Integrative Analysis of Biomarkers Through Machine Learning Identifies Stemness Features in Colorectal Cancer

Cancer Stemness Online: A Resource for Investigating Cancer Stemness and Associations with Immune Response

Cancer progression involves the gradual loss of a differentiated phenotype and the acquisition of progenitor and stem cell-like features, which are potential culprits of immunotherapy resistance. Although the state-of-the-art predictive computational methods have facilitated the prediction of cancer stemness, there remains a lack of efficient resources to accommodate various usage requirements. Here, we present the Cancer Stemness Online, an integrated resource for efficiently scoring cancer stemness potential at both bulk and single-cell levels. This resource integrates eight robust predictive algorithms as well as 27 signature gene sets associated with cancer stemness for predicting stemness scores. Downstream analyses were performed from five distinct aspects: identifying the signature genes of cancer stemness; exploring the associations with cancer hallmarks and cellular states; exploring the associations with immune response and the communications with immune cells; investigating the contributions to patient survival; and performing a robustness analysis of cancer stemness among different methods. Moreover, the pre-calculated cancer stemness atlas for more than 40 cancer types can be accessed by users. Both the tables and diverse visualizations of the analytical results are available for download. Together, Cancer Stemness Online is a powerful resource for scoring cancer stemness and expanding downstream functional interpretation, including immune response and cancer hallmarks. Cancer Stemness Online is freely accessible at http://bio-bigdata.hrbmu.edu.cn/CancerStemnessOnline.

KIF2C/MCAK a prognostic biomarker and its oncogenic potential in malignant progression, and prognosis of cancer patients: a systematic review and meta-analysis as biomarker

KIF2C/MCAK (KIF2C) is the most well-characterized member of the kinesin-13 family, which is critical in the regulation of microtubule (MT) dynamics during mitosis, as well as interphase. This systematic review briefly describes the important structural elements of KIF2C, its regulation by multiple molecular mechanisms, and its broad cellular functions. Furthermore, it systematically summarizes its oncogenic potential in malignant progression and performs a meta-analysis of its prognostic value in cancer patients. KIF2C was shown to be involved in multiple crucial cellular processes including cell migration and invasion, DNA repair, senescence induction and immune modulation, which are all known to be critical during the development of malignant tumors. Indeed, an increasing number of publications indicate that KIF2C is aberrantly expressed in multiple cancer entities. Consequently, we have highlighted its involvement in at least five hallmarks of cancer, namely: genome instability, resisting cell death, activating invasion and metastasis, avoiding immune destruction and cellular senescence. This was followed by a systematic search of KIF2C/MCAK's expression in various malignant tumor entities and its correlation with clinicopathologic features. Available data were pooled into multiple weighted meta-analyses for the correlation between KIF2Chigh protein or gene expression and the overall survival in breast cancer, non-small cell lung cancer and hepatocellular carcinoma patients. Furthermore, high expression of KIF2C was correlated to disease-free survival of hepatocellular carcinoma. All meta-analyses showed poor prognosis for cancer patients with KIF2Chigh expression, associated with a decreased overall survival and reduced disease-free survival, indicating KIF2C's oncogenic potential in malignant progression and as a prognostic marker. This work delineated the promising research perspective of KIF2C with modern in vivo and in vitro technologies to further decipher the function of KIF2C in malignant tumor development and progression. This might help to establish KIF2C as a biomarker for the diagnosis or evaluation of at least three cancer entities.

Transcription factor ZEB1 coordinating with NuRD complex to promote oncogenesis through glycolysis in colorectal cancer

Background

Colorectal cancer (CRC) is an aggressive primary intestinal malignancy with the third-highest incidence and second-highest mortality among all cancer types worldwide. Transcription factors (TFs) regulate cell development and differentiation owing to their ability to recognize specific DNA sequences upstream of genes. Numerous studies have demonstrated a strong correlation between TFs, the etiology of tumors, and therapeutic approaches. Here, we aimed to explore prognosis-related TFs and comprehend their carcinogenic mechanisms, thereby offering novel insights into the diagnosis and management of CRC.

Materials and Methods

Differentially expressed TFs between CRC and normal tissues were identified leveraging The Cancer Genome Atlas database, Weighted correlation network analysis and Cox regression analysis were performed to identify prognosis-related TFs. The cellular functions of hub TF zinc finger E-box binding homeobox 1 (ZEB1) were determined using by 5-ethynyl-2'-deoxyuridine and cell invasion assays in CRC cells. RNA-sequencing, Kyoto Encyclopedia of Genes and Genomes enrichment, and gene set enrichment analyses were used to identify the cellular processes in which ZEB1 participates. Immunoaffinity purification, silver staining mass spectrometry, and a chromatin immunoprecipitation assay were conducted to search for proteins that might interact with ZEB1 and the target genes they jointly regulate.

Results

Thirteen central TFs related to prognosis were identified through bioinformatics analysis techniques. Among these TFs, ZEB1 emerged as the TF most closely associated with CRC, as determined through a combination of regulatory network diagrams, survival curves, and phenotype analyses. ZEB1 promotes CRC cell growth by recruiting the NuRD(MTA1) complex, and the ZEB1/NuRD(MTA1) complex transcriptionally represses glycolysis-associated tumor suppressor genes.

Conclusion

Our study not only identified a hub biomarker related to CRC prognosis but also revealed the specific molecular mechanisms through which ZEB1 affects cancer progression. These insights provide crucial evidence for the diagnosis of CRC and potential treatment opportunities.

Interaction of the intestinal cytokines-JAKs-STAT3 and 5 axes with RNA N6-methyladenosine to promote chronic inflammation-induced colorectal cancer

Colorectal cancer (CRC) is one of the most common cancers, with a high mortality rate worldwide. Mounting evidence indicates that mRNA modifications are crucial in RNA metabolism, transcription, processing, splicing, degradation, and translation. Studies show that N6-methyladenosine (m6A) is mammalians' most common epi-transcriptomic modification. It has been demonstrated that m6A is involved in cancer formation, progression, invasion, and metastasis, suggesting it could be a potential biomarker for CRC diagnosis and developing therapeutics. Cytokines, growth factors, and hormones function in JAK/STAT3/5 signaling pathway, and they could regulate the intestinal response to infection, inflammation, and tumorigenesis. Reports show that the JAK/STAT3/5 pathway is involved in CRC development. However, the underlying mechanism is still unclear. Signal Transducer and Activator of Transcription 3/5 (STAT3, STAT5) can act as oncogenes or tumor suppressors in the context of tissue types. Also, epigenetic modifications and mutations could alter the balance between pro-oncogenic and tumor suppressor activities of the STAT3/5 signaling pathway. Thus, exploring the interaction of cytokines-JAKs-STAT3 and/or STAT5 with mRNA m6A is of great interest. This review provides a comprehensive overview of the characteristics and functions of m6A and JAKs-STAT3/5 and their relationship with gastrointestinal (GI) cancers.

Leveraging an immune cell signature to improve the survival and immunotherapy response of lung adenocarcinoma

Background: Immune cells play a critical role in the prognosis of cancer. However, the function of different immune cell types in lung adenocarcinoma (LUAD) and the development of a prognostic signature based on immune cell types have not been comprehensively investigated. Methods: We collected and included a total of 2499 LUAD patients and performed calculations to determine the penetration level of 24 immune cells. This examination was conducted using the macro-gene-based approach provided by ImmuCellAI. We performed a meta-analysis using Lasso-Cox analysis to establish the immune cell pair score (ICPS). We conducted a survival analysis to measure differences in survival across ICPS-risk groups. Wilcox test was used to measure the difference in expression level. Spearman correlation analysis was used for the relevance assessment. Results: We collected a total of 24 immune cell types to construct cell pairs. Utilizing 17 immune cell pairs, we constructed and validated the ICPS, which plays a critical role in stratifying survival and dynamically monitoring the effectiveness of immunotherapy. Additionally, we identified several candidate drugs that target ICPS. Conclusions: The ICPS shows promise as a valuable tool for identifying suitable candidates for immunotherapy among patients. Our comprehensive assessment of immune cell interactions in LUAD contributes to a deeper understanding of infiltration patterns and functions, thereby guiding the development of more efficacious immunotherapy strategies.

Deciphering Brain Metastasis Stem Cell Properties From Colorectal Cancer Highlights Specific Stemness Signature and Shared Molecular Features

BACKGROUND & AIMS

Brain metastases (BMs) from colorectal cancer (CRC) are associated with significant morbidity and mortality, with chemoresistance and short overall survival. Migrating cancer stem cells with the ability to initiate BM have been described in breast and lung cancers. In this study, we describe the identification and characterization of cancer stem cells in BM from CRC.

METHODS

Four brain metastasis stem cell lines from patients with colorectal cancer (BM-SC-CRC1 to BM-SC-CRC4) were obtained by mechanical dissociation of patient's tumors and selection of cancer stem cells by appropriate culture conditions. BM-SC-CRCs were characterized in vitro by clonogenic and limiting-dilution assays, as well as immunofluorescence and Western blot analyses. In ovo, a chicken chorioallantoic membrane (CAM) model and in vivo, xenograft experiments using BALB/c-nude mice were realized. Finally, a whole exome and RNA sequencing analyses were performed.

RESULTS

BM-SC-CRC formed metaspheres and contained tumor-initiating cells with self-renewal properties. They expressed stem cell surface markers (CD44v6, CD44, and EpCAM) in serum-free medium and CRC markers (CK19, CK20 and CDX-2) in fetal bovine serum-enriched medium. The CAM model demonstrated their invasive and migratory capabilities. Moreover, mice intracranial xenotransplantation of BM-SC-CRCs adequately recapitulated the original patient BM phenotype. Finally, transcriptomic and genomic approaches showed a significant enrichment of invasiveness and specific stemness signatures and highlighted KMT2C as a potential candidate gene to potentially identify high-risk CRC patients.

CONCLUSIONS

This original study represents the first step in CRC BM initiation and progression comprehension, and further investigation could open the way to new therapeutics avenues to improve patient prognosis.

Integrated analysis of multi-omics data for the discovery of biomarkers and therapeutic targets for colorectal cancer

The considerable burden of colorectal cancer and the rising trend in young adults emphasize the necessity of understanding its underlying mechanisms, providing new diagnostic and prognostic markers, and improving therapeutic approaches. Precision medicine is a new trend all over the world and identification of novel biomarkers and therapeutic targets is a step forward towards this trend. In this context, multi-omics data and integrated analysis are being investigated to develop personalized medicine in the management of colorectal cancer. Given the large amount of data from multi-omics approach, data integration and analysis is a great challenge. In this Review, we summarize how statistical and machine learning techniques are applied to analyze multi-omics data and how it contributes to the discovery of useful diagnostic and prognostic biomarkers and therapeutic targets. Moreover, we discuss the importance of these biomarkers and therapeutic targets in the clinical management of colorectal cancer in the future. Taken together, integrated analysis of multi-omics data has great potential for finding novel diagnostic and prognostic biomarkers and therapeutic targets, however, there are still challenges to overcome in future studies.

mRNAsi-related metabolic risk score model identifies poor prognosis, immunoevasive contexture, and low chemotherapy response in colorectal cancer patients through machine learning

Colorectal cancer (CRC) is one of the most fatal cancers of the digestive system. Although cancer stem cells and metabolic reprogramming have an important effect on tumor progression and drug resistance, their combined effect on CRC prognosis remains unclear. Therefore, we generated a 21-gene mRNA stemness index-related metabolic risk score model, which was examined in The Cancer Genome Atlas and Gene Expression Omnibus databases (1323 patients) and validated using the Zhongshan Hospital cohort (200 patients). The high-risk group showed more immune infiltrations; higher levels of immunosuppressive checkpoints, such as CD274, tumor mutation burden, and resistance to chemotherapeutics; potentially better response to immune therapy; worse prognosis; and advanced stage of tumor node metastasis than the low-risk group. The combination of risk score and clinical characteristics was effective in predicting overall survival. Zhongshan cohort validated that high-risk score group correlated with malignant progression, worse prognosis, inferior adjuvant chemotherapy responsiveness of CRC, and shaped an immunoevasive contexture. This tool may provide a more accurate risk stratification in CRC and screening of patients with CRC responsive to immunotherapy.

Development and validation of a hypoxia-stemness-based prognostic signature in pancreatic adenocarcinoma

Background: Pancreatic adenocarcinoma (PAAD) is one of the most aggressive and fatal gastrointestinal malignancies with high morbidity and mortality worldwide. Accumulating evidence has revealed the clinical significance of the interaction between the hypoxic microenvironment and cancer stemness in pancreatic cancer progression and therapies. This study aims to identify a hypoxia-stemness index-related gene signature for risk stratification and prognosis prediction in PAAD.

Methods: The mRNA expression-based stemness index (mRNAsi) data of PAAD samples from The Cancer Genome Atlas (TCGA) database were calculated based on the one-class logistic regression (OCLR) machine learning algorithm. Univariate Cox regression and LASSO regression analyses were then performed to establish a hypoxia-mRNAsi-related gene signature, and its prognostic performance was verified in both the TCGA-PAAD and GSE62452 corhorts by Kaplan-Meier and receiver operating characteristic (ROC) analyses. Additionally, we further validated the expression levels of signature genes using the TCGA, GTEx and HPA databases as well as qPCR experiments. Moreover, we constructed a prognostic nomogram incorporating the eight-gene signature and traditional clinical factors and analyzed the correlations of the risk score with immune infiltrates and immune checkpoint genes.

Results: The mRNAsi values of PAAD samples were significantly higher than those of normal samples (p < 0.001), and PAAD patients with high mRNAsi values exhibited worse overall survival (OS). A novel prognostic risk model was successfully constructed based on the eight-gene signature comprising JMJD6, NDST1, ENO3, LDHA, TES, ANKZF1, CITED, and SIAH2, which could accurately predict the 1-, 3-, and 5-year OS of PAAD patients in both the training and external validation datasets. Additionally, the eight-gene signature could distinguish PAAD samples from normal samples and stratify PAAD patients into low- and high-risk groups with distinct OS. The risk score was closely correlated with immune cell infiltration patterns and immune checkpoint molecules. Moreover, calibration analysis showed the excellent predictive ability of the nomogram incorporating the eight-gene signature and traditional clinical factors.

Conclusion: We developed a hypoxia-stemness-related prognostic signature that reliably predicts the OS of PAAD. Our findings may aid in the risk stratification and individual treatment of PAAD patients.

Identification and Validation of a Potential Stemness-Associated Biomarker in Hepatocellular Carcinoma

Background

Cancer stem cells (CSCs) are typically related to metastasis, recurrence, and drug resistance in malignant tumors. However, the biomarker and mechanism of CSCs need further exploration. This study is aimed at comprehensively depicting the stemness characteristics and identify a potential stemness-associated biomarker in hepatocellular carcinoma (HCC).

Methods

The data of HCC patients from The Cancer Genome Atlas (TCGA) were collected and divided based on the mRNA expression-based stemness index (mRNAsi) in this study. Weighted gene coexpression network analysis (WGCNA) and the protein-protein interaction (PPI) network were performed, and the genes were screened through the Cytoscape software. Then, we constructed a prognostic expression signature using the multivariable Cox analysis and verified using the GEO and ICGC databases. Even more importantly, we used the three-dimensional (3D) fibrin gel to enrich the tumor-repopulating cells (TRCs) to validate the expression of the signature in CSCs by quantitative RT-PCR.

Results

mRNAsi was significantly elevated in tumor and high-mRNAsi score was associated with poor overall survival in HCC. The positive stemness-associated (blue) module with 737 genes were screened based on WGCNA, and Budding uninhibited by benzimidazoles 1 (BUB1) was identified as the hub gene highly related to stemness in HCC. Then, the prognostic value and stemness characteristics were well validated in the ICGC and GSE14520 cohorts. Further analysis showed the expression of BUB1 was elevated in TRCs.

Conclusion

BUB1, as a potential stemness-associated biomarker, could serve as a therapeutic CSCs-target and predicted the clinical outcomes of patients with HCC.

Characterization of stem cell landscape and identification of stemness-relevant prognostic gene signature to aid immunotherapy in colorectal cancer

Background

It is generally accepted that colorectal cancer (CRC) originates from cancer stem cells (CSCs), which are responsible for CRC progression, metastasis and therapy resistance. The high heterogeneity of CSCs has precluded clinical application of CSC-targeting therapy. Here, we aimed to characterize the stemness landscapes and screen for certain patients more responsive to immunotherapy.

Methods

Twenty-six stem cell gene sets were acquired from StemChecker database. Consensus clustering algorithm was applied for stemness subtypes identification on 1,467 CRC samples from TCGA and GEO databases. The differences in prognosis, tumor microenvironment (TME) components, therapy responses were evaluated among subtypes. Then, the stemness-risk model was constructed by weighted gene correlation network analysis (WGCNA), Cox regression and random survival forest analyses, and the most important marker was experimentally verified.

Results

Based on single-sample gene set enrichment analysis (ssGSEA) enrichments scores, CRC patients were classified into three subtypes (C1, C2 and C3). C3 subtype exhibited the worst prognosis, highest macrophages M0 and M2 infiltrations, immune and stromal scores, and minimum sensitivity to immunotherapies, but was more sensitive to drugs like Bosutinib, Docetaxel, Elesclomol, Gefitinib, Lenalidomide, Methotrexate and Sunitinib. The turquoise module was identified by WGCNA that it was most positively correlated with C3 but most negatively with C2, and five hub genes in turquoise module were identified for stemness model construction. CRC patients with higher stemness scores exhibited worse prognosis, more immunosuppressive components in TME and lower immunotherapeutic responses. Additionally, the model’s immunotherapeutic prediction efficacy was further confirmed from two immunotherapy cohorts (anti-PD-L1 in IMvigor210 cohort and anti-PD-1 in GSE78220 cohort). Mechanistically, Gene Set Enrichment Analysis (GSEA) results revealed high stemness score group was enriched in interferon gamma response, interferon alpha response, P53 pathway, coagulation, apoptosis, KRAS signaling upregulation, complement, epithelial–mesenchymal transition (EMT) and IL6-mediated JAK-STAT signaling gene sets.

Conclusions

Our study characterized three stemness-related subtypes with distinct prognosis and TME patterns in CRC patients, and a 5-gene stemness-risk model was constructed by comprehensive bioinformatic analyses. We suggest our stemness model has prospective clinical implications for prognosis evaluation and might facilitate physicians selecting prospective responders for preferential use of current immune checkpoint inhibitors.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13287-022-02913-0.