Biomarker for colorectal cancer detection and application

Abstract

The invention provides a biological marker for colorectal cancer detection and an application of the biological marker. The biological marker comprises the following components: hsacirc0008039, hsacirc0004865, hsacirc0077837, hsacirc0019223, hsacirc0000110, hsacirc0073244, hsacirc0087960, hsacirc0001461, hsacirc0001900, hsacirc0000467 and hsacirc0001663, and the biological marker can be used for detecting colorectal cancer. Research finds that the biomarkers play an important role in the progress of CRC, a CRC detection model is formed by using the biomarkers, and CRC diagnosis and staging can be accurately evaluated.

Description

Technical Field

[0001] The present invention relates to the field of biomedicine technology, and in particular to biomarkers for colorectal cancer detection and applications. Background Art

[0002] Early diagnosis is the key to improving the survival rate of CRC patients. The 5-year survival rates of CRC stage I and II patients are 90% and 84%, respectively, while the 5-year survival rates of CRC stage III and IV patients are 70% and 11%, respectively. Although countries continue to promote intestinal screening, early diagnosis remains challenging due to the complex pathogenesis of CRC and the lack of effective biomarkers. Therefore, finding RNA molecules whose expression levels change significantly during the development of CRC and exploring their regulatory networks are crucial for the development of new early diagnosis technologies, which can effectively curb the incidence of CRC and improve the survival rate and quality of life of patients.

[0003] Currently commonly used CRC detection methods: 1. Fecal occult blood test (FOBT): FOBT is a non-invasive, easy-to-operate screening method that detects occult blood in stool samples to determine whether there are signs of bleeding. Although FOBT is widely used in large-scale screening, it has certain limitations. For example, FOBT may be interfered with by food (such as red meat and certain vegetables) and vitamin C, resulting in false negative results. In addition, FOBT cannot accurately locate the source of bleeding, and even if the test result is positive, further diagnostic methods (such as colonoscopy) are required to determine the cause of bleeding. 2. Colonoscopy: Colonoscopy is the gold standard method for directly observing the internal conditions of the colon and rectum, which can detect and remove precancerous lesions and early cancers. However, colonoscopy is an invasive procedure that may cause discomfort and complications to patients, such as bleeding and perforation. In addition, colonoscopy requires professional medical personnel to operate, and patients need to prepare their intestines, which may affect patient compliance.

[0004] With the ever-accelerating pace of life, non-invasive diagnosis has become the first choice for most young people to conduct early cancer screening. Therefore, as a biomarker that is stably present in extracellular fluids (such as plasma, exosomes, and urine), circRNA provides a new and valuable direction for the non-invasive diagnosis of CRC with its sensitivity and specificity. Although circRNA has achieved some exciting results as a biomarker for CRC liquid biopsy, due to the influence of its sensitivity and specificity, the reliability and accuracy of most circRNAs in the clinical diagnosis of colorectal cancer are not satisfactory. Despite some challenges, circRNA still has good application prospects as a diagnostic biomarker for CRC based on its stability and its advantages of being widely present in body fluids. The present invention aims to further explore new circRNAs in order to improve their accuracy and reliability in the clinical diagnosis of colorectal cancer. Summary of the invention

[0005] The purpose of the present invention is to provide biomarkers and applications for colorectal cancer detection, so as to discover more reliable and accurate circRNA as a diagnostic biomarker for rapid CRC detection, provide precise treatment for patients, and ensure or improve patient survival rate and quality of life.

[0006] In order to achieve the above technical objectives, the technical solutions adopted in this application are as follows:

[0007] In a first aspect, the present invention provides biomarkers for detecting colorectal cancer, including: hsa_circ_0008039, hsa_circ_0004865, hsa_circ_0077837, hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663.

[0008] Preferably, the hsa_circ_0008039, the hsa_circ_0004865 and the hsa_circ_0077837 are used to detect and distinguish between early and late patients with colorectal cancer; the hsa_circ_0019223, the hsa_circ_0000110, the hsa_circ_0073244, the hsa_circ_0087960, the hsa_circ_0001461, the hsa_circ_0001900, the hsa_circ_0000467 and the hsa_circ_0001663 are used to detect and distinguish between colorectal cancer malignancies and normal tissues.

[0009] In a second aspect, the present invention provides the use of the biomarker described in the first aspect in preparing a colorectal cancer detection model.

[0010] In a third aspect, the present invention provides a method for constructing a colorectal cancer detection model, wherein the colorectal cancer detection model comprises the biomarker described in the first aspect, and the construction method comprises the following steps:

[0011] Step 1: DESeq2 method was used to identify and screen circRNAs in normal colorectal tissues, paracancerous tissues, and malignant tumor tissues to obtain circRNA dataset 1;

[0012] Step 2, using the WGCNA method to identify and screen the circRNAs in the normal colorectal tissues, paracancerous tissues, and malignant tumor tissues that were the same as in step 1, to obtain circRNA dataset 2;

[0013] Step 3, screen out the circRNAs that intersect in circRNA dataset 1 and circRNA dataset 2 to obtain a candidate circRNA dataset, and calculate the ROC value and P value of the candidate circRNA dataset;

[0014] Step 4, the expression trend consistency of the candidate circRNA datasets was verified by using the two circRNA datasets GSE221240 and GSE235850, and the ROC value and P value of the candidate circRNA datasets were combined to further obtain the biomarkers and construct a colorectal cancer detection model.

[0015] Preferably, in step 1, the control parameters for identification and screening are: |log2Fold change|≥1, baseMean≥3.51, P-value≤0.05 and Padj≤0.05.

[0016] Preferably, in step 1, the WGCNA method is used to identify and screen the circRNAs in the normal colorectal tissues, paracancerous tissues and malignant tumor tissues that are the same as those in step 1, comprising the following steps:

[0017] Step 2-1, primary screening of all tissue samples;

[0018] Step 2-2, calculate the correlation of all tissue circRNAs by soft thresholding, and determine the circRNA dataset 2 by hierarchical clustering;

[0019] Further preferably, the soft threshold β=3.

[0020] Preferably, in step 4, the candidate circRNA data set is verified by GSE221240 and GSE235850 data sets, and the Log2FoldChange value of circRNA in the GSE221240 and GSE235850 data sets and the candidate circRNA data set is calculated using the DESeq2 method, and the consistency of expression trends is compared.

[0021] In a fourth aspect, the present invention provides a colorectal cancer detection model obtained by the construction method described in the third aspect.

[0022] In a fifth aspect, the present invention provides the use of the biomarker described in the first aspect or the colorectal cancer detection model described in the fourth aspect in the preparation of a colorectal cancer detection product.

[0023] In a sixth aspect, the present invention provides a kit for detecting and distinguishing colorectal cancer malignant tumors and normal tissues, including the biomarkers hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663 The invention relates to a colorectal cancer detection model formed by the aforementioned biomarkers, and a reagent for detecting the expression of the biomarkers hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663.

[0024] Preferably, the primers for hsa_circ_0019223 are shown in SEQ ID NO.1 and SEQ ID NO.2; the primers for hsa_circ_0000110 are shown in SEQ ID NO.5 and SEQ ID NO.6; the primers for hsa_circ_0073244 are shown in SEQ ID NO.7 and SEQ ID NO.8; the primers for hsa_circ_0087960 are shown in SEQ ID NO.13 and SEQ ID NO.14; the primers for hsa_circ_0001461 are shown in SEQ ID NO.3 and SEQ ID NO.4; the primers for hsa_circ_0001900 are shown in SEQ ID NO.9 and SEQ ID NO.10; the primers for hsa_circ_0000467 are shown in SEQ ID NO.11 and SEQ ID NO.12; the primers for hsa_circ_0001663 are shown in SEQ ID NO. NO.15 and SEQ ID NO.16.

[0025] In the seventh aspect, the present invention provides specific primers for detecting the biomarkers described in the first aspect, the primers for hsa_circ_0019223 are shown in SEQ ID NO.1 and SEQ ID NO.2; the primers for hsa_circ_0000110 are shown in SEQ ID NO.5 and SEQ ID NO.6; the primers for hsa_circ_0073244 are shown in SEQ ID NO.7 and SEQ ID NO.8; the primers for hsa_circ_0087960 are shown in SEQ ID NO.13 and SEQ ID NO.14; the primers for hsa_circ_0001461 are shown in SEQ ID NO.3 and SEQ ID NO.4; the primers for hsa_circ_0001900 are shown in SEQ ID NO.9 and SEQ ID NO.10; the primers for hsa_circ_0000467 are shown in SEQ ID NO.11 and SEQ ID NO.12. NO.12; the primers for hsa_circ_0001663 are shown in SEQ ID NO.15 and SEQ ID NO.16.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] The present invention has discovered new biomarkers for colorectal cancer detection. Studies have found that these biomarkers play an important role in the progression of CRC. Using these biomarkers to form a CRC detection model can accurately evaluate CRC diagnosis and staging. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 The analysis and identification results of circRNA in normal tissues, paracancerous tissues and malignant tumor tissues of 30 CRC patients in Example 1 of the present invention, wherein (A)-(B) are the numbers of circRNAs identified from 30 normal tissues (N, red), 30 paracancerous tissues (P, green) and 30 malignant tumor tissues (T, yellow) using bwa-CIRI2 ((A)) and STAR-DCC ((B)), respectively; (C) is a box plot showing the numbers of circRNAs identified in the three tissues; (D)-(E) are bar graphs showing the numbers of circRNAs identified in the three tissues after removing circRNAs with low-quality BSJ reads; (F) is a box plot showing the numbers of circRNAs identified in the three tissues after removing low-quality BSJ reads.

[0029] Figure 2 The results of differential analysis of 2389 circRNAs identified in Example 1 using DEseq2 in Example 2 of the present invention, wherein (A)-(C) are volcano plots showing the DEcircRNAs identified between three tissue types, N vs. P (A), N vs. T (B), and P vs. T (C); red dots indicate upregulated circRNAs, blue dots indicate downregulated circRNAs, and black dots indicate non-significant circRNAs (P value or FC has no set standard); (D) is a Venn diagram showing circRNAs that meet specific conditions; (E) is a heat map showing the log2TPM values ​​of circRNAs in three tissue types; (F) is a chromosomal distribution showing the chromosomal distribution of 55 DEcircRNAs, with mutation-susceptible regions highlighted in red; (G)-(H) are scatter plots showing the fold change (Log2FoldChange) of circRNAs in N vs. T and N vs. T and their host genes in N vs. T and P Correlation between fold change (Log2FoldChange) in vs. T; (I) The figure shows the potential functional relationship between host genes of CircRNA.

[0030] Figure 3PCA analysis was performed using the 50 circRNAs (Figure (A)) and 55 DEcircRNAs (Figure (B)) with the highest average expression levels in Example 2 of the present invention as significant genes.

[0031] Figure 4 The results of analysis and identification of co-expressed circRNAs in malignant tumor tissues in Example 3 of the present invention are shown in Figure 3, wherein (A) shows the screening and removal of outlier samples; (B) shows the scale independence and average connectivity of different soft thresholds in each tissue type; (C) shows a gene tree diagram based on dendrogram analysis; (D) shows the three modules remaining after merging; (E)-(F) are scatter plots of the correlation between MM and GS of clustering modules significantly associated with malignant tumor tissues.

[0032] Figure 5 The results of merging the 55 DEcircRNAs obtained by DESeq2 analysis and the 100 tumor-specific circRNAs obtained by WGCNA analysis in Example 4 of the present invention, wherein (A)-(B) are 36 circRNAs overlapping between the DEcircRNAs displayed by the Venn diagram and the tumor-specific circRNAs identified by WGCNA, including 24 significantly upregulated circRNAs and 12 significantly downregulated circRNAs.

[0033] Figure 6 The clinical diagnostic evaluation results of circRNA in Example 5 of the present invention, wherein (A) shows the verification of DEcircRNA in two GEO datasets (GSE235850 and GSE221240); (B) shows the ROC curve showing the potential of 8 DEcircRNAs (based on AUC scores) to distinguish between malignant tumors and normal tissues; (C) shows the ability of hsa_circ_0008039, hsa_circ_0004865, and hsa_circ_0077837 to distinguish between early CRC patients and late CRC patients; *p<0.05, **p<0.01, n=28.

[0034] Figure 7 This is the qRT-PCR verification of the DEcircRNA in Example 6 of the present invention and the Sanger sequencing verification of its BSJ segment, wherein (A) to (H) correspond to 8 circRNAs respectively; *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001; n=27; N, normal tissue; P, paracancerous tissue; T, malignant tumor tissue; the bases marked in red represent the circRNA breakpoint junction (BSJ) region.

[0035] Figure 8 The difference in the concentrated expression levels of 8 DEcircRNAs in normal tissues and malignant tumor tissues in Example 6 of the present invention, wherein (A) to (H) correspond to the 8 circRNAs respectively.

[0036] Fig. 9 The analysis results of the single-cell sequencing data in Example 7 of the present invention, wherein Figure A is a t-SNE diagram of the single-cell sequencing data (including 13 tissue samples from four patients), and different cell types are distinguished by different colors (a total of 11 cell types); Figure B is a tSNE diagram based on tissue source (left) and patient source (right); Figure C is a bubble chart showing the marker genes of 11 different cells, where the size of the circle in the figure indicates the percentage of cells expressing the gene, and the color indicates the average expression level within the cluster; Figure D shows the distribution of host genes corresponding to 11 circRNAs in different cell types in single-cell sequencing data from plasma samples; Figure E shows the distribution of host genes corresponding to 11 circRNAs in different cell types in single-cell sequencing data of three different types of tissue samples.

[0037] Fig.10 The expression levels of 8 DEcircRNAs in normal tissues (n=10) and tumor tissues (n=10) were analyzed by qPCR in Example 7 of the present invention.

[0038] Fig.11 These are the AUC curves of hsa_circ_0003761 (Figure (A)) and hsa_circ_0068464 (Figure (B)) in Comparative Example 1 of the present invention. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments, which are intended to explain the present invention rather than to limit it.

[0040] Example 1

[0041] This example analyzes and identifies circRNAs in three different types of tissue samples, as follows:

[0042] 1. Whole transcriptome sequencing of tissue samples:

[0043] In this study, whole transcriptome sequencing was performed on normal tissues, paracancerous tissues, and malignant tumor tissues of 30 CRC patients from a hospital in Shenzhen (clinical information is shown in Table 1). The collected tissue samples were constructed by a Shenzhen technology company for RNA-seq library construction and high-throughput sequencing. The sequencing platform was DNBSEQ-T7, the sequencing specification was double-end 150bp (PE-150), and each library generated about 40M reads.

[0044] Table 1 Clinical information of CRC patient tissue samples

[0045]

[0046] 2. Identification of circRNA:

[0047] The present invention uses MD5summer software to detect the MD5 value of the original sequencing data to ensure the integrity of the source file; cutadapt (v2.10) and trimmomatic (v0.39) are used to remove the adapter sequence in the original file and filter the low-quality data. For each sample, the present invention established two circRNA identification processes (bwa-CIRI2 and STAR-DCC), both of which use STAR (v2.7.8) and bwa (v0.7.17) to align the clean reads with the standard reference genome of humans (GRCh38), thereby generating BAM files (such as Figure 1 (A) and (B) in the figure). samtools (v1.7) was used to extract reads that were not mapped to the genome (unmapped reads) from the BAM file. Then, CIRI2 (v2.0.6) and DCC (v0.5.0) software were used to identify circRNAs for the unmapped reads corresponding to each sample. The results are shown in Figure 1 As shown in Figure 2, the study found that at the same sequencing depth, the average number of circRNAs identified in normal tissues, paracancerous tissues, and malignant tumor tissues was 3334, 3300, and 2479, respectively. Obviously, the number of circRNAs identified in malignant tumor tissues is significantly less than that in the other two tissues (see Figure 1 (C) Figure). This result shows that there is a potential correlation between circRNA and CRC. Given that circRNA accounts for a low proportion of total RNA (usually about 1-5%), in order to avoid the impact of false positive results on subsequent analysis, the present invention strictly screened the identified circRNA. Specifically, only circRNA detected in half of the CRC patient samples (n=15, 50%) were included in the subsequent analysis. After screening, it can be clearly observed that the individual differences within the group were significantly reduced (see Figure 1 (D), (E)), and the quantitative differences of circRNAs showed more significant tissue specificity (see Figure 1(F) Figure). In order to explore the relationship between sequencing depth and the number of circRNAs, the present invention performed a correlation analysis on the number of circRNAs identified in each tissue sample and the number of unaligned genome reads corresponding to it. This finding shows that there is a close correlation between the identification of circRNAs and the sequencing depth. Specifically, 1981 circRNAs were identified in normal tissues, 2006 circRNAs were identified in paracancerous tissues, and 1272 circRNAs were identified in malignant tumor tissues. In this embodiment, it can also be observed that there is a positive correlation between higher sequencing depth and more circRNAs. Finally, a total of 2389 circRNAs were identified in 90 CRC tissues in this embodiment.

[0048] Example 2

[0049] In order to screen out circRNAs significantly associated with CRC, this example used DEseq2 to perform differential analysis on the 2389 circRNAs obtained in Example 1, and used |log2Fold change|≥1, baseMean≥3.51, P-value≤0.05 and Padj≤0.05 as the criteria for screening. The results are as follows Figure 2 and Figure 3 shown.

[0050] Figure 2 The results showed that in the malignant tumor group (T) vs. the normal group (N), a total of 67 significantly changed circRNAs were screened, including 43 significantly upregulated circRNAs and 24 significantly downregulated circRNAs; in the malignant tumor group (T) vs. the paracancerous group (P), a total of 71 significantly changed circRNAs were screened, including 45 significantly upregulated circRNAs and 26 significantly downregulated circRNAs (see Figure 2 (B)-(C) in the middle); however, no circRNAs meeting the criteria were screened in the normal group (N) vs. the adjacent cancer group (P) (see Figure 2 (A) in the figure); the consistency of the comparison between N vs.T and P vs.T, as well as the specificity of the comparison between N vs.P, suggest that the significant change in the expression of circRNA only occurs after the intestinal tissue has completely become malignant. Therefore, this example selected the intersection of N vs.T and P vs.T (n=55), and only focused on the 55 circRNAs in the intersection in the following study (see Figure 2 (D)

[0051] In order to evaluate the ability of the 55 circRNAs in the intersection to distinguish malignant tumor tissues from the other two types of tissues, the following analysis was performed in this example: (1) The RPM (BSJ reads / total reads*106) of the 55 DEcircRNA BSJs was calculated, and a heat map was drawn using the RPM values. In the heat map, the modules representing malignant tumor tissues (yellow) and the modules representing the other two tissues (red and green) can be clearly distinguished (see Figure 2 (E) Figure); (2) The average expression of each circRNA in 90 tissues was calculated by TPM (BSJ count / 90), and the 50 circRNAs with the highest expression abundance (Top 50) were screened out. Subsequently, the principal component analysis was performed using the 55 circRNAs and the Top 50 as significant genes. The results showed that the discrimination effect of the 55 DEcircRNAs was significantly better than that of the Top 50 circRNAs (see Figure 3 (A)-(B) in the figure); (3) The chromosomal distribution of 55 DEcircRNAs was separately counted and a chromosomal distribution map was drawn. The results showed that these DEcircRNAs did not show obvious clustering on specific chromosomes, but there was a significant overlap with known CRC sensitive mutation fragments, such as RP11-492D6.3, SYNE1, PCBP1 and APC. This suggests that the differences in the expression levels of these DEcircRNAs may be potentially related to the differences in the expression levels of mRNAs in these regions (see Figure 2 (middle (F) figure).

[0052] In addition, based on the characteristics of circRNA derived from its host gene, this example further explored the correlation between the Log2FoldChange values ​​of 55 circRNAs and their host genes in N vs.T and P vs.T comparisons. The results showed that in these two groups, the expression changes of circRNAs and their host genes showed a significant positive correlation (see Figure 2 This finding further verified the correlation between the changes in circRNA expression and the changes in the expression of its host genes. In addition, this example also conducted a comprehensive analysis of the host genes of 55 circRNAs: a PPI network was constructed using Cytoscape; and 18 potential interacting proteins were identified through network analysis, suggesting that they may have a synergistic effect in the development of CRC (see Figure 2 (Fig. I).

[0053] Example 3

[0054] In order to reveal the unique circRNA co-expression network in malignant tumor tissues, this example performed WGCNA analysis on all samples (n=90, the same samples as in Example 1) and identified modules related to cancer specificity. The specific steps are:

[0055] (1) The software's built-in quality control function was used to perform a primary screening of the samples. After quality control, 88 of the 90 samples remained (see Figure 4 (A) in the middle); (2) the correlation between circRNAs was calculated by using the selected optimal soft threshold (β = 3), and three co-expression modules were identified by hierarchical clustering (see Figure 4 (B)-(D) Figure); Among them, the cyan module (containing a total of 1734 circRNAs) showed a significant negative correlation with the malignant tumor phenotype (Cor = -0.66, P-value = 3e^-12), indicating that this module may play an important role in malignant tumor suppression; while the blue module (containing a total of 181 circRNAs) showed a significant positive correlation with the malignant tumor phenotype (Cor = 0.46, P-value = 5e^-6), indicating that this module may play an important role in tumor promotion. For each cancer-specific module, 50 circRNAs significantly correlated with it were screened (screened based on the correlation between MM and Gs) (see Figure 4 (E)-(F)).

[0056] Example 4

[0057] The 55 DEcircRNAs identified by DESeq2 in Example 2 and the 100 tumor-specific circRNAs obtained by WGCNA analysis in Example 3 were combined to obtain 36 candidate circRNAs. Among them, 12 circRNAs with significantly upregulated expression in CRC and 24 circRNAs with significantly downregulated expression in CRC (see Figure 5 (A)-(B) in the figure).

[0058] Example 5

[0059] The purpose of this example is to evaluate the clinical value of 36 candidate circRNAs in CRC diagnosis, as follows:

[0060] (1) In this example, two independent circRNA datasets were obtained from the GEO database: GSE221240 (including 8 normal tissues and 6 malignant tumor tissues) and GSE235850 (including 5 normal tissues and 5 malignant tumor tissues). The Log2FoldChange values ​​calculated by DESeq2 reflect the changes of these circRNAs in the two external datasets, and the dumbbell plot shows the consistency of the change trends of these 36 circRNAs in the two datasets (see Figure 6 (A) in the middle.

[0061] (2) In addition, the ROC curve drawn by GraphPad Prism software explored the diagnostic potential of these circRNAs in distinguishing CRC malignancies. The results showed that a total of 8 circRNAs showed excellent distinguishing efficacy: hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663, especially hsa_circ_0073244 (AUC = 0.9717, 95% CI = 0.9398-1.000) (see Figure 6 (middle (B) figure).

[0062] (3) To further explore the potential role of these circRNAs in CRC progression, this example divided 30 CRC patients (same as Example 1) into an early CRC group (n=14) and a late CRC group (n=14) according to their clinical stage (clinical stage information was not collected for 2 patients). By comparing the expression differences of circRNAs in the two groups, CRC stage-specific circRNAs were screened out. The results showed that hsa_circ_0008039 was significantly upregulated in late CRC malignancies, while hsa_circ_0004865 and hsa_circ_0077837 were downregulated (see Figure 6 These findings suggest that these three circRNAs may play an important role in the progression of CRC and may serve as biomarkers for diagnosis and staging.

[0063] Example 6

[0064] In order to verify the accuracy of the results of bioinformatics analysis and RNA sequencing, the 8 DEcircRNAs screened out in this example were verified by qRT-PCR in 10 CRC malignant tumor tissues and their corresponding normal tissue samples (the clinical information of the samples is shown in Table 4), and the PCR products were sequenced by Sanger to verify the accuracy of the BSJ sequence. The primers involved are shown in Table 2, and the reaction system is shown in Table 3. The results showed that in malignant tumor tissues, the expression levels of hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244 and hsa_circ_0087960 were significantly downregulated, while the expression levels of hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663 were significantly upregulated (see Table 4). Figure 7 (A)-(H)), the results are consistent with the results of RNA-seq. And the Sanger sequencing results are consistent with the splicing sites predicted by CIRI2 and DCC software. These results further verify the accuracy of previous data analysis and highlight the potential clinical application value of circRNA as a candidate biomarker for CRC diagnosis.

[0065] Table 2 Primer sequences

[0066]

[0067]

[0068] Table 3 qRT-PCR reaction system

[0069]

[0070] Table 4 Clinical information of 10 tissue samples

[0071]

[0072] Example 7

[0073] The purpose of this example is to explore the expression trend of circRNA in different cell types, as follows:

[0074] Using the GSE161277 single-cell transcriptome dataset in the GEO database (including: 4 normal tissue samples, 4 adjacent cancer tissue samples, 4 malignant tumor tissue samples and 1 colorectal cancer plasma sample), the cellular composition and expression characteristics of colorectal cancer at the single-cell level were explored.

[0075] The single-cell sequencing data (ScRNA-seq) was subjected to rigorous cell screening and clustering analysis using built-in functions, and 21 different cell populations containing 43,851 cells were finally obtained. The 21 different cell populations were annotated into 11 different cell types using specific marker genes of different cells, including CD8+T cells, endothelial cells, intestinal epithelial cells, adenoma epithelial cells, carcinoma epithelial cells, fibroblasts, follicular B cells, goblet cells, NKT cells, plasma B cells, and Tcm+Tfh+Treg cells (see Fig. 9 The marker genes for each cell type were visualized using bubble plots (see Figures (A)-(B)). Fig. 9 (C) Figure). Subsequently, single-cell analysis was performed on the expression patterns of maternal genes corresponding to the 11 previously screened circRNAs (including 8 malignant tumor-specific circRNAs and 3 stage-specific circRNAs) in different cell types. The results showed that the maternal genes corresponding to multiple circRNAs showed significant enrichment in fibroblasts and epithelial cells of the three tissue types, such as CAMSAP1, EDIL3, EPB41L2, FAT1, LPAR1, SKA3, TMEM181, etc. In addition, further studies found that these genes showed obvious dynamic gradient expression changes in normal, peritumoral and malignant tumor tissues. At the same time, some other host genes also showed significant enrichment in specific cell types, such as PTPN22 and PLCE1 in CD8+T cells, PRKAR1B in endothelial cells, SPATA13 in Tcm+Tfh+Treg cells, etc. (see Fig. 9 Middle (E) Figure and Fig.10 Among them, PTPN22 (corresponding to the maternal gene of hsa_circ_0000110) is particularly noteworthy, as it is significantly enriched in plasma B cells in the blood (see Fig. 9 Middle (D), suggesting its possibility as a potential non-invasive biomarker for CRC diagnosis. In-depth analysis of single-cell RNA sequencing data revealed the dynamic changes in the expression of host genes corresponding to DEcircRNA in different cell types during tumor progression, highlighting the potential role of circRNA in cell transformation.

[0076] In the single-cell analysis, the Seurat R package (v4.2.3) in R (v4.3) was used for data processing and analysis. The specific analysis and screening conditions were as follows: (1) Data preprocessing: The screening conditions included that each gene was expressed in at least 3 cells, each cell expressed at least 250 genes, and the gene expression level was between 100 and 7500, the mitochondrial proportion was less than 35%, and the UMI>1000; (2) Data normalization and feature selection: The “NormalizeData” function was used to normalize the data. The “FindVariableFeatures” function was used to identify variable genes based on the variance stabilization transformation. The variable genes were standardized by the “ScaleData” function; (3) Principal component analysis (PCA) and clustering: PCA was used to reduce the dimensionality of the data, and the optimal dimensionality reduction parameters were determined by the “JackStraw” and “ElbowPlot” functions. Cell clustering was performed using the “FindNeighbors” and “FindClusters” functions; (4) Cell type annotation: cell types were annotated using human cell marker genes in the CellMarker database to identify and classify different cell populations.

[0077] Comparative Example 1

[0078] The RPMs of hsa_circ_0003681 disclosed in CN117568476A, hsa_circ_0003761 disclosed in CN113667750A, and hsa_circ_0068464 disclosed in CN113249478A were used to distinguish tumor tissue (T) from the other two tissues (N and P), and the ROC curves were drawn using Graphpad Pism9.5 software.

[0079] Among them, hsa_circ_0003681 disclosed in CN117568476A could not be detected in 90 clinical samples, which may be due to low abundance of circRNA or incorrect data in the patent. The AUC curves of hsa_circ_0003761 disclosed in CN113667750A and hsa_circ_0068464 disclosed in CN113249478A are shown in Figure 2. Fig.11 As shown, the areas under the curve of these two circRNAs are significantly lower than those of the 8 circRNAs in this application.

[0080] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A biomarker for colorectal cancer detection, characterized in that: include: hsa_circ_0008039, hsa_circ_0004865, hsa_circ_0077837, hsa_circ_0019223, hsa_circ_0000110, hsa_circ _0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663.

2. The biomarker according to claim 1, characterized in that in, hsa_circ_0008039, hsa_circ_0004865 and hsa_circ_0077837 are used to detect and distinguish early and late stage patients of colorectal cancer; hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663 are used to detect and distinguish colorectal cancer malignancies and normal tissues.

3. Use of the biomarker described in claim 1 or 2 in preparing a colorectal cancer detection model.

4. A method for constructing a colorectal cancer detection model, characterized in that: The colorectal cancer detection model comprises the biomarker according to claim 1 or 2, and the construction method comprises the following steps: Step 1: DESeq2 method was used to identify and screen circRNAs in normal colorectal tissues, paracancerous tissues, and malignant tumor tissues to obtain circRNA dataset 1; Step 2, using the WGCNA method to identify and screen the circRNAs in the normal colorectal tissues, paracancerous tissues, and malignant tumor tissues that were the same as in step 1, to obtain circRNA dataset 2; Step 3, screen out the circRNAs that intersect in circRNA dataset 1 and circRNA dataset 2 to obtain a candidate circRNA dataset, and calculate the ROC value and P value of the candidate circRNA dataset; Step 4, the expression trend consistency of the candidate circRNA datasets was verified by using the two circRNA datasets GSE221240 and GSE235850, and the ROC value and P value of the candidate circRNA datasets were combined to further obtain the biomarkers and construct a colorectal cancer detection model.

5. The construction method according to claim 4, characterized in that: In step 1, the control parameters for identification and screening are: |log2Fold change|≥1, baseMean≥3.51, P-value≤0.05 and Padj≤0.

05.

6. The construction method according to claim 4, characterized in that: In step 1, the WGCNA method is used to identify and screen the circRNA of the normal colorectal tissue, paracancerous tissue and malignant tumor tissue that are the same as step 1, including the following steps: Step 2-1, primary screening of all tissue samples; Step 2-2, calculate the correlation of all tissue circRNAs by soft thresholding, and determine the circRNA dataset 2 by hierarchical clustering; Preferably, the soft threshold β=3.

7. The construction method according to claim 4, characterized in that: In step 4, the candidate circRNA dataset is verified by GSE221240 and GSE235850 datasets, and the Log2FoldChange values ​​of circRNAs in the GSE221240 and GSE235850 datasets and the candidate circRNA dataset are calculated using the DESeq2 method, and the consistency of expression trends is compared.

8. A colorectal cancer detection model obtained by the construction method according to any one of claims 4 to 7.

9. Use of the biomarker according to claim 1 or 2 or the colorectal cancer detection model according to claim 8 in the preparation of a colorectal cancer detection product.

10. A kit for detecting and distinguishing colorectal cancer malignant tumors and normal tissues, characterized in that: Including hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663 biomarkers or colorectal cancer detection models formed by the aforementioned biomarkers, and reagents for detecting the expression of hsa_circ_0019223, hsa_circ_0000110, hsa_circ_0073244, hsa_circ_0087960, hsa_circ_0001461, hsa_circ_0001900, hsa_circ_0000467 and hsa_circ_0001663 biomarkers.

11. The kit according to claim 10, characterized in that Primers for hsa_circ_0019223 are shown in SEQ ID NO.1 and SEQ ID NO.2; primers for hsa_circ_0000110 are shown in SEQ ID NO.5 and SEQ ID NO.6; primers for hsa_circ_0073244 are shown in SEQ ID NO.7 and SEQ ID NO.8; primers for hsa_circ_0087960 are shown in SEQ ID NO.13 and SEQ ID NO.14; primers for hsa_circ_0001461 are shown in SEQ ID NO.3 and SEQ ID NO.4; primers for hsa_circ_0001900 are shown in SEQ ID NO.9 and SEQ ID NO.10; primers for hsa_circ_0000467 are shown in SEQ ID NO.11 and SEQ ID NO.12; primers for hsa_circ_0001663 are shown in SEQ ID NO. NO.15 and SEQ ID NO.

16.

12. A specific primer for detecting the biomarker according to claim 1 or 2, characterized in that: Primers for hsa_circ_0019223 are shown in SEQ ID NO.1 and SEQ ID NO.2; primers for hsa_circ_0000110 are shown in SEQ ID NO.5 and SEQ ID NO.6; primers for hsa_circ_0073244 are shown in SEQ ID NO.7 and SEQ ID NO.8; primers for hsa_circ_0087960 are shown in SEQ ID NO.13 and SEQ ID NO.14; primers for hsa_circ_0001461 are shown in SEQ ID NO.3 and SEQ ID NO.4; primers for hsa_circ_0001900 are shown in SEQ ID NO.9 and SEQ ID NO.10; primers for hsa_circ_0000467 are shown in SEQ ID NO.11 and SEQ ID NO.12; primers for hsa_circ_0001663 are shown in SEQ ID NO. NO.15 and SEQ ID NO.16.

Images