We obtained a unified and standardized pan-cancer dataset (TCGA TARGET GTEx, https://xenabrowser.net/) from the UCSC database. Furthermore, we extracted the ORC6 gene values from each sample and applied log₂ (x+1) transformation for each value. The Sangerbox (13) online tool was used for visualization. In addition, we acquired validation cohorts from external sources, including the International Cancer Genome Consortium (ICGC), Chinese Glioma Genome Atlas Project (CGGA), and Gene Expression Omnibus (GEO) databases. These cohorts included the ICGC-LIRI-JP cohort, CGGA-301 cohort, CGGA-325 cohort, CGGA-693 cohort, and GSE13041 cohort. Cancer-type abbreviations are listed in Table 1 .

The gene mutation type and frequency of ORC6 in the TCGA pan-cancer dataset were explored by accessing cBioPortal (https://www.cbioportal.org/). We obtained the copy number variation (CNV) dataset at gene level 4 from all TCGA samples processed by GISTIC software (14) through GDC (https://portal.gdc.cancer.gov/). We used the unpaired Wilcoxon rank sum test or the Wilcoxon signed rank test to compare the values between the two groups and the Kruskal–Wallis test for differences among multiple groups.

The Human Protein Atlas (HPA, https://www.proteinatlas.org/) was utilized to obtain images of the subcellular localization of ORC6 protein in cancer cells (HEK293 and PC-3) by immunofluorescence staining of cells. Furthermore, the subcellular localization of the ORC6 gene was obtained through the Genecards database.

The comPPI website (http://comppi.linkgroup.hu/) was utilized to analyze the protein-protein interaction network of ORC6. The minimum interaction score was 1, and the edge width was scaled based on the interaction score.

The correlation of ORC6 expression with clinical features was assessed by Spearman correlation analysis. We performed univariate Cox regression analysis to investigate the prognostic significance of ORC6 expression in predicting the disease-free interval (DFI), progression-free interval (PFI), overall survival (OS), and disease-specific survival (DSS) in pan-cancer cohorts. We then utilized forest plots for a graphical representation of these results.

TCGA data were curated to extract ORC6 expression levels in transcripts per million (TPM) format, followed by data normalization using log2(TPM+1). Survival data of matched samples were integrated and subsequently subjected to optimal grouping truncation using the ‘surv_cutpoint’ function from the ‘survminer’ package. The aim was to distinguish between the high and low ORC6 expression groups. Prognostic differences between the high- and low-expression groups were evaluated using the log-rank test. RNA-seq data in TPM format from TCGA and GTEx were uniformly processed through the Toil pipeline, as sourced from UCSC XENA (https://xenabrowser.net/datapages/). ORC6 expression levels corresponding to TCGA cancer samples and GTEx normal tissue samples for each cancer type were extracted. The data were normalized using Log2(TPM+1). To assess the diagnostic accuracy for tumor detection, we employed the ‘pROC’ package to calculate sensitivity and specificity. Diagnostic value was quantified by the area under the curve (AUC), with a value of 1.0 indicating perfect diagnostics and 0.5 representing no diagnostic value. An AUC greater than 0.85 was considered to possess a high diagnostic value.

The correlation of ORC6 expression with DNA promoter methylation levels in cancer was explored by UALCAN(https://ualcan.path.uab.edu/) (15). The correlation of ORC6 expression with marker gene expression associated with three classes of RNA modifications (N1-methyladenosine (m1A), 5-methylcytosine (m5C) and N6-methyladenosine (m6A)) (16) across cancers was assessed using Spearman correlation analysis.

To clarify the expression of ORC6 and immune-related characteristics, we employed Spearman correlation analysis to calculate the correlation between ORC6 and 5 types of immune-related genes (chemokines, chemokine receptors, immunosimulators, immunoinhibitors, and MHC). TISIDB(http://cis.hku.hk/TISIDB/) (17) to assess the immune cell infiltration status of ORC6.

To assess the impact of immunotherapy on ORC6 expression, we analyzed the immunotherapy advanced urothelial carcinoma cohort (IMvigor210 cohort) (18). The R package ‘limma’ was utilized for differential expression analysis of the target gene in the different groups. Additionally, we accessed the CAMOIP database (https://www.camoip.net/) (19) to obtain the prognostic information of the Auslander-Melanoma (20) immunotherapy cohort and assessed the effect of ORC6 expression on the survival time of patients after immunotherapy.

The possibility of ORC6 expression as a predictive marker for immunotherapy response was analyzed using the TISMO (http://tismo.cistrome.org/) (21) and TIDE (http://tide.dfci.harvard.edu/) (22) databases. To examine the correlation between ORC6 expression and the half-inhibitory concentration (IC50) of the drug, we employed the R package ‘pRRophetic’ (23) for the analysis.

Tumor Immune Single-cell Hub (TISCH, http://tisch.comp-genomics.org/) is a scRNA-seq database that has been specifically developed to investigate the single-cell landscape of the tumor microenvironment (TME) (24). We screened single-cell datasets, including ALL_GSE132509, BRCA_GSE161529, CESC_GSE168652, CHOL_GSE138709, CRC_GSE166555, ESCA_GSE160269, HNSC_GSE103322, LIHC_GSE166635, LSCC_GSE150321, OV_GSE154600, PAAD_ CRA001160, PRAD_GSE141445, STAD_GSE134520, THCA_GSE148673 and UVM_GSE139829. UMAP plots were used for the visualization of cell types and ORC6 expression levels.

We utilized single-cell sequence data obtained from CancerSEA(http://biocc.hrbmu.edu.cn/CancerSEA/) (25) to examine the relationship between ORC6 and 14 distinct cancer functional states.

To investigate the mechanisms underlying the impact of ORC6 expression on the prognosis of tumor patients, we performed gene set enrichment analysis (GSEA) to explore the ORC6-related signaling pathways, as previously described in the literature (26, 27). We performed differentially expressed gene (DEG) analysis on the ORC6-low and ORC6-high subgroups of each cancer using the “limma” R package. The threshold was set at 30%, and genes with adjusted P values <0.05 were considered DEGs. We selected the h.all.v7.2.symbols.gmt gene set as our reference and employed it to determine the normalized enrichment score (NES) and false discovery rate (FDR). By examining the correlation between the ORC6 gene expression matrix and the known functional genome, we evaluated the impact of coordinated changes in genes within the genome on phenotypic alterations. The presented findings were visualized as bubble plots with the aid of the R package “ggplot2”. The CAMOIP (19) network server was employed to perform Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) analyses based on ORC6 expression in TCGA-LIHC transcriptome data using the R package “clusterProfiler”.

The LIHC cell lines HCCLM3 and MHCC97-H and the hepatic epithelial cell line THLE-2 were acquired from BeNa Culture Collection. The HepG2, U-251 MG, and LN229 cell lines were obtained from Procell and cultured according to the manufacturer’s instructions. Transfection was carried out in 6-well plates (NEST Biotechnology) using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol in HepG2, HCCLM3, U-251 MG, and LN229 cells. The siRNA used in this study was synthesized by GenePharma. Supplementary Table S1 lists the sequences of the siRNAs used in this study. The Western blot experimental steps were described in a previous study (28). The antibodies used in this study were anti-ORC6 (Proteintech, 17784-1-AP, 1:1000) and anti-alpha tubulin (Proteintech, 11224-1-AP, 1:5000).

Control and experimental cells were placed in 96-well plates at cell densities of 5,000 (HepG2 and HCCLM3) or 3,000 (U251 MG and LN229) cells per well, respectively. After incubation for 0, 24, 48, and 78 hours, cell viability was assessed by using the CCK-8 assay (GlpBio), and the optical density (OD) was measured at 450 nm with a microplate reader.

Cell proliferation was assessed using EdU (5-ethynyl-2’-deoxyuridine) staining. Briefly, control and experimental cells were seeded in 96-well plates and incubated overnight. After incubation with 10 μM EdU (RiboBio) for 2 hours, cells were fixed with 4% paraformaldehyde for 20 minutes, permeabilized with 0.5% Triton X-100 for 15 minutes, incubated with EdU reaction solution for 30 minutes, and finally incubated with Hoechst 33342 for 10 minutes. Images were taken using an inverted fluorescence microscope (Olympus).

Control and experimental cells were seeded onto a 6-well plate and cultured until reaching a confluence of 70% before undergoing transfection and continuing to be cultured normally. Scratch assays were performed when cell confluence reached approximately 90%, with images taken at 0 and 36 hours thereafter. For Transwell migration assays, 50,000 (HepG2 and HCCLM3) or 15,000 (U251 MG and LN229) cells were seeded into the chamber and cultured with serum-free medium within the insert and with 10% complete medium outside of the insert. After 24 hours, cells were fixed and stained, and then the cells within the insert were removed by gently swabbing with a cotton tip before imaging.

We used Student’s t-test to determine the statistical significance of differences between the two groups. Paired t-tests were conducted to compare the expression levels of ORC6 in tumor tissues with those in their paired normal tissues. We evaluated the prognostic significance of ORC6 by conducting log-rank and univariate Cox regression analyses. Spearman correlation analysis was employed to assess the correlations between ORC6 and its corresponding features. A p value < 0.05 was considered to indicate statistical significance.

We conducted a comparative analysis of ORC6 expression in tumor vs. normal tissues by merging the TCGA and GTEx databases. ORC6 was found to be significantly upregulated in 33 tumors (all p<0.05), as illustrated in Figure 1A . However, no significant changes were observed in TCGT. Then, our analysis of tumor and matched normal tissue samples from the TCGA database revealed that ORC6 expression was significantly elevated in tumor samples from BLCA, BRCA, CHOL, COAD, ESCA, HNSC, KICH, KIRC, KIRP, LIHC, LUSC, LUAD, PRAD, READ, STAD, THCA, and UCEC ( Figure 1B , all p<0.05). Further protein score hints were provided by the HPA online database. ORC6 showed high protein scores in the stomach, duodenum, colon, pancreas, lymph nodes, testis, and bone marrow but low protein scores in the liver ( Figure 1C ). Regarding its protein expression in tumors, we observed that ORC6 was moderately/highly expressed in 100% of head and neck cancer (3/3) and testicular cancer (11/11) tissues and was moderately/highly expressed in 54.5% (6/11) of liver cancer tissues ( Figure 1D ). To clarify the localization of ORC6 protein expression, we obtained immunofluorescence staining images of ORC6 protein expression in HEK 293 and PC-3 cells through the HPA database ( Supplementary Figure S1A ). Further access to the Genecards database was performed for validation ( Supplementary Figure S1B ). We found that ORC6 was mainly concentrated in the nucleus and cytoplasm. Finally, we constructed a PPI network using interaction data sourced from the ComPPI website. The results of our analysis showed that proteins found to be in close interaction with ORC6 were primarily localized within the nucleus, as depicted in Supplementary Figure S1C . In summary, we identified that ORC6 was generally highly expressed in tumors.

The frequency and type of ORC6 gene genetic alterations across cancers were analyzed by the cBioPortal platform. As illustrated in Supplementary Figure S1D , the most frequent type of genetic alteration in ORC6 was “amplification”, followed by “deep deletion”, “mutation” and “structural variation”. SARC exhibited the greatest frequency of ORC6 genetic mutations. These alterations included “deep deletions” in 2.35% of genes and “structural variants” in 0.39% of genes. In PRAD, the gene alteration frequency of ORC6 was 2.63%, of which the frequency of “amplification” reached 2.43%. In both DLBC and UCS, ORC6 genetic alterations were “deeply deleted”. In UCEC and SKCM, the frequency of ORC6 gene “mutation” reached 1.13%. In five different cancer types (ACC, ESCA, LIHC, KIRP, and PAAD), the ORC6 gene only exhibited “amplified” genetic variants. The mutation frequency of the ORC6 gene is generally low, at less than 3%. This could be due to the high conservation of genes within the ORC family (29). Subsequently, to explore the relationship between ORC6 expression and genomic variations across different cancer types, we employed either the Wilcoxon rank-sum test or the Kruskal–Wallis rank-sum test. We observed differential expression of ORC6 across three distinct variant groups (gain-variant, loss-variant, and no-variant) in 14 different cancers: BLCA, BRCA, CESC, COAD, HNSC, KIPAN, LIHC, LUAD, LUSC, OV, PRAD, STAD, STES, and UCS ( Supplementary Figure S1E , all p<0.05). Specifically, ORC6 expression was generally higher in the gain-variant group than in the loss/neutral-variant group.

Numerous reports suggest that abnormal DNA methylation in the promoter region of genes can induce changes in chromatin structure and DNA stability, ultimately leading to the dysregulation of gene expression within the body (30). Therefore, we analyzed differences in the DNA promoter methylation levels of ORC6 between tumor and normal tissues using UALCAN. As depicted in Supplementary Figure S2A , methylation levels were lower in BLCA, BRCA, HNSC, THYM, UCEC, and PRAD than in normal tissues (all p<0.05). In contrast, methylation levels were higher in PAAD, KIRC, LUSC, and SARC than in normal tissues ( Supplementary Figure S2A , all p<0.05). Moreover, RNA modifications are critical in selectively regulating the expression of genes (31). Our analysis, as illustrated in Supplementary Figure S2B , reveals a strong positive correlation between ORC6 expression and m1A-, m5C-, and m6A-related genes across almost all tumor types. These findings suggest that the ubiquitous overexpression of ORC6 in tumors may be closely associated with its epigenetic modifications and genetic variations. This correlation further supports the potential of ORC6 as a cancer regulatory factor and provides valuable clues for further exploring its role in cancer.

We also investigated the correlations between ORC6 expression and various clinicopathological features. According to the results presented in Figure 2A , there was a positive correlation between ORC6 expression and lymph node metastasis in several tumor types, and the correlations in HNSC (p=2.2e-4), KIPAN (p=6.8e-8), KIRC (p=5.6e-3), PRAD (p=4.0e-8), THCA (p=0.02) and other tumor types were the most robust. Figure 2B shows that the increase in ORC6 expression was positively correlated with tumor metastasis in ACC (p=8.2e-3), KIPAN (p=1.9e-3), KIRC (p=4.3e-4), LUAD (p=9.5e-3), PRAD (p=0.01), and SKCM (p=0.03). Furthermore, increased ORC6 expression was positively correlated with the T stage of ACC (p=4.8e-8), KIPAN (p=1.1e-9), KIRC (p=1.4e-5), KIRP (p=2.4e-8) and PRAD (p =7.0e-12) ( Figure 2C ). Figure 2D shows that the increased expression of ORC6 was positively correlated with the histological grade of GBMLGG (p=9.2e-25), HNSC (p=5.6e-9), LGG (p=9.2e-25), LIHC (p=3.0e-15) and PAAD (p=2.9e-9) but negatively correlated with the histological grade of STES (p=6.3e-5). Similarly, increased ORC6 expression was positively correlated with clinical staging ( Figure 2E ), and typical tumor types were ACC (p=6.3e-5), HNSC (p=1.5e-3), KIPAN (p=3.3e-9), KIRC (p=2.4e-4), and LUAD (p=0.01). From the above findings, it can be inferred that ORC6 might play a role in tumor progression and metastasis.

To investigate the effect of ORC6 on tumor prognosis, we plotted survival curves and assessed OS using the Kaplan–Meier method. As shown in Figure 3 , in ACC, BRCA, GBMLGG, HNSC, KICH, KIRC, KIRP, LGG, LIHC, LUAD, LUADLUSC, MESO, OSCC, PAAD, PCPG, SARC and UCEC patients, high ORC6 levels were highly correlated with poorer OS ( Figure 3A ; all p<0.05). Moreover, it is worth noting that high ORC6 expression was exclusively correlated with improved OS in OV ( Figure 3B ; p<0.05).

Following this, we conducted Cox regression analysis to assess the correlation between ORC6 expression and several survival outcomes, including OS, DSS, DFI, and PFI, for each tumor type. The results were then presented in the form of a forest plot. As shown in Figure 4A , our findings indicate that high ORC6 expression was significantly associated with shorter OS in GBMLGG, KIPAN, KIRP, LGG, ACC, KIRC, MESO, LIHC, PCPG, PRAD, PAAD, KICH, LUAD, UVM, BRCA and HNSC patients (all p<0.05). High ORC6 expression in GBMLGG, KIPAN, KIRP, KIRC, ACC, LGG, MESO, LIHC, PRAD, KICH, PCPG, BRCA, PAAD, UVM, and LUAD patients was associated with poorer DSS ( Figure 4B , all p<0.05). Regarding DFI, there was a significant association between high ORC6 expression and lower DFI in KIRP, KIPAN, LIHC, BRCA, PRAD, PAAD, SARC, THCA, MESO, and LUAD patients, whereas high ORC6 expression in OV patients was associated with improved prognosis for DFI ( Figure 4C , all p<0.05). Furthermore, as illustrated in Figure 4D , high ORC6 levels were strongly correlated with poorer PFI in GBMLGG, KIPAN, PRAD, KIRP, ACC, LIHC, LGG, UVM, KIRC, KICH, PAAD, BRCA, MESO, BLCA and HNSC (all p<0.05).

Moreover, we assessed the diagnostic accuracy of ORC6 in different types of cancer using ROC curves. As shown in Figure 5 , in ACC, BLCA, BRCA, CESC, CHOL, COAD, COADREAD, ESAD, ESCA, GBM, HNSC, LAML, LIHC, LUAD, LUADLUSC, LUSC, OSCC, OV, PAAD, READ, STAD, UCEC and UCS, ORC6 could be used as a highly accurate diagnostic marker ( Figures 5A–W , all AUC>0.85). In DLBC, GBMLGG, KICH, KIRC, KIRP, SKCM, and THYM, ORC6 had moderate diagnostic performance ( Supplementary Figures S3A–E, H, K ; all AUC=0.7~0.85). In LGG, PRAD, TGCT, and THCA, ORC6 had poor diagnostic values ( Supplementary Figures S3F, G, I, J ; all AUC=0.5~0.7). In conclusion, our study revealed that high expression of ORC6 is generally linked to unfavorable prognosis in the majority of cancer types, and it has good diagnostic value.

To explore the possible biological pathways influenced by ORC6 that may contribute to tumorigenesis and progression, we conducted GSEA on data obtained from 33 tumors from TCGA. As illustrated in Figure 6A , we observed that immune-related pathways, including TNFα signaling via NFκB, IFN-α response, IFN-γ response, inflammatory response, IL-6/JAK/STAT3, IL-2/STAT5, complement and coagulation cascades, and allograft rejection pathways, were significantly enriched across a diverse range of tumors. Moreover, we observed a positive correlation between ORC6 expression and MYC target V2, MYC target V1, MTORC1, mitotic spindle, G2 checkpoint, E2F target, DNA repair, and other pathways across cancers. Furthermore, KEGG analysis revealed that ORC6 was mainly involved in the synthesis and degradation of various substances, drug metabolism, the cell cycle, ferroptosis, and neuroactive ligand-receptor interactions in LIHC ( Figure 6B ). GO analysis, including the BP, CC, and MF categories, indicated that ORC6 was mainly related to immune response regulation and biological enzyme activity in LIHC ( Figures 6C–E ).

To explore ORC6 expression in diverse TMEs, including ALL, BRCA, CESC, CHOL, CRC, ESCA, HNSC, LIHC, LSCC, OV, PAAD, PRAD, STAD, THCA, and UVM, we investigated their expression distribution ( Figures 7A–O ). The results were interesting, as they showed that ORC6 was primarily expressed at high levels in the malignant cells of these cancers. In STAD, it was predominantly expressed in pit mucus cells ( Figure 7M ). It is worth noting that in LIHC, ORC6 was also found to be expressed in T-cell proliferation, which demonstrates its potential role in this immune response ( Figures 7H, P ). To further investigate the relationship between ORC6 and the functional status of different cancers, we analyzed single-cell sequencing data obtained from CancerSEA for 14 types of cancer. In most tumors, ORC6 showed a positive correlation with the cell cycle, proliferation, DNA damage, and DNA repair ( Supplementary Figure S4 ). In contrast, ORC6 was negatively associated with apoptosis, hypoxia, metastasis, and quiescence in most tumors ( Supplementary Figure S4 ). These findings suggest a correlation between abnormal expression of ORC6 and the advancement of cancer as well as the immune response of cancer.

Investigating the possible gene expression within tumors and its connection to immune cells can greatly aid in predicting the clinical outcome for patients with tumors and selecting appropriate diagnostic targets and intervention strategies (32). To gain further insights into the correlation between ORC6 and immune cells in the TME, we used the TISIDB tool for analysis. The pan-cancer analysis indicated that the expression level of ORC6 displayed an inverse correlation with the infiltration abundance of various immune cells, including Tem CD8 cells, Th1 cells, NK cells, pDCs, iDCs, eosinophils, monocytes, and neutrophils, while showing a positive correlation with the infiltration abundance of Act CD8 cells and Th2 cells ( Figure 8A ). This phenomenon was especially evident in LIHC ( Figures 8B–K ). Furthermore, in gliomas, which include GBM and LGG, ORC6 expression was negatively correlated with the abundance of infiltrating immune cells, including Tem CD8, Tcm CD4, Tfh, Th1, Th17, Act B, lmm B, NK, MDSC, NKT, Act DC, pDC, iDC, macrophage, eosinophil, mast, monocyte, and neutrophil cells. In contrast, ORC6 expression levels in THCA and KIRC were positively correlated with the majority of immune cell infiltration ( Figure 8A ).

Furthermore, we examined the association between ORC6 expression and the expression of genes related to immune regulation ( Figures 9A–E ). The heatmap indicated that ORC6 was coexpressed with most immune-related genes across cancers. Especially in DLBC, UVM, LIHC, KIRC, and THCA, ORC6 was roughly positively correlated with 5 immune-related genes. In TGCT, GBM, and LUSC, ORC6 was roughly negatively correlated with immune-related genes. Furthermore, the chemokine CCL14 was negatively correlated with ORC6 expression across cancers ( Figure 9A ). Among the immune activation-related genes, MICB, PVR, ULBP1, CD276, and TNFRSF25 were positively correlated with ORC6 expression in most tumors ( Figure 9C ). There was a positive correlation between the expression of ORC6 and genes related to immunosuppression in several cancer types, such as DLBC, UVM, LIHC, KIRC, THCA, GBMLGG, LGG, and PRAD, as indicated in Figure 9D . Notably, in the advanced urothelial carcinoma cohort with immunotherapy, the response group had significantly higher ORC6 expression ( Figure 9F , p=0.00099). Furthermore, we observed that in the melanoma immunotherapy cohort, the high ORC6 expression group had prolonged survival after immunotherapy ( Figure 9G , p=0.027). Taken together, these results suggest that ORC6 may be involved in immune cell infiltration and the expression of immunomodulatory genes and that high ORC6 expression may indicate a better response to immunotherapy.

To elucidate the predictive function of ORC6 expression in immunotherapy response, we evaluated it using the TISMO database. As depicted in Figure 10A , ORC6 expression was markedly different in 5 subjects before and after ICB treatment and between responder and nonresponder cohorts. Moreover, ORC6 expression was significantly different in the six cell lines before and after cytokine treatment ( Figure 10B ). Furthermore, we performed a biomarker assessment of ORC6 by TIDE. The findings indicated that ORC6 had a better predictive effect in 7 immunotherapy cohorts ( Figure 10C ). In addition, we performed a sensitivity analysis of chemotherapy drugs commonly used to treat LIHC. As shown in Figure 10D , the ORC6 high-expression group was closely correlated with the reduction of IC50 of 5-fluorouracil, doxorubicin, gemcitabine, and imatinib (all p<0.001).

After collating and analyzing the pan-cancer data mentioned above, it was observed that the elevated expression of ORC6 was significantly associated with the unfavorable prognosis and malignancy of LIHC and GBMLGG patients ( Figures 1 – 4 ). Consequently, our study will concentrate on LIHC and GBMLGG.

We collected clinical information and ORC6 expression profiles of patients belonging to the TCGA-LIHC cohort. After performing a chi-square test analysis, we discovered a significant correlation between ORC6 expression and tumor histological grade, alpha-fetoprotein (AFP) content, and vascular invasion in LIHC ( Supplementary Table S2 ). Furthermore, via univariate and multivariate Cox regression analysis, we identified ORC6 and pathological stage as independent prognostic risk factors for LIHC ( Supplementary Table S3 ). Subsequently, a nomogram was developed to estimate the survival likelihood of patients at intervals of 1, 3, and 5 years, and its prediction efficiency was confirmed by the calibration curve, as illustrated in Figures 11A , B . This indicates that the model had a high accuracy in its predictive ability. Given that our analysis of ORC6 was solely based on the TCGA database, we conducted external verification by collating clinical information and ORC6 expression profiles of patients belonging to the ICGC-LIRI-JP cohort. The results showed that ORC6 expression in LIHC was significantly higher than that in normal tissues ( Figure 11C ). It was positively correlated with the clinical stage and associated with poor overall survival ( Figures 11D, E ). Furthermore, univariate and multivariate Cox regression analyses revealed ORC6, sex, and clinical stage as independent risk factors for poor prognosis in LIHC ( Figures 11F, G ). To investigate the function of ORC6 in LIHC cells, we carried out in vitro experiments. Initially, we assessed the basal expression of ORC6 in LIHC cell lines and normal hepatocytes. Our findings indicate that LIHC cell lines have noticeably increased ORC6 expression compared to normal hepatocytes, as illustrated in Figure 12A . To assess the impact of ORC6 downregulation in LIHC cells, we chose two LIHC cell lines with high expression (HepG2 and HCCLM3) and conducted an ORC6 knockout assay ( Figures 12B, C ). Using the EdU cell proliferation test, a significant decrease in the proliferation of LIHC cells was observed after the knockout of ORC6 expression ( Figure 12D ). In addition, similar results were obtained through CCK-8 analysis ( Figures 12E, F ). Subsequently, we conducted wound healing and Transwell tests to examine the impact of ORC6 downregulation on the migratory capacity of LIHC cells. The results indicate that compared to the control cells, the downregulation of ORC6 significantly inhibited the migration ability of LIHC cells ( Figures 12G–L ).

Subsequently, we compiled the ORC6 expression and clinical information of patients in the TCGA-GBMLGG cohort. The chi-square test revealed a significant association between ORC6 expression and age, histological type, WHO grade, IDH status, and 1p/19q codeletion in GBMLGG patients ( Supplementary Table S4 ). Furthermore, our investigation revealed that ORC6, age, WHO grade, and 1p/19q codeletion were independent prognostic risk factors for GBMLGG patients, as confirmed by univariate and multivariate Cox regression analyses ( Supplementary Table S5 ). Time-dependent ROC curve analysis further identified age, WHO grade, and ORC6 as the top three effective predictors for 1-, 3-, and 5-year patient survival ( Figure 13A , all AUC>0.7). Our results were validated in an external cohort. As demonstrated in Figures 13B–D , ORC6 expression increased concomitantly with WHO grade in the CGGA-301, CGGA-325, and CGGA-693 cohorts. Moreover, high levels of ORC6 expression were significantly associated with an unfavorable prognosis among GBMLGG patients in the CGGA-301, CGGA-325, CGGA-693, and GSE13041 cohorts ( Figures 13E–H , all p < 0.05). In addition, the expression profile of ORC6 in the GBMLGG cell line was analyzed using the CCLE database ( Figure 13I ). Further, the expression of ORC6 protein was effectively inhibited in U251 MG and LN229 cells ( Figures 14A, B ). Consistent with findings in LIHC, the inhibition of ORC6 expression led to decreased proliferation and migration of U251 MG and LN229 cells ( Figures 14C–J ). In conclusion, our in vitro results strengthen the evidence supporting the carcinogenic effects of ORC6 in LIHC and GBMLGG.