding (NES = −1.808; p adjust = 0.040; FDR = 0.029) and G alpha (i) signaling events (NES = −1.560; p adjust = 0.040; FDR = 0.029), as shown in Figs. 7H and 7I. Altogether, these findings imply that CDC45 plays a crucial role in the regulation of muscle development and function, which may affect the tumor cell invasion and metastasis mechanism of GC.