Open Access

Potential prognostic and predictive value of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 in colorectal cancer stool specimens

  • Authors:
    • Yu-Nung Chen
    • Cheng-Yen Shih
    • Shu-Lin Guo
    • Chih-Yi Liu
    • Ming-Hung Shen
    • Shih-Chang Chang
    • Wei-Chi Ku
    • Chi-Cheng Huang
    • Chi-Jung Huang
  • View Affiliations

  • Published online on: February 7, 2023     https://doi.org/10.3892/br.2023.1604
  • Article Number: 22
  • Copyright: © Chen et al. This is an open access article distributed under the terms of Creative Commons Attribution License.

Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )
Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )


Abstract

Colorectal cancer (CRC) is the most common gastrointestinal malignancy worldwide. The poor specificity and sensitivity of the fecal occult blood test has prompted the development of CRC‑related genetic markers for CRC screening and treatment. Gene expression profiles in stool specimens are effective, sensitive and clinically applicable. Herein, a novel advantage of using cells shed from the colon is presented for cost‑effective CRC screening. Molecular panels were generated through a series of leave‑one‑out cross‑validation and discriminant analyses. A logistic regression model following reverse transcription‑quantitative polymerase chain reaction (RT‑qPCR) and immunohistochemistry was used to validate a specific panel for CRC prediction. The panel, consisting of ubiquitin‑conjugating enzyme E2 N (UBE2N), inosine monophosphate dehydrogenase 1 (IMPDH1), dynein cytoplasmic 1 light intermediate chain 1 (DYNC1LI1) and phospholipase A and acyltransferase 2 (HRASLS2), accurately recognized patients with CRC and could thus be further investigated as a potential prognostic and predictive biomarker for CRC. UBE2N, IMPDH1 and DYNC1LI1 expression levels were upregulated and HRASLS2 expression was downregulated in CRC tissues. The predictive power of the panel was 96.6% [95% confidence interval (CI), 88.1‑99.6%] sensitivity and 89.7% (95% CI, 72.6‑97.8%) specificity at a predicted cut‑off value at 0.540, suggesting that this four‑gene panel testing of stool specimens can faithfully mirror the state of the colon. On the whole, the present study demonstrates that screening for CRC or cancer detection in stool specimens collected non‑invasively does not require the inclusion of an excessive number of genes, and colonic defects can be identified via the detection of an aberrant protein in the mucosa or submucosa.

Introduction

Gastrointestinal cancers encompass a variety of malignant diseases, with only colorectal cancer (CRC) ranked among the most common tumors (1). The integrated conventional staging system (T, tumor size; N, lymph node status; M, distant metastasis) and molecular classifications of CRC may aid in reliable personalized treatments and may contribute to the prediction of the prognosis of cancer (2-4) Moreover, the success of CRC screening and detection must depend significantly on molecular parameters and not exclusively on clinical stage (5,6).

Based on findings concerning peripheral blood possibly reflecting changes that occur in tissues (7,8), numerous molecular parameters from blood specimens have been reported for CRC detection (9-11). As this type of test does not directly include samples of colonic cell origin, the molecular parameters of numerous genes should be composed to achieve adequate sensitivity and specificity (9,11). Compared with the results of other groups that have used blood specimens for screening, a previous study by the authors involving the extraction of expression profiles from stool specimens revealed a good association of these profiles with CRC progression and recurrence (12-14). The direct detection of changes in gene expression in colonic tissues may contribute to a further understanding of CRC progression and may allow the development of biomarkers and drug targets for this malignant disease (15). Human stool has been studied for CRC screening for several years (16-18), and several lines of evidence have indicated that cells shed from the colonic tract may reflect localized diseases (19-21). Thus, either DNA or RNA extracted from stool specimens can be used to detect colorectal neoplasia accurately (22-24). Genes that are actively expressed in human stool specimens have emerged as specific molecular signatures of CRC (25,26). These genetic molecules may aid in the understanding of the process of the development of CRC (27,28). Moreover, the signatures concurrently reflect CRC biology, and inform prognoses and treatment responses in a non-invasive manner (29). The gene expression status of colonic cells that pass into stool specimens has been considered to faithfully represent CRC manifestations (30-33).

Herein, the novel, to the best of our knowledge, advantage of using cells shed from the colonic tract for cost-effective CRC screening is presented. Previously, genes expressed in human stool specimens that were used to identify patients with differentially staged CRC distinguishing them from healthy donor control samples were acquired using whole-genome microarrays. Expressed genes were further filtered via custom-made microarrays. Specific gene sets were tested with a small number of testing samples using reverse transcription-quantitative polymerase chain reaction (RT-qPCR) analyses, trained with other samples set by a series of leave-one-out cross-validation (LOOCV) and discriminant analyses (34,35), and then confirmed with the third sample set for the probabilities of group membership (healthy donor or CRC). The corresponding proteins of target genes were assessed based on CRC tissue arrays. Furthermore, a logistic regression model was used to predict diseases for a specific molecular panel (36,37).

Materials and methods

Study participants and ethical approval

Human stool specimens were collected from 29 healthy donor controls (age range, 23-78 years; 8 males and 21 females) and 58 patients with well-diagnosed CRC (age range, 29-87 years; 33 males and 25 females) at the Sijhih Cathay General Hospital for predictive model training. All the participating subjects provided their written informed consent. The research was conducted with the obtained approval (approval no. CGH-P101014) of the Institutional Review Board of the Cathay General Hospital and according to the Principles of the Declaration of Helsinki.

All 87 cases (29 healthy donors and 58 CRC cases) were randomly divided into three independent sets that were used for different purposes as follows: Set I (n=11, five healthy donor and six CRC cases); set II (n=56, 20 healthy donor and 36 CRC cases); and set III (n=20, four healthy donor and 16 CRC cases). The initial tumor stage of patients with CRC was classified using the 8th Edition of the American Joint Committee on Cancer (AJCC) staging system (38), and the healthy donor controls were requested to undergo a colonoscopy examination. All patients enrolled in the present study were managed according to standard guidelines, with regular follow-up. To characterize the targets of interest in CRC samples, the proteins encoded by these genes were immunostained in colon cancer tissue arrays (COC1021; total 102 available cores, including two non-CRC colonic tissues, one congenital megacolon, two colon adenomas, four papillary adenocarcinomas, seven mucinous adenocarcinomas, and 86 colon adenocarcinomas; Pantomics, Inc.). According to the provided TNM classification from Pantomics, Inc., 22 tissue cores with adenocarcinoma were diagnosed as AJCC stage I, 39 were AJCC stage II, and 36 were AJCC stage III (Fig. 1) (39). To validate a predictive model, other stool specimens and one cDNA array of colonic tissues were used. Briefly, stool specimens of two female patients (59 and 71 years of age) during pre- and post-surgical treatments at the Sijhih Cathay General Hospital (New Taipei City, Taiwan) and 119 individuals (age range, 26-76 years; 67 males and 52 females) following a colonoscopy at Cathay Healthcare Management (Taipei City, Taiwan) were collected. In addition, one cDNA array (HCRT104; OriGene Technologies, Inc.) of 48 colon tissues covering non-CRC status (n=8) and four CRC stages (n=40) was purchased to examine the predictive model.

LOOCV, discriminant analyses and predictive model for CRC risk

The stool specimens from set I were initially applied to screen out genes with a differential expression in previous microarray hybridizations performed by the authors (14,40). A LOOCV analysis was then performed for the additional screened genes on a given set (set II) as a learning system. Briefly, each stool specimen was excluded in turn and classified using a defined model based on the non-excluded samples (41,42). Using the cases in set II as the well-defined group, significant LOOCV-estimated molecular panels were used to test the probabilities of group membership (healthy donor or CRC) in the testing sets (sets I and III) via discriminant analysis. As a result, a predictive model of an optimal molecular panel for CRC risk was produced from all 87 well-known cases (sets I, II, and III) via the analysis of the logistic regression model (43,44), and a receiver operating characteristic (ROC) curve was generated to assess model discrimination (45).

Relative gene expression quantification

The method of purifying total RNA from stool mud [0.5 g stool in 1 ml guanidinium thiocyanate buffer; 10 mM Tris (pH 7.4), 200 mM NaCl, 1 mM EDTA (pH 8.0), 4 M guanidinium thiocyanate, and 1% β-mercaptoethanol] was largely described in previous reports by the authors (40,46). An appropriate supernatant was then extracted using a MagCore Nucleic Acid Extract kit in a MagCore Nucleic Acid Extractor (RBC Bioscience Corp.). The eluted fecal total RNA was then quantified using a NanoDrop ND 1000 spectrophotometer (Thermo Fisher Scientific, Inc.) and reverse-transcribed to generate single-stranded cDNAs using random primers and a PowerScript Reverse Transcriptase Kit (cat. no. RR037B; Takara Bio USA, Inc.), according to the manufacturer's instructions.

The genes of interest were quantified by a program (10 min at 95˚C, proceeding with 60 cycles at 95˚C for 10 sec and at 60˚C for 20 sec) in the presence of a TaqMan probe and TaqMan Master Mix using a Roche LightCycler nano system (Roche Diagnostics GmbH) according to the manufacturer's instructions. Briefly, all genes were quantified relative to the level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) for each quantification. All primers and probes used in the present study are listed in Table I. LightCycler Software (version 4.05; Roche Diagnostics GmbH) was used to analyze the PCR kinetics. Expression levels were quantified using the 2-ΔΔCq method (47) and normalized to the expression level of GAPDH (48). Each run also included an appropriate and predetermined diluted human reference cDNA (Takara Bio USA, Inc.), which was used as a standard to estimate relative expression levels.

Table I

RT-qPCR primers and probe numbers for gene expression quantification.

Table I

RT-qPCR primers and probe numbers for gene expression quantification.

Gene nameAccession no.Sequence (from 5' to 3')UPL no.
GAPDHNM_002046Fw: CTCTGCTCCTCCTGTTCGAC#60
  Rv: ACGACCAAATCCGTTGACTC 
UBE2NNM_003348.3Fw: AAGCCCAAGCCATAGAAACA#2
  Rv: ATGCAAACAAAGAGGAGGAAGT 
AKIRIN1NM_024595.1Fw: ACTCCTCAGCACTCACAGCA#80
  Rv: CCAACTTGTCGGAGGGTAAA 
IMPDH1NM_000883.3Fw: GTCCATGGCCTGCACTCT#22
  Rv: GTGGACACTGGGGTGCAT 
SLC15A4NM_145648.3Fw: GAGCAGTCACACAGACTTTGGT#71
  Rv: CAGGAGGGTAGCTCCTTGAA 
DYNC1LI1NM_016141.2Fw: CTGGTGTGAGTGGTGGTAGC#10
  Rv: TCTGCATGAACATCTAAGACAGG 
HRASLS2NM_017878.1Fw: ATCTGCGCTATGGCGTCT#74
  Rv: CAGCAGGATCCCCACAAG 
APOA1NM_000039.1Fw: CCTTGGGAAAACAGCTAAACC#39
  Rv: CCAGAACTCCTGGGTCACA 
STK17BNM_004226.2Fw: GGAAATCATGGGAACACCAG#50
  Rv: TTGCTGTGGTAATGGGATCAT 

[i] UPL, Roche universal probe library; UBE2N, ubiquitin-conjugating enzyme E2N; AKIRIN1, akirin 1; IMPDH1, inosine monophosphate dehydrogenase 1; SLC15A4, solute carrier family 15, member 4; DYNC1LI1, dynein, cytoplasmic 1 light intermediate chain 1; HRASLS2, phospholipase A and acyltransferase 2; APOA1, apolipoprotein A-I; STK17B, serine/threonine kinase 17b.

Immunohistochemistry

COC1021 tissue arrays (Pantomics, Inc.), which have defined clinical diagnosis and clinicopathological information, were used for immunohistochemical staining (49). Firstly, tissue arrays with 4 µm core thickness were deparaffinated and hydrated using routine protocols. Antigen retrieval was performed by steaming in Tris-EDTA buffer (pH 9.0) for 20 min, and arrays were then blocked in 1.5% (v/v) Normal Horse Serum Blocking Solution (Vector Laboratories, Inc.) for 2 h at room temperature. The specific target protein was immunodetected using an adequate primary antibody concentration [ubiquitin-conjugating enzyme E2 N (UBE2N): anti-Ube2N, cat. no. ab117090, dilution 1:20, Abcam; inosine monophosphate dehydrogenase 1 (IMPDH1): anti-IMPDH1, cat. no. ab84957, 1:20, Abcam; dynein cytoplasmic 1 light intermediate chain 1 (DYNC1LI1): anti-DLC-A, cat. no. ab154251, 1:20, Abcam; phospholipase A and acyltransferase 2 (HRASLS2): anti-HRASLS2, cat. no. bs-6013R, 1:1,000, Thermo Fisher Scientific, Inc.] in the blocking solution mentioned above at 4˚C overnight. Endogenous peroxidases in tissue sections were removed by incubation with 0.3% H2O2 for 15 min, and these peroxidase-free arrays were further incubated with a biotinylated secondary antibody (either goat anti-rabbit, cat. no. BA-1000-1.5, 1:200, Vector Laboratories, Inc. or rabbit anti-goat immunoglobulin G, cat. no. BA-5000-1.5, 1:200, Vector Laboratories, Inc.) at room temperature for 60 min. The VECTASTAIN ABC system and DAB Substrate kit (both from Vector Laboratories, Inc.) were used to develop the secondary antibodies captured on arrays, according to the manufacturer's instructions. Following hematoxylin (GHS3; 50 ml, Merck KGaA) counterstaining at room temperature for 5 min and slide mounting with Malinol (Muto Pure Chemicals, Co., Ltd.), images were digitalized using a high-resolution scanner (Mirax Scan, Carl Zeiss AG) at the Taiwan Mouse Clinic (Academia Sinica, Taipei City, Taiwan). Two independent pathologists reviewed and evaluated the imaging results. QuPath (Version 0.3.0; https://qupath.github.io) was employed to produce the cell densities of immunoreactive cells per mm2 for all target proteins (50).

Statistical analyses

Differences in the relative levels of the target mRNAs among the various samples were determined using the unpaired non-parametric two-sample Mann-Whitney U test. The equation for the probability of CRC was acquired via a logistic regression model. In addition, one-way analysis of variance (ANOVA) was performed to compare the densities of positive signaling cells for the immunohistochemical staining of target proteins. All ANOVA analyses were followed by a Bonferroni post hoc test. These deduced probabilities and identifications based on 119 individuals after colonoscopy were compared using the chi-squared test. To assess the effectiveness of the predictive model of CRC risk, a ROC curve was plotted between the sensitivity and (1-specificity), and the total area under the ROC curve (AUC) was calculated. All statistical analyses were performed with IBM SPSS Statistics for Windows (Version 22.0; IBM Corp.). P<0.05 was considered to indicate a statistically significant difference.

Results

Development of gene panels for patients with CRC

All genes were selected according to different clinical statuses from two microarray hybridizations, whole-genome oligonucleotides and custom-made microarrays with ~4,000 genes from previous studies by the authors (14,40,51). Three independent sets of stool specimens (sets I, II and III) were then used to validate gene expression by using a series of RT-qPCR assays in order to select the optimal stool-based panels to predict CRC probability. From sample set I, eight differentially expressed genes [UBE2N, akirin 1 (AKIRIN1), IMPDH1, solute carrier family 15 member 4 (SLC15A4), DYNC1LI1, HRASLS2, apolipoprotein A1 (APOA1) and serine/threonine kinase 17b (STK17B)] (P<0.2; Mann-Whitney U test) were in accordance with the corresponding expression trends obtained from the custom-made microarray hybridization (Table SI). However, six genes only (UBE2N, IMPDH1, SLC15A4, DYNC1LI1, HRASLS2 and STK17B) were differentially expressed in sample set II, which was used as the training set with statistical significance (P≤0.05; Mann-Whitney U test) (Fig. 1A). These six genes comprised 63 different molecular panels. Moreover, using the same samples of set II, each panel was used to predict the healthy donor or CRC disease status using LOOCV analysis. As demonstrated in Fig. 1B, the diagnostic accuracy of each panel for healthy donor controls or patients with CRC was ≥50%, and seven different panels (UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 genes; UBE2N, IMPDH1, HRASLS2 and STK17B genes; UBE2N, DYNC1LI1, HRASLS2 and STK17B genes; IMPDH1, SLC15A4, HRASLS2 and STK17B genes; UBE2N, IMPDH1, SLC15A4, DYNC1LI1 and STK17B genes; UBE2N, IMPDH1, SLC15A4, HRASLS2 and STK17B genes; UBE2N, IMPDH1, SLC15A4, DYNC1LI1, HRASLS2 and STK17B genes) yielded higher sensitivity (≥90%) and specificity (≥85%). A four-gene panel composed of only three upregulated genes (UBE2N, IMPDH1 and DYNC1LI1) and one downregulated (HRASLS2) in patients with CRC exhibited the highest accuracy for healthy donor controls (88.9%, eight of nine) and patients with CRC (95.4%, 21 of 22) in predicting the testing sets (sets I and III) and using sample set II as the training set, as assessed by discriminant analyses.

Relative protein expression levels of the four-gene panel in CRC tissues

Subsequently, it was investigated whether the corresponding proteins of four genes were differentially expressed in tumor samples from patients with CRC. An array method based on established immunohistochemistry protocols was used to detect the corresponding protein expression in human colonic tissues (Fig. 2 for overview of COC1021 tissue array with 102 cores (x-axis, nos. 1 to 13; y-axis, a to h); Fig. 3 for immunohistochemical staining of HRASLS2; Fig. 4 for immunohistochemical staining of UBE2N; Fig. 5 for immunohistochemical staining of IMPDH1; Fig. 6 for immunohistochemical staining of DYNC1LI1). Moreover, CRC tissues at different stages with CRC and non-CRC fractions in the same donor of the COC1021 tissue array revealed that the signal density (positive cells/mm2) of UBE2N (Fig. 4B: 2,139 cells/mm2 in c12, 2,802 cells/mm2 in e7, and 3,156 cells/mm2 in g9), IMPDH1 (Fig. 5B: 1,729 cells/mm2 in c12, 732 cells/mm2 in e7, and 2,234 cells/mm2 in g9), or DYNC1LI1 (Fig. 6B: 3,905 cells/mm2 in c12, 2,272 cells/mm2 in e7, and 2,061 cells/mm2 in g9) was high in the CRC fractions and low in the adjacent non-CRC fractions (Fig. 4B: 1,050 cells/mm2 in c12, 1,411 cells/mm2 in e7, and 556 cells/mm2 in g9; Fig. 5B: 452 cells/mm2 in c12, 625 cells/mm2 in e7, and 1,011 cells/mm2 in g9; Fig. 6B: 1,523 cells/mm2 in c12, 1,814 cells/mm2 in e7, and 785 cells/mm2 in g9). By contrast, HRASLS2 was expressed with a high density of positive cells in the non-CRC fractions (1,993 cells/mm2 in c12, 3,538 cells/mm2 in e7, and 545 cells/mm2 in g9), but relatively less densely expressed in the adjacent CRC fractions (1,228 cells/mm2 in c12, 3,153 cells/mm2 in e7, and 403 cells/mm2 in g9) (Fig. 3B).

A logistic regression model to predict CRC disease

Three sets of stool specimens (n=87) were then pooled to produce a binary logistic regression equation as follows: ProbCRC=1/1 + e-z, where ProbCRC is the probability of CRC disease, ‘e’ denotes the exponential function, and ‘z’ is equal to 5.468 + 1.394 x log(UBE2N) + 0.859 x log(IMPDH1) + 1.129 x log(DYNC1LI1)-1.471 x log(HRASLS2). The overall accuracy of this model to predict CRC in the participants was 94.3%. The sensitivity was 96.6% (56 out of 58), and the specificity was 89.7% (26 out of 29). The positive predictive value (PPV) was 95.1% (56 out of 59), and the negative predictive value (NPV) was 93.9% (26 out of 28). Thus, this logistic regression model was used for the prediction of increased risk for CRC in patients, among new study participants. Furthermore, the AUC for the assessment of the model discrimination was 0.957 [95% confidence interval (CI), 0.914-1.000; P<0.01] (Fig. 7). The cut-off value for predicting CRC (ProbCRC) was 0.540, with 96.6% (95% CI, 88.1-99.6%) sensitivity and 89.7% (95% CI, 72.6-97.8%) specificity (Table II).

Table II

Sensitivity and specificity for predicting colorectal cancer.

Table II

Sensitivity and specificity for predicting colorectal cancer.

CriterionSensitivity (%)95% CISpecificity (%)95% CIAccuracy (%)
0.00698.390.8-100.06.90.8-22.869.2
0.26498.390.8-100.075.956.5-89.790.1
0.54096.688.1-99.689.772.6-97.894.5
0.71482.870.6-91.493.177.2-99.286.8
0.95062.148.4-74.596.682.2-99.974.7

[i] CI, confidence interval.

Model validity of patients with CRC and healthy individuals

First, pre- and post-surgical stool specimens were obtained from two patients with CRC (pCCG1: AJCC stage III; pCCG2: AJCC stage II). In addition, stool specimens were obtained from another patient with CRC (a 49-year-old female) as a positive CRC control and a healthy donor individual (a 72-year-old male) with negative colonoscopy results as a negative control. As demonstrated in Fig. 8A, the highest positive predicted values for CRC disease were detected for the two pre-surgical samples (pCCG1, 0.905; pCCG2, 1.000) as compared with the healthy donor control (0.199), as assessed using the logistic regression model of the four-gene panel. By contrast, markedly lower predicted values were detected in post-surgical samples (0.093 for pCCG1; 0.284 for pCCG2), which were prepared from the feces of two patients whose tumors were completely removed. It is notable that the timepoint of collection of the postsurgical samples was at least 1 month after surgery, with patient pCCG1 receiving further chemotherapy.

This specific four-gene panel was then used to test the 119 stool specimens that were obtained before colonoscopy examination. Briefly, 112 cases presented with various hemorrhoids, 51 with polyps, eight with colitis, four with CRC, and only one was entirely healthy. Among the 43 cases (36.1%, 43 out of 119) with a higher ProbCRC (>0.540), four were proven to have CRC (mean ProbCRC=0.968), five had colitis (mean ProbCRC=0.991) and 16 had polyps (mean ProbCRC=0.834) following a colonoscopy examination (Table III). Conversely, up to 62.5% (5 out of 8) of the cases with colitis or proctitis were predicted to be positive, with only 31.4% (16 out of 51) of the cases with polyps exhibiting increased predictive rates.

Table III

Prediction of CRC by the four-gene panel for the 119 patients with CRC.

Table III

Prediction of CRC by the four-gene panel for the 119 patients with CRC.

FeaturesCase no.Probability of CRC (%) >0.540 (positive no.)P-value
Age, years   
     <495736.8% (21)0.878
     ≥496235.5% (22) 
Sex   
     Male6541.5% (27)0.178
     Female5429.6% (16) 
CRC   
     Negative11533.9% (39)0.016
     Positive4100.0% (4) 
Colitis/proctitis   
     Negative10731.8% (34)0.077
     Positive862.5% (5) 
Polyp   
     Negative6435.9% (23)0.607
     Positive5131.4% (16) 
Hemorrhoid   
     Negative333.3% (1)0.265
     Internal9437.2% (35) 
     External250.0% (1) 
     Mixed1612.5% (2) 

[i] CRC, colorectal cancer.

Finally, the ProbCRC of each case was calculated in the HCRT104 cDNA array composed of 48 colonic tissues. The mean ProbCRC of 40 patients with CRC was 0.913 (range: 0.535-0.997), which was significantly higher than that of eight non-CRC colon tissues (0.459; range: 0.042-0.740) (Fig. 8B). In particular, the mean ProbCRC of 10 patients with CRC with distant metastasis was up to 0.973 (range, 0.924-0.997) (Table SII). Using this predictive model, 39 patients with CRC were identified as positive cases (ProbCRC >0.540), and the PPV was up to 97.5% (39 of 40). In addition, five non-CRC cases were correctly diagnosed as negative cases (ProbCRC ≤0.540), with the corresponding NPV being 62.5% (5 of 8) (Table SIII).

Discussion

The early detection of CRC would permit timely surgical intervention in patients and this in turn would halt tumor progression, thus enabling effective therapy (52,53). Sigmoidoscopy and colonoscopy are simultaneously used at present for the early detection of CRC. However, these are invasive techniques (54-56). By contrast, clinical examinations are limited due to their risk and inconvenience. Consequently, a non-invasive fecal occult blood test is currently and widely applied for CRC screening, even though it has poor specificity and sensitivity (57).

CRC is believed to develop slowly via the progressive accumulation of genetic mutations (58,59). Genes involved in human diseases, tumorigenesis, or even those with unrecognized functions are potential markers that may aid in the diagnosis of CRC. A variety of molecular approaches are emerging to improve the effectiveness and user-friendliness of non-invasive CRC screening (60,61). In the present study, it was demonstrated that limited genetic markers rarely discussed in CRC were able to screen or detect CRC. In other words, it is not necessary to examine additional genes in the field of clinical practice for CRC when the sample type used for examination is stool instead of blood (61,62). For example, assays that aim to determine early alterations in gene expression in stool specimens due to malignancy, appear to be a promising alternative to the detection methods currently used (23,63). In fact, stool specimens have been used as a representative of CRC manifestations for a number of years (30,64,65). Either CRC tissues or human stool specimens can serve as an appropriate object to explore CRC development (27,28). Genetic information from these two clinical materials is significantly interrelated (56,66). This is also reflected in the results of the present study, where the differentially expressed genes detected in the stool specimens of patients with CRC were also detected at the mRNA and protein levels, in CRC tissues. These findings suggest that a high gene expression in stool specimens may indicate the presence of colorectal neoplasia in the colonic tract and may provide an instrument for understanding CRC development (67-69). Moreover, this non-invasive approach can greatly enhance screening acceptance and be implemented in public health services due to its relative ease of use and cost-effectiveness (52).

Specific gene expression profiles have been used to classify tumors or predict prognoses (70,71). In CRC research, whole-genome analyses of human stool specimens have been performed by several groups (14,72-74). Miyake et al (2) revealed that a discriminator gene set could predict metastases from tissues of patients with early-stage CRC. As an example of translation from basic science to clinical care, these results suggested that understanding the molecular mechanisms of CRC tumorigenesis may lead to new approaches for the diagnosis of CRC. The results of the present study provide evidence to support this concept (75). Actually, eight differentially expressed genes (UBE2N, AKIRIN1, IMPDH, SLC15A4, DYNC1LI1, HRASLS2, APOA1 and STK17B) were originally selected. However, six genes (UBE2N, IMPDH, SLC15A4, DYNC1LI1, HRASLS2 and STK17B) were used for the final LOOCV analysis in the present study, after validating all corresponding relative expression using a series of RT-qPCR assays for sample set II. Finally, only four genes (UBE2N, IMPDH1, DYNC1LI1 and HRASLS2) presented with high sensitivity (96.6%) and high sensitivity (89.7%). As a result, it was ultimately decided to include only four, rather than eight or six genes, for subsequent experiments in the present study. Stool-based assays were used to detect CRC and proposed this specific gene panel (four-gene panel: UBE2N, IMPDH1, DYNC1LI1 and HRASLS2) that exhibited the best predictive power. Furthermore, apart from achieving a high sensitivity for the detection of CRC in patients using this four-gene panel, it also identified cases with a high risk for CRC, including cases presenting with inflammation (e.g., colitis). This finding may indicate that individuals with colitis are at an increased risk of developing CRC (76). In addition, the panel used in the present study may be applied to disease tracking and treatment strategies due to the reduced positive predicted values obtained for the post-surgical stool specimens. This implies that the results of the present study may be used for the evaluation and selection of the best treatment option for individual patients and may allow for more intensive follow-up examinations for post-surgical or post-chemotherapeutic patients with CRC.

Varying levels of gene expression in stool specimens were reflected at the respective protein and mRNA levels in the tissues using immunohistochemistry and the CRC cDNA array, respectively. As previously reported by the authors, DYNC1LI1 was expressed at a significantly higher level in cDNA samples from patients with distant metastasis than in cDNA samples from non-metastatic patients and in normal colonic tissues (77). In addition, increased (UBE2N and IMPDH1) and decreased (HRASLS2) cDNA levels were observed to be statistically significant for CRC tissues of all stages (P=0.020 for UBE2N, P=0.032 for IMPDH1 and P<0.001 for HRASLS2) compared with those detected in eight cDNA samples of non-CRC tissues in the HCRT104 cDNA array (Table SIV). The densities of positive signaling cells for HRASLS2 (Fig. S1A), UBE2N (Fig. S1B), IMPDH1 (Fig. S1C), and DYNC1LI1 (Fig. S1D) varied with the grade of differentiation. Briefly, poorly differentiated (G3) CRC had relatively low HRASLS2- and high UBE2N-expressing cell densities, compared to that of normal/benign tissues or well-differentiated (G1) and moderately differentiated (G2) CRCs with statistical significance (P<0.05 for HRASLS2, G2 vs. G3; P<0.0001 for UNE2N, G1 vs. G3 and G2 vs. G3). Although there was no statistically significant difference in the expression of IMPDH1 and DYNC1LI1 among the groups, their expression levels in differently differentiated CRC tissues still exhibited a tendency to increase gradually with a poorer differentiation. However, further analyses using a greater number of samples are required to confirm these results. The product of UBE2N, also known as UBC13, is a member of the E2 ubiquitin-conjugating enzyme family (78). These ubiquitin ligase components may function as oncogenes in several malignancies, including CRC (79,80). This may also apply to diffuse large B-cell lymphoma, in which the reduced expression of UBE2N inhibits tumor cell survival (81). Accordingly, it was inferred that this aberrant expression in CRC may upregulate the UBC13-p53 complex, which may impede the normal function of the p53 tumor suppressor (82,83). The second upregulated gene, IMPDH1, is also a crucial factor in p53-dependent growth regulation (84). IMPDH1 upregulation is tightly associated with proliferative phenotypes, including malignancy (85). Taken together with the results of the present study, it may be suggested that the increased expression of IMPDH1 in colonic cells is closely related to the occurrence of CRC. Moreover, the overexpression of the protein products of UBE2N and IMPDH1 in the mucosa and submucosa of the colonic tract were also observed. This may imply that the invasion of CRC cells into the inner cellular layers can also be encountered in stool specimens. The results of the present study may provide evidence that human stool specimens are suitable for CRC detection (14,86). In addition to the genes that were upregulated, HRASLS2, which is the only gene that was downregulated in this four-gene panel, was reported to suppress the growth of cancer cells and might be a tumor suppressor (87,88). Nevertheless, the present study is the first report, to the best of our knowledge, discussing the association between the downregulation of HRASLS2 and CRC. Moreover, the loss of repressing RAS activities may lead to CRC tumorigenesis, as HRASLS2 was downregulated in colonic cells (89).

It is hoped that new insights into colorectal carcinogenesis and personalized prediction can be achieved in a non-invasive manner. One of the key elements in CRC therapy is the ability to detect the disease easily, including using non-invasive screening methods (90,91). The results of the present study suggest that stool specimens may faithfully reproduce the actual health status of the colonic tract. Using this four-gene panel, it was demonstrated that CRC screening or detection in stool specimens can be performed without the inclusion of an excessive number of genes and that this type of sample, which is collected non-invasively, being also able to reflect colonic defects via the determination of aberrant protein expression in the colonic mucosa or submucosa. It is possible to evaluate chemotherapeutic efficacy or further elucidate the selection of the optimal treatment option from stool specimens of individual cases, anytime and anywhere. However, since RNA quality is critical for obtaining strong prognostic and predictive values, fresh stool specimens or stool in qualified storage buffers are of utmost importance and a limitation of this assay. By contrast, to improve the accuracy of this analysis, further studies are necessary to expand the sample size of training set.

In conclusion, this specific four-gene panel, consisting of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2, represents a clinical tool as potential prognostic and predictive targets in stool specimens for CRC. The overexpression of UBE2N, IMPDH1 and DYNC1LI1, and the downregulation of HRASLS2 was also detected in CRC tissues. This molecular panel can be non-invasively and easily detected from stool specimens and may allow for the assessment of CRC screening, treatments, and follow-ups. The present study comprehensively demonstrates the translation of basic science into clinical practice.

Supplementary Material

Cell densities of immunoreactive cells per mm2 from COC1021 colon cancer tissue array. (A) HRASLS2, (B) UBE2N, (C) IMPDH1, (D) DYNC1LI1. QuPath (Version 0.3.0; https://qupath.github.io) was employed to produce the cell densities of immunoreactive cells per mm2 for all target proteins. One-way analysis of variance was performed to compare the densities of positive signaling cells forimmunohistochemical staining of target proteins and followed by the Bonferroni post hoc test. *P<0.05 and ****P<0.0001. HRASLS2, phospholipase A and acyltransferase 2; UBE2N, ubiquitin-conjugating enzyme E2N; IMPDH1, inosine monophosphate dehydrogenase 1; DYNC1LI1, dynein, cytoplasmic 1 light intermediate chain 1. Normal/Benign, two normal colonic tissues, one congenital megacolon, and two adenomas; G1, well-differentiated CRC cells; G2, moderately differentiated CRC cells; G3, poorly differentiated CRC cells.
Analysis of eight differentially expressed genes in stool specimensof 11 cases using theMann-Whitney U test.
ProbCRC in the HCRT104 cDNA array.
Numbers of cases with positive and negative predictive values from the HCRT104 cDNA array.
Analysis of four genes for non-CRC and CRC in the HCRT104 cDNA arrayusing theMann-Whitney U test.

Acknowledgements

The authors would like to thank Dr John Chung-Liang Lee (Vie Longue Biotech, New Taipei 22102, Taiwan) for his assistance with total RNA preparation from stool specimens in the present study.

Funding

Funding: This research was funded by the Ministry of Science and Technology, Taiwan, R.O.C., grant no. MOST109-2314-B-281-001.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

The conceptualization of the present study was performed by CCH and YNC. CYS, CYL, MHS and CJH curated all data analyzed and used in the present study. SCC, WCK and CJH performed the gene quantifications. SLG performed the statistical analysis. CCH, CYL and CJH performed and analyzed the pathological data from the experiments. CCH and SCC acquired resources. CJH supervised the study. CYS, CYL and MHS drafted the original manuscript. CCH, YNC, CYS and CJH reviewed and edited the manuscript. All authors have read and approved the final manuscript. CCH and CJH confirmed the authenticity of all the raw data.

Ethics approval and consent to participate

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Cathay General Hospital, Taiwan, R.O.C. (approval no. CGH-P101014). All the participating subjects provided their written informed consent.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References

1 

Taniguchi H, Moriya C, Igarashi H, Saitoh A, Yamamoto H, Adachi Y and Imai K: Cancer stem cells in human gastrointestinal cancer. Cancer Sci. 107:1556–1562. 2016.PubMed/NCBI View Article : Google Scholar

2 

Miyake M, Takemasa I, Matoba R, Tanino M, Niijima S, Ikeda M, Yamamoto H, Sekimoto M, Kuhara S, Okayama T, et al: Heterogeneity of colorectal cancers and extraction of discriminator gene signatures for personalized prediction of prognosis. Int J Oncol. 39:781–789. 2011.PubMed/NCBI View Article : Google Scholar

3 

Pitroda SP and Weichselbaum RR: Integrated molecular and clinical staging defines the spectrum of metastatic cancer. Nat Rev Clin Oncol. 16:581–588. 2019.PubMed/NCBI View Article : Google Scholar

4 

Wang S, Guan X, Ma M, Zhuang M, Ma T, Liu Z, Chen H, Jiang Z, Chen Y, Wang G and Wang X: Reconsidering the prognostic significance of tumour deposit count in the TNM staging system for colorectal cancer. Sci Rep. 10(89)2020.PubMed/NCBI View Article : Google Scholar

5 

Das V, Kalita J and Pal M: Predictive and prognostic biomarkers in colorectal cancer: A systematic review of recent advances and challenges. Biomed Pharmacother. 87:8–19. 2017.PubMed/NCBI View Article : Google Scholar

6 

Sudoyo AW: Biomolecular markers as determinants of patients selection for adjuvant chemotherapy of sporadic colorectal cancers. Acta Med Indones. 42:45–50. 2010.PubMed/NCBI

7 

Dong L and Ren H: Blood-based DNA methylation biomarkers for early detection of colorectal cancer. J Proteomics Bioinform. 11(120)2018.PubMed/NCBI View Article : Google Scholar

8 

Wilhelmsen M, Christensen IJ, Rasmussen L, Jørgensen LN, Madsen MR, Vilandt J, Hillig T, Klaerke M, Nielsen KT, Laurberg S, et al: Detection of colorectal neoplasia: Combination of eight blood-based, cancer-associated protein biomarkers. Int J Cancer. 140:1436–1446. 2017.PubMed/NCBI View Article : Google Scholar

9 

Marshall KW, Mohr S, Khettabi FE, Nossova N, Chao S, Bao W, Ma J, Li XJ and Liew CC: A blood-based biomarker panel for stratifying current risk for colorectal cancer. Int J Cancer. 126:1177–1186. 2010.PubMed/NCBI View Article : Google Scholar

10 

Rodriguez-Casanova A, Costa-Fraga N, Bao-Caamano A, López-López R, Muinelo-Romay L and Diaz-Lagares A: Epigenetic landscape of liquid biopsy in colorectal cancer. Front Cell Dev Biol. 9(622459)2021.PubMed/NCBI View Article : Google Scholar

11 

Xu Y, Xu Q, Yang L, Ye X, Liu F, Wu F, Ni S, Tan C, Cai G, Meng X, et al: Identification and validation of a blood-based 18-gene expression signature in colorectal cancer. Clin Cancer Res. 19:3039–3049. 2013.PubMed/NCBI View Article : Google Scholar

12 

Huang CC, Shen MH, Chen SK, Yang SH, Liu CY, Guo JW, Chang KW and Huang CJ: Gut butyrate-producing organisms correlate to Placenta specific 8 protein: Importance to colorectal cancer progression. J Adv Res. 22:7–20. 2020.PubMed/NCBI View Article : Google Scholar

13 

Huang CJ, Lee CL, Yang SH, Chien CC, Huang CC, Yang RN and Chang CC: Upregulation of the growth arrest-specific-2 in recurrent colorectal cancers, and its susceptibility to chemotherapy in a model cell system. Biochim Biophys Acta. 1862:1345–1353. 2016.PubMed/NCBI View Article : Google Scholar

14 

Lee CL, Huang CJ, Yang SH, Chang CC, Huang CC, Chien CC and Yang RN: Discovery of genes from feces correlated with colorectal cancer progression. Oncol Lett. 12:3378–3384. 2016.PubMed/NCBI View Article : Google Scholar

15 

Corbo C, Cevenini A and Salvatore F: Biomarker discovery by proteomics-based approaches for early detection and personalized medicine in colorectal cancer. Proteomics Clin Appl. 11(1600072)2017.PubMed/NCBI View Article : Google Scholar

16 

Imperiale TF, Kisiel JB, Itzkowitz SH, Scheu B, Duimstra EK, Statz S, Berger BM and Limburg PJ: Specificity of the multi-target stool DNA test for colorectal cancer screening in average-risk 45-49 year-olds: A cross-sectional study. Cancer Prev Res (Phila). 14:489–496. 2021.PubMed/NCBI View Article : Google Scholar

17 

Lin J, Zhang L, Chen M, Chen J, Wu Y, Wang T, Lu Y, Ba Z, Cheng X, Xu R, et al: Evaluation of combined detection of multigene mutation and SDC2/SFRP2 methylation in stool specimens for colorectal cancer early diagnosis. Int J Colorectal Dis. 37:1231–1238. 2022.PubMed/NCBI View Article : Google Scholar

18 

Vega-Benedetti AF, Loi E, Moi L, Orrù S, Ziranu P, Pretta A, Lai E, Puzzoni M, Ciccone L, Casadei-Gardini A, et al: Colorectal cancer early detection in stool samples tracing CpG islands methylation alterations affecting gene expression. Int J Mol Sci. 21(4494)2020.PubMed/NCBI View Article : Google Scholar

19 

Chen J, Sun H, Tang W, Zhou L, Xie X, Qu Z, Chen M, Wang S, Yang T, Dai Y, et al: DNA methylation biomarkers in stool for early screening of colorectal cancer. J Cancer. 10:5264–5271. 2019.PubMed/NCBI View Article : Google Scholar

20 

De Wijkerslooth T, Bossuyt P and Dekker E: Strategies in screening for colon carcinoma. Neth J Med. 69:112–119. 2011.PubMed/NCBI

21 

Young GP and Bosch LJW: Fecal tests: From blood to molecular markers. Curr Colorectal Cancer Rep. 7:62–70. 2011.PubMed/NCBI View Article : Google Scholar

22 

Ahlquist DA, Zou H, Domanico M, Mahoney DW, Yab TC, Taylor WR, Butz ML, Thibodeau SN, Rabeneck L, Paszat LF, et al: Next-generation stool DNA test accurately detects colorectal cancer and large adenomas. Gastroenterology. 142:248–256.e25-e26. 2012.PubMed/NCBI View Article : Google Scholar

23 

Miller S and Steele S: Novel molecular screening approaches in colorectal cancer. J Surg Oncol. 105:459–467. 2012.PubMed/NCBI View Article : Google Scholar

24 

Pox C: Colon cancer screening: Which non-invasive filter tests? Dig Dis. 29:56–59. 2011.PubMed/NCBI View Article : Google Scholar

25 

Komor MA, Bosch LJ, Coupé VM, Rausch C, Pham TV, Piersma SR, Mongera S, Mulder CJ, Dekker E, Kuipers EJ, et al: Proteins in stool as biomarkers for non-invasive detection of colorectal adenomas with high risk of progression. J Pathol. 250:288–298. 2020.PubMed/NCBI View Article : Google Scholar

26 

Choi HH, Cho YS, Choi JH, Kim HK, Kim SS and Chae HS: Stool-based miR-92a and miR-144* as noninvasive biomarkers for colorectal cancer screening. Oncology. 97:173–179. 2019.PubMed/NCBI View Article : Google Scholar

27 

Steele RJ, Kostourou I, McClements P, Watling C, Libby G, Weller D, Brewster DH, Black R, Carey FA and Fraser C: Effect of repeated invitations on uptake of colorectal cancer screening using faecal occult blood testing: Analysis of prevalence and incidence screening. BMJ. 341(c5531)2010.PubMed/NCBI View Article : Google Scholar

28 

Tonus C, Sellinger M, Koss K and Neupert G: Faecal pyruvate kinase isoenzyme type M2 for colorectal cancer screening: A meta-analysis. World J Gastroenterol. 18:4004–4011. 2012.PubMed/NCBI View Article : Google Scholar

29 

Osborn NK and Ahlquist DA: Stool screening for colorectal cancer: Molecular approaches. Gastroenterology. 128:192–206. 2005.PubMed/NCBI View Article : Google Scholar

30 

Hamaya Y, Yoshida K, Takai T, Ikuma M, Hishida A and Kanaoka S: Factors that contribute to faecal cyclooxygenase-2 mRNA expression in subjects with colorectal cancer. Br J Cancer. 102:916–921. 2010.PubMed/NCBI View Article : Google Scholar

31 

Mojtabanezhad Shariatpanahi A, Yassi M, Nouraie M, Sahebkar A, Varshoee Tabrizi F and Kerachian MA: The importance of stool DNA methylation in colorectal cancer diagnosis: A meta-analysis. PLoS One. 13(e0200735)2018.PubMed/NCBI View Article : Google Scholar

32 

Robertson DJ and Imperiale TF: Stool testing for colorectal cancer screening. Gastroenterology. 149:1286–1293. 2015.PubMed/NCBI View Article : Google Scholar

33 

Yang SH, Huang CJ, Lee CL, Liu CC, Chien CC and Chen SH: Fecal RNA detection of cytokeratin 19 and ribosomal protein L19 for colorectal cancer. Hepatogastroenterology. 57:710–715. 2010.PubMed/NCBI

34 

Song L, Lu H and Yin J: Preliminary study on discriminating HER2 2+ amplification status of breast cancers based on texture features semi-automatically derived from pre-, post-contrast, and subtraction images of DCE-MRI. PLoS One. 15(e0234800)2020.PubMed/NCBI View Article : Google Scholar

35 

Juhola M, Joutsijoki H, Penttinen K, Shah D and Aalto-Setälä K: On computational classification of genetic cardiac diseases applying iPSC cardiomyocytes. Comput Methods Programs Biomed. 210(106367)2021.PubMed/NCBI View Article : Google Scholar

36 

Yu Z, Yu H, Zou Q, Huang Z, Wang X, Tang G, Bai L, Zhou C, Zhuang Z, Xie Y, et al: Nomograms for prediction of molecular phenotypes in colorectal cancer. Onco Targets Ther. 13:309–321. 2020.PubMed/NCBI View Article : Google Scholar

37 

Liu CY, Huang CS, Huang CC, Ku WC, Shih HY and Huang CJ: Co-occurrence of differentiated thyroid cancer and second primary malignancy: Correlation with expression profiles of mismatch repair protein and cell cycle regulators. Cancers (Basel). 13(5486)2021.PubMed/NCBI View Article : Google Scholar

38 

Weiser MR: AJCC 8th edition: Colorectal cancer. Ann Surg Oncol. 25:1454–1455. 2018.PubMed/NCBI View Article : Google Scholar

39 

Treanor D and Quirke P: Pathology of colorectal cancer. Clin Oncol (R Coll Radiol). 19:769–776. 2007.PubMed/NCBI View Article : Google Scholar

40 

Chien CC, Chang CC, Yang SH, Chen SH and Huang CJ: A homologue of the Drosophila headcase protein is a novel tumor marker for early-stage colorectal cancer. Oncol Rep. 15:919–926. 2006.PubMed/NCBI

41 

Dowling P, Clarke C, Hennessy K, Torralbo-Lopez B, Ballot J, Crown J, Kiernan I, O'Byrne KJ, Kennedy MJ, Lynch V and Clynes M: Analysis of acute-phase proteins, AHSG, C3, CLI, HP and SAA, reveals distinctive expression patterns associated with breast, colorectal and lung cancer. Int J Cancer. 131:911–923. 2012.PubMed/NCBI View Article : Google Scholar

42 

Ginestier C, Cervera N, Finetti P, Esteyries S, Esterni B, Adélaïde J, Xerri L, Viens P, Jacquemier J, Charafe-Jauffret E, et al: Prognosis and gene expression profiling of 20q13-amplified breast cancers. Clin Cancer Res. 12:4533–4544. 2006.PubMed/NCBI View Article : Google Scholar

43 

Bair E and Tibshirani R: Semi-supervised methods to predict patient survival from gene expression data. PLoS Biol. 2(E108)2004.PubMed/NCBI View Article : Google Scholar

44 

Chou HL, Yao CT, Su SL, Lee CY, Hu KY, Terng HJ, Shih YW, Chang YT, Lu YF, Chang CW, et al: Gene expression profiling of breast cancer survivability by pooled cDNA microarray analysis using logistic regression, artificial neural networks and decision trees. BMC Bioinformatics. 14(100)2013.PubMed/NCBI View Article : Google Scholar

45 

Hughes G: On the binormal predictive receiver operating characteristic curve for the joint assessment of positive and negative predictive values. Entropy (Basel). 22(593)2020.PubMed/NCBI View Article : Google Scholar

46 

Yang SH, Chien CC, Chen CW, Li SY and Huang CJ: Potential of faecal RNA in diagnosing colorectal cancer. Cancer Lett. 226:55–63. 2005.PubMed/NCBI View Article : Google Scholar

47 

Livak KJ and Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) method. Methods. 25:402–408. 2001.PubMed/NCBI View Article : Google Scholar

48 

Chang CC, Kao WY, Liu CY, Su HH, Kan YA, Lin PY, Ku WC, Chang KW, Yang RN and Huang CJ: Butyrate supplementation regulates expression of chromosome segregation 1-like protein to reverse the genetic distortion caused by p53 mutations in colorectal cancer. Int J Oncol. 60(64)2022.PubMed/NCBI View Article : Google Scholar

49 

Wei Q, Jiang H, Baker A, Dodge LK, Gerard M, Young MR, Toledano MB and Colburn NH: Loss of sulfiredoxin renders mice resistant to azoxymethane/dextran sulfate sodium-induced colon carcinogenesis. Carcinogenesis. 34:1403–1410. 2013.PubMed/NCBI View Article : Google Scholar

50 

Bankhead P, Loughrey MB, Fernández JA, Dombrowski Y, McArt DG, Dunne PD, McQuaid S, Gray RT, Murray LJ, Coleman HG, et al: QuPath: Open source software for digital pathology image analysis. Sci Rep. 7(16878)2017.PubMed/NCBI View Article : Google Scholar

51 

Huang CJ, Chien CC, Yang SH, Chang CC, Sun HL, Cheng YC, Liu CC, Lin SC and Lin CM: Faecal ribosomal protein L19 is a genetic prognostic factor for survival in colorectal cancer. J Cell Mol Med. 12:1936–1943. 2008.PubMed/NCBI View Article : Google Scholar

52 

Bernal G: Use of RNA isolated from feces as a promising tool for the early detection of colorectal cancer. Int J Biol Markers. 27:e82–e89. 2012.PubMed/NCBI View Article : Google Scholar

53 

Grizzi F, Celesti G, Basso G and Laghi L: Tumor budding as a potential histopathological biomarker in colorectal cancer: Hype or hope? World J Gastroenterol. 18:6532–6536. 2012.PubMed/NCBI View Article : Google Scholar

54 

Belletti B, Nicoloso MS, Schiappacassi M, Chimienti E, Berton S, Lovat F, Colombatti A and Baldassarre G: p27(kip1) functional regulation in human cancer: A potential target for therapeutic designs. Curr Med Chem. 12:1589–1605. 2005.PubMed/NCBI View Article : Google Scholar

55 

Blázquez C, Geelen MJ, Velasco G and Guzmán M: The AMP-activated protein kinase prevents ceramide synthesis de novo and apoptosis in astrocytes. FEBS Lett. 489:149–153. 2001.PubMed/NCBI View Article : Google Scholar

56 

Nikolouzakis TK, Vassilopoulou L, Fragkiadaki P, Mariolis Sapsakos T, Papadakis GZ, Spandidos DA, Tsatsakis AM and Tsiaoussis J: Improving diagnosis, prognosis and prediction by using biomarkers in CRC patients (Review). Oncol Rep. 39:2455–2472. 2018.PubMed/NCBI View Article : Google Scholar

57 

Cheng YW and Li YC: Factors affecting the follow-up time after a positive result in the fecal occult blood test. PLoS One. 16(e0258130)2021.PubMed/NCBI View Article : Google Scholar

58 

Bardhan K and Liu K: Epigenetics and colorectal cancer pathogenesis. Cancers (Basel). 5:676–713. 2013.PubMed/NCBI View Article : Google Scholar

59 

Mundade R, Imperiale TF, Prabhu L, Loehrer PJ and Lu T: Genetic pathways, prevention, and treatment of sporadic colorectal cancer. Oncoscience. 1:400–406. 2014.PubMed/NCBI View Article : Google Scholar

60 

Ahlquist DA: Molecular detection of colorectal neoplasia. Gastroenterology. 138:2127–2139. 2010.PubMed/NCBI View Article : Google Scholar

61 

Bailey JR, Aggarwal A and Imperiale TF: Colorectal cancer screening: Stool DNA and other noninvasive modalities. Gut Liver. 10:204–211. 2016.PubMed/NCBI View Article : Google Scholar

62 

Youssef O, Sarhadi V, Ehsan H, Böhling T, Carpelan-Holmström M, Koskensalo S, Puolakkainen P, Kokkola A and Knuutila S: Gene mutations in stool from gastric and colorectal neoplasia patients by next-generation sequencing. World J Gastroenterol. 23:8291–8299. 2017.PubMed/NCBI View Article : Google Scholar

63 

Januszewicz W and Fitzgerald RC: Early detection and therapeutics. Mol Oncol. 13:599–613. 2019.PubMed/NCBI View Article : Google Scholar

64 

Pickhardt PJ: Emerging stool-based and blood-based non-invasive DNA tests for colorectal cancer screening: The importance of cancer prevention in addition to cancer detection. Abdom Radiol (NY). 41:1441–1444. 2016.PubMed/NCBI View Article : Google Scholar

65 

Shin HY, Suh M, Choi KS, Hwang SH, Jun JK, Han DS, Lee YK, Oh JH, Lee CW and Lee DH: Higher satisfaction with an alternative collection device for stool sampling in colorectal cancer screening with fecal immunochemical test: A cross-sectional study. BMC Cancer. 18(365)2018.PubMed/NCBI View Article : Google Scholar

66 

de Wit M, Fijneman RJ, Verheul HM, Meijer GA and Jimenez CR: Proteomics in colorectal cancer translational research: Biomarker discovery for clinical applications. Clin Biochem. 46:466–479. 2013.PubMed/NCBI View Article : Google Scholar

67 

Montrose DC, Zhou XK, Kopelovich L, Yantiss RK, Karoly ED, Subbaramaiah K and Dannenberg AJ: Metabolic profiling, a noninvasive approach for the detection of experimental colorectal neoplasia. Cancer Prev Res (Phila). 5:1358–1367. 2012.PubMed/NCBI View Article : Google Scholar

68 

Slaby O: Non-coding RNAs as biomarkers for colorectal cancer screening and early detection. Adv Exp Med Biol. 937:153–170. 2016.PubMed/NCBI View Article : Google Scholar

69 

Tsai TJ, Lim YP, Chao WY, Chen CC, Chen YJ, Lin CY and Lee YR: Capping actin protein overexpression in human colorectal carcinoma and its contributed tumor migration. Anal Cell Pathol (Amst). 2018(8623937)2018.PubMed/NCBI View Article : Google Scholar

70 

Pan F, Chen T, Sun X, Li K, Jiang X, Försti A, Zhu Y and Lai M: Prognosis prediction of colorectal cancer using gene expression profiles. Front Oncol. 9(252)2019.PubMed/NCBI View Article : Google Scholar

71 

Xu G, Zhang M, Zhu H and Xu J: A 15-gene signature for prediction of colon cancer recurrence and prognosis based on SVM. Gene. 604:33–40. 2017.PubMed/NCBI View Article : Google Scholar

72 

Eckmann JD, Ebner DW and Kisiel JB: Multi-target stool DNA testing for colorectal cancer screening: Emerging learning on real-world performance. Curr Treat Options Gastroenterol: Jan 21, 2020 (Epub ahead of print).

73 

Kisiel JB, Klepp P, Allawi HT, Taylor WR, Giakoumopoulos M, Sander T, Yab TC, Moum BA, Lidgard GP, Brackmann S, et al: Analysis of DNA methylation at specific loci in stool samples detects colorectal cancer and high-grade dysplasia in patients with inflammatory bowel disease. Clin Gastroenterol Hepatol. 17:914–921.e5. 2019.PubMed/NCBI View Article : Google Scholar

74 

Naber SK, Knudsen AB, Zauber AG, Rutter CM, Fischer SE, Pabiniak CJ, Soto B, Kuntz KM and Lansdorp-Vogelaar I: Cost-effectiveness of a multitarget stool DNA test for colorectal cancer screening of medicare beneficiaries. PLoS One. 14(e0220234)2019.PubMed/NCBI View Article : Google Scholar

75 

Jung G, Hernández-Illán E, Moreira L, Balaguer F and Goel A: Epigenetics of colorectal cancer: Biomarker and therapeutic potential. Nat Rev Gastroenterol Hepatol. 17:111–130. 2020.PubMed/NCBI View Article : Google Scholar

76 

Olén O, Erichsen R, Sachs MC, Pedersen L, Halfvarson J, Askling J, Ekbom A, Sørensen HT and Ludvigsson JF: Colorectal cancer in ulcerative colitis: A Scandinavian population-based cohort study. Lancet. 395:123–131. 2020.PubMed/NCBI View Article : Google Scholar

77 

Chang CC, Chao KC, Huang CJ, Hung CS and Wang YC: Association between aberrant dynein cytoplasmic 1 light intermediate chain 1 expression levels, mucins and chemosensitivity in colorectal cancer. Mol Med Rep. 22:185–192. 2020.PubMed/NCBI View Article : Google Scholar

78 

Hodge CD, Ismail IH, Edwards RA, Hura GL, Xiao AT, Tainer JA, Hendzel MJ and Glover JN: RNF8 E3 ubiquitin ligase stimulates Ubc13 E2 conjugating activity that is essential for DNA double strand break signaling and BRCA1 tumor suppressor recruitment. J Biol Chem. 291:9396–9410. 2016.PubMed/NCBI View Article : Google Scholar

79 

Wu Z, Neufeld H, Torlakovic E and Xiao W: Uev1A-Ubc13 promotes colorectal cancer metastasis through regulating CXCL1 expression via NF-кB activation. Oncotarget. 9:15952–15967. 2018.PubMed/NCBI View Article : Google Scholar

80 

Zhang P, Wang L, Rodriguez-Aguayo C, Yuan Y, Debeb BG, Chen D, Sun Y, You MJ, Liu Y, Dean DC, et al: miR-205 acts as a tumour radiosensitizer by targeting ZEB1 and Ubc13. Nat Commun. 5(5671)2014.PubMed/NCBI View Article : Google Scholar

81 

Pulvino M, Liang Y, Oleksyn D, DeRan M, Van Pelt E, Shapiro J, Sanz I, Chen L and Zhao J: Inhibition of proliferation and survival of diffuse large B-cell lymphoma cells by a small-molecule inhibitor of the ubiquitin-conjugating enzyme Ubc13-Uev1A. Blood. 120:1668–1677. 2012.PubMed/NCBI View Article : Google Scholar

82 

Cheng J, Fan YH, Xu X, Zhang H, Dou J, Tang Y, Zhong X, Rojas Y, Yu Y, Zhao Y, et al: A small-molecule inhibitor of UBE2N induces neuroblastoma cell death via activation of p53 and JNK pathways. Cell Death Dis. 5(e1079)2014.PubMed/NCBI View Article : Google Scholar

83 

Ishtiaq A, Bakhtiar A, Silas E, Saeed J, Ajmal S, Mushtaq I, Ali T, Wahedi HM, Khan W, Khan U, et al: Pistacia integerrima alleviated Bisphenol A induced toxicity through Ubc13/p53 signalling. Mol Biol Rep. 47:6545–6559. 2020.PubMed/NCBI View Article : Google Scholar

84 

Pelletier J, Riaño-Canalias F, Almacellas E, Mauvezin C, Samino S, Feu S, Menoyo S, Domostegui A, Garcia-Cajide M, Salazar R, et al: Nucleotide depletion reveals the impaired ribosome biogenesis checkpoint as a barrier against DNA damage. EMBO J. 39(e103838)2020.PubMed/NCBI View Article : Google Scholar

85 

Ouchida M, Kanzaki H, Ito S, Hanafusa H, Jitsumori Y, Tamaru S and Shimizu K: Novel direct targets of miR-19a identified in breast cancer cells by a quantitative proteomic approach. PLoS One. 7(e44095)2012.PubMed/NCBI View Article : Google Scholar

86 

Zhao G, Liu X, Liu Y, Li H, Ma Y, Li S, Zhu Y, Miao J, Xiong S, Fei S and Zheng M: Aberrant DNA methylation of SEPT9 and SDC2 in stool specimens as an integrated biomarker for colorectal cancer early detection. Front Genet. 11(643)2020.PubMed/NCBI View Article : Google Scholar

87 

Gu Y, Lin X, Kapoor A, Chow MJ, Jiang Y, Zhao K and Tang D: The oncogenic potential of the centromeric border protein FAM84B of the 8q24.21 gene desert. Genes (Basel). 11(312)2020.PubMed/NCBI View Article : Google Scholar

88 

Uyama T, Jin XH, Tsuboi K, Tonai T and Ueda N: Characterization of the human tumor suppressors TIG3 and HRASLS2 as phospholipid-metabolizing enzymes. Biochim Biophys Acta. 1791:1114–1124. 2009.PubMed/NCBI View Article : Google Scholar

89 

Shyu RY, Hsieh YC, Tsai FM, Wu CC and Jiang SY: Cloning and functional characterization of the HRASLS2 gene. Amino Acids. 35:129–137. 2008.PubMed/NCBI View Article : Google Scholar

90 

Bi F, Wang Q, Dong Q, Wang Y, Zhang L and Zhang J: Circulating tumor DNA in colorectal cancer: Opportunities and challenges. Am J Transl Res. 12:1044–1055. 2020.PubMed/NCBI

91 

Lansdorp-Vogelaar I, Goede SL, Bosch LJW, Melotte V, Carvalho B, van Engeland M, Meijer GA, de Koning HJ and van Ballegooijen M: Cost-effectiveness of high-performance biomarker tests vs fecal immunochemical test for noninvasive colorectal cancer screening. Clin Gastroenterol Hepatol. 16:504–512.e11. 2018.PubMed/NCBI View Article : Google Scholar

Related Articles

Journal Cover

March-2023
Volume 18 Issue 3

Print ISSN: 2049-9434
Online ISSN:2049-9442

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
x
Spandidos Publications style
Chen Y, Shih C, Guo S, Liu C, Shen M, Chang S, Ku W, Huang C and Huang C: Potential prognostic and predictive value of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 in colorectal cancer stool specimens. Biomed Rep 18: 22, 2023.
APA
Chen, Y., Shih, C., Guo, S., Liu, C., Shen, M., Chang, S. ... Huang, C. (2023). Potential prognostic and predictive value of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 in colorectal cancer stool specimens. Biomedical Reports, 18, 22. https://doi.org/10.3892/br.2023.1604
MLA
Chen, Y., Shih, C., Guo, S., Liu, C., Shen, M., Chang, S., Ku, W., Huang, C., Huang, C."Potential prognostic and predictive value of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 in colorectal cancer stool specimens". Biomedical Reports 18.3 (2023): 22.
Chicago
Chen, Y., Shih, C., Guo, S., Liu, C., Shen, M., Chang, S., Ku, W., Huang, C., Huang, C."Potential prognostic and predictive value of UBE2N, IMPDH1, DYNC1LI1 and HRASLS2 in colorectal cancer stool specimens". Biomedical Reports 18, no. 3 (2023): 22. https://doi.org/10.3892/br.2023.1604