Biomarker for predicting the prognosis of colorectal cancer

Abstract

The present invention relates to a biomarker for predicting the prognosis of colorectal cancer by using changes in expression level of NNT and/or OSBPL3, and a prognosis prediction method using same. The expression level of NNT and/or OSBPL3 is analyzed so that the prognosis of colorectal cancer patients in clinical practice can be predicted, and if the analysis is performed in combination with TNM stage, more accurate prediction can be made such that individualized and customized strategies can be designed.

Claims

1. A method for predicting the survival prognosis of an obese colorectal cancer patient, wherein the obese colorectal cancer patient has a body mass index (BMI) of 25 or higher, comprising the following steps: (a) a step of measuring the protein level of nicotinamide nucleotide transhydrogenase (NNT) gene and OSBPL3 (oxysterol binding protein like 3) gene in a tissue sample isolated from an obese colorectal cancer patient, comprising: constructing a tissue microarray (TMA) from the tissue sample isolated from the obese colorectal cancer patient, and measuring the protein levels of the NNT gene and OSBPL3 gene by immunohistochemical (IHC) staining; (b) a step of comparing the measured mRNA or protein level with the protein level of a control sample; and (c) a step of determining the survival prognosis of the obese colorectal cancer patient, wherein the survival prognosis is determined to be good if the protein level of the NNT gene is higher than that of the control sample and the protein level of the OSBPL3 gene is lower than that of the control sample.

2. The method for predicting the prognosis of an obese colorectal cancer patient according to claim 1, wherein the sample of step (a) above is colon tissue.

3. A method for predicting the survival prognosis of an obese colorectal cancer patient, wherein the obese colorectal cancer patient has a body mass index BMI) of 25 or higher, comprising the following steps: (a) a step of measuring the protein level nicotinamide nucleotide transhydrogenase (NNT) gene and OSBPL3 (oxysterol binding protein like 3) gene in a tissue sample isolated from an obese colorectal cancer patient, comprising: constructing a tissue microarray (TMA) from the tissue sample isolated from the obese colorectal cancer patient, and measuring the protein levels of the NNT gene and OSBPL3 gene by immunohistochemical (IHC) staining; (b) a step of comparing the measured protein level with the protein level of a control sample; (c) a step of combining and analyzing the information classified according to the TNM stage (tumor, node, metastasis stage); and (d) a step of determining the survival prognosis of the obese colorectal cancer patient, wherein the survival prognosis is determined to be good if the NNT level is higher than that of a control group in the same TNM stage and the OSBPL3 level is lower than that of the control group in the same TNM stage.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1A is a graph showing the gene expression level of OSBPL3 discovered from the TCGA COAD (The Cancer Genome Atlas Colon Adenocarcinoma) data set.

(2) FIG. 1B is a graph showing the gene expression level of FAM220A discovered from the TCGA COAD data set.

(3) FIG. 1C is a graph showing the gene expression level of TMEM9 discovered from the TCGA COAD data set.

(4) FIG. 1D is a graph showing the gene expression level of NNT discovered from the TCGA COAD data set.

(5) FIG. 1E is a graph showing the gene expression level of PANK3 discovered from the TCGA COAD data set.

(6) FIG. 1F is a graph showing the gene expression level of PPARGC1B discovered from the TCGA COAD data set.

(7) FIG. 1G is a graph showing the gene expression level of RTL6 discovered from the TCGA COAD data set.

(8) FIG. 2A is a graph showing the analysis of overall survival rate according to the expression pattern of OSBPL3.

(9) FIG. 2B is a graph showing the analysis of overall survival rate according to the expression pattern of FAM220A.

(10) FIG. 2C is a graph showing the analysis of overall survival rate according to the expression pattern of TMEM9.

(11) FIG. 2D is a graph showing the analysis of overall survival rate according to the expression pattern of NNT.

(12) FIG. 2E is a graph showing the analysis of overall survival rate according to the expression pattern of PANK3.

(13) FIG. 2F is a graph showing the analysis of overall survival rate according to the expression pattern of PPARGC1B.

(14) FIG. 2G is a graph showing the analysis of overall survival rate according to the expression pattern of RTL6.

(15) FIG. 3 is a set of photographs showing the results of immunohistochemical staining for NNT in normal colon epithelial cells and tumor cells (A: normal colon epithelial cells showing weak cytoplasmic staining, B: unstained tumor cells, C: tumor cells showing weak cytoplasmic staining, D: tumor cells showing moderate cytoplasmic staining, E: tumor cells showing strong cytoplasmic staining) (magnification: 400).

(16) FIG. 4 is a set of photographs showing the results of immunohistochemical staining for OSBPL3 in normal colon epithelial cells and tumor cells (A: unstained normal colon epithelial cells, B: unstained tumor cells, C: tumor cells showing weak cytoplasmic staining, D: tumor cells showing moderate cytoplasmic staining, E: tumor cells showing strong cytoplasmic staining) (magnification: 400).

(17) FIG. 5A is a graph showing the event free survival analysis according to the NNT protein expression level in the OW group in the GMC (Gachon University Gil Medical Center) colorectal cancer cohort.

(18) FIG. 5B is a graph showing the event free survival analysis according to the NNT protein expression level in the HW group in the GMC colorectal cancer cohort.

(19) FIG. 5C is a graph showing the event free survival analysis according to the NNT protein expression level in the OW&HW group in the GMC colorectal cancer cohort.

(20) FIG. 6A is a graph showing the event free survival analysis according to the OSBPL3 protein expression level in the OW group in the GMC colorectal cancer cohort.

(21) FIG. 6B is a graph showing the event free survival analysis according to the OSBPL3 protein expression level in the HW group in the GMC colorectal cancer cohort.

(22) FIG. 6C is a graph showing the event, free survival analysis according to the OSBPL3 protein expression level in the OW&HW group in the GMC colorectal cancer cohort.

(23) FIG. 7 is a set of graphs showing the event free survival analysis according to the NNT and OSBPL3 protein expression levels in the OW (FIG. 7A), HW (FIG. 7B), and OW&HW groups (FIG. 7C) in the GMC colorectal cancer cohort.

(24) FIG. 8 is a set of graphs showing the disease specific survival analysis according to the NNT and OSBPL3 protein expression levels in the OW (FIG. 8A), HW (FIG. 8B), and OW&HW groups (FIG. 8C) in the GMC colorectal cancer cohort.

(25) FIG. 9A is a graph showing the event free survival analysis based solely on TNM (tumor, node, metastasis stage) classification in the GMC colorectal cancer cohort.

(26) FIG. 9B is a diagram showing the hazard ratio analysis of FIG. 9A.

(27) FIG. 9C is a graph showing the event free survival analysis by combining the NNT protein expression level and TNM stage in the GMC colorectal cancer cohort.

(28) FIG. 9D is a diagram showing the hazard ratio analysis of FIG. 9C.

(29) FIG. 9E is a graph showing the event free survival analysis according to the NNT protein expression level among patients with late TNM stages (stages 3 and 4) in the GMC colorectal cancer cohort.

(30) FIG. 9F is a diagram showing the hazard ratio analysis of FIG. 9E.

(31) FIG. 10A is a graph showing the disease specific survival analysis based solely on TNM stage in the GMC colorectal cancer cohort.

(32) FIG. 10B is a diagram showing the hazard ratio analysis of FIG. 10A.

(33) FIG. 10C is a graph showing the disease specific survival analysis by combining the NNT protein expression level and TNM stage in the GMC colorectal cancer cohort.

(34) FIG. 10D is a diagram showing the hazard ratio analysis of FIG. 10B.

(35) FIG. 10E is a graph showing the disease specific survival analysis according to the NNT protein expression level among patients with late TNM stages (stages 3 and 4) in the GMC colorectal cancer cohort.

(36) FIG. 10F is a diagram showing the hazard ratio analysis of FIG. 10E.

(37) FIG. 11 is a set of graphs showing the disease specific survival analysis according to the OSBPL3 protein expression level in the OW (FIG. 11A), HW (FIG. 11B), and OW&HW groups (FIG. 11C) with late TNM stages (stages 3 and 4) in the GMC colorectal cancer cohort.

(38) FIG. 12 is a diagram showing the biological functions closely correlated with the NNT gene.

(39) FIG. 13A is a diagram showing the network of proteins closely related to the NNT protein.

(40) FIG. 13B is a diagram showing the biological functions associated with the proteins participating in the network of proteins closely related to the NNT protein.

(41) FIG. 14 is a diagram showing the biological functions that differ between the obese colorectal cancer patients with high NNT gene expression and the obese colorectal cancer patients with low NNT gene expression compared to the normal tissue of the normal weight colorectal cancer patients.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

(42) Hereinafter, the present invention is described in detail.

(43) In a specific experimental example of the present invention, the present inventors discovered seven genes (NNT, PANK3, PPARGC1B, RTL6, OSBPL3, FAM220A, and TMEM9) whose expression levels were consistently increased or decreased in relation to colorectal cancer, and the differences in expression levels were statistically significant. OSBPL3, FAM220A, and TMEM9 were confirmed to be increased in expression in the normal weight colorectal cancer patient group compared to the normal population, and in the obese colorectal cancer patient group compared to the normal weight colorectal cancer patient group. In addition, NNT, PANK3, PPARGC1B, and RTL6 were confirmed to be decreased in expression in the normal weight colorectal cancer patient group compared to the normal population, and in the obese colorectal cancer patient group compared to the normal weight colorectal cancer patient group (FIGS. 1A to 1G). As a result of analyzing the overall survival according to the expression levels of these genes, in the survival analysis targeting a group of patients with obese colorectal cancer, it was confirmed that the patient group with high expression levels of OSBPL3, FAM220A, and TMEM9 had a disadvantage in survival, and the patient group with low expression levels of NNT, PANK3, PPARGC1B, and RTL6 had a disadvantage in survival (FIGS. 2A to 2G). To perform a survival analysis according to the expression levels of the proteins encoded by these genes, immunohistochemical staining for NNT and OSBPL3 was performed on normal colon epithelial cells and tumor cells, and the expression levels of the proteins encoded by these gene candidates was quantified (FIGS. 3A to 3E, and FIGS. 4A to 4E). In addition, the present inventors performed a survival analysis based on the NNT and OSBPL3 protein expression levels, and confirmed that if the expression level of the NNT gene is higher than that of the control group, the prognosis of colorectal cancer patients is favorable (FIGS. 5A to 5C). On the other hand, it was confirmed that the prognosis of colorectal cancer patients can be judged to be good when the expression level of the OSBPL3 gene is lower than that of the control group (FIGS. 6A to 6C). In addition, it was confirmed that the patients with lower expression level of the NNT gene and higher expression level of the OSBPL3 gene than the control group had a poorer prognosis (FIGS. 7A to 7C, and FIGS. 8A to 8C). In addition, the present inventors confirmed that better survival prediction is possible when event free survival analysis and disease specific survival analysis are performed by combining the TNM stage (tumor, node, metastasis stage) and NNT expression level (FIGS. 9A to 9F, FIGS. 10A to 10F, and FIGS. 11A to 11C). In addition, the biological functions or protein networks that were closely correlated with the NNT gene and protein were identified to nicotinate and nicotinamide metabolism, pyruvate metabolism, and TCA cycle (FIGS. 12, 13A and 13B). The obese and NNT.sub.low groups, which were identified as being disadvantageous for survival, were confirmed to have statistically significant suppression of nicotinate and nicotinamide metabolism, inositol phosphate metabolism, and purine metabolism compared to the normal group (FIG. 14).

(44) The present invention provides a composition for predicting the prognosis of a colorectal cancer patient, comprising an agent for measuring the level of mRNA or protein of the NNT and/or OSBPL3 gene.

(45) The NNT is an enzyme that catalyzes the following chemical reaction.
NADPH+NAD.sup.+.Math.NADP.sup.++NADH

(46) The above enzyme is an oxidoreductase that acts on NADH or NADPH using NAD+ or NADP+ as an acceptor. This enzyme participates in nicotinate and nicotinamide metabolism.

(47) The above enzymes can be referred to as NADPH:NAD.sup.+ oxidoreductase, pyridine nucleotide transhydrogenase, transhydrogenase, NAD(P).sup.+ transhydrogenase, nicotinamide adenine dinucleotide (phosphate) transhydrogenase, NAD.sup.+ transhydrogenase, NADH transhydrogenase, NADPH-NAD.sup.+ transhydrogenase, pyridine nucleotide transferase, NADPH-NAD.sup.+ oxidoreductase, NADH-NADP.sup.+-transhydrogenase, NADPH:NAD.sup.+ transhydrogenase, H.sup.+-Thase, and energy-linked transhydrogenase.

(48) The said OSBPL3 is a member of the oxysterol-binding protein (OSBP) family, a group of intracellular lipid receptors.

(49) The above colorectal cancer patients may be obese.

(50) The obesity can be determined by measuring the obesity-related indices including the patient's body mass index (BMI), waist-hip ratio (WHR), waist circumference (WC), waist-stature ratio (WSR), body fat percentage (BF %), and relative fat mass (RFM), preferably based on the patient's BMI.

(51) BMI is calculated by dividing weight (kg) by the square of height (m), and is one of the methods to determine obesity. The criteria for determining obesity may vary slightly, but the World Health Organization considers the BMI of 25 or higher to be overweight and 30 or higher to be obese. The World Health Organization Asia-Pacific Region and the Korean Society of Obesity define the BMI of 23 or higher to be overweight and 25 or higher to be obese, taking into account racial characteristics. In one embodiment of the present invention, TCGA data is mostly based on Western patients, so the Western standard of BMI 25 or more is applied as overweight and 30 or more as obese, and TMA analysis is based on Korean patients, so the Korean standard of BMI 23 or more is applied as overweight and 25 or more as obese.

(52) The present invention provides a composition for predicting the prognosis of a colorectal cancer patient, comprising an agent for measuring the level of mRNA or protein of the NNT and/or OSBPL3 gene.

(53) The agent for measuring the level of mRNA may include a primer or a probe, but not always limited thereto.

(54) The above primer is a nucleic acid sequence having a short free 3 hydroxyl group that can form complementary base pairs with a template and functions as a starting point for replicating the template strand. The primers can initiate DNA synthesis in the presence of the reagents for the polymerization reaction (i.e., DNA polymerase or reverse transcriptase) and the different four nucleoside triphosphates in an appropriate buffer and at a proper temperature.

(55) The probe may be a probe capable of binding complementarily to a target gene, and the nucleotide sequence of the probe is not limited as long as it is capable of binding complementarily to each gene.

(56) The method for measuring the mRNA expression level includes RT-PCR (reverse transcription polymerase chain reaction), competitive RT-PCR, real-time RT-PCR, RNase protection assay (RPA), Northern blotting, and a method utilizing DNA chips, but not always limited thereto.

(57) The agent for measuring the level of the protein may be an antibody, a peptide, an aptamer or a compound specific for the protein.

(58) The said antibody refers to a protein molecule capable of specifically binding to an antigenic site on a protein or peptide molecule. Such antibodies can be produced by cloning each gene into an expression vector according to a conventional method to obtain a protein encoded by the marker gene, and by a conventional method from the obtained protein. The form of the above antibody is not particularly limited, and a polyclonal antibody, a monoclonal antibody, or a part thereof having antigen binding property may be included in the antibody of the present invention, and any immunoglobulin antibody may be included. In addition, the antibody of the present invention may include a special antibody such as a humanized antibody. Furthermore, the antibody comprises a functional fragment of an antibody molecule as well as a complete form having two full-length light chains and two full-length heavy chains. The functional fragment of an antibody molecule means a fragment that possesses at least an antigen-binding function, and can be Fab, F(ab), F(ab) 2, and Fv, etc.

(59) The above peptide has the advantage of high binding to the target substance and does not undergo denaturation even when treated with heat or chemicals. In addition, since the above peptide has a small molecular size, it can be attached to other protein and used as a fusion protein. Specifically, it can be attached to a polymer protein chain, which can be used as a diagnostic kit and a drug delivery material.

(60) The above aptamer refers to a single-stranded oligonucleotide, which is a nucleic acid molecule having binding activity for a specific target molecule. The above aptamer can have various three-dimensional structures depending on its nucleotide sequence and can have a high affinity for a specific substance, such as an antigen-antibody reaction. Aptamers can inhibit the activity of a specific target molecule by binding to the target molecule. Aptamers can be used as a replacement for antibodies because they are composed of polynucleotides that can specifically bind to antigenic substances in the same way as antibodies, but are more stable than proteins, have a simpler structure, and are easy to synthesize.

(61) The method for measuring the protein level includes Western blot, enzyme linked immunosorbent assay (ELISA), radioimmunoassay (RIA), radioimmunodiffusion, ouchterlony immunodiffusion, rocket immunoelectrophoresis, tissue immunostaining, immunoprecipitation assay, complement fixation assay, FACS, and a method utilizing DNA chips, but not always limited thereto.

(62) The present invention also provides a kit for predicting the prognosis of a colorectal cancer patient, comprising a composition containing an agent for measuring the level of mRNA or protein of the NNT and/or OSBPL3 gene.

(63) The above kit may include an RT-PCR kit, a DNA chip kit, an ELISA kit, a protein chip kit, a rapid kit, or an MRM (multiple reaction monitoring) kit, but not always limited thereto.

(64) The present invention also provides a method for predicting the prognosis of a colorectal cancer patient comprising the following steps: (a) a step of measuring the mRNA level of the NNT and/or OSBPL3 gene or the protein expression level thereof in a sample isolated from a colorectal cancer patient; and (b) a step of comparing the measured mRNA level or protein level thereof with the mRNA level or protein level thereof of a control sample.

(65) If the mRNA expression level or the protein expression level of the NNT gene is higher than that of the control group, it can be judged that the prognosis will be good.

(66) If the mRNA expression level or the protein expression level of the OSBPL3 gene is lower than that of the control group, it can be judged that the prognosis will be good. The above judgment of a good prognosis may mean that the event free survival, disease specific survival, and overall survival are higher than in the group with a poor prognosis.

(67) The sample of the above step (a) may be colon tissue.

(68) In addition, the present invention provides a method for predicting the prognosis of a colorectal cancer patient comprising the following steps: (a) a step of measuring the mRNA level of the NNT and/or OSBPL3 gene or the protein level thereof in a sample isolated from a colorectal cancer patient; (b) a step of comparing the measured mRNA level or protein level thereof with the mRNA level or protein level thereof of a control sample; and (c) a step of combining and analyzing information classified according to TNM stage.

(69) The TNM stage is one way to determine the stage of a tumor. T stands for tumor, and depending on the depth of invasion into the organ wall, it is divided into TO (no primary tumor), Tis (carcinoma in situ), and T1 to T4 (the higher the number, the more invasion around it), etc. N stands for node (lymph node), and depending on the number, size, and location of the invaded lymph nodes, it is divided into NO (no lymph node metastasis), and N1 to N3, etc. M stands for metastasis, and depending on the presence or absence of remote metastasis, it is divided into M0 (no metastasis), and M1 (metastasized), etc. Once T, N, and M are determined using the above methods they are combined to determine the final stage of a disease. The stage determined in this way is very important in determining treatment plan and prognosis.

(70) The information categorized according to the TNM stage above is the most common way to provide prognostic information for cancer and can be divided into stages 1 through 4. However, there is a problem that predicting the prognosis of an individual patient is incomplete because the TNM stage does not include all the important variables that affect the prognosis in order to simplify the stage setting.

(71) In the present invention, the event free survival period refers to the period of time after a patient undergoes curative surgery during which death occurs due to inoperable disease progression, recurrence of cancer, complications, etc.

(72) In the disease specific survival analysis of the present invention, deaths due to colorectal cancer were categorized as death (event), and censored deaths due to causes other than colorectal cancer were categorized as survival.

(73) In one embodiment of the present invention, when the event free survival was analyzed based solely on the TNM stage, the event free survival risk rate of the patients with TNM stage 3/4 was 12.07 times higher than that of the patients with TNM stage 1/2. On the other hand, when the NNT and TNM stage were combined for analysis, the event free survival risk rate of the patients with high NNT expression and TNM stage 3/4 was 8.39 times higher than that of the patients with high NNT expression and TNM stage 1/2, and the event free survival risk rate of the patients with low NNT expression and TNM stage 3/4 was 21.2 times higher than that of the patients with high NNT expression and TNM stage 1/2. In addition, in the patients with the same stage (TNM stage 3/4), the event free survival risk rate of the group with low NNT expression was 2.52 times higher than that of the group with high NNT expression, which was statistically significantly worse prognosis.

(74) In one embodiment of the present invention, when the disease specific survival was analyzed based solely on the TNM stage, the event free survival risk rate of the patients with TNM stage 3/4 was 8.33 times higher than that of the patients with TNM stage 1/2. On the other hand, when NNT and TNM stages were combined for analysis, the event free survival risk rate of the patients with high NNT expression and TNM stage 3/4 was 6.25 times higher than that of the patients with high NNT expression and TNM stage 1/2, and the event free survival risk rate of the patients with low NNT expression and TNM stage 3/4 was 14.05 times higher than that of the patients with high NNT expression and TNM stage 1/2. In addition, in the patients at the same stage (TNM stage 3/4), the event free survival risk rate of the group with low NNT expression was 2.23 times higher than that of the group with high NNT expression, which was statistically significantly worse prognosis. In addition, the present inventors confirmed that the event free survival rate of the group with low OSBPL3 expression was higher than that of the group with high OSBPL3 expression in the patients at the same stage (TNM stage 3/4). Therefore, a combined analysis of the expression levels of NNT and/or OSBPL3 and the TNM stage may provide a more precise prognosis prediction.

(75) Hereinafter, the present invention will be described in detail by the following experimental examples.

(76) However, the following experimental examples are only for illustrating the present invention, and the contents of the present invention are not limited thereto.

Experimental Example 1: Discovery of Biomarkers for Predicting Colorectal Cancer Prognosis

(77) Using a genomic database, the present inventors identified genes whose expression levels were consistently increased or decreased in association with colorectal cancer, and whose differences in expression were statistically significant.

(78) Specifically, to discover the genes specifically expressed in the obese colorectal cancer patients, the gene expression data set and clinical information data set among the colon adenocarcinoma (COAD) data sets of TCGA (The Cancer Genome Atlas) project were downloaded from the UCSC cancer genomics browser. From the clinical information data set, 417 samples with an initial diagnosis age of 40 years or older were extracted from a total of 434 colorectal cancer samples, and 414 primary tumor and adjacent normal tissue samples were extracted. Of these, 268 samples for which weight and height could be confirmed at the time of initial diagnosis were used in the experiment. In the process of discovering genes specifically expressed in the obese colorectal cancer patients, the expression levels of specific genes were divided into the following three groups. 1) Gene expression levels in normal tissues of normal weight people (hereinafter referred to as Normal), 2) Expression levels in tissues of normal weight colorectal cancer patients (hereinafter referred to as HW), 3) Expression levels in tissues of overweight and obese colorectal cancer patients (BMI>=25, hereinafter referred to as OW). Among these genes, only those whose expression levels were in the order of OW>HW>Normal (hereinafter, UP-ward) or Normal>HW>OW (hereinafter, DOWN-ward), and whose gene expression level differences in each group were statistically significant were found (normal sample number=9, HW sample number=78, OW sample number=184). To test statistical significance, one-way ANOVA was performed for group comparisons, and Tukey's multiple comparison analysis was performed for intergroup comparisons. R version 4.0 was used for the statistical test.

(79) As a result, as shown in FIGS. 1A to 1G, it was confirmed that the expression levels of OSBPL3 (Oxysterol Binding Protein Like 3), FAM220A (Family With Sequence Similarity 220 Member A), and TMEM9 (Transmembrane Protein 9) genes were statistically significantly increased in the HW group compared to the Normal group, and in the HW group compared to the OW group. In addition, it was confirmed that the expression levels of NNT (Nicotinamide Nucleotide Transhydrogenase), PANK3 (Pantothenate Kinase 3), PPARGC1B (Peroxisome Proliferator Activated Receptor Gamma Coactivator 1 Beta), and RTL6 (Retrotransposon Gag Like 6) genes were statistically significantly decreased in the HW group compared to the Normal group, and in the OW group compared to the HW group (*: p-value<0.05, **: p-value<0.01, ***: p-value<0.001).

(80) Survival analysis was performed on the seven gene candidates found in this way. Herein, the present inventors confirmed that the genes showing statistically significant differences as candidate biomarkers for prognosis prediction.

(81) Specifically, the survival analysis according to the gene expression level in the TCGA COAD data set was performed using the log-rank test, a univariate analysis for overall survival (HW sample number=78, OW sample number=184).

(82) As a result, as shown in FIGS. 2A to 2G, in the survival analysis on the OW group, the patients with high expression of OSBPL3, FAM220A, and TMEM9 were unfavorable for survival. Furthermore, in the survival analysis of the OW group, patients with low expression of NNT, PANK3, PPARGC1B, and RTL6 were unfavorable for survival.

Experimental Example 2: Survival Analysis According to NNT and/or OSBPL3 Protein Expression Level

(83) In order to perform a survival analysis according to the expression level of the proteins encoded by NNT and OSBPL3, among the seven genes discovered from the TCGA colorectal cancer gene expression data set, a tissue microarray was constructed from the tumor tissues derived from colorectal cancer patients, and the expression level of the proteins encoded by the candidate genes specifically expressed in obese colorectal cancer patients was quantified by immunohistochemical staining.

(84) Specifically, the study was conducted on patients who underwent colorectal cancer surgery at Gachon University Gil Medical Center (GMC) from April 2010 to January 2013. Patients who underwent surgery for primary colorectal cancer and patients whose tumors were preserved in paraffin blocks were included in the study, and a total of 476 patients were analyzed. Patients with recurrent colorectal cancer, patients whose normal colonic structure had been altered by previous surgery, patients who had received chemotherapy or abdominal radiation therapy prior to colorectal cancer surgery, and patients who had been treated for other cancers prior to colorectal cancer surgery were excluded. After microdissecting the paraffin blocks, Hematoxyling and Eosin (H&E) staining was performed, the pathological findings were reviewed, and two tumor cores were marked on the corresponding paraffin blocks. Cylindrical tumor tissue with a diameter of 2 mm was extracted using a tissue microarray machine and transferred to a new paraffin block. A new tissue microarray (TMA) block was created by inserting each cylindrical tissue from 69 patients into a single paraffin block. The above TMA block was cut into 4 m thick sections using a microtome, flattened by pulling out wrinkles, and attached to a slide in a certain orientation to dry. Next, immunohistochemical staining was performed using the dried slides. The slides were treated at 601 for 10 minutes, deparaffinized with xylene, rehydrated using different concentrations of alcohol (100% alcohol, 95% alcohol, 80% alcohol, and 70% alcohol), and then washed with distilled water. Then, for antigen recovery, 10 mM citrate buffer solution (citric acid 2.1 g in H.sub.2O 1 L, pH 6.0) was heated and when it started to boil, the slides were placed therein, boiled for another 10 minutes, and the slides were washed in cold water for 10 minutes, taken out, and washed in PBS buffer solution (NaCl 8 g, KCl 200 mg, Na.sub.2HPO.sub.4 1.44 g, KH.sub.2PO.sub.4 24 mg in 1 L H.sub.2O). Afterwards, the endogenous peroxidase activity was inhibited by treatment with 3% hydrogen peroxide for 10 minutes, and the antigen was recovered in 0.01 M sodium citrate buffer (pH 6.0) using a microwave oven. To prevent non-specific reactions, the slides were reacted with a blocking antibody (DAKO #X0909; Glostrup., Denmark) for 10 minutes at room temperature, and the samples were reacted in a humidified container at 4 C. with anti-NNT (1:250, #NBP1-32109, Novus Biological Inc., Littleton, CO, USA) and anti-OSBPL3 (1:200, #NBP1-82968, Novus Biological Inc., Littleton, CO, USA). The tissue slides were processed with a non-biotinylated horseradish-peroxidase (HRP) detection system according to the manufacturer's instructions (Gene Tech).

(85) Meanwhile, statistical analysis was performed as follows. In the present invention, the survival-related factors and survival rates at 5 years after colorectal cancer surgery were the focus of event free survival and disease specific survival analyses. In the event free survival analysis, events were defined as deaths from cancer progression, cancer recurrence, and colorectal cancer. Disease specific survival was defined as the time from surgery to death due to colorectal cancer, excluding deaths due to diseases other than colorectal cancer. The log-rank test and CPH (Cox proportional hazard) modeling for survival analysis were performed using the survival package of R version 4.0. Hazard rates for each variable obtained from the CPH model were visualized with the forestmodel package and used under R version 4.0.

(86) The results of immunohistochemical staining for NNT in normal colon epithelial cells and tumor cells are shown in FIG. 3. FIG. 3A is a photograph showing the normal colon epithelial cells showing weak cytoplasmic staining, and FIGS. 3B to 3E are photographs showing the unstained tumor cells, the tumor cells showing weak cytoplasmic staining, the tumor cells showing moderate cytoplasmic staining, and the tumor cells showing strong cytoplasmic staining, respectively. In addition, the results of immunohistochemical staining for OSBPL3 in normal colon epithelial cells and tumor cells are shown in FIG. 4. FIG. 4A is a photograph showing the unstained normal colon epithelial cells, and FIGS. 4B to 4E are photographs showing the unstained tumor cells, the tumor cells showing weak cytoplasmic staining, the tumor cells showing moderate cytoplasmic staining, and the tumor cells showing strong cytoplasmic staining, respectively.

(87) As a result of statistical analysis, as shown in FIGS. 5A to 5C, it was confirmed that the event free survival of colorectal cancer patients significantly differed according to the expression level of NNT protein in the OW group. The event free survival risk rate of the OW group patients with low expression of NNT protein was 2.2 times higher than that of the patients with high expression of NNT protein (95% confidence interval [CI]: 1.01-4.80), and the p-value of the CPH model corrected with gender and age was 0.047, which was statistically significant. Meanwhile, as shown in FIGS. 6A to 6C, in the case of OSBPL3 protein, in the OW group patients obtained from the GMC colorectal cancer cohort, the group with high OSBPL3 expression tended to have a higher event free survival risk rate than that of the group with low OSBPL3 expression. In addition, when all patients were targeted regardless of weight status, the event free survival risk rate of patient group with high expression of OSBPL3 protein was 1.7 times higher than that of the patient group with low expression of OSBPL3 protein (95% CI: 1.04-2.79), and the p-value of the CPH model corrected with gender and age was 0.034, which was statistically significant.

(88) In addition, as shown in FIGS. 7A to 7C, when the expression levels of NNT and OSBPL3 proteins were analyzed in combination, the OW group patients with low expression of NNT protein and high expression of OSBPL3 had a 2.02 times higher event free survival risk rate compared to the other groups (95% CI: 1.05-3.90) and the p-value of the CPH model corrected with gender and age was 0.035, which was statistically significant. Meanwhile, as shown in FIGS. 8A to 8C, even in the event free survival analysis, it was confirmed that the OW group patients with low expression of NNT protein and high expression of OSBPL3 had a tendency to have a disadvantageous survival compared to other patient groups.

(89) The above results indicate that it is possible to predict survival of colorectal cancer patients using NNT and/or OSBPL3 gene.

Experimental Example 3: Survival Prediction Combining TNM Stage and NNT Protein Expression Level

(90) Survival prediction was analyzed by combining the TNM stage (tumor, node, metastasis stage) and NNT protein expression level, which are accepted as the best performing biomarkers for predicting colorectal cancer patient survival.

(91) Specifically, the OW group of the GMC colorectal cancer cohort was divided into 4 groups (group 1: patients with high NNT protein expression (hereinafter referred to as NNT.sub.High) and TNM stage 1/2, group 2: patients with low NNT protein expression (hereinafter referred to as NNT.sub.low) and TNM stage 1/2, group 3: patients with NNT.sub.High and TNM stage 3/4, and group 4: patients with NNT.sub.Low and TNM stage 3/4), and the event free survival analysis was performed. As a comparison, the OW colorectal cancer patients were grouped into the patients of TNM stage 1/2 and the patients of TNM stage 3/4 for the event free survival analysis.

(92) In addition, survival analysis of gene candidates selected by TCGA COAD data analysis in the GMC colorectal cancer cohort according to their protein expression determined by IHC staining was tested using the log-rank test, a univariate analysis, and the CPH modeling, a multivariate analysis, for the event free survival and disease specific survival. The log-rank test and CPH modeling for survival analysis were performed using the survival package in R version 4.0. The hazard rate of each variable obtained from the CPH modeling was visualized using the forestmodel package and used under R version 4.0.

(93) As a result, as shown in FIGS. 9A to 9D, in the event free survival analysis using only the TNM stage, the event free survival risk rate of the patient group with TNM stage 3/4 was 12.07 times (95% CI: 4.28-34.08, log-rank test p-value<0.001) higher than that of the patient group with TNM stage 1/2. On the other hand, when the NNT and TNM stage were combined for analysis, the event free survival risk rate of the group 3 was 8.39 times (95% CI: 1.03-68.39, log-rank test p-value: 0.047) higher than that of the group 1, and the event free survival risk rate of the group 4 was 21.2 times (95% CI: 2.88-156.08, log-rank test p-value: 0.003) higher than that of the group 1. As shown in FIGS. 9E and 9F, in the patients with the same stage (TNM stage 3/4), the event free survival risk rate of the group with low NNT expression was 2.52 times higher than that of the group with high NNT expression, which was statistically significantly worse prognosis.

(94) In addition, as shown in FIGS. 10A to 10D, in the disease specific survival analysis using only the TNM stage, the disease specific survival risk rate of the patient group with TNM stage 3/4 was 8.33 times (95% CI: 2.9-23.9, log-rank test p-value<0.001) higher than that of the patient group with TNM stage 1/2. On the other hand, when the NNT and TNM stage were combined for analysis, the disease specific survival risk rate of the group 3 was 6.25 times higher than that of the group 1, and the event free survival risk rate of the group 4 was 14.05 times (95% CI: 1.89-104.38, log-rank test p-value: 0.01) higher than that of the group 1. As shown in FIGS. 10E and 10F, in the patients with the same stage (TNM stage 3/4), the disease specific survival risk rate of the group with low NNT expression was 2.23 times higher than that of the group with high NNT expression, which was statistically significantly worse prognosis.

(95) From the above results, it was confirmed that when the TNM stage and NNT expression were combined to predict survival, better survival prediction was possible than when the only TNM stage was used to predict survival. This suggests that better performance in survival prediction is possible when the NNT expression is combined with the TNM stage and used as a biomarker.

Experimental Example 4: Survival Prediction Combining TNM Stage and OSBPL3 Expression Level

(96) Survival prediction was analyzed by combining the TNM stage (tumor, node, metastasis stage) and OSBPL3 protein expression level.

(97) Specifically, the OW group of the GMC colorectal cancer cohort was divided into 2 groups (group 1: patients with high OSBPL3 protein expression (hereinafter referred to as OSBPL3.sub.High) and TNM stage 3/4, and group 2: patients with low OSBPL3 protein expression (hereinafter referred to as OSBPL3.sub.Low) and TNM stage 3/4), and the event free survival analysis was performed.

(98) As a result, as shown in FIGS. 11A to 11C, when event-free survival was determined in patients with the same stage (TNM stage 3/4), it was confirmed that the group with low OSBPL3 expression tended to have a better chance of survival than the group with high OSBPL3T expression. This suggests that better performance in survival prediction is possible when the OSBPL3T expression is combined with the TNM stage and used as a biomarker.

Experimental Example 5: Analysis of Signaling Pathways, Protein Networks, and Biological Functions

(99) Network and biological function analyses were performed to identify the differences in biological functions of the NNT protein found to be significantly associated with survival prediction in the OW group within the GMC colorectal cancer cohort.

(100) Specifically, to confirm the differences in biological functions of proteins specifically expressed in obese colorectal cancer patients, using the TCGA and COAD gene expression data sets, patients were divided into two groups: obese colorectal cancer patients with high expression of the corresponding gene and obese colorectal cancer patients with low expression of the corresponding gene, and compared with the expression of the gene in the normal tissues of normal weight colorectal cancer patients. Biological functions with high biological correlation to genes were searched in the Comparative Toxicogenomics Database (CTD). A network of highly related proteins was constructed using the String database, and the biological functions of the proteins in the network were confirmed using Reactome, a biological pathway database provided by the String database. The biological functions according to the differences in expression levels of genes encoding proteins specifically expressed in obese colorectal cancer patients were analyzed with the Gene Set Enrichment Analysis (GSEA), an analysis tool.

(101) As a result, as shown in FIG. 12, the biological functions with high biological correlation with the NNT gene were confirmed as nicotinate and nicotinamide metabolism, pyruvate metabolism, citric acid (TCA) cycle, etc.

(102) In addition, as shown in FIGS. 13A and 13B, the network of proteins closely related to the NNT protein and the biological functions associated with the proteins participating in this network were confirmed to be statistically significant in nicotinate and nicotinamide metabolism and TCA cycle.

(103) In addition, from the TCGA colorectal cancer gene expression dataset, only the obese patient group and the normal group were extracted and analyzed with GSEA. As a result, as shown in FIG. 14, it was confirmed that there were statistically significant differences in signaling pathways between the obese colorectal cancer patient group with NNT.sub.High and the obese colorectal cancer patient group with NNT.sub.Low. Specifically, the obese NNT.sub.Low group, which was confirmed as being disadvantageous for survival, showed statistically significant suppression of nicotinate and nicotinamide metabolism, inositol phosphate metabolism, and purine metabolism compared to the normal group. On the other hand, the obese NNT.sub.High group, which was confirmed to have an advantage in survival among the obese colorectal cancer patients, showed statistically significant suppression of glycosphingolipid biosynthesisganglio series compared to the normal group.