Single-Cell RNA Sequencing Reveals the Role of Epithelial Cell Marker Genes in Predicting the Prognosis of Colorectal Cancer Patients

Shen, Kai-yu; Chen, Bin-yu; Gao, Wen-cang

doi:https://doi.org/10.1155/2022/8347125

Disease Markers

On this page

Abstract Introduction Materials and Methods Results Discussion Conclusion Data Availability Conflicts of Interest Acknowledgments Supplementary Materials References Copyright Related Articles

Special Issue

High Resolution Omics for Biomarker Discovery

View this Special Issue

Research Article | Open Access

Volume 2022 | Article ID 8347125 | https://doi.org/10.1155/2022/8347125

Single-Cell RNA Sequencing Reveals the Role of Epithelial Cell Marker Genes in Predicting the Prognosis of Colorectal Cancer Patients

Kai-yu Shen,¹Bin-yu Chen,¹and Wen-cang Gao²

Academic Editor: Qiaoli Wang

Received13 May 2022

Revised09 Jul 2022

Accepted21 Jul 2022

Published01 Aug 2022

Abstract

Single-cell RNA sequencing (scRNA-seq) is increasingly used in studies on gastrointestinal cancers. This study investigated the prognostic value of epithelial cell-associated biomarkers in colorectal cancer (CRC) using scRNA-seq data. We downloaded and analysed scRNA-seq data from four CRC samples from the Gene Expression Omnibus (GEO), and we identified marker genes of malignant epithelial cells (MECs) using CRC transcriptome and clinical data downloaded from The Cancer Genome Atlas (TCGA) and GEO as training and validation cohorts, respectively. In the TCGA training cohort, weighted gene correlation network analysis, univariate Cox proportional hazard model (Cox) analysis, and least absolute shrinkage and selection operator regression analysis were performed on the marker genes of MEC subsets to identify a signature of nine prognostic MEC-related genes (MECRGs) and calculate a risk score based on the signature. CRC patients were divided into high- and low-risk groups according to the median risk score. We found that the MECRG risk score was significantly correlated with the clinical features and overall survival of CRC patients, and that CRC patients in the high-risk group showed a significantly shorter survival time. The univariate and multivariate Cox regression analyses showed that the MECRG risk score can serve as an independent prognostic factor for CRC patients. Gene set enrichment analysis revealed that the MECRG signature genes are involved in fatty acid metabolism, p53 signalling, and other pathways. To increase the clinical application value, we constructed a MECRG nomogram by combining the MECRG risk score with other independent prognostic factors. The validity of the nomogram is based on receiver operating characteristics and calibration curves. The MECRG signature and nomogram models were well validated in the GEO dataset. In conclusion, we established an epithelial cell marker gene-based risk assessment model based on scRNA-seq analysis of CRC samples for predicting the prognosis of CRC patients.

1. Introduction

Colorectal cancer (CRC) is among the most common malignant tumours of the digestive system worldwide, and its high incidence and mortality are second only to those of lung cancer and breast cancer [1, 2]. Unhealthy lifestyle habits, such as smoking, drinking, and consuming a high-fat diet, have led to an increase in CRC incidence [3], with 1.5% annual increase reported in people aged 30 to 39 years from 2007 to 2016 [4]. At present, tumour node metastasis (TNM) staging, histopathology, and completion of surgical resection are mainly evaluated to determine CRC prognosis, and molecular markers are widely used for CRC diagnosis and treatment. Various medical treatments, including surgery, postoperative adjuvant chemoradiotherapy, and molecular targeted therapy, have been used to treat CRC [5]. CRC symptoms are generally minor in early disease stages, and therefore, patients are often at an advanced stage of the disease by the time they seek medical attention. Furthermore, the optimal treatment timing can easily be missed in cases of advanced, metastatic, and recurrent CRC, and the effects of conventional treatment are usually inconsistent, resulting in variable prognoses.

As a malignant tumour of epithelial origin, the development and progression of CRC are closely related to epithelial tissue [6, 7]. Normal epithelial cells have antitumor activities and are able to eliminate oncogenic transformed cells by regulating cytoskeletal proteins [8]. According to Lv et al., epithelial cells secrete periostin, which inhibits the growth of gastric cancer cells by stabilising P53 and E-cadherin proteins [9]. Epithelial cells exhibit apical-basal polarity and cell–cell adhesion. Correct regulation of polarity is essential to inhibit tumour growth [10]. Royer and Lu suggested that the malignant transformation of epithelial cells in the presence of oncogene activation is generally closely associated with the loss of cell polarity and disorganisation and that the disruption of epithelial cell polarity promotes epithelial-mesenchymal transition (EMT), which is a key step in the invasion of the surrounding stroma by epithelial tumour cells [11]. In addition, under hypoxic conditions, pulmonary epithelial cells can downregulate connexin (CX)26 and CX43 via the P53 signalling pathway to promote lung carcinogenesis [12]. Epithelial cells are both the structural basis of most organ tissues in the body and the source of most human tumours [13]. With the progressive research on epithelial cells, epithelial cell-related biomarkers have become a research hotspot in recent years. Keratins are widely used as epithelial cell biomarkers in the pathological diagnosis of tumours and in predicting survival prognosis. In CRC patients, decreased expression of keratin (K)8 and K20 is closely related to EMT and suggestive of poor survival and high tumour invasion [14]. Increased serum levels of a cleaved K18 fragment produced by apoptotic epithelial cells suggest a high risk of CRC recurrence within 3 years [15]. However, because of the uncontrollable pathological types, stages, and metastases of CRC [16], using conventional biomarkers to predict prognosis does not achieve adequate results. Therefore, to explore more epithelial cell-related biomarkers with clinical application potential, we used high-resolution omics tools to perform a more accurate analysis of CRC epithelial cells.

Single-cell RNA sequencing (scRNA-seq) allows the construction of a gene-regulatory network at the cellular level [17]. Analysis of the genome, transcriptome, or epigenome of single cells individually or simultaneously enables the detection of gene expression profiles and tracking of cell development at the single-cell level [18]. The RNA-seq technology has been widely used to evaluate the tumour immune microenvironment [19, 20]; however, few studies have applied this method to CRC epithelial cells to predict CRC prognosis.

In this study, we identified marker genes of CRC epithelial cell subsets using scRNA-seq analysis, determined the prognostic significance of nine malignant epithelial cell- (MEC-) related genes (MECRGs) using data from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO), and integrated the MECRG-based risk score and clinical traits to construct a nomogram for the prediction of CRC patient prognosis.

2. Materials and Methods

2.1. Data Availability

ScRNA-seq data of four CRC samples were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) (GSE161277). Relevant clinical information related to the CRC samples is provided in Supplementary Table S1. The dataset comprises four CRC samples, four adenoma samples, three normal tissue samples, one paracancer sample, and one blood sample [21]. Tissue samples other than the CRC samples were not considered in this study. Transcriptome data of 473 CRC samples and 41 nontumour tissues were obtained from the TCGA database (https://portal.gdc.cancer.gov/), including 435 CRC samples with matched clinical data. The extracted clinical information included sex, age, and stage. TCGA-CRC samples with complete clinical information were used to construct the training set used to develop a prediction score. A total of 232 CRC samples with complete clinical information obtained from the GEO database (GSE17538) were used as the external validation set.

2.2. scRNA-seq Data Processing

The scRNA-seq data of the four CRC samples were processed using the R language (v4.1.0; https://www.r-project.org/) and the “Seurat” package [22] (https://cran.r-project.org/web/packages/Seurat/index.html). The “harmony” package was used to remove batch effects from the scRNA-seq data of the four CRC samples. First, we determined the percentage of mitochondrial genes in each cell using the “PercentageFeatureSet” function with the parameter set to pattern = “^MT-”. Subsequently, using the “subset” function, genes expressed in <10 cells and cells expressing <200 genes were eliminated. Further, we excluded noncells and cell aggregates. Cell samples with a mitochondrial gene % were included in the subsequent analysis and logarithmically normalized. Principal component analysis (PCA) was used for unsupervised clustering, and the “JackStraw” function was used to determine and visualise the number of principal components. We used nonlinear dimension reduction (t-distributed stochastic neighbour embedding; t-SNE) clustering and the “FindAllMarkers” function (parameters: , log function ) to identify marker genes based on between-cluster differences. The cell clusters were then annotated based on reported cell-specific marker genes [23–29].

2.3. Analysis of Chromosome Copy Number Variation (CNV) and Screening of MEC Marker Genes

To distinguish epithelial cell malignancy, we identified chromosome CNV in each sample using the “infercnv” package [30] (https://github.com/broadinstitute/inferCNV; default parameters). Using immune cells as a reference [31], we calculated CNV scores for epithelial cell subsets and defined the epithelial cell subset showing a median CNV as the MEC subset, which was used for subsequent analysis. Differences in CNV scores between epithelial cell clusters were compared using the Kruskal–Wallis test. We then identified marker genes in the MEC subset based on a and a false discovery rate .

2.4. Analysis of Differential Gene Enrichment of Epithelial Cell Subsets

The potential biological mechanisms of the MEC marker genes were determined using the “clusterProfiler” (https://bioconductor.org/packages/release/bioc/html/clusterProfiler.html) and “org.Hs.eg.db” (https://www.bioconductor.org/packages/release/data/annotation/html/org.Hs.eg.db.html) packages. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed, using a value < 0.05 to determine statistically significant enrichment.

2.5. MEC Marker Gene Analysis Using Weighted Gene Correlation Network Analysis (WGCNA)

WGCNA can be used to find modules of correlated genes and identify disease-related biomarkers. We used the WGCNA R package [32] (https://horvath.genetics.ucla.edu/html/CoexpressionNetwork/Rpackages/WGCNA/) to identify MEC marker genes related to CRC. We generated a similarity matrix between MEC marker genes using the Pearson correlation analysis and then calculated an adjacency matrix and constructed a topological overlap matrix. Next, we plotted a tree diagram of modules and clustered closely related MEC marker genes within the modules using hierarchical clustering. The MEC marker genes in the final correlation module () were used for subsequent analysis.

2.6. Construction and Validation of the MECRG Signature

The univariate Cox analysis was used to find MECRGs significantly related to overall survival (OS) in the TCGA-CRC cohort, and a forest map was plotted. MECRGs were identified using least absolute shrinkage and selection operator (LASSO) analysis, using 10-fold cross-validation and 1,000 iterations to determine the minimum value of the penalty parameter (λ) and construct a MECRG signature. The regression coefficients of the nine MECRGs thus identified were calculated using the “predict” function. The following formula was used to calculate the risk score: where “” is the regression coefficient and “” describes the expression levels of the MECRGs. Patients from the TCGA-CRC training cohort were divided into high- and low-risk groups according to the median value of the MECRG risk score. A Kaplan–Meier survival curve was used to explore differences in survival and prognosis between the two groups. We then used a receiver operating characteristic (ROC) curve to evaluate the predictive value of the MECRG signature. Finally, the signature was validated using the GSE17538 GEO dataset.

2.7. Analysis of the Prognostic Accuracy of the MECRG Risk Score

Combining the MECRG risk score with clinical features, we used univariate and multivariate Cox analyses to assess whether the risk score could serve as an independent prognostic factor. Using the training and validation sets, we then performed a survival analysis of the MECRG signature for patients of different clinical subgroups. Additionally, we identified the relationships between MECRG risk groups and clinical traits (including sex, age, and TNM stage) and generated a heat map.

2.8. Gene Set Enrichment Analysis (GSEA) of the MECRG Signature

Patients in the TCGA-CRC training cohort were grouped according to the median value of the risk score and we used GSEA to evaluate the functions of and signalling pathways associated with MECRGs in the high- and low-risk groups, using as a threshold. The “AUCell” package (https://www.rdocumentation.org/packages/msigdbr/versions/7.4.1) was used to present the results of enrichment analysis according to cell subsets.

2.9. Construction of a MECRG Nomogram

Using the training set, we integrated the three independent prognostic factors of age, stage, and MECRG risk group to plot a MECRG nomogram capable of predicting the 1-, 3-, and 5-year OS of CRC patients for clinical application. ROC and calibration curves were used to evaluate the predictive value of the MECRG nomogram, and feasibility was confirmed by external validation using the GSE17538 GEO dataset.

2.10. Statistical Analysis

The “survival” (https://cran.r-project.org/web/packages/survival/index.html) and “survminer” (https://cran.r-project.org/web/packages/survminer/index.html) packages were used to construct Kaplan–Meier survival curves. The log-rank test was used to determine significant differences in survival between the high- and low-risk groups according to the training and validation datasets. LASSO regression was performed using the “glmnet” package (https://cran.r-project.org/web/packages/glmnet/index.html), and the “timeROC” package (https://cran.r-project.org/web/packages/timeROC/index.html) was used to generate the ROC curve to evaluate model accuracy. The univariate and multivariate Cox regression analyses were used to determine the independent predictors of OS outcomes in CRC patients. The “rms” package (https://cran.r-project.org/web/packages/rms/) was used to construct the nomogram model. Comparisons between two groups were made using the Wilcoxon rank sum test, and comparisons between multiple groups were made using the Kruskal–Wallis test. Correlation analysis was conducted using the Pearson method. was considered significant.

3. Results

3.1. CRC scRNA-seq Data Analysis

We obtained 15,465 cells from the four CRC samples. After applying the screening criteria, 8,798 high-quality cell samples were obtained. Gene numbers (nFeature_RNA), sequencing depth (nCount_RNA), and mitochondrial gene percentage (percent.mt) are shown in Supplementary Figure S1. The Pearson correlation coefficient between gene count and sequencing depth was 0.92 (Supplementary Figure S2), suggesting a positive correlation. PCA used to classify the high-quality cells identified 40 principal components (Supplementary Figure S3), and t-SNE of the first 11 principal components () (Supplementary Figure S4) allowed visualisation of the high-dimensional CRC scRNA-seq data and the distribution of the cell subsets, as well as classification of the cells into 18 subclasses (Figures 1(a) and 1(b)). Among the 18 subclasses, clusters 0, 4, 7, 9, 10, and 11 were identified as epithelial cell subtypes 1, 2, 3, 4, 5, and 6, respectively, based on the presence of epithelial cell marker genes (EPCAM, KRT19, and CDH1), clusters 2, 12, 13, 15, and 17 were identified as B cell subsets (CD79A, MS4A1, and CD19), clusters 1, 3, 6, and 8 were identified as T cell subsets (CD3D, CD8A, and CD3E), clusters 5 and 16 were identified as macrophage subsets (CD14, CD68, and CD163), and cluster 14 was identified as an endothelial cell subset (IL3RA, SERPINF1, and CCDC88A). For the purpose of selecting epithelial cell subsets with high malignancy, we did not merge the epithelial cell subsets. Based on the above annotation effects, we summarised the final cell subpopulation annotations in Figure 2(a). In addition, we calculated the proportions of the five cell types in the four CRC samples, which revealed that the epithelial cell subpopulation accounted for a relatively large fraction (Figure 2(b)). The five most strongly expressed marker genes in each cell subset are shown in Figure 2(c).

(a)

(b)

(a)

(b)

(c)

3.2. Evaluation of Chromosome CNV in the Epithelial Cell Subsets

We next determined the chromosome CNV in each sample based on the transcriptome data to evaluate the degree of malignancy in the epithelial cell subsets. We observed low CNV in the immune cell subsets (macrophages, B cells, and T cells) in control samples, whereas high CNV was observed in epithelial cells. Chromosome amplification mainly occurred in chromosomes 7, 8, 9, 12, 13, 16, 19, 20, and 21, and deletions were most prevalent in chromosomes 4, 6, 8, 12, 14, 15, 17, 18, 19, and 22 (Figure 3(a)). The MEC subpopulation was screened using a median CNV as the threshold. Notably, the CNV scores for epithelial cell subtypes 1 through 5 were more significant than epithelial cell subtype 6 (Figure 3(b) and Supplementary Table S2), suggesting a higher degree of malignancy of CRC lesions associated with these cell subsets. Therefore, we defined these as MEC subsets, and, using and as thresholds, we identified 1,259 marker genes (Supplementary Table S3), and we speculated that 1259 MEC marker genes function differently in CRC than in normal colonic epithelial cells and therefore require further analysis.

(a)

(b)

3.3. GO and KEGG Enrichment Analyses of MEC Subset Marker Genes

We performed GO function (Figure 4(a)) and KEGG pathway (Figure 4(b)) enrichment analyses of the 1,259 marker genes in the five MEC subsets to determine their possible biological functions. We found that the differentially expressed genes in the MEC subsets were mainly involved in processes associated with ATP metabolism, focal adhesion, formation of cell–substrate junctions, cadherin binding, formation of adherens junctions, reactive oxygen species (ROS), and thyroid cancer. Notably, cell–substrate junction formation, calmodulin binding, and adhesion junctions are closely related to the characteristics of epithelial cells themselves, ROS is strongly associated with tumor progression, and as thyroid cancer and CRC are both epithelial-derived malignancies, we hypothesised that common epithelial cell-related genes might be involved in the development of both. These results tentatively suggest that marker genes of the MEC subpopulation are involved in the occurrence and progression of CRC mainly through the above-mentioned biological mechanisms.

(a)

(b)

3.4. WGCNA of MEC Marker Genes

Using the TCGA-CRC cohort, we performed WGCNA of the expression profiles of the identified 1259 marker genes. We clustered the MEC genes into modules associated with clinical traits (“tumour” and “normal”) based on a soft threshold of . As shown in Figure 5, to prevent meaningful MEC marker genes from being missed, with the exception of the turquoise module (), all genes in the remaining modules were significantly associated with clinical characteristics (“tumour” and “normal”) () and were used in subsequent analyses. To build a clinical prediction model by linking to the clinical characteristics of CRC patients, we initially identified 787 genes significantly associated with clinical characteristics (“tumour” and “normal”) and further identified MEC marker genes associated with survival prognosis.

(a)

(b)

3.5. Establishment and Validation of a MECRG Signature for Predicting CRC Patient Survival

The univariate Cox analysis (Figure 6(a)) and LASSO regression analysis (Supplementary Figure S7) of the TCGA-CRC training cohort identified 47 MEC marker genes capable of predicting OS, the regression coefficients of the nine MECRGs were calculated using the “predict” function, and nine MECRGs were identified based on the minimum λ (): galectin 2 (LGALS2), glycerophosphodiester phosphodiesterase 1 (GDE1), monocyte chemoattractant protein 1 (MPC1), bone marrow stromal cell antigen 2 (BST2), tropomyosin 2 (TPM2), PRELI domain-containing 2 (PRELID2), G protein subunit gamma 5 (GNG5), calcyphosin (CAPS), and calcium voltage-gated channel subunit alpha 1D (CACNA1D). Validation using the GSE17538 data identified BST2, TPM2, and CAPS as risk genes (hazard ratio ) and LGALS2, GDE1, MPC1, GNG5, PRELID2, and CACNA1D as protective genes (). The MECRG risk score was defined as follows: . The final risk scores for all CRC samples in the training and validation cohorts are shown in Supplementary Tables S4 and S5. The MECRGs were predominantly distributed in the epithelial cell subsets (Figure 6(b)).

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

Figure 6

Identification and validation of MECRGs in the TCGA-CRC training and GSE17538 validation cohorts. (a) Univariate Cox analysis was used to screen MECRGs with prognostic significance. (b) Distribution of MECRGs in cell subsets. (c and d) Kaplan–Meier survival curve showing the prognostic value of the MECRG signature in the training cohort (c) and the validation cohort (d). (e and f) Distribution of the MECRG risk scores and survival status of CRC patients in the training cohort (e) and validation cohort (f). ROC curve representing the efficiency of the MECRG signature in predicting 1-, 3-, and 5-year OS in CRC patients in the training cohort (g) and validation cohort (h).

Next, we divided the patients into high- and low-risk groups according to the median MECRG risk score. The Kaplan–Meier survival analysis demonstrated that in the TCGA-CRC training and GSE17538 validation cohorts, the survival time of the high-risk group was shorter than that of the low-risk group () (Figures 6(c) and 6(d)). In both cohorts, an increase in the risk score was accompanied by an increase in patient mortality (Figures 6(e) and 6(f)). The area under the ROC curve (AUC) at 1, 3, and 5 years was 0.746, 0.726, and 0.710, respectively, in the training cohort, and 0.618, 0.668, and 0.658, respectively, in the validation cohort (Figures 6(g) and 6(h)). These data indicated that the MECRG risk score shows good sensitivity and specificity for predicting the prognosis of CRC patients.

3.6. Independent Prognostic Value of the MECRG Risk Score

To determine whether the MECRG risk score can predict prognosis independently of traditional clinical features, such as age, sex, and TNM stage, we performed univariate and multivariate Cox regression analyses. In the training cohort, age ( and 1.040, 95% confidence interval (CI): 1.010–1.048 and 1.021–1.060, respectively; and ), TNM stage ( and 2.106, 95% CI: 1.628–2.627 and 1.642–2.701, respectively; ), and the MECRG risk score ( and 1.234, 95% CI: 1.200–1.396 and 1.139–1.336, respectively; ) were independent predictors of OS. Analysis of the GSE17538 validation cohort confirmed that TNM stage ( and 3.027, 95% CI: 2.077–3.541 and 2.323–3.945, respectively; ) and the MECRG risk score ( and 1.298, 95% CI: 1.005–1.289 and 1.127–1.496, respectively; and ) were independent predictors of OS (Figure 7). The above results indicated that the MECRG risk score was an independent prognostic factor in both the training and validation sets.

(a)

(b)

(c)

(d)

3.7. Correlation between MECRG Risk Groups and Clinical Features

We evaluated the correlation between clinical traits and risk groups among the 435 patients in the TCGA-CRC training cohort and the 232 CRC patients in the GSE17538 validation cohort (with complete clinical information). The patients were grouped according to sex (male or female), age (≤65 or >65 years), and TNM stage (I–II or III–IV), and analyses were performed using the log-rank test and Kaplan–Meier analysis. We found that among female and male patients (), patients aged >65 years and ≤65 years (), and patients at TNM stage III–IV (), those in the high-risk group had a shorter survival time than those in the low-risk group (Figures 8(a) and 8(b)). A heat map revealed significant differences in tumour staging between the high- and low-risk groups (TCGA-CRC: ; GSE17538: ) (Figures 8(c) and 8(d)). These results suggested that the MECRG risk score can significantly affect the survival prognosis of CRC patients across gender, age, and stage.

(a)

(b)

(c)

(d)

3.8. Functional Enrichment Analysis of the MECRG Signature

We used GSEA to evaluate the functions of and signalling pathways associated with the genes in the high- and low-risk groups (Figure 9(a)). The results showed that the high-risk group was significantly enriched in functions related to the positive regulation of migration involved in sprouting angiogenesis, the phosphoinositide 3-kinase (PI3K)-AKT-mammalian target of rapamycin (mTOR) signalling pathway, basal cell carcinoma, extracellular matrix (ECM) receptor interaction, and angiogenesis, whereas the low-risk group was significantly enriched in apoptosis and the p53 signalling pathway. Additionally, enrichment analysis of the epithelial cell subsets indicated significant enrichment of activities involving collagen-containing ECM, fatty acid metabolism, basal cell carcinoma, PI3K-AKT signalling, and p53 signalling (Figure 9(b)). We found that the above signalling pathways are more likely to mediate the process of tumor progression in CRC epithelial cells than in other cell subpopulations.

(a)

(b)

3.9. Nomogram for Predicting the Prognosis of CRC Patients

To establish a practical method for predicting the probability of CRC patient survival, we constructed a MECRG nomogram using the TCGA-CRC training cohort to predict OS at 1, 3, and 5 years (Figure 10(a)). The predictors included three clinical features (age, sex, and TNM stage) and the risk score. ROC curve analysis of nomogram reliability revealed AUC values at 1, 3, and 5 years of 0.787, 0.806, and 0.777, respectively (Figure 10(b)). Analysis of the GSE17538 validation cohort yielded AUC values of 0.843, 0.812, and 0.827, respectively (Figure 10(c)). The calibration curve showed that the predicted survival rate agreed well with the actual survival rate (Figures 10(d) and 10(e)). Our results demonstrated that the MECRG nomogram constructed in combination with the clinical characteristics of CRC can provide good prediction.

(a)

(b)

(c)

(d)

(e)

4. Discussion

Epithelial adenocarcinoma represents the most prevalent type of CRC and arises from benign colon adenoma [33]. Studies have shown that in epithelial malignant tumours, unstable adhesion between epithelial cells strongly correlates with an increased invasiveness of tumour cells. Recently, numerous mechanisms underlying cell–cell junctions regulated by E-cadherin expression in epithelial cells have been discovered [34–36]. Apical-basal polarity is the main characteristic of epithelial cells. Epithelial cell-associated polarity proteins are associated with the origin and poor prognosis of colorectal tumours, hepatocellular carcinoma, and endometrial cancer [37–39]. Therapeutic targets based on epithelial cell apical-basal polarity complexes have been reported. For example, partitioning-defective 6 (Par6) is expected to be a therapeutic target for breast cancer [37], and atypical protein kinase C (aPKC) has been suggested as a possible therapeutic target for gastric cancer [40]. Based on the above findings, we believed that the discovery of CRC epithelial cell-related biomarkers would facilitate the development of new therapeutic and predictive targets for CRC prognosis. ScRNA-seq technology allows the sequencing of RNA from individual cells to systematically track the dynamic changes in individual cells and deepen the understanding of cellular states and gene expression regulation in pathological disease processes [41]. scRNA-seq analysis provides analytical detail on the cellular level [42]. The technology has been recently applied to study epithelial cells in serous epithelial ovarian cancer and endometrial cancer [43, 44]. In nontumour research, scRNA-seq has been used for gene expression profiling of breast epithelial cells during their development [45]. Xu et al. revealed the biological mechanisms implicated in the involvement of lung epithelial cells in idiopathic fibrosis using scRNA-seq [46]. Additionally, scRNA-seq technology has received increasing attention for predicting the prognosis of cancer patients. For example, Wang et al. constructed a model based on 10 biomarkers of pancreatic ductal epithelial cells to predict the prognosis of patients with pancreatic adenocarcinoma [47], Zheng et al. [48] screened nine fibroblast marker genes from scRNA-seq CRC data as potential prognostic markers, and Li et al. [49] identified seven macrophage marker genes from breast cancer scRNA-seq data as promising diagnostic and prognostic biomarkers.

In our study, we constructed a CRC prognosis model by analysing scRNA-seq data from CRC epithelial cells. Specifically, we analysed scRNA-seq data of 15,465 cells from four CRC patients, followed by PCA for unsupervised clustering and the identification of MEC subsets based on chromosome CNV analysis. We then used WGCNA, univariate Cox, and LASSO regression analyses to construct a MECRG signature of nine genes related to CRC prognosis. We validated the MECRG signature using both training and validation cohorts. The Kaplan–Meier analysis confirmed a significantly shortened OS for CRC patients with a high MECRG risk score. Furthermore, the univariate and multivariate Cox regression analyses showed that the MECRG risk score may be an independent predictor of OS. Based on the MECRG risk score coupled with the clinical characteristics of sex, age, and TNM stage (III–IV), low-risk patients showed significantly longer survival times than high-risk patients. It is worth noting that there were significant differences between the MECRG signature-based risk groups at different CRC stages. Furthermore, the MECRG nomogram showed excellent prediction in both the training and validation sets, suggesting that it may efficiently predict CRC patient survival in the clinical setting.

Elevated expression of LGALS2 reportedly inhibits the development of CRC and lymph node metastasis of gastric cancer [50, 51]. GDE1 expression is significantly reduced in drug-resistant ovarian cancer samples [52]. Consistent with these findings, our results suggested that high expression of LGALS2 and GDE1 in CRC patients implies a good prognosis. However, the relationship between GDE1 and CRC requires further study. Lysine demethylase 5A (KDM5A) regulates MPC1 expression, and the KDM5A–MPC1 axis is involved in regulating the mesenchymal characteristics of cancer cells during EMT [53]. Schell et al. [54] found that loss of MPC1 expression enhances the Warburg effect and promotes the proliferation of CRC cells. Deletion of MPC1 is related to a poor prognosis in glioblastoma [55]. Furthermore, BST2 activates the nuclear factor-κB-Snail-Raf kinase inhibitor protein axis to promote tumour invasion and metastasis via EMT [56, 57]. BST2 overexpression correlates with poor prognosis in CRC, stomach cancer, and oesophageal cancer [58]. BST2 expression is specifically upregulated in oral squamous cell carcinoma and is responsible for drug resistance [59]. BST2 is more highly expressed in breast cancer cells derived from patients presenting bone metastasis than in human primary breast cancer cells [60]. Furthermore, Mukai et al. [58] showed that BST2 knockout in vitro inhibited the proliferation of gastric cancer cells. Together, these findings suggest that MPC1 is a protective gene and BST2 a risk gene in CRC, which is consistent with the results of the present study. CAPS is a calcium-binding protein related to cell proliferation and differentiation signals [61] and associated with poor prognosis in CRC and gliomas and drug resistance in breast cancer [62–64], which is consistent with our results. Abnormal expression of TPM2 is involved in actin cytoskeleton remodelling during EMT of lens epithelial cells [65]. Shibata et al. [66] showed that downregulation of TPM2 expression decreased EMT in injured mouse lens epithelium, resulting in delayed lens wound healing. Additionally, TPM2 is upregulated in ovarian cancer, liver cancer, and breast cancer [67–69]. Zhou et al. [70] found that elevated TPM2 expression in CRC patients was predictive of poor prognosis, which is in line with the findings of the present study. However, Ma et al. [71] showed that TPM2 expression was downregulated in CRC; therefore, the precise mechanism of TPM2 in CRC requires clarification. Patients with glioma exhibiting elevated GNG5 expression have a shorter survival time [72], and patients with head and neck squamous cell carcinoma and elevated PRELID2 expression have a poor prognosis [73]. Tan et al. [74] reported that excessive aldosterone secretion from aldosteronoma is related to a CACNA1D mutation. We identified GNG5, PRELID2, and CACNA1D as protective genes in CRC patients; however, further basic experimental studies of these three genes in CRC are needed. Additionally, we showed that MPC1, BST2, and TPM2 are closely related to EMT, suggesting that these molecules are potentially important EMT-related therapeutic targets.

GSEA suggested that genes related to collagen-containing ECM, fatty acid metabolism, PI3K-AKT signalling, p53 signalling, EMT, and other related pathways were enriched in epithelial cell subsets. Among these, EMT-related pathways were more significantly enriched in epithelial cell clusters 7, 9, and 10. Activation of the PI3K-AKT signalling pathway is a key feature of the EMT programme during tumour progression [75]. The expression of genes involved in fatty acid synthesis is upregulated in CRC epithelial cells, where the accumulation of polyunsaturated fatty acids contributes to the development of CRC [76]. Targeting fatty acid synthesis genes may become a new strategy for the treatment of CRC in the future. ECM remodelling can affect the signalling in the tumour microenvironment [77]. The P53 signalling pathway is a common tumour suppressor pathway [78], and we speculate that epithelial cells may promote CRC progression by interfering with this pathway. Thus, in the CRC microenvironment, epithelial cells may mediate the development of CRC through collagen-containing ECM, fatty acid metabolism, PI3K-AKT signalling, p53 signalling, EMT, and other pathological mechanisms.

This was a retrospective study that used scRNA-seq data and bulk data from public databases to construct a model for predicting survival in CRC patients. However, this study had some limitations. First, the mechanisms of GDE1, PRELID2, GNG5, and CACNA1D in CRC have not been clarified; therefore, the data suggesting their prognostic value need to be validated. Second, we did not evaluate tumour size, metastasis, surgery, postoperative chemoradiotherapy, and other prognostic factors in this study. This may have affected the predictive accuracy of the model. In future studies, we will include additional data to increase the accuracy of the model.

5. Conclusion

We evaluated epithelial cell marker genes with prognostic significance in CRC using scRNA-seq data and bulk data and generated a MECRG signature and risk score, which was confirmed to show independent prognostic efficacy for CRC patients. A nomogram based on the MECRG signature along with specific clinical features demonstrated accurate prediction of CRC patient survival, suggesting its potential utility for clinical application.

Data Availability

The datasets generated during and/or analysed in this study are available from the corresponding author upon reasonable request.

Conflicts of Interest

The authors declare no conflicts of interest.

Acknowledgments

This study was supported by the Zhejiang Chinese Medicine 13th Five-Year Key Discipline Construction Program (2017-XK-A12), the Zhejiang Provincial Traditional Chinese Medicine Science and Technology Programme Projects (2021ZB128 and 2018ZB064), and the Pang Dexiang Famous Veteran Chinese Medicine Specialist Inheritance Studio (2A12012014).

Supplementary Materials

Supplementary 1. Figure S1: quality control of the scRNA-seq data from the four CRC samples. Figure S2: gene numbers in the four CRC samples are positively correlated with sequencing depth. Figure S3: PCA classification of the cells based on the scRNA-seq data. Figure S4: eleven statistically significant principal components as identified by PCA. Figure S5: dot diagram showing the expression of specific marker genes in the cell subsets. Figure S6: distribution of specific marker genes in the four cell subsets. Figure S7: LASSO regression was used to screen MECRGs with prognostic significance.

Supplementary 2. Table S1: clinical characteristics of the four CRC patients. Table S2: CNV scores of epithelial cell subclusters. Table S3: MEC subclusters of marker genes. Table S4: MECRG signature using the TCGA-CRC training cohort. Table S5: MECRG signature using the GSE17538 validation cohort.

References

Y. N. Peng, M. L. Huang, and C. H. Kao, “Prevalence of depression and anxiety in colorectal cancer patients: a literature review,” International Journal Environmental Research and Public Health, vol. 16, no. 3, p. 411, 2019.
View at: Publisher Site | Google Scholar
H. Sung, J. Ferlay, R. L. Siegel et al., “Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries,” CA: a Cancer Journal for Clinicians, vol. 71, no. 3, pp. 209–249, 2021.
View at: Publisher Site | Google Scholar
F. Islami, A. Goding Sauer, K. D. Miller et al., “Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States,” CA: a Cancer Journal for Clinicians, vol. 68, no. 1, pp. 31–54, 2018.
View at: Publisher Site | Google Scholar
K. D. Miller, M. Fidler-Benaoudia, T. H. Keegan, H. S. Hipp, A. Jemal, and R. L. Siegel, “Cancer statistics for adolescents and young adults, 2020,” CA: a Cancer Journal for Clinicians, vol. 70, no. 6, pp. 443–459, 2020.
View at: Publisher Site | Google Scholar
S. B. Edge and C. C. Compton, “The American Joint Committee on Cancer: the 7th edition of the AJCC cancer staging manual and the future of TNM,” Annals of Surgical Oncology, vol. 17, no. 6, pp. 1471–1474, 2010.
View at: Publisher Site | Google Scholar
A. Jemal, R. Siegel, E. Ward et al., “Cancer statistics, 2008,” CA: a Cancer Journal for Clinicians, vol. 58, no. 2, pp. 71–96, 2008.
View at: Publisher Site | Google Scholar
V. Karantza, “Keratins in health and cancer: more than mere epithelial cell markers,” Oncogene, vol. 30, no. 2, pp. 127–138, 2011.
View at: Publisher Site | Google Scholar
N. Tanimura and Y. Fujita, “Epithelial defense against cancer (EDAC),” Seminars in Cancer Biology, vol. 63, pp. 44–48, 2020.
View at: Publisher Site | Google Scholar
H. Lv, R. Liu, J. Fu et al., “Epithelial cell-derived periostin functions as a tumor suppressor in gastric cancer through stabilizing p53 and E-cadherin proteins via the Rb/E2F1/p14ARF/Mdm2 signaling pathway,” Cell Cycle, vol. 13, no. 18, pp. 2962–2974, 2014.
View at: Publisher Site | Google Scholar
L. M. McCaffrey and I. G. Macara, “Epithelial organization, cell polarity and tumorigenesis,” Trends in Cell Biology, vol. 21, no. 12, pp. 727–735, 2011.
View at: Publisher Site | Google Scholar
C. Royer and X. Lu, “Epithelial cell polarity: a major gatekeeper against cancer?” Cell Death & Differentiation, vol. 18, no. 9, pp. 1470–1477, 2011.
View at: Publisher Site | Google Scholar
S. G. Zeng, X. Lin, J.-C. Liu, and J. Zhou, “Hypoxia-induced internalization of connexin 26 and connexin 43 in pulmonary epithelial cells is involved in the occurrence of non-small cell lung cancer via the P53/MDM2 signaling pathway,” International Journal of Oncology, vol. 55, no. 4, pp. 845–859, 2019.
View at: Publisher Site | Google Scholar
J. Varga and F. R. Greten, “Cell plasticity in epithelial homeostasis and tumorigenesis,” Nature Cell Biology, vol. 19, no. 10, pp. 1133–1141, 2017.
View at: Publisher Site | Google Scholar
T. Knösel, V. Emde, K. Schlüns, P. M. Schlag, M. Dietel, and I. Petersen, “Cytokeratin profiles identify diagnostic signatures in colorectal cancer using multiplex analysis of tissue microarrays,” Cell Oncology, vol. 28, no. 4, pp. 167–175, 2006.
View at: Publisher Site | Google Scholar
C. Ausch, V. Buxhofer-Ausch, U. Olszewski et al., “Caspase-cleaved cytokeratin 18 fragment (M30) as marker of postoperative residual tumor load in colon cancer patients,” European Journal of Surgical Oncology, vol. 35, no. 11, pp. 1164–1168, 2009.
View at: Publisher Site | Google Scholar
C. K. Cheong, K. R. Y. Nistala, C. H. Ng et al., “Neoadjuvant therapy in locally advanced colon cancer: a meta-analysis and systematic review,” Journal of Gastrointestinal Oncology, vol. 11, no. 5, pp. 847–857, 2020.
View at: Publisher Site | Google Scholar
D. Lambrechts, E. Wauters, B. Boeckx et al., “Phenotype molding of stromal cells in the lung tumor microenvironment,” Nature Medicine, vol. 24, no. 8, pp. 1277–1289, 2018.
View at: Publisher Site | Google Scholar
F. Tang, C. Barbacioru, Y. Wang et al., “mRNA-Seq whole-transcriptome analysis of a single cell,” Nature Methods, vol. 6, no. 5, pp. 377–382, 2009.
View at: Publisher Site | Google Scholar
E. Azizi, A. J. Carr, G. Plitas et al., “Single-cell map of diverse immune phenotypes in the breast tumor microenvironment,” Cell, vol. 174, no. 5, pp. 1293–1308.e36, 2018.
View at: Publisher Site | Google Scholar
Q. Zhang, Y. He, N. Luo et al., “Landscape and dynamics of single immune cells in hepatocellular carcinoma,” Cell, vol. 179, no. 4, pp. 829–845.e20, 2019.
View at: Publisher Site | Google Scholar
X. Zheng, J. Song, C. Yu et al., “Single-cell transcriptomic profiling unravels the adenoma-initiation role of protein tyrosine kinases during colorectal tumorigenesis,” Signal Transduction and Targeted Therapy, vol. 7, no. 1, 2022.
View at: Publisher Site | Google Scholar
S. Mangiola, M. A. Doyle, and A. T. Papenfuss, “Interfacing Seurat with the R tidy universe,” Bioinformatics, vol. 37, no. 22, pp. 4100–4107, 2021.
View at: Google Scholar
Z. Chen, H. Zhang, Y. Bai et al., “Single cell transcriptomic analysis identifies novel vascular smooth muscle subsets under high hydrostatic pressure,” Science China Life Sciences, vol. 64, no. 10, pp. 1677–1690, 2021.
View at: Publisher Site | Google Scholar
J. H. Pan, H. Zhou, L. Cooper et al., “LAYN is a prognostic biomarker and correlated with immune infiltrates in gastric and colon cancers,” Frontiers in Immunology, vol. 10, p. 6, 2019.
View at: Publisher Site | Google Scholar
G. Lombardo, M. Gili, C. Grange et al., “IL-3R-alpha blockade inhibits tumor endothelial cell-derived extracellular vesicle (EV)-mediated vessel formation by targeting the β-catenin pathway,” Oncogene, vol. 37, no. 9, pp. 1175–1191, 2018.
View at: Publisher Site | Google Scholar
H. Ichimiya, K. Maeda, A. Enomoto, L. Weng, M. Takahashi, and T. Murohara, “Girdin/GIV regulates transendothelial permeability by controlling VE-cadherin trafficking through the small GTPase, R-Ras,” Biochemical and Biophysical Research Communications, vol. 461, no. 2, pp. 260–267, 2015.
View at: Publisher Site | Google Scholar
O. Gires, M. Pan, H. Schinke, M. Canis, and P. A. Baeuerle, “Expression and function of epithelial cell adhesion molecule EpCAM: where are we after 40 years?” Cancer and Metastasis Reviews, vol. 39, no. 3, pp. 969–987, 2020.
View at: Publisher Site | Google Scholar
G. Corso, J. Figueiredo, S. P. De Angelis et al., “E-cadherin deregulation in breast cancer,” Journal of Cellular and Molecular Medicine, vol. 24, no. 11, pp. 5930–5936, 2020.
View at: Publisher Site | Google Scholar
L. Fang, G. Yu, W. Yu, G. Chen, and B. Ye, “The correlation of WDR76 expression with survival outcomes and immune infiltrates in lung adenocarcinoma,” PeerJ, vol. 9, article e12277, 2021.
View at: Publisher Site | Google Scholar
A. P. Patel, I. Tirosh, J. J. Trombetta et al., “Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblastoma,” Science, vol. 344, no. 6190, pp. 1396–1401, 2014.
View at: Publisher Site | Google Scholar
Q. Xu, S. Chen, Y. Hu, and W. Huang, “Single-cell RNA transcriptome reveals the intra-tumoral heterogeneity and regulators underlying tumor progression in metastatic pancreatic ductal adenocarcinoma,” Cell Death Discovery, vol. 7, no. 1, p. 331, 2021.
View at: Publisher Site | Google Scholar
P. Langfelder and S. Horvath, “WGCNA: an R package for weighted correlation network analysis,” BMC Bioinformatics, vol. 9, no. 1, p. 559, 2008.
View at: Publisher Site | Google Scholar
A. S. Takhar, O. Eremin, and S. A. Watson, “The role of gastrin in colorectal carcinogenesis,” The Surgeon, vol. 2, no. 5, pp. 251–257, 2004.
View at: Publisher Site | Google Scholar
D. M. Bryant and J. L. Stow, “The ins and outs of E-cadherin trafficking,” Trends in Cell Biology, vol. 14, no. 8, pp. 427–434, 2004.
View at: Publisher Site | Google Scholar
A. P. Kowalczyk and A. B. Reynolds, “Protecting your tail: regulation of cadherin degradation by p120-catenin,” Current Opinion in Cell Biology, vol. 16, no. 5, pp. 522–527, 2004.
View at: Publisher Site | Google Scholar
G. Swaminathan and C. A. Cartwright, “Rack1 promotes epithelial cell-cell adhesion by regulating E-cadherin endocytosis,” Oncogene, vol. 31, no. 3, pp. 376–389, 2012.
View at: Publisher Site | Google Scholar
Y. Kamei, K. Kito, T. Takeuchi et al., “Human scribble accumulates in colorectal neoplasia in association with an altered distribution of β-catenin,” Human Pathology, vol. 38, no. 8, pp. 1273–1281, 2007.
View at: Publisher Site | Google Scholar
Z. Ouyang, W. Zhan, and L. Dan, “hScrib, a human homolog of Drosophila neoplastic tumor suppressor, is involved in the progress of endometrial cancer,” Oncology Research, vol. 18, no. 11, pp. 593–599, 2009.
View at: Publisher Site | Google Scholar
S. Wan, A.-S. Meyer, S. M. E. Weiler et al., “Cytoplasmic localization of the cell polarity factor scribble supports liver tumor formation and tumor cell invasiveness,” Hepatology, vol. 67, no. 5, pp. 1842–1856, 2018.
View at: Publisher Site | Google Scholar
R. Takagawa, K. Akimoto, Y. Ichikawa et al., “High expression of atypical protein kinase C lambda/iota in gastric cancer as a prognostic factor for recurrence,” Annals of Surgical Oncology, vol. 17, no. 1, pp. 81–88, 2010.
View at: Publisher Site | Google Scholar
A. A. Kolodziejczyk, J. K. Kim, V. Svensson, J. C. Marioni, and S. A. Teichmann, “The technology and biology of single-cell RNA sequencing,” Molecular Cell, vol. 58, no. 4, pp. 610–620, 2015.
View at: Publisher Site | Google Scholar
X. Ren, B. Kang, and Z. Zhang, “Understanding tumor ecosystems by single-cell sequencing: promises and limitations,” Genome Biology, vol. 19, no. 1, p. 211, 2018.
View at: Publisher Site | Google Scholar
A. J. Shih, A. Menzin, J. Whyte et al., “Identification of grade and origin specific cell populations in serous epithelial ovarian cancer by single cell RNA-seq,” PLoS One, vol. 13, no. 11, article e0206785, 2018.
View at: Publisher Site | Google Scholar
Y.-E. Guo, Y. Li, B. Cai et al., “Phenotyping of immune and endometrial epithelial cells in endometrial carcinomas revealed by single-cell RNA sequencing,” Aging, vol. 13, no. 5, pp. 6565–6591, 2021.
View at: Publisher Site | Google Scholar
K. Bach, S. Pensa, M. Grzelak et al., “Differentiation dynamics of mammary epithelial cells revealed by single-cell RNA sequencing,” Nature Communications, vol. 8, no. 1, p. 2128, 2017.
View at: Publisher Site | Google Scholar
Y. Xu, T. Mizuno, A. Sridharan et al., “Single-cell RNA sequencing identifies diverse roles of epithelial cells in idiopathic pulmonary fibrosis,” JCI Insight, vol. 1, no. 20, article e90558, 2016.
View at: Publisher Site | Google Scholar
X. Wang, X. Dou, X. Ren et al., “A ductal-cell-related risk model integrating single-cell and bulk sequencing data predicts the prognosis of patients with pancreatic adenocarcinoma,” Frontiers in Genetics, vol. 12, article 763636, 2021.
View at: Google Scholar
H. Zheng, H. Liu, Y. Ge, and X. Wang, “Integrated single-cell and bulk RNA sequencing analysis identifies a cancer associated fibroblast-related signature for predicting prognosis and therapeutic responses in colorectal cancer,” Cancer Cell International, vol. 21, no. 1, 2021.
View at: Publisher Site | Google Scholar
Y. Li, X. Zhao, Q. Liu, and Y. Liu, “Bioinformatics reveal macrophages marker genes signature in breast cancer to predict prognosis,” Annals of Medicine, vol. 53, no. 1, pp. 1019–1031, 2021.
View at: Publisher Site | Google Scholar
H. Li, L. Zhao, Y. S. Lau, C. Zhang, and R. Han, “Genome-wide CRISPR screen identifies LGALS2 as an oxidative stress-responsive gene with an inhibitory function on colon tumor growth,” Oncogene, vol. 40, no. 1, pp. 177–188, 2021.
View at: Publisher Site | Google Scholar
J.-H. Jung, H.-J. Kim, J. Yeom et al., “Lowered expression of galectin-2 is associated with lymph node metastasis in gastric cancer,” Journal of Gastroenterology, vol. 47, no. 1, pp. 37–48, 2012.
View at: Publisher Site | Google Scholar
J. Lu, Y. Li, Y. A. Li et al., “In vivo detection of dysregulated choline metabolism in paclitaxel-resistant ovarian cancers with proton magnetic resonance spectroscopy,” Journal of Translational Medicine, vol. 20, no. 1, p. 92, 2022.
View at: Publisher Site | Google Scholar
J. H. You, J. Lee, and J. L. Roh, “Mitochondrial pyruvate carrier 1 regulates ferroptosis in drug-tolerant persister head and neck cancer cells via epithelial-mesenchymal transition,” Cancer Letters, vol. 507, pp. 40–54, 2021.
View at: Publisher Site | Google Scholar
J. C. Schell, K. A. Olson, L. Jiang et al., “A role for the mitochondrial pyruvate carrier as a repressor of the Warburg effect and colon cancer cell growth,” Molecular Cell, vol. 56, no. 3, pp. 400–413, 2014.
View at: Publisher Site | Google Scholar
Y. Chai, C. Wang, W. Liu, Y. Fan, and Y. Zhang, “MPC1 deletion is associated with poor prognosis and temozolomide resistance in glioblastoma,” Journal of Neurooncology, vol. 144, no. 2, pp. 293–301, 2019.
View at: Publisher Site | Google Scholar
A. Matsuda, Y. Suzuki, G. Honda et al., “Large-scale identification and characterization of human genes that activate NF-κB and MAPK signaling pathways,” Oncogene, vol. 22, no. 21, pp. 3307–3318, 2003.
View at: Publisher Site | Google Scholar
K. Wu and B. Bonavida, “The activated NF-κB-Snail-RKIP circuitry in cancer regulates both the metastatic cascade and resistance to apoptosis by cytotoxic drugs,” Critical Reviews of Immunology, vol. 29, no. 3, pp. 241–254, 2009.
View at: Publisher Site | Google Scholar
S. Mukai, N. Oue, T. Oshima et al., “Overexpression of transmembrane protein BST2 is associated with poor survival of patients with esophageal, gastric, or colorectal cancer,” Annals of Surgical Oncology, vol. 24, no. 2, pp. 594–602, 2017.
View at: Publisher Site | Google Scholar
H. Jin, L. Zhang, S. Wang, and L. Qian, “BST2 promotes growth and induces gefitinib resistance in oral squamous cell carcinoma via regulating the EGFR pathway,” Archives of Medicinal Science, vol. 17, no. 6, pp. 1772–1782, 2021.
View at: Google Scholar
D. Cai, J. Cao, Z. Li et al., “Up-regulation of bone marrow stromal protein 2 (BST2) in breast cancer with bone metastasis,” BMC Cancer, vol. 9, no. 1, p. 102, 2009.
View at: Publisher Site | Google Scholar
R. Lecocq, F. Lamy, C. Erneux, and J. E. Dumont, “Rapid purification and identification of calcyphosine, a Ca²⁺-binding protein phosphorylated by protein kinase A,” Biochemistry Journal, vol. 306, no. 1, pp. 147–151, 1995.
View at: Publisher Site | Google Scholar
W. Shao, Q. Wang, F. Wang, Y. Jiang, M. Xu, and J. Xu, “Abnormal expression of calcyphosine is associated with poor prognosis and cell biology function in colorectal cancer,” Onco Targets Therapy, vol. 9, pp. 477–487, 2016.
View at: Publisher Site | Google Scholar
Z. Zhu, J. Wang, J. Tan et al., “Calcyphosine promotes the proliferation of glioma cells and serves as a potential therapeutic target,” Journal of Pathology, vol. 255, no. 4, pp. 374–386, 2021.
View at: Publisher Site | Google Scholar
H. J. Johansson, B. C. Sanchez, J. Forshed et al., “Proteomics profiling identify CAPS as a potential predictive marker of tamoxifen resistance in estrogen receptor positive breast cancer,” Clinical Proteomics, vol. 12, no. 1, p. 8, 2015.
View at: Publisher Site | Google Scholar
E. Kubo, N. Hasanova, N. Fatma, H. Sasaki, and D. P. Singh, “Elevated tropomyosin expression is associated with epithelial-mesenchymal transition of lens epithelial cells,” Journal of Cellular and Molecular Medicine, vol. 17, no. 1, pp. 212–221, 2013.
View at: Publisher Site | Google Scholar
T. Shibata, S. Shibata, Y. Ishigaki et al., “Tropomyosin 2 heterozygous knockout in mice using CRISPR-cas9 system displays the inhibition of injury-induced epithelial-mesenchymal transition, and lens opacity,” Mechanisms of Ageing and Development, vol. 171, pp. 24–30, 2018.
View at: Publisher Site | Google Scholar
D. Q. Li, L. Wang, F. Fei et al., “Identification of breast cancer metastasis-associated proteins in an isogenic tumor metastasis model using two-dimensional gel electrophoresis and liquid chromatography-ion trap-mass spectrometry,” Proteomics, vol. 6, no. 11, pp. 3352–3368, 2006.
View at: Publisher Site | Google Scholar
Z. Tian, J. Zhao, and Y. Wang, “The prognostic value of TPM1–4 in hepatocellular carcinoma,” Cancer Medicine, vol. 11, no. 2, pp. 433–446, 2022.
View at: Publisher Site | Google Scholar
H.-Y. Tang, L. A. Beer, J. L. Tanyi, R. Zhang, Q. Liu, and D. W. Speicher, “Protein isoform-specific validation defines multiple chloride intracellular channel and tropomyosin isoforms as serological biomarkers of ovarian cancer,” Journal of Proteomics, vol. 89, pp. 165–178, 2013.
View at: Publisher Site | Google Scholar
Y. Zhou, S. Bian, X. Zhou et al., “Single-cell multiomics sequencing reveals prevalent genomic alterations in tumor stromal cells of human colorectal cancer,” Cancer Cell, vol. 38, no. 6, pp. 818–828.e5, 2020.
View at: Publisher Site | Google Scholar
Y. Ma, T. Xiao, Q. Xu, X. Shao, and H. Wang, “iTRAQ-based quantitative analysis of cancer-derived secretory proteome reveals TPM2 as a potential diagnostic biomarker of colorectal cancer,” Frontiers in Medicine, vol. 10, no. 3, pp. 278–285, 2016.
View at: Publisher Site | Google Scholar
W. Zhang, Z. Liu, B. Liu et al., “GNG5 is a novel oncogene associated with cell migration, proliferation, and poor prognosis in glioma,” Cancer Cell International, vol. 21, no. 1, 2021.
View at: Publisher Site | Google Scholar
G. Liu, X. Zeng, B. Wu, J. Zhao, and Y. Pan, “RNA-seq analysis of peripheral blood mononuclear cells reveals unique transcriptional signatures associated with radiotherapy response of nasopharyngeal carcinoma and prognosis of head and neck cancer,” Cancer Biology and Therapy, vol. 21, no. 2, pp. 139–146, 2020.
View at: Publisher Site | Google Scholar
G. C. Tan, G. Negro, A. Pinggera et al., “Aldosterone-producing adenomas: histopathology-genotype correlation and identification of a novel CACNA1D mutation,” Hypertension, vol. 70, no. 1, pp. 129–136, 2017.
View at: Publisher Site | Google Scholar
L. Larue and A. Bellacosa, “Epithelial-mesenchymal transition in development and cancer: role of phosphatidylinositol 3 kinase/AKT pathways,” Oncogene, vol. 24, no. 50, pp. 7443–7454, 2005.
View at: Publisher Site | Google Scholar
J. Hofmanová, J. Slavík, M. Ciganek et al., “Complex alterations of fatty acid metabolism and phospholipidome uncovered in isolated colon cancer epithelial cells,” International Journal of Molecular Science, vol. 22, no. 13, p. 6650, 2021.
View at: Publisher Site | Google Scholar
V. Mohan, A. Das, and I. Sagi, “Emerging roles of ECM remodeling processes in cancer,” Seminars in Cancer Biology, vol. 62, pp. 192–200, 2020.
View at: Publisher Site | Google Scholar
L. J. Hernández Borrero and W. S. El-Deiry, “Tumor suppressor p53: biology, signaling pathways, and therapeutic targeting,” Biochimica Biophysica Acta Reviews on Cancer, vol. 1876, no. 1, article 188556, 2021.
View at: Google Scholar

Copyright

Copyright © 2022 Kai-yu Shen et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies

Views

1233

Downloads

1069

Citations