KR101829450B1 - Evaluation of drug-targetable genes by defining modes of abnormality in gene expression - Google Patents
Evaluation of drug-targetable genes by defining modes of abnormality in gene expression Download PDFInfo
- Publication number
- KR101829450B1 KR101829450B1 KR1020160005101A KR20160005101A KR101829450B1 KR 101829450 B1 KR101829450 B1 KR 101829450B1 KR 1020160005101 A KR1020160005101 A KR 1020160005101A KR 20160005101 A KR20160005101 A KR 20160005101A KR 101829450 B1 KR101829450 B1 KR 101829450B1
- Authority
- KR
- South Korea
- Prior art keywords
- gene
- drug
- genes
- expression
- region
- Prior art date
Links
Images
Classifications
-
- G06F19/18—
-
- G06F19/16—
-
- G06F19/28—
Abstract
The present invention relates to a method for easily verifying recoverability of genes useful as drug targets in order to study abnormal diversity in addition to discrimination of abnormal genes, and to evaluate various other types of abnormal genes through prediction of results after drug treatment And introduced the concept of "drug targetability". Specifically, the method of the present invention was applied to breast cancer, and PTPRF, PRKAR2B, MAP4K3, and RICTOR were found to be high-fitness genes for breast cancer. The siRNAs were used to inhibit genes having high drug target fitness and then evaluated by the method of the present invention using cell death and cell migration test. As a result, it was confirmed that the inhibition of RICTOR or PTPRF could prolong the life of breast cancer patients , And the method of the present invention is widely applied to the selection and evaluation of novel drug targets and ultimately contributes to the improvement of the efficiency of disease treatment.
Description
The present invention relates to a method for evaluating a drug-targetable gene through identification of a gene abnormal expression mode.
Systematic identification of new drug targets is one of the most common applications of large gene expression profiles. As one of the conventional methods, differentially expressed genes (DEGs), which basically express expression differences obtained from microarray experiments, have been studied. However, since microarray data is inconsistent and less reproducible, a large number of samples are required, which limits the practical application of DEG in this field [Ein-Dor, L., Zuk, O. & Domany, E. Thousands of samples needed to generate a robust gene list for predicting outcome in cancer. Proc Natl Acad Sci USA 103, 5923-8 (2006) / Kim, SY Effects of sample size on robustness and prediction accuracy of a prognostic gene signature. BMC Bioinformatics 10, 147 (2009)]. In addition, the gene expression level does not show a significant difference in specific groups, and the relationship between DEG and phenotype is unclear. It is uncertain whether DEG can be a good biomarker or drug target for a particular phenotype. Despite various data normalization, sample addition and gene-set approaches [Steinhoff, C. & Vingron, M. Normalization and quantification of differential expression in gene expression microarrays. Brief Bioinform 7, 166-77 (2006)], expression level analysis alone is not sufficient to identify good drug targets. Therefore, a method that introduces characteristics different from expression difference is needed.
In previous studies, the inventors have succeeded in identifying a number of causative genes related to the chemosensitivity of tamoxifen and epirubicin using a transcriptional reaction to develop a platform for identifying phenotypic determining genes [Lee, J. et al. , Park, J. & Choi, C. Identification of phenotype deterministic genes using systemic analysis of transcriptional response.
Therefore, simply identifying an abnormal gene is insufficient for use as an effective drug target. Time predicted drug targets specific to effective breast cancer subtypes, using integrated sequencing analysis and RNAi screening data. They identified genes essential for cell proliferation and survival and applied this information to human signaling networks representing key signaling networks (Zaman, N. et al. Signaling network assessment of mutations and copy number variations predict breast cancer subtype-specific drug targets. Cell
Many scientists have attempted to develop effective methods for the systematic discovery of functional gene genes in various diseases. The inventors have successfully shown in previous studies that the degree of transcriptional response can be indicative of the determined genes for different phenotypes. However, most of these abnormalities in clinical drug therapy are difficult to recover. In addition, it has been confirmed that in practical applications, the discovery of abnormal genes is not sufficient. In this study, the present inventors aimed at simply identifying restorable genes as drug targets.
A wide selection of effective drug targets is important in target therapy in the post-genomic era. To achieve this, the present inventors classified and ranked abnormal genes into several different modes. In fact, the precise modulation of gene expression or activity is highly limited in practice. Most available drugs merely reduce the expression or activity of the target gene [Gerber, D. E. Targeted therapies: a new generation of cancer treatments. Am Fam Physician 77, 311-9 (2008) / Bailly, C. Ready for a comeback of natural products in oncology. Biochem Pharmacol 77, 1447-57 (2009)].
Thus, among the many genes that exhibit abnormal transcriptional responses, it is necessary to select genes that will regain normal levels when expression levels decrease. The characteristics of these genes were defined as drug targetability and applied to breast cancer samples. The present inventors predicted PTPRF, PRKAR2B, MAP4K3, and RICTOR as genes that are thought to be related to the phenotype of cancer, and each gene was experimentally evaluated through apoptosis or migration experiment.
In a previous study, we investigated drug candidate candidates, RICTOR and MAP4K3. RICTOR, a component of the well-known
Thanks to high-throughput technology, many researchers have attempted to systematically identify abnormal genes in various diseases. However, only a few of these genes can be used as drug targets. It is not enough to identify only abnormal genes. Evaluation of various modes of abnormality is required. The present invention attempts to predict events after inhibition of each gene by calculating the drug target fitness.
In the present invention, the concept of drug targetability is introduced into the 'degree of similarity to a normal sample after inhibition of an abnormal gene', and this is applied to identify a breast cancer gene having a high target compatibility with a drug. it was confirmed that the present invention's computer prediction is effective by suppressing the highest-ranking gene (high suitability for drug target) using siRNA and performing cell death and cell migration evaluation. From the successful results of the present invention, the method of the present invention will be widely applicable to the evaluation of novel drug targets for various diseases.
It is an object of the present invention to provide a method for evaluating a drug-targetable gene through identification of a gene abnormal expression mode.
According to an aspect of the present invention,
1) extracting gene expression data from a disease sample and a normal control sample from a gene expression information database;
2) The disease samples were classified into the low region (LR) and the high region (HR) according to the expression level of the pathway gene, and the expression level of the path gene and the target gene Calculating a P-score indicating a degree of similarity of gene expression patterns between two samples in each region by comparing the normal control sample and the normal control sample; And
3) selecting a path gene having a high absolute value of the P-index (HP) in the high region group and a low absolute value of the p-index (LP) in the low region group; (drug targetability).
The present invention relates to a method for evaluating a drug-targetable gene through identification of a gene abnormal expression mode, wherein the method of the present invention is applied to breast cancer and PTPRF, PRKAR2B, MAP4K3, and RICTOR are used for breast cancer And the inhibition of RICTOR and PTPRF, which had high drug targeting suitability as a result of analysis of clinical data, were found to be inhibited by siRNA, It has been confirmed that the method of the present invention is widely applicable to the selection and evaluation of novel drug targets and ultimately contributes to the improvement of the efficiency of disease treatment.
1 shows a concept of the present invention:
1a. Several abnormal genes (genes A and B) can cause disease. Here, when gene A is inhibited by Drug A, the patient's phenotype is converted to another kind of abnormal state (the left indicates poor target with poor drug target fitness). On the other hand, when gene B is inhibited by Drug B, the patient's phenotype is restored to a normal state (the right indicates an excellent target with a high drug target fitness).
1b. FIG. 4 is a graph showing the relationship between a single path unit and a target gene. FIG.
For one gene related to the pathway (pathway gene), the target gene of the transcription factors (TFs) associated with the same pathway unit is considered the target gene of the pathway gene. For the path information, the UnitPath database was used and MSigDB was used to obtain the target gene information of TFs.
1C. And a P-index (score) in the low region and the high region. Expression levels of the target gene (Y axis) in each region were compared according to the pathway gene expression level (X axis), by dividing the subregion (≤0.5) and the high region (> 0.5).
Figure 2 shows the calculation of drug target fitness and various types of abnormal situations:
(A) An illustration of drug target fitness and its two-dimensional thermal mapping in terms of LP (P-score in the low region) and HP (P-score in the high region).
(BE) A diagram showing various types of drug target fitness shown in (A), together with gene expression patterns:
B and C: Modes of good drug targets; And
D and E: Examples of bad drug targets.
3 is an illustration of evaluation of drug target fitness in breast cancer.
(A) Two-dimensional heat map for the number of path gene-target gene pairs (log 10 ) in terms of LP and HP.
(AP4M1, MMP11, IGFBP6, and LAMB3) of the BE pathway genes (NCK1, ACVR2A, PSEN1 and GSTP1) and their target genes
B is high for HP;
C for high LP and high for HP;
D for low LP and low HP; And
E for low HP.
(F) Distribution of drug target fitness values for all pathway gene-target gene pairs.
(G) distribution of drug target fitness values integrated for each pathway gene, which was derived from path gene-target gene pairs with the top 5% drug target fitness.
(H) The X-axis represents the rank of the drug target fitness from the entire data set, and the Y-axis represents the rank of the drug target fitness from the 100 random samples. The slope of the linear regression was not significantly zero (t = 44.78), and Pearson's coefficient of the two ranks was also statistically significant (P <0.0001, R 2 = 0.3084).
Figure 4 is a measure of cell viability when target genes (PTPRF, PRKAR2B, MAP4K3, and RICTOR) are knocked down:
4a: After 48 hours of transformation, the cells were treated with doxorubicin (5 mg / L), epirubicin (5 mg / L) and 5-fluorouracil And the cell viability was measured by WST-assay.
Figure 4b: After 48 hours of transformation, cells were cultured in EGF or serum-free medium for 24 hours and cell viability was measured by WST-assay. Results are expressed as means ± SEM, n = 4.
Each group (control, doxorubicin, and EGF) was compared by one-way ANOVA using Tukey's post hoc test for multiple comparisons. In all groups, the ANOVA was P <0.0001 and the asterisk indicates significant differences when compared to the control group in the post-mortem analysis (* P <0.05, ** P <0.01, *** P <0.001). (B), the degree of significance of comparison between the control and EGF groups was determined using a student's T-test (NS, meaning not significant).
Fig. 5 is a diagram showing cell migration evaluation of a knockdown target gene. Fig. Knockdown of PTPRF, PRKAR2B, MAP4K3, and RICTOR was performed by siRNA transfection and cell migration was analyzed by wound-healing assay:
5a: After 48 hours of transformation, MDA-MB-231 cells underwent wound healing assays. The horizontal axis (white dotted line) represents the wounded region. Photographic images were taken using phase contrast microscopy at 0, 12 and 24 hours.
Figure 5b: Cell migration was quantified in terms of two functions using CellProfiler. The black bars represent the percentage of the area occupied by the transferred cells relative to the total wounded area.
Figure 6 shows the results of Kaplan-Meier survival analysis for grouped data sets. The entire breast cancer data set was divided into two groups according to the expression levels of the drug target genes (PTPRF, PRKAR2B, MAP4K3, and RICTOR). The grouped data set was analyzed by Kaplan-Meier survival analysis to compare distant metastasis free survival (DMFS), relapse free survival (RFS), and OS (survival time) Respectively. P <0.001, P (0.0096), and P (P) were significantly different according to the log-rank test according to the amount of RICTOR and PTPRF expression = 0.023).
Hereinafter, the present invention will be described in detail.
According to the present invention,
1) extracting gene expression data from a disease sample and a normal control sample from a gene expression information database;
2) The disease samples were classified into the low region (LR) and the high region (HR) according to the expression level of the pathway gene, and the expression level of the path gene and the target gene Calculating a P-score indicating a degree of similarity of gene expression patterns between two samples in each region by comparing the normal control sample and the normal control sample; And
3) selecting a path gene having a high absolute value of the P-index (HP) in the high region group and a low absolute value of the p-index (LP) in the low region group; (drug targetability).
The gene expression information database of the above step 1) may be classified into a gene expression omnibus database (GEO), an ArrayExpress, a cancer genome atlas (TCGA), and an International Cancer Genome Database (ICGC) Consortium database). However, the present invention is not limited to this, and any database may be used as long as it can obtain gene expression data information.
It is preferable to distinguish the high-level region group and the low-level region group according to the expression level of the pathway gene in the step 2). In the present invention, normalization is performed using the Universal exPression Code method (SCAN.UPC package of bioconductor) When the expression level is above 0.5, it is classified as a high region group. When the expression level is higher than 0.5, it is classified as a sub-region group. When other correction methods are used, the average value or median value of each gene expression value can be used as a value to distinguish between the sub- .
Preferably, the gene having high fitness for drug target is a gene that restores the expression of the transcription target gene of the protein when the function of the protein encoded by the gene is inhibited.
The P-score of the above step 2) is preferably derived from the following equation (P-value of the expression below is the expression level of the target gene of the normal sample and disease sample in the sub-region group or the high- Lt; / RTI >t-test);
The drug targetability (DT) is preferably but not limited to be quantified in the following formula:
(Where HP is the P-index of the high-level region (HR) and LP is the P-index of the sub-region (LP)).
The disease is preferably cancer, but the present invention is not limited thereto, and the present inventors have conducted breast cancer.
The present inventors have defined the drug target fitness as follows.
1A, several abnormal genes (gene A and gene B) can cause disease. Here, when gene A is inhibited by Drug A, the patient's phenotype is converted to another kind of abnormal state. On the other hand, when gene B is inhibited by Drug B, the patient's phenotype is restored to its normal state. Thus, the gene B has a high drug target fitness and is a good drug target. In addition, the relationship between the single path unit and the target gene is shown in Fig. 1B. Here, the target gene of the transcription factors (TFs) related to the same pathway unit for one gene related to the path unit (path gene) It is considered to be the target gene of path gene. For the path information, the UnitPath database was used and MSigDB was used to obtain target gene information of TFs (FIG. 1B). In addition, the expression level of the target gene (Y axis) in each region was compared by sub-region and high-level region according to the pathway gene expression level (X axis) (FIG.
In a specific embodiment, the inventors have downloaded a raw CEL file from the Gene Expression Omnibus database using the Affymetrix Human Genome U133 Plus 2.0 Array (GPL570) platform to obtain a public microarray database, (Piccolo, SR et al., Proc Natl Acad Sci USA 110, 17778-83, 2013) using the exPression Code method (SCAN.UPC package of bioconductor).
We also used the UnitPath database (www.unitpath.com), which has a well-defined path and an interactive modularity for computational analysis. In order to evaluate the index of statistical significance, a P-score was defined according to the P-value:
We have shown that for all pathway gene-target gene pairs, the expression levels of the low region (LR) and the high region , HR) (HR for pathway genes with an expression level greater than 0.5, and LR otherwise). Expression of target genes between two groups (e. G., Normal and cancer patient samples) were compared separately for each region and the P-score and P-value ) Were calculated. The value of drug targetability (DT), which is large when HP is large and low when LP is small, is defined as:
The present inventors have shown a heat map of the defined equation and drug target compatibility (see A in FIG. 2), and the central region of the thermal map (both LP and HP values have very small values) response genes. In the regions B and C (small LP, large HP) on the thermogram, the target gene expression in HR was remarkably different from that in the normal group, but was similar in LR (Fig. 2B and C). As a result, it can be expected that drugs that target or inhibit the pathway gene will restore a transcriptional response similar to that of the normal group. Thus, by classifying the types of abnormality by determining the drug target fitness for all pathway gene-target gene pairs and when the inhibition of specific genes in disease patients is expected to be restored to normal transcription , Drug target fitness was defined as having a high value.
The present inventors investigated drug target fitness in breast cancer. The values of LP (P-score in the low region) and HP (P-score in the high region) were calculated for all the pathway gene-target gene pairs as described above and represented by a 2D thermal map histogram As a result, in the case of B or E on the thermogram, the target gene expression in the cancer was confirmed to be abnormal only in the high region group (HR) (see B and E in FIG. 3). Therefore, if expression of NCK1 or GSTP1 can be inhibited in cancer, it can be expected that the expression of the target gene will be similar to that of a normal control.
In order to verify the accuracy of prediction by a computer, the present inventors classified all path genes according to drug target fitness and knockdown four top ranking genes, namely, PTPRF, PRKAR2B, MAP4K3 and RICTOR, The viability of the cells was analyzed. As a result, in MCF-7 and MDA_MB-231 cells, the knockdown of MAP4K3 or RICTOR decreased cell viability and increased sensitivity to cell death induced by anticancer drugs (doxorubicin, epirubicin and 5-fluorouracil) (See Fig. 4A). In addition, EGF treatment induced cell proliferation under serum-free medium conditions. As a result, MAP4K3 or RICTOR knockdown canceled the proliferative effect of EGF in all cell lines (see FIG. 4B). These results indicate that MAP4K3 and RICTOR are responsible for proliferation and growth in breast cancer cells.
The present inventors also examined cell movement, which is another characteristic of cancer. As a result, it was confirmed that cell migration was significantly reduced in cells transformed with PRKAR2B, MAP4K3, or RICTOR siRNA as compared with the control group (see FIG. In particular, when MAP4K3 was knocked down, it was confirmed that cell migration was almost completely blocked. These results were confirmed by quantification using the number of cells and occupied areas in the wounded region (see FIG. 5B).
In order to examine the effect of the proposed drug target-compatible gene transcription on the transcription, RICTOR was knocked down after siRNA and the mRNA expression profile was compared with the control group. Since the present invention identifies pathway genes of target genes that vary similarly to normal after inhibition, the fold changes in siRICTOR: control microarray results are compared in normal and cancer patients. It was confirmed that the microarray data had a statistically significant Pearson correlation with a 20% cutoff value (multiple, 1.2 <or 0.8) (R = 0.6836, P-value = 0.0204). These results imply that inhibition of RICTOR results in a similar level of expression of the target gene to normal.
The present inventors also conducted Kaplan-Meier survival analysis using patient information in each microarray data to verify the experimental results. The entire breast cancer data set was divided into low or high expression levels of a specific drug target gene, and in each group of data, distant metastasis-free survival (DMFS), relapse-free survival RFS) and overall survival (OS). As a result, patients with lower levels of RICTOR exhibited significantly longer overall survival (OS) (see FIG. 6A), patients with lower PTPRF levels had significantly longer distant metastasis-free survival (DMFS) ) (See B and C in Fig. 6). These results suggest that inhibition of these genes may improve medical outcome and prolong survival of breast cancer patients.
In addition, the inventors of the present invention found that GNA15, PPP2R3A, LAMB3, HSPA2, and LPS were found as a result of comparing between the stool cancer sample and the corresponding normal sample, in order to confirm whether the method proposed by the present invention can be widely applied to many other diseases MAP4K3 has been identified as the top five-ranking genes, among which GNA15, LAMB3, HSPA2, and MAP4K3 have been previously reported as prognostic markers in a variety of cancer types that indirectly support the methods of the present invention.
The method of the present invention is not limited to cancer and is applicable in any case that is interested in two different phenotypes. The approach according to the invention will improve the therapeutic efficacy of many diseases.
Hereinafter, the present invention will be described in detail with reference to Examples and Experimental Examples.
However, the following Examples and Experimental Examples are merely illustrative of the present invention, and the content of the present invention is not limited by the following Examples and Experimental Examples.
< Example 1> Microarray Data set and analysis
The raw CEL file was downloaded from the public microarray database Gene Expression Omnibus database using the Affymetrix Human Genome U133 Plus 2.0 Array (GPL570) platform, and the Universal exPression Code method (SCAN.UPC package of the bioconductor) (Piccolo, SR et al., Proc Natl Acad Sci USA 110, 17778-83, 2013). The samples were grouped into normal and cancer according to an annotation in the original GSM file. During microarray experiments, total RNA was isolated from MCF-7 cells using RNeasy Mini Kit (Qiagen, Hilden, Germany) and Affymetrix human 2.0 ST array was purchased from commercial microarray service (eBiogen, Seoul, Analysis was carried out.
< Example 2 > predefined pathways and target genes Data set
We used the UnitPath database (Unpublished Data, www.unitpath.com), which has a well-defined path and an interactive modularity for computational analysis. UnitPath uses all of these pathways for analysis, as the user can select one gene of interest and acquire other genes that affect the selected gene in each pathway. In the search of each unit, the transcription factor was set to the end point. The transcription factors (TF) (the target gene of each TF) were obtained from MSigDB c3 TFT v4.0. Finally, all genes associated with at least one transcription factor (TF) in the corresponding target gene were plotted (FIG. 1B). Finally, the analysis included 1,191 pathways and 10,305 target genes from the pathway genes.
< Example 3> Statistical significance
To evaluate the index for statistical significance, a P-score was defined according to the P-value ( P -value). The P-index evaluation takes the following form:
The larger the p-value, the smaller the p-score. Multiple cut calibration thresholds were excluded because arbitrary cutoff thresholds were not used at significant levels in the calculation of drug target fitness. To estimate the P-index to be 0.5 when the P-value is 0.01, a =
< Example 4> Search for genes with drug target compatibility
For all the pathway gene-target gene pairs obtained in the above Example 2, the expression levels of the low region (LR) and the low region (HR for pathway genes with an expression level of greater than 0.5, and LR if not). Expression of target genes between two groups (for example, normal and cancer patient samples) was separately compared for each region, and P-score and P-value were calculated.
As shown in Figure 1c, the P-score (HP) in HR is high if one target gene shows a markedly higher expression level in cancer than in the control (HR, blue). However, the P-score (LP) in LR is small if any target gene shows a similar expression level in the normal group and cancer. When HP is large and LP is small, the value of large drug targetability (DT) is defined as:
Figure 2 A shows the heat map of the defined equation and drug target fitness. According to the defined equation, the central region of the thermogram (LP and HP values all have very small values) indicated normal genes in terms of transcriptional response.
As shown in Figs. 2B and 2C, in the regions B and C (small LP, large HP) on the thermogram, the target gene expression in HR is remarkably different as compared with the normal group, but similar in LR B and C in Fig. 2). As a result, it can be expected that drugs that target or inhibit the pathway gene will restore a transcriptional response similar to that of the normal group.
However, as shown in Figures 2 D and E, target gene expression differs significantly only in LR (D and E in Figure 2) in regions D and E on the thermal map (large LP, small HP). In this case, suppression of the pathway gene is unlikely to restore the transcriptional response similar to that of the normal group.
As described above, various types of abnormality were classified by determining drug target fitness for all path gene-target gene pairs. Drug target fitness was defined as having a high value when the inhibition of a particular gene in a patient with the disease would be expected to restore to a normal group of transcriptional responses.
< Experimental Example 1> Evaluation of drug target fitness in breast cancer
The present inventors investigated drug target fitness in breast cancer. The values of LP (P-score in the low region) and HP (P-score in the high region) were calculated for all path gene-target gene pairs in the same manner as in Example 4, Are also shown in the histogram.
As a result, as shown in Fig. 3, although a number of path gene-target gene pairs existed in the central region having a normal transcription reaction (Fig. 3A), pairs indicating an abnormal transcription reaction showing high LP or HP could also be confirmed .
In the case of C or D on the thermogram, the target gene expression was found to be essentially high (C in Fig. 3) or low (Fig. 3 D) for cancer. These gene pairs (ACVR2A or PSEN1) exhibit abnormal responses, but their target genes can not be good drug targets because they still exhibit abnormal expression levels when their expression is low.
However, in the B or E case on the thermogram, the target gene expression in the cancer was confirmed to be abnormal only in the high region group (HR) (FIGS. 3B and 3E). Therefore, if expression of NCK1 or GSTP1 can be inhibited in cancer, it can be expected that the expression of the target gene will be similar to that of a normal control.
The drug target fitness frequency distribution was shown for all pathway gene-target gene pairs (Fig. 3F). In order to use potential path genes as novel drug targets, each pair of drug target suitability was averaged to a single value so that one path gene had one value for drug target suitability. In order to prominently display the abnormal pathway gene-target gene pair, only the upper 5% pairs of each pathway gene were used (G in FIG. 3). The distribution of these integrated drug target fitness values (finally used values) is similar to that of each pair. For a systematic analysis of drug target fitness, it is possible to easily identify drug target compatible genes in the overall pathway diagram by listing these values in the cancer pathway from UnitPath.
To confirm the usefulness of the method, 100 random sampling runs were performed and ranks from the drug target fitness rank and random samples from the entire data set were compared. The ranks showed statistically significant Pearson correlation coefficients, especially at high ranks (H in Fig. 3).
< Experimental Example 2> Experimental Verification of Drug Target Suitability in Breast Cancer
<2-1> Cell Survival Confirm
In order to verify the accuracy of the prediction by the computer, all path genes were classified according to the drug target fitness, and four top ranking genes, namely, PTPRF, PRKAR2B, MAP4K3 and RICTOR were selected in the <Experimental Example 1>. After knockdown of these genes, cell viability was analyzed by various anti-cancer drugs.
Specifically, MCF-7 and MDA-MB-231 cells were incubated with 10% fetal bovine serum (Gibco, Gaithersburg, MD), 100 U / ml penicillin, and 100 μg / ml streptomycin (Invitrogen, Carlsbad, CA) Dulbecco's modified Eagle's medium (Welgene, Seoul, Republic of Korea) was maintained at 37 ° C in an atmosphere containing 5% CO 2 . Anticancer drugs (doxorubicin, epirubicin and 5-fluorouracil) were purchased from Sigma-Aldrich (St. Louis, Mo.).
Targeted siRNA duplexes for PTPRF, PRKAR2B, MAP4K3, and RICTOR, and siRNAs with a random sequence (negative control) were synthesized by Bioneer (Daejeon, Republic of Korea). Transformation was carried out using lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol.
WST-1 assays were performed to determine cell viability. MCF-7 and MDAMB-231 cells were inoculated into 24-well plates at a density of 4 × 10 4 cells / well four times and WST-1 reagent (Nalgene, Rochester, NY) Well. After incubation for 2 hours at 37 ° C in a 5% CO 2 incubator, the absorbance was measured at 450 nm using a microplate reader (Bio-Rad, Richmond, Calif.). Cell death was confirmed by staining with 100 nM tetramethylrhodamine ethyl ester (TMRE, Sigma-Aldrich). 10,000 cells were analyzed using a Caliber flow cytometer.
As a result, as shown in Fig. 4, in MCF-7 and MDA_MB-231 cells, the knockdown of MAP4K3 or RICTOR decreased cell viability and induction by anticancer drugs (doxorubicin, epirubicin and 5-fluorouracil) Lt; RTI ID = 0.0 > (FIG. 4A). ≪ / RTI > In addition, proliferation of cells was induced by treatment with EGF under serum-free medium conditions. As a result, the knockdown of MAP4K3 or RICTOR canceled the proliferative effect of EGF in all cell lines (Fig. 4B). These results indicate that MAP4K3 and RICTOR are responsible for proliferation and growth in breast cancer cells.
<2-2> Confirmation of cell migration
The present inventors investigated cell migration, another characteristic of cancer. Each of the siRNAs was transformed as described in Experimental Example <2-1>, and wound-healing assay was performed 48 hours later with MDA-MD-231 cells.
Specifically, to create a cell-free (wounded) site, adhesive MDA-MB-231 cells were scraped from the bottom of the culture surface (6-well plate) using a pipette tip. To remove cell debris, each well was washed with PBS and then cultured in serum-free medium. Various characteristics of cell migration were evaluated using CellProfiler (Carpenter, A. E. et al.,
As a result, as shown in FIG. 5, cell migration was significantly reduced in cells transformed with PRKAR2B, MAP4K3, or RICTOR siRNA as compared with the control group (FIG. 5A). In particular, when MAP4K3 was knocked down, it was confirmed that cell migration was almost completely blocked. These results were confirmed by quantification using the number of cells and the occupied area in the wounded region (Fig. 5B).
<2-3> Verification of experimental results
In order to examine the effect of the proposed drug target-compatible gene on transcription, direct validation cDNA microarrays were performed. After RICTOR was knocked down the siRNA, the mRNA expression profile was compared to the control.
Since the present invention identifies pathway genes of target genes that vary similarly to normal after inhibition, the fold changes in siRICTOR: control microarray results are compared in normal and cancer patients. It was confirmed that the microarray data had a statistically significant Pearson correlation with a 20% cutoff value (multiple, 1.2 <or 0.8) (R = 0.6836, P-value = 0.0204). These results imply that inhibition of RICTOR results in a similar level of expression of the target gene to normal.
To verify the experimental results, Kaplan-Meier survival analysis was performed using patient information in each microarray data.
First, we divide the entire breast cancer data set into groups of lower or higher expression levels of a particular drug target gene and determine the distant metastasis-free survival (DMFS), non-recurrence-free survival relapse-free survival (RFS), and overall survival (OS).
As a result, patients with lower RICTOR levels showed significantly longer overall survival (OS) as shown in Fig. 6 (Fig. 6A), patients with lower PTPRF levels experienced significantly longer distant metastasis free survival (DMFS) (RFS) without recurrence (B and C in Fig. 6).
Although all other cases are not statistically significant, these results suggest that inhibition of these genes may improve medical outcomes and prolong survival of breast cancer patients.
< Experimental Example 3> Evaluation of Drug Target Suitability in Sulcus
Other examples were investigated to ascertain whether the method proposed by the present invention could be applied to a wide variety of other diseases.
Comparison of the stool samples and the corresponding normal samples revealed that GNA15, PPP2R3A, LAMB3, HSPA2, and MAP4K3 were the top five-ranked genes.
Of these, GNA15, LAMB3, HSPA2, and MAP4K3 have been previously reported as prognostic markers in a variety of cancer types that indirectly support the methods of the present invention.
The method of the present invention is not limited to cancer and is applicable in any case that is interested in two different phenotypes. The approach according to the invention is expected to improve the therapeutic efficacy of many diseases.
Claims (7)
2) The disease samples were classified into the low region (LR) and the high region (HR) according to the expression level of the pathway gene, and the expression level of the path gene and the target gene Calculating a P-score indicating a degree of similarity of gene expression patterns between two samples in each region by comparing the normal control sample and the normal control sample; And
3) selecting a path gene having a large absolute value of the P-index (HP) in the high region group and a small absolute value of the P-index (LP) in the low region group;
In a method for screening a gene having high drug targetability for a disease,
Wherein the P-score of the step 2) is derived by the following equation: < RTI ID = 0.0 >
(Wherein,
The P-value represents the value calculated by t-testing the expression level of the target gene in the normal control sample and the disease sample in the low region (LR) or high region (HR).
Here, HP is the P-index of the high-level region (HR) and LP is the P-index of the sub-region (LP).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020160005101A KR101829450B1 (en) | 2016-01-15 | 2016-01-15 | Evaluation of drug-targetable genes by defining modes of abnormality in gene expression |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020160005101A KR101829450B1 (en) | 2016-01-15 | 2016-01-15 | Evaluation of drug-targetable genes by defining modes of abnormality in gene expression |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| KR20170086149A KR20170086149A (en) | 2017-07-26 |
| KR101829450B1 true KR101829450B1 (en) | 2018-02-20 |
Family
ID=59427209
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| KR1020160005101A KR101829450B1 (en) | 2016-01-15 | 2016-01-15 | Evaluation of drug-targetable genes by defining modes of abnormality in gene expression |
Country Status (1)
| Country | Link |
|---|---|
| KR (1) | KR101829450B1 (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100692319B1 (en) | 2006-02-20 | 2007-03-12 | 한국생명공학연구원 | The finding method of new disease-associated genes through analysis of protein-protein interaction network |
| JP2015525886A (en) | 2012-08-08 | 2015-09-07 | エフ．ホフマン−ラ ロシュ アーゲーＦ． Ｈｏｆｆｍａｎｎ−Ｌａ Ｒｏｃｈｅ Ａｋｔｉｅｎｇｅｓｅｌｌｓｃｈａｆｔ | Determination of binding kinetics in solution based on immunoassay method |
-
2016
- 2016-01-15 KR KR1020160005101A patent/KR101829450B1/en active IP Right Grant
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100692319B1 (en) | 2006-02-20 | 2007-03-12 | 한국생명공학연구원 | The finding method of new disease-associated genes through analysis of protein-protein interaction network |
| JP2015525886A (en) | 2012-08-08 | 2015-09-07 | エフ．ホフマン−ラ ロシュ アーゲーＦ． Ｈｏｆｆｍａｎｎ−Ｌａ Ｒｏｃｈｅ Ａｋｔｉｅｎｇｅｓｅｌｌｓｃｈａｆｔ | Determination of binding kinetics in solution based on immunoassay method |
Non-Patent Citations (2)
| Title |
|---|
| Lee 등, Identification of phenotype deterministic genes using systemic analysis of transcriptional response, Scientific Reports, 2014, 4:4413 |
| 이정설, Defining transcriptional response with gene expression profile and its application, 박사학위논문, KAIST, 2013 |
Also Published As
| Publication number | Publication date |
|---|---|
| KR20170086149A (en) | 2017-07-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Yang et al. | Transcription factor family‐specific DNA shape readout revealed by quantitative specificity models | |
| Aune et al. | Expression of long non-coding RNAs in autoimmunity and linkage to enhancer function and autoimmune disease risk genetic variants | |
| Lembo et al. | Shortening of 3′ UTRs correlates with poor prognosis in breast and lung cancer | |
| Araya et al. | Identification of significantly mutated regions across cancer types highlights a rich landscape of functional molecular alterations | |
| Zhang et al. | Apparently low reproducibility of true differential expression discoveries in microarray studies | |
| Shepherd et al. | The SOX11 transcription factor is a critical regulator of basal-like breast cancer growth, invasion, and basal-like gene expression | |
| Borisov et al. | Quantitation of molecular pathway activation using RNA sequencing data | |
| Jin et al. | Integrative multiomics analysis of human atherosclerosis reveals a serum response factor‐driven network associated with intraplaque hemorrhage | |
| Zheng et al. | Genetic characteristics and prognostic implications of m1A regulators in pancreatic cancer | |
| Kuosmanen et al. | The effects of sequence variation on genome-wide NRF2 binding—new target genes and regulatory SNPs | |
| Ching et al. | Pan-cancer analyses reveal long intergenic non-coding RNAs relevant to tumor diagnosis, subtyping and prognosis | |
| Wang et al. | Computational annotation of miRNA transcription start sites | |
| Zhan et al. | A five-gene signature predicts prognosis in patients with kidney renal clear cell carcinoma | |
| Li et al. | Toward predicting metastatic progression of melanoma based on gene expression data | |
| Hyeon et al. | Proteogenomic landscape of human pancreatic ductal adenocarcinoma in an Asian population reveals tumor cell-enriched and immune-rich subtypes | |
| Liu et al. | Molecular characterization clinical and immunotherapeutic characteristics of m5C regulator NOP2 across 33 cancer types | |
| Sircar et al. | Biphasic components of sarcomatoid clear cell renal cell carcinomas are molecularly similar to each other, but distinct from, non‐sarcomatoid renal carcinomas | |
| Haakensen et al. | Molecular characterisation of TP53 mutated squamous cell carcinomas of the lung to identify putative targets for therapy | |
| Koo et al. | Prediction of tumor purity from gene expression data using machine learning | |
| KR101829450B1 (en) | Evaluation of drug-targetable genes by defining modes of abnormality in gene expression | |
| Dadiani et al. | Tumor evolution inferred by patterns of microRNA expression through the course of disease, therapy, and recurrence in breast cancer | |
| US20120173160A1 (en) | Methods and systems for quantitatively assessing biological events using energy-paired scoring | |
| Lourdusamy et al. | Transcriptomic analysis in pediatric spinal ependymoma reveals distinct molecular signatures | |
| Fu et al. | The activity and expression of microRNAs in prostate cancers | |
| Xu et al. | m6A-related lncRNAs are potential biomarkers for the prognosis of COAD patients |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| A201 | Request for examination | ||
| E902 | Notification of reason for refusal | ||
| E701 | Decision to grant or registration of patent right | ||
| GRNT | Written decision to grant |