A predictive system for driving heterogeneous molecules important in tumor metastasis
The prediction system identifies and targets heterogeneous molecules driving tumor metastasis through quantitative protein analysis and sequential interventions, overcoming tumor heterogeneity to inhibit drug resistance and metastasis effectively.
Patent Information
- Application Number
- JP2024543426
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-11-30
- Filing Date
- 2022-03-15
- Publication Date
- 2024-11-01
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Current methods fail to accurately identify and intervene on heterogeneous molecules driving tumor drug resistance and metastasis, due to tumor heterogeneity, leading to ineffective treatments.
A prediction system comprising an input module, analysis module, and output module to determine key heterogeneous molecules through quantitative protein analysis before and after intervention, using high-throughput sequencing, protein interaction networks, and Hazard Ratio values to guide sequential interventions.
Accurately identifies and intervenes on heterogeneous molecules, effectively inhibiting tumor drug resistance and metastasis, prolonging patient survival by targeting independent molecules that current drugs miss.
Smart Images

Figure 2024540777000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a prediction system for driving important heterologous molecules in tumor metastasis, and belongs to the field of medical biology technology. [Background technology]
[0002] Malignant tumors are a major public health problem that seriously harms human health, and tumor drug resistance and metastasis are one of the fundamental reasons for the lack of treatment and low survival rate of tumor patients. Tumor therapy faces the biggest challenge of how to overcome tumor drug resistance and metastasis due to tumor heterogeneity. New therapies, including biological therapy and immunotherapy, have given new hope to tumor therapy, but due to tumor heterogeneity, it is difficult to defeat tumors by therapy against a single target in the end, and although there is a certain therapeutic effect in the early stage of treatment, the drug resistance or metastasis of tumor cells appears after a while, and the treatment is ineffective. Heterogeneous tumor cell subpopulations are independent of each other, and even if one of the oncogenes or tumor inhibitor genes can be favorably responded to, other heterogeneous molecules are not affected, and these heterogeneous tumor cell subpopulations adapt to the environment in the host body under appropriate conditions and play a leading role in driving the occurrence of tumor metastasis.
[0003] It is a consensus in the field of tumor research that tumor drug resistance and metastasis driven by tumor heterogeneity lead to tumor progression, but most researchers still seek out differentially expressed molecules and try to prove the precise upstream and downstream regulatory networks and mechanisms, but there is no method to accurately identify the tumor heterogeneity molecules of individual patients. There is an urgent need for technical means to accurately identify the tumor heterogeneity molecules of individual patients.
[0004] According to the search results, there is no similar technical solution in the prior art. In the prior art, the application for an invention patent with application number CN202010811273.X and disclosure number CN111763740A discloses a prediction system (including a system for detecting the expression quantities of three lncRNAs, SCAT1, PRKAG2-AS1 and FLG-AS1) that can predict the treatment effect and prognosis of new supportive chemoradiotherapy for patients with esophageal squamous cell carcinoma based on lncRNA molecular model, and can predict the treatment effect of new supportive chemoradiotherapy for patients with esophageal squamous cell carcinoma and the prognosis (prognosis overall survival rate, prognosis recurrence-free survival rate) of new supportive chemoradiotherapy for patients with esophageal squamous cell carcinoma, but there is no technical solution with the goal and technical effect of discovering and guiding tumor therapy intervention. Summary of the Invention [Problem to be solved by the invention]
[0005] The present invention overcomes the problems in the prior art and provides a prediction system for determining the important heterogeneous molecules driving tumor metastasis. Based on the results of quantitative analysis of proteins in metastasis-focused tumor tissues before and after tumor therapy, in combination with bioinformatics and computational biology tools, it is possible to select and order, and then determine a ranking table of tumor heterogeneous molecules, so as to effectively distinguish the heterogeneous molecules driving tumor metastasis in different patients, and provide accurate guidance for individualized measures for tumor metastasis treatment. [Means for solving the problem]
[0006] A prediction system for driving important heterologous molecules in tumor metastasis, comprising an input module, an analysis module and an output module. The input module is for inputting the results of quantitative analysis of proteins before and after the intervention, the results of quantitative analysis of proteins before the intervention are a collection of quantitative analysis results of the expression quantity of each protein in the tumor metastasis lesions of the subject patient before drug intervention, and the results of quantitative analysis of proteins after the intervention are a collection of quantitative analysis results of the expression quantity of each protein in the residual tumor metastasis lesions of the subject patient after drug intervention. The analysis module includes a primary analysis sub-module, a secondary analysis sub-module and a calculation and ordering sub-module. The primary analysis submodule is used for initial selection and analysis, which includes comparing the quantitative analysis results of proteins before and after the intervention and entering proteins whose expression levels are within a predetermined range into a candidate protein set. The secondary analysis submodule is used for reselection analysis, and the reselection analysis includes: using a protein interaction network analysis tool to construct a protein interaction network based on experimental verification by the candidate protein set, linking with the signal pathways related to the candidate protein set in the signal pathway database to organize the protein interaction network; putting the most important proteins in each protein interaction network into a collection of independent molecules as independent heterogeneous molecules; using a protein interaction network prediction tool to construct a protein interaction prediction network based on the protein interaction discipline by the candidate protein set; and listing the proteins in the candidate protein set that have not participated in any protein interaction prediction network as independent heterogeneous molecules in the collection of independent molecules. The calculation and ordering module is used for calculation and ordering, and the calculation and ordering includes calculating an HR value for each heterogeneous molecule in the collection of independent molecules, ordering each heterogeneous molecule according to its HR value, and forming an ordered list of each heterogeneous molecule. The output module is used to output an ordered list of each heterogeneous molecule.
[0007] Note: HR values are Hazard Ratio values.
[0008] If we can accurately identify novel heterogeneity targets that drive tumor metastasis and intervene in time until tumor cells fully adapt to current treatments, we can overcome the drug resistance and metastasis of heterogeneity-driven tumors, effectively control tumor progression, and prolong the lives of patients. Since the key to driving tumor metastasis or recurrence lies in heterogeneity molecules that are not controlled by current drugs but are closely related to tumor progression, precise intervention of heterogeneity molecules in tumor residual foci may solve the problems of drug resistance, metastasis, and recurrence of heterogeneity-driven tumors.
[0009] In the prediction system of the present invention, the relevant proteins in tumor metastasis foci before and after intervention are selected, and then further selection is performed by means of bioinformatics and computational biology, and the importance of the heterogeneous molecules that drive tumor metastasis is ranked, and the theoretical intervention order of the target molecules is determined. This prediction system can accurately identify and drive the heterogeneous molecules of tumor metastasis, and provide accurate intervention strategies to effectively inhibit the drug resistance and metastasis of tumors driven by heterogeneous molecules.
[0010] The prediction system of the present invention further has a complete technical solution as follows:
[0011] Preferably, quantitative analysis results on proteins before and after the intervention are obtained by high-throughput sequencing of proteins, which includes protein profile analysis.
[0012] This preferred solution allows optimizing the details of obtaining quantitative analytical results for proteins.
[0013] Preferably, in the initial selection and analysis, the expression level of a protein whose change range is within a preset range is an expression level after the intervention of the protein that is 0.95 to 1.05 times the expression level before the intervention of the protein.
[0014] Preferably, in the initial selection and analysis, a stepwise selection is performed for proteins whose expression level changes are within a preset range, and then the proteins are added to the candidate protein set. The stepwise selection is then performed to select proteins that satisfy the following conditions: the number of specific peptides is ≥ 2, the protein is not a predicted protein, the coverage of the identified peptide amino acid in the current protein is ≥ 10%, the number of times of the current protein quantification is ≥ 2, the total score of the sequence quality of the current protein is ≥ 200 points, and there is evidence in the database that proves that the protein is involved in tumor progression. The specific peptide refers to a specific peptide fragment for identifying the current protein, the predicted protein refers to a predicted protein whose reliability properties are defined in the Uniport or NCBI database, the identified peptide fragment refers to a specific peptide fragment used in an indication and detected by the current protein, and the number of times of protein quantification refers to the number of times that the current protein is quantified.
[0015] The above preferred solution allows further optimization of the initial selection and analysis details of the primary analysis submodule.
[0016] Preferably, in the reselection analysis, the protein interaction network analysis tool includes the STRING network tool, the signal pathway database includes KEGG, SMPDB, WikiPathways, Cell Signaling Technology, BioCarta Pathway, Pathway Commons and PID, and the protein interaction network prediction tool includes the Genemania network tool, the Unihi network tool and the Hitpredict network tool.
[0017] More preferably, a protein interaction network is constructed and organized using a Cytoscape tool, and a predicted protein interaction network is constructed.
[0018] Preferably, in the reselection analysis, the most important proteins in each protein interaction network are identified by quality characterization, the quality characterization including an analysis of the importance and radial extent of each protein in the protein interaction network, wherein the protein with the highest importance and radial extent is the most important protein in the protein interaction network.
[0019] The above preferred solution further allows optimizing the details of the reselection analysis of the secondary analysis submodule.
[0020] Preferably, in the calculation and ordering, the HR value of the heterogeneous molecule is calculated as follows: obtain sequence data on the tumor from a public database, obtain a normalized count matrix of the gene corresponding to the heterogeneous molecule from the sequence data, perform a batch single-factor COX regression on the gene, and calculate the HR value.
[0021] More preferably, said public databases include TCGA, ICGC and GEO.
[0022] Preferably, in the calculation and ordering, the ordering rule is that heterogeneous molecules with HR values ≧1 are directly ordered in reverse order according to their HR values, and heterogeneous molecules with HR values <1 are ordered in reverse order according to 1 / HR value.
[0023] The above preferred solution further allows optimizing the computation and ordering details of the computation and ordering module.
[0024] The sequential intervention system for the execution of tumor-driving metastasis-driving heterogeneous molecules includes the prediction system for identifying important tumor-driving metastasis-driving heterogeneous molecules and an execution module. The execution module is for determining the sequential intervention sequence according to the ordering list of each heterogeneous molecule outputted from the prediction system and the predetermined rule. The predetermined rule is a selection and ordering rule determined by a user according to the actual situation (e.g., when the heterogeneous molecule is unavailable due to the absence of an inhibitor or agonist or the high drug price), so that the sequential intervention sequence can be more practical. Effect of the Invention
[0025] Compared with the conventional analytical methods for identifying tumor heterogeneity molecules, the present invention has the following advantages: (1) Uniqueness: Compared with other identification and analysis methods, the present invention focuses on the target of mutation, alteration and abnormal expression, to identify and select heterogeneous molecules whose protein expression levels are stable in the foci of tumor metastasis before and after drug intervention, but which have the potential to drive tumor metastasis. (2) Practicality: Verification and research on sequential interventions in gastric cancer and melanoma were conducted based on the heterogeneous molecules that drive and metastasize tumors predicted and selected by the present invention, and the expected effects were obtained. Therefore, the strategy of selecting heterogeneous molecules and sequential interventions determined by the present invention is accurate and feasible. (3) Efficacy: According to the results of the verification study of the present invention on gastric cancer and melanoma heterogeneous molecules, the outcome of the model rats was significantly improved and the survival time was extended with each additional intervening heterogeneous molecule. Therefore, the heterogeneous molecules determined and predicted by the method of the present invention can indeed independently drive and metastasize gastric cancer or melanoma. By carrying out sequential intervention according to the order of the heterogeneous molecules determined by the method of the present invention, tumor drug resistance and recurrence / metastasis caused by heterogeneity can be effectively overcome.
[0026] In summary, the present invention can more accurately and precisely identify and sequence the key heterogeneous molecules that drive tumor metastasis, and then carry out sequential interventions according to the results according to actual practice, and determine the therapeutic effect. The discoveries and evidence from these studies provide innovative ideas, specific algorithms and practical guidance for the clinical treatment of tumor heterogeneity-induced drug resistance and metastasis. [Brief description of the drawings]
[0027] [Figure 1] The relationship between tumor inhibitor gene JWA and prognosis and metastasis characteristics of gastric cancer in embodiment 1 of the present invention 1.1 [Diagram 2] The relationship between tumor inhibitor gene JWA and prognosis and metastasis characteristics of gastric cancer in embodiment 1 of the present invention 1.1 [Diagram 3] The relationship between tumor inhibitor gene JWA and prognosis and metastasis characteristics of gastric cancer in embodiment 1 of the present invention 1.1 [Figure 4] The relationship between tumor inhibitor gene JWA and prognosis and metastasis characteristics of gastric cancer in embodiment 1 of the present invention 1.1 [Diagram 5] The relationship between the oncogene MDM2 and the prognosis and metastatic characteristics of gastric cancer in Example 1, paragraph 1.2 of the present invention [Figure 6] The relationship between the oncogene MDM2 and the prognosis and metastatic characteristics of gastric cancer in Example 1, paragraph 1.2 of the present invention [Figure 7] The relationship between the oncogene MDM2 and the prognosis and metastatic characteristics of gastric cancer in Example 1, paragraph 1.2 of the present invention [Figure 8] The relationship between the oncogene MDM2 and the prognosis and metastatic characteristics of gastric cancer in Example 1, paragraph 1.2 of the present invention [Figure 9] Model design and results demonstrating that JWA and MDM2 are functionally independent heterogeneous molecules and can cooperate to inhibit the in vivo and in vivo metastasis of gastric cancer in Example 1, paragraph 1.3 of the present invention [Figure 10] Model design and results demonstrating that JWA and MDM2 are functionally independent heterogeneous molecules and can cooperate to inhibit the in vivo and in vivo metastasis of gastric cancer in Example 1, paragraph 1.3 of the present invention [Figure 11] Model design and results demonstrating that JWA and MDM2 are functionally independent heterogeneous molecules and can cooperate to inhibit the in vivo and in vivo metastasis of gastric cancer in Example 1, paragraph 1.3 of the present invention [Figure 12] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 13] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 14] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 15] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 16] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 17] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 18] Model design and results of identifying other heterogeneous molecules that drive and metastasize gastric cancer after joint intervention of JWA and MDM2 in Example 1, 1.3 of the present invention [Figure 19] In Example 1, 1.5 of the present invention, the model design and results show that JWA, MDM2 and FASN can be sequentially intervened to effectively inhibit metastasis of cancer cells in the body. [Figure 20] In Example 1, 1.5 of the present invention, the model design and results show that JWA, MDM2 and FASN can be sequentially intervened to effectively inhibit metastasis of cancer cells in the body. [Figure 21]In Example 1, 1.5 of the present invention, the model design and results show that JWA, MDM2 and FASN can be sequentially intervened to effectively inhibit metastasis of cancer cells in the body. [Figure 22] In Example 1, 1.6 of the present invention, the model design and results of how to identify other heterogeneous molecules that further drive and metastasize gastric cancer after sequential intervention in JWA, MDM2 and FASN [Figure 23] In Example 1, 1.6 of the present invention, the model design and results of how to identify other heterogeneous molecules that further drive and metastasize gastric cancer after sequential intervention in JWA, MDM2 and FASN [Figure 24] In Example 1, 1.7 of the present invention, JWA, MDM2, FASN, USP9X and PFN1 are sequentially intervened to effectively inhibit the metastasis of gastric cancer in the body, and the time of recurrence and metastasis of gastric cancer can be significantly delayed with each increase in the intervening heterogeneous molecule. [Diagram 25] In Example 1, 1.7 of the present invention, JWA, MDM2, FASN, USP9X and PFN1 are sequentially intervened to effectively inhibit the metastasis of gastric cancer in the body, and the time of recurrence and metastasis of gastric cancer can be significantly delayed with each increase in the intervening heterogeneous molecule. [Figure 26] In Example 1, 1.7 of the present invention, JWA, MDM2, FASN, USP9X and PFN1 are sequentially intervened to effectively inhibit the metastasis of gastric cancer in the body, and the time of recurrence and metastasis of gastric cancer can be significantly delayed with each increase in the intervening heterogeneous molecule. [Figure 27] Model design and results of JWA mimicking peptide JP1 inhibiting melanoma metastasis in vivo in Example 2-2.1 of the present invention [Figure 28] Model design and results of JWA mimicking peptide JP1 inhibiting melanoma metastasis in vivo in Example 2-2.1 of the present invention [Figure 29] Model design and results of MDM2 inhibitor SP141 inhibiting melanoma metastasis in the present invention in Example 2-2.2 [Diagram 30]Model design and results demonstrating in Examples 2-2.3 of the present invention that JWA and MDM2 are functionally independent heterogeneous molecules that can cooperate to inhibit melanoma metastasis in vivo [Diagram 31] Model design and results demonstrating in Examples 2-2.3 of the present invention that JWA and MDM2 are functionally independent heterogeneous molecules that can cooperate to inhibit melanoma metastasis in vivo [Diagram 32] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Diagram 33] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Diagram 34] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Diagram 35] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Diagram 36] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 37] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 38A] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 38B]Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 38C] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 38D] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Figure 39] Model design and results on how to identify the heterogeneous molecules that drive and metastasize melanoma after joint intervention of JWA and MDM2 in the present invention example 2-2.3 [Diagram 40] In Examples 2-2.5 of the present invention, model design and results showing that JWA, MDM2, SMAD3, IDH1 / 2 and mTOR can be sequentially intervened to effectively inhibit melanoma metastasis in vivo [Diagram 41] In Examples 2-2.5 of the present invention, model design and results showing that JWA, MDM2, SMAD3, IDH1 / 2 and mTOR can be sequentially intervened to effectively inhibit melanoma metastasis in vivo [Figure 42A] English abbreviations and Japanese names of the main genes related to each example of the present invention [Figure 42B] English abbreviations and Japanese names of the main genes related to each example of the present invention DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0028] The present invention provides a prediction system for identifying important heterologous molecules that drive tumor metastasis, comprising an input module, an analysis module and an output module.
[0029] The input module is for inputting the results of quantitative analysis of proteins before and after the intervention, the results of quantitative analysis of proteins before the intervention are a collection of the results of quantitative analysis of the expression quantity of each protein in the tumor metastasis lesions of the subject patient before the drug intervention, and the results of quantitative analysis of proteins after the intervention are a collection of the results of quantitative analysis of the expression quantity of each protein in the residual tumor metastasis lesions of the subject patient after the drug intervention.
[0030] Among them, high-throughput protein sequencing is used to obtain quantitative analysis results of proteins before and after intervention, and high-throughput protein sequencing includes protein profile analysis.
[0031] The analysis module includes a primary analysis sub-module, a secondary analysis sub-module and a calculation and ordering sub-module.
[0032] The primary analysis submodule is used for initial selection and analysis, which involves comparing the quantitative analysis results of proteins before and after the intervention and selecting proteins whose expression levels change within a pre-defined range as a candidate protein set.
[0033] Among them, the protein whose expression amount is within a preset range is 0.95-1.05 times the expression amount before and after the protein intervention.
[0034] Proteins with expression level changes within a predefined range are selected stepwise and then added to the candidate protein set. The stepwise selection is performed according to the following criteria: specific peptide level ≥ 2, not predicted protein, coverage of identified peptide amino acids in the current protein ≥ 10%, current protein quantification count ≥ 2, total score of sequence quality of the current protein ≥ 200, and evidence in the database that proves involvement in tumor progression. Specific peptide refers to a specific peptide fragment for identifying the current protein, predicted protein refers to the current protein reliability property defined as predicted protein in Uniport or NCBI database, identified peptide fragment refers to a specific peptide fragment used for indications and detected in the current protein, and protein quantification count refers to the overlapping number of quantification detections of the current protein.
[0035] The secondary analysis submodule is used for reselection analysis, in which the reselection analysis uses a protein interaction network analysis tool to construct a protein interaction network based on experimental verification by the candidate protein set, and combines it with the signal pathways related to the candidate protein set in the signal pathway database to improve the protein interaction network, and puts the most important proteins in each protein interaction network into a collection of independent molecules as independent heterogeneous molecules, and uses a protein interaction network prediction tool to construct a protein interaction prediction network based on the protein interaction discipline by the candidate protein set, and puts the proteins in the collection of candidate proteins that have not participated in any protein interaction prediction network into a collection of independent molecules as independent heterogeneous molecules.
[0036] Among them, the protein interaction network analysis tools include STRING network tools, the signal pathway databases include KEGG, SMPDB, WikiPathways, Cell Signaling Technology, BioCarta Pathway, Pathway Commons and PID, the protein interaction network prediction tools include Genemania network tools, Unihi network tools and Hitpredict network tools, and the Cytoscape tool is used to construct and organize the protein interaction network, and to construct the protein interaction prediction network.
[0037] The most important proteins in each protein interaction network can be identified by characterization, which includes analysis of the importance and radial extent of each protein in the protein interaction network, among which the protein with the highest importance and radial extent is the most important protein in this protein interaction network.
[0038] The calculation and ordering module is used for calculation and ordering, and the calculation and ordering includes calculating the HR value of each heterogeneous molecule in a set of independent molecules, ordering each heterogeneous molecule according to the HR value, and forming an ordered list of each heterogeneous molecule (Note: the HR value is the Hazard Ratio value of risk ratio).
[0039] Wherein, the HR value of the heterogeneous molecule is calculated as follows: obtain the sequence data of the tumor from the public database, obtain the normalized count matrix of the gene corresponding to the heterogeneous molecule from the sequence data, perform batch single factor COX regression for this gene, and calculate the HR value. The public database includes TCGA, ICGC, and GEO.
[0040] As a rule for ordering, heterogeneous molecules with HR values ≥ 1 are ordered in reverse order directly according to their HR values, while heterogeneous molecules with HR values < 1 are ordered in the reverse order of 1 / HR value.
[0041] An output module is used to output an ordered list of each heterogeneous molecule.
[0042] Note: All of the above modules operate based on computer readable media.
[0043] In addition, the sequential intervention system for heterogeneous molecules that drive tumors and cause metastasis, which is specifically implemented in the present invention, includes a prediction system that identifies important heterogeneous molecules that drive tumors and cause metastasis, and an execution module, and the execution module determines the order in which to perform sequential interventions according to the ordering list of each heterogeneous molecule output to the prediction system.
[0044] In order to verify the effectiveness of each of the above systems, verification is carried out as follows.
[0045] A sequential intervention is carried out on the predicted heterogeneous molecules, that is, a verification method is verified that intervening in the heterogeneous molecules effectively inhibits tumor metastasis and prolongs survival time. This method (taken as an example) is as follows. 1) Based on the principle that tumor heterogeneous molecules do not regulate each other, joint intervention is carried out with two of the molecules (e.g., tumor inhibitor gene JWA and oncogene MDM2) as first-round intervention targets. 2) Extract residual tumor tissue from metastatic lesions following joint intervention of JWA and MDM2 and perform high-throughput protein sequencing. 3) According to the results of high-throughput protein sequencing, the heterogeneous molecules that drive and metastasize tumors are selected by the above prediction system, and then ranked by HR value to become candidate targets for sequential intervention. 4) Develop and implement a sequential intervention plan to inhibit tumor metastasis using candidate targets. 5) After intervention according to the implementation plan in 4), extract the residual foci of tumor metastasis and perform high-throughput protein sequencing. Again, select the heterogeneous molecules that drive tumor metastasis using the above prediction system, and formulate and implement the next round of sequential intervention plans to inhibit tumor metastasis. 6) Repeat step 5) and perform several rounds of sequential intervention treatments until tumor metastasis is completely and effectively inhibited.
[0046] The present invention will now be described in more detail with reference to the accompanying drawings in conjunction with the examples, however the present invention is not limited to the examples given. Unless otherwise specified, ordinary experimental procedures are used.
[0047] The present examples mainly describe how to identify the algorithm of heterogeneous molecular systems that drive tumor metastasis, and the method is suitable for various organisms and tumor types, and is preferably applied to humans, but also to other mammals, and the applicable tumor types include, but are not limited to, gastric cancer and melanoma.
[0048] In the present embodiment, the sequencing method for identifying heterogeneous molecules that drive tumor metastasis is high-throughput protein sequencing. The application of the results of the present invention is based on, but not limited to, the results of high-throughput protein sequencing.
[0049] In an embodiment of the present invention, sequential interventions for gastric cancer and melanoma include five independent heterogeneous molecular targets, and the accelerated application of the results of the present invention includes, but is not limited to, the five target molecules listed in the present invention. EXAMPLES
[0050] This example demonstrates how to identify heterogeneous molecules that drive and metastasize gastric cancer and perform sequential interventions.
[0051] The main purpose of this example is to predict and identify the heterogeneous molecular strategies that drive tumor metastasis using the present invention. Taking gastric cancer as an example, JWA and MDM2 are jointly intervened to identify the heterogeneous molecules that drive gastric cancer metastasis, and then sequential intervention strategies are implemented to verify the effectiveness of the strategies of the present invention.
[0052] 1.1 Relationship between the tumor-inhibitory gene JWA and prognosis and metastatic characteristics of gastric cancer The relationship between JWA and the prognosis of gastric cancer was analyzed using the TCGA database. As shown in Figure 1, JWA mRNA levels were lower in paired and non-paired cancer tissues than in paracancerous tissues (Figure 1-A), and patients with low JWA expression had a poor prognosis (Figure 1-B).
[0053] si-JWA was transfected into gastric cancer BGC823 and SGC7901 cells to interfere with the expression levels of JWA mRNA and its protein. The results of the perforation experiment showed that gastric cancer cells with low expression of JWA were able to pass more tumor cells through the chamber in the same time (Fig. 2-A, B), and the results of the scratch experiment showed that gastric cancer cells with low expression of JWA were able to heal the scratch faster (Fig. 2-C, D). Therefore, as shown in Fig. 2, lowering the expression level of JWA can enhance the migration ability of gastric cancer cells.
[0054] Similarly, gastric cancer BGC823 and MFC cells were transfected with Flag-JWA to increase the expression level of JWA protein. The results of the perforation experiment showed that gastric cancer cells with high expression of JWA had fewer tumor cells passing through the chamber in the same time (Figure 3-A, B), and the results of the scratch experiment showed that gastric cancer cells with high expression of JWA had a significantly slower post-healing rate (Figure 3-C, D). Therefore, as shown in Figure 3, increasing the expression level of JWA can weaken the migration ability of gastric cancer cells.
[0055] A passive transfer model was established by injecting BGC823 cells into the tail vein of nude rats. Interventions were performed by intraperitoneal injection of polypeptides in groups on the second day after cell inoculation (JP3 or Ctrl-PH). As shown in Figure 4, compared with the control peptide Ctrl-PH intervention group, rats in the JP3 intervention group had significantly smaller metastatic lesions in the lungs (Note: For JP3 peptide and Ctrl-PH peptide, see Chen J et al. J Exp Clin Cancer Res. 2020, 39:118).
[0056] Therefore, the mimetic peptide JP3 based on the tumor-inhibiting gene JWA can effectively inhibit the metastasis of gastric cancer in vivo and in vivo.
[0057] 1.2 Relationship between the oncogene MDM2 and gastric cancer prognosis and metastatic characteristics The relationship between MDM2 and the prognosis of gastric cancer was analyzed using the TCGA database. As shown in Figure 5, when comparing the MDM2 mRNA levels of paired and unpaired cancer tissues and paracancerous tissues, the cancer tissues had higher levels than the paracancerous tissues (Figure 5-A), and patients with low MDM2 expression had a better prognosis than patients with high MDM2 expression (Figure 5-B).
[0058] As shown in Figure 6, the half inhibitory concentration (IC50) of MDM2 inhibitor SP141 (a product of conventional technology) against gastric cancer cells MGC803, AGS, MFC, SGC7901 and BGC823 was detected by CCK-8 method and was 0.616μM, 0.335μM, 0.345μM, 0.687μM and 0.599μM, respectively (Figure 6-A). According to the results of the perforation experiment, when gastric cancer cells BGC823, SGC7901 and MFC were treated with different doses of SP141 for 24 hours, the number of perforated cells gradually decreased with the increase in the dose of SP141, showing a dose-effect relationship (Figure 6-B, C).
[0059] Similar results were obtained from cell scratch experiments, in which the rate of cell migration was significantly slowed after treatment of gastric cancer cells BGC823 and SGC790MFC with SP141 (see Figure 7).
[0060] A peritoneal seeding metastasis model was established using mouse-derived gastric cancer cell line MFC. Different doses of SP141 were intraperitoneally administered from the second day of cell inoculation as an intervention. After the model experiment was completed, the number of peritoneal metastasis nodules was recorded. As shown in Figure 8, SP141 could inhibit the number of gastric cancer peritoneal seeding metastasis, and there was a dose-effect relationship. Compared with the control group, the difference in the 40 mg / kg group was statistically significant. The residual tumor tissues from the control group and 40 mg / kg group were subjected to total protein immunoblot analysis, and the results showed that the level of MDM2 protein was significantly reduced in the SP141 intervention group.
[0061] Therefore, lowering the expression of the oncogene MDM2 can effectively inhibit the metastasis of gastric cancer both in vivo and in vivo.
[0062] 1.3 JWA and MDM2 are functionally independent heterogeneous molecules that can cooperate to inhibit gastric cancer metastasis. The correlation between JWA and MDM2 mRNA was analyzed in 413 gastric cancer tissue samples using the TCGA database, and the results showed that there was no correlation between the two (Figure 9-A). Furthermore, the mutual regulation between JWA and MDM2 protein was examined using a cell model. BGC823 cells were transfected with different doses of Flag-JWA plasmid to gradually increase the level of JWA protein, and there was no significant change in MDM2 protein (Figure 9-B). BGC823 gastric cancer cells were treated with 0, 0.25, and 0.5 μM SP141 for 24 hours to inhibit the expression level of MDM2. As a result, the expression level of MDM2 protein decreased with increasing dose of SP141, but there was no significant change in the expression level of JWA protein (Figure 9-C). Therefore, as shown in Figure 9, JWA and MDM2 have no mutual regulatory relationship at the protein level or RNA level in gastric cancer, and their biological functions on gastric cancer cells are independent.
[0063] According to the results of the cell perforation experiment, AGS cells were transfected with Flag-JWA plasmid and treated with SP141. Compared with the control group, the number of cell perforations was lower in both the single and combined treatment groups, and the combined treatment group had the lowest cell number, and the difference was statistically significant. Therefore, as shown in Figure 10, the combined intervention of JWA and MDM2 can cooperate to inhibit the metastasis of gastric cancer cells.
[0064] In the in vivo model, a passive metastasis model of gastric cancer was established by injecting fluorescently labeled gastric cancer cells BGC823-luc into the tail vein of nude rats, and drug intervention was performed from the second day after cell inoculation (Figure 11-A). According to the results of weight observation, in the early stage of the model, the rats' weight steadily increased, and from the 33rd day, the rats in the control group began to lose weight, while the rats in the drug intervention group had stable weight and were in good condition (Figure 11-B). According to the in vivo fluorescence photographs, the saline group and the solvent control group had similar fluorescence intensities in the metastatic lesions in the lungs, the SP141 and JP3 single intervention group had weaker fluorescence intensities than the control group, and the difference was statistically significant, and the joint intervention group had the weakest fluorescence in the lungs, and the difference from the fluorescence intensity of the single intervention group was statistically significant (Figure 11-C, D). Therefore, as shown in Figure 11, joint intervention with JWA and MDM2 can cooperate to inhibit the in vivo metastasis of gastric cancer.
[0065] Therefore, joint intervention of JWA and MDM2, the heterologous molecules that drive and metastasize gastric cancer, can cooperate to inhibit the in vivo metastasis of gastric cancer.
[0066] 1.4 Identify other heterogeneous molecules that jointly interact with JWA and MDM2 to drive and metastasize gastric cancer. In this section, a theory for identifying heterogeneous molecules that drive tumor metastasis is proposed, and verified by performing sequential intervention. Tumor heterogeneous molecules are partially independent in function and can drive a variety of malignant phenotypes, including metastasis. Even if genes (whether oncogenes or tumor inhibitor genes) can be favorably responded to, the expression and function of other heterogeneous molecules are not affected. Thus, under appropriate conditions, these heterogeneous molecules can adapt and play a leading role in the fate of tumor cells, driving tumors to recur and metastasize. Here, the inventor replaces the idea of searching for and discovering differentially expressed molecules in conventional studies and performing targeted interventions by searching for heterogeneous molecules whose expression was not affected after targeted drug intervention, but which drive tumors to metastasize. It is assumed that sequential intervention of these molecules can effectively prolong the survival time of tumor patients (see Figure 12).
[0067] As shown in 1.3, the joint intervention of JWA and MDM2 can significantly inhibit the metastasis of gastric cancer. However, even after the joint intervention, the lung metastasis foci in the rat body still adapt to the microenvironment and develop resistance and metastasis after a while. This indicates that there are heterogeneous molecules that are not regulated by JWA and MDM2 after the intervention of JP3 and SP141. In order to predict and identify the future progress of these unknown heterogeneous molecules in driving gastric cancer, protein profile analysis is performed on the metastatic foci that still exist after the joint intervention of JP3 and SP141, and protein molecules with more stable expression levels than the control group are selected from the quantitative analysis of a total of more than 7,000 proteins (specific indicator: the change in expression level of the target molecule in the metastatic residual foci and the control metastatic residual foci after the joint intervention of JP3 and SP141 is in the range of 0.95 to 1.05). Furthermore, candidate heterogeneous molecules that roughly meet the following conditions are selected based on the sequence-related quality control data and the association between the molecule and the tumor. (1) A protein molecule having a specific peptide quantity of ≥ 2 (2) removing predicted protein molecules; (3) Molecules whose identified peptide fragments have a coverage of less than 10% of the total protein are removed. (4) Remove molecules with protein (129C / 129N polypeptide) quantification less than twice. (5) Molecules with a total protein sequence score of 200 or less are removed. (6) According to the keywords, i.e., the proteins and their corresponding gene names and cancer names, search relevant databases (e.g., PubMed, Web of science, cnki.net, Wanfang Database, VIP.com, etc.) to determine their correlation with tumor progression, and eliminate irrelevant molecules.
[0068] After the above selection, there were 2066 protein molecules with a quantity of at least 2 that met the specific peptide (referring to a specific peptide fragment for identifying the current protein), 1581 were left after performing protein elimination prediction (the reliability properties of the current protein are defined for the predicted protein in the Uniport or NCBI database), 1187 molecules with a coverage of 10% or more of the amino acid identified by peptide fragments (used for indications and referring to a specific peptide fragment detected in the current protein) in the whole protein, 1169 molecules with 129C / 129N (ion ratio in the mass spectrometry report) polypeptide quantification times (referring to the overlapping number of times the current protein is quantified) more than twice, 117 molecules with a total score of 200 points or more were left, and database analysis proved that there were 38 protein molecules involved in tumor progression (see Figure 13). Figure 14 and Figure 42 show the information of each candidate heterogeneity molecule relationship.
[0069] There may be completely independent single heterogeneous molecules and mutually regulated heterogeneous molecules in the initially selected heterogeneous molecules. In order to select independent molecules in the heterogeneous molecules, a network of protein interaction relationships known so far was constructed on the STRING website. Among them, the criteria for determining protein interactions were "experimental verification" and "known in the database". Then, 38 of the above heterogeneous molecules were input, and the information on the signal pathways involved in obtaining the heterogeneous molecules was analyzed based on KEGG, SMPDB, WikiPathways, Cell Signaling Technology, BioCarta Pathway, Pathway Commons and PID databases to find out whether there are protein molecules in the same signal pathway, and new protein points and protein interaction relationships were added using Cytoscape software, and further modifications were made (Figure 15). Finally, the network properties (connectivity, clustering coefficient, tightness and topological coefficient) were calculated by script based on the molecular point relationships in this interaction network, and the importance and radial degree of the network were analyzed using the network analyzer module in Cytoscape software (Figures 16-18). Based on the heterogeneous molecules with the highest selection importance and radial degree in the independent interaction network, the molecules that can form mutually independent networks (the number of proteins included in the network is 2 or more) and the most important heterogeneous molecules in this network (HSP90AB1, ENO1 / PKM, HNRNPK / PCBP2 and RAB1A / SPTBN1) were determined. For the heterogeneous molecules determined to form a regulatory network, the HR cumulative value of each molecule in this network was calculated, and this HR value was assigned to the representative heterogeneous molecule of this network (Figures 15-18).Meanwhile, using the Genemania website, Unihi website and Hitpredict website, we predicted algorithmic thinking about protein interactions based on text mining, co-expression, domain and gene fusion, and discovered 38 independent heterogeneous molecules (FASN, SLC3A2, ANXA4, VIM, USP9X and AGRN) that do not form an interaction network with other molecules. We then calculated the cumulative HR values of these molecules in the same way. Then, we compiled the HR values for overall survival corresponding to the mutually independent heterogeneous molecules and ordered the intervention order (ordering above 1 was reversed by HR value, and HR values below 1 were reversed by 1 / HR value). After deciding to jointly intervene in JWA and MDM2, the HR order of the independent heterogeneous molecules that drive and metastasize gastric cancer is as follows: FASN(2.21(1.82-2.86), p=1.6e-16), HNRNPK(0.46(0.37-0.56), p=3.9e-14), AGRN(2.07(1.75- 2.46), p<1e-16), HSP90AB1(1.73(1.46-2.05), p=1.7e-10), RAB1A(0.64(0.53-0.76), p=5.2e-0 7), ANXA4 (0.65 (0.55-0.78), p=1.4e-06), SLC3A2 (1.47 (1.27-1.74), p=1.1e-05), ENO1 (1.32 (1.11-1.56), p=0.0014), VIM (1.3 (1.09-1.53), p=0.0027) and USP9X (1.25 (0.99-1.59), p=0.061). Therefore, whether it is a regulatory related heterogeneous molecule or a completely independent single heterogeneous molecule, the corresponding HR value was obtained through the above analysis, calculation and assignment. The final intervention order of all heterogeneous molecules is determined by the HR value. Figure 15 shows the selection of some gastric cancer heterogeneous molecules and the ranking of HR values.
[0070] 1.5 Sequential intervention of JWA, MDM2 and FASN can effectively inhibit the metastasis of gastric cancer in the body. Based on the results of 1.4, FASN was made the next intervention target.
[0071] The relationship between FASN and the prognosis of gastric cancer was analyzed using the TCGA database. As shown in Figure 19, the FASN mRNA levels in both paired and unpaired cancer tissues and paracancerous tissues were higher than those in the paracancerous tissues, and the difference was statistically significant (Figure 19-A). Patients with low FASN expression had a better prognosis than those with high FASN expression (Figure 19-B).
[0072] The results of FASN protein level detection in the residual metastatic lesions of rat lung tissues in the vehicle control group and the JWA+MDM2 joint intervention group in the animal model showed that FASN expression was stable in the two groups, which was consistent with the protein profile results (Figure 20-A). Furthermore, the relationship between FASN and gastric cancer metastasis was verified by perforation and scratch experiments. BGC823 cells and BGC823-LPC1 (animal model SP141 and JP3 joint intervention of primary cells in the residual metastatic lesions of the lungs in the group) were treated with different concentrations of the FASN inhibitor EGCG for 24 hours and then perforation experiments were performed. The results showed that the number of perforated cells in the EGCG treatment group was significantly reduced, and there was a relationship between the dose and the effect, and the inhibition rate of perforated cells was similar in the two types of cells (Figure 20-B). Similar results were obtained from the scratch experiment (Figure 20-C). The immunoblotting experiment results (Figure 20-D) showed that after cells were treated with EGCG, the expression level of FASN protein decreased, and the metastasis-related protein FAK also decreased accordingly, with a dose-effect relationship. Therefore, as shown in Figure 20, intervention in FASN can inhibit the metastasis of gastric cancer cells.
[0073] A passive transfer model was constructed by extracting primary cells BGC823-LPC1 from the residual lung metastasis foci after joint intervention with JWA and MDM2. The model rats were divided into groups on the second day after the cells were injected into the tail vein, and then the drugs were administered. The rats were weighed regularly during the administration period, and the lung water content of each group was measured after the model experiment was completed, and the status of the lung metastasis foci was observed. According to the results (see Figure 21), the rats in the group were then jointly intervened with SP141 and JP3, and the lung water content was not significantly reduced compared to the solvent control group, and the number of metastasis foci was not significantly reduced, but the lung water content of the rats in the FASN inhibitor EGCG group and the FASN inhibitor Orlistat group was significantly reduced and the number of metastasis foci was significantly reduced, and the difference was statistically significant compared to the control group.
[0074] Therefore, the primary gastric cancer cells BGC823-LPC1 extracted by treating the lung metastatic lesions of the group jointly with SP141 and JP3 were inoculated into nude rats for modeling, and then SP141 and JP3 were used to inhibit the lung metastasis of gastric cancer cells. Since the effect of SP141 and JP3 intervention was very small, the rats were subjected to the first round of SP141+JP3 intervention and showed drug resistance in the second round, but the novel heterogeneous molecule FASN was intervened with the inhibitor EGCG or Orlistat (both of which are products of conventional technology) to significantly inhibit the lung metastasis of gastric cancer cells. Therefore, intervention according to the algorithm results (i.e., the results of 1.4) of identifying the heterogeneous molecules that drive tumor metastasis proposed by the present invention is effective.
[0075] 1.6 We selected heterogeneous molecules that sequentially intervene in JWA, MDM2 and FASN and further drive gastric cancer metastasis. Based on the above model, primary cells were extracted from the residual metastatic foci of lung gastric cancer cells in the EGCG treatment group of rats and named "BGC823-LPC2". First, protein profile analysis was performed on the residual metastatic foci in the lungs of the EGCG group and the SP141 and JP3 combined treatment group, mainly to observe the changes in protein expression of each independent molecule determined in 1.4 above. The results showed that FASN in the EGCG group was inhibited by 22%, and the remaining protein inhibition was below 20%. As shown in Figure 22, the heterogeneous molecules that drive tumor metastasis are functionally independent of each other, and even if one of the heterogeneous target molecules is intervened, the expression levels of other heterogeneous molecules remain stable.
[0076] According to the strategy of identifying the heterogeneous molecules that drive and progress gastric cancer described in 1.4 above and the results of this protein profile analysis, and considering the accessibility of comprehensive intervention inhibitors (many heterogeneous molecules can be selected and intervention can be performed on individual molecules in theory, but in reality, molecules must be selected from the candidate heterogeneous molecules according to the actual situation when there are no inhibitors or agonists or cannot be obtained due to high drug prices, etc.), the oncogenes USP9X and PFN1 were selected as the next intervention targets from the selected heterogeneous molecules. The relationship between USP9X and PFN1 and the prognosis of gastric cancer groups was analyzed using the TCGA database, and as shown in Figure 23, patients with low expression of USP9X or PFN1 have a better prognosis than those with high expression.
[0077] 1.7 Sequential intervention of JWA, MDM2, FASN, USP9X and PFN1 effectively inhibits the metastasis of gastric cancer in the body. To verify the relationship between USP9X and gastric cancer metastasis, BGC823-LPC2 cells cultured with USP9X inhibitor WP-1130 (a product of conventional technology) were treated for 8 hours and then perforation experiments were performed. The results (see Figure 24) showed that WP-1130 could significantly inhibit the number of cell perforations, and there was a relationship between the dose and the effect (Figure 24-A). Immunoblotting results showed that the expression of USP9X protein in cells treated with WP-1130 was decreased, and the expression of the metastasis-related protein p-AKT was also decreased accordingly, and there was a relationship between the dose and the effect (Figure 24-B). The results of the scratch experiment and the small chamber perforation experiment were consistent (Figure 24-C). Therefore, inhibiting USP9X expression can effectively reduce the metastasis of gastric cancer cells.
[0078] In the case of the heterologous molecule PFN1, since there is no commercially available inhibitor compound, we constructed a PFN1 low-expression lentiviral plasmid sh-PFN1, transfected sh-con and sh-PFN1 into BGC823-LPC2, and performed a perforation experiment 48 hours after transfection. The results (see Figure 25) showed that the number of gastric cancer cell perforations in low-expression PFN1 was significantly reduced. Therefore, by intervening in PFN1 expression, it is possible to effectively inhibit the metastasis of gastric cancer cells.
[0079] After obtaining positive results in the cell model, we conducted an animal model validation. A passive transfer model was established by intravenous injection of BGC823-LPC2 cells into the tail of nude rats. On the second day, the rats were randomly divided into groups and single interventions were performed with the control solvent, EGCG, and WP-1130, and sequential interventions were performed with WP-1130 and sh-PFN1 lentivirus. After the model experiment was completed, the rats were observed for the transfer status in the lungs (Figure 26-A). According to the results of weight observation, the rats' weight increased steadily in the early model period and reached a plateau on the 24th day. From the 40th day, the weight of the sequential group (lentivirus plasmid intervention group) increased steadily, while the weight of the remaining groups began to decrease (Figure 26-B). Compared with the solvent control group, the rats in the EGCG treatment group did not show a significant reduction in lung metastasis lesions or a significant drop in lung water volume, while the WP-1130 group and the sequential group showed a significant reduction in lung metastasis lesions and a significant drop in lung water volume (Figure 26-C, D). The inhibitory effect of sequential intervention on tumor metastasis is even better. Therefore, as shown in Figure 26, the sequential intervention of JWA, MDM2, FASN, USP9X and PFN1 effectively inhibited gastric cancer metastasis in vivo and in vivo.
[0080] Summary of Example 1 In this example, animal model verification was conducted for the inhibition of gastric cancer drug resistance and metastasis. The core of the research plan was to use a targeted drug to jointly intervene in two independent tumor heterogeneous molecules, and then sequencing (protein profile analysis) was performed on the remaining metastatic tumor lesion tissue to select the heterogeneous molecules that drive the tumor to metastasize and order them by HR value to obtain the next heterogeneous molecule to intervene in. Intervention was then performed on the new heterogeneous molecule, and then sequencing and selection was performed on the remaining metastatic tumor lesion tissue, and this process was repeated until all of the expected heterogeneous molecules were found and intervention was performed.
[0081] In this case, the first round of sequential intervention plans to inhibit gastric cancer metastasis were JWA and MDM2, and after joint intervention, protein sequence analysis was performed on the residual metastatic lesions in the lungs of experimental animals to determine that the third heterogeneous molecule was the fatty acid synthase FASN. After joint intervention with JWA and MDM2, a rat model of passive metastasis of gastric cancer was constructed using primary cells extracted from the residual metastatic lesions of tumor metastasis, and sequential intervention and treatment were performed with FASN inhibitors. After intervention, resequencing was performed on the residual metastatic lesion tissues of gastric cancer in the lungs of experimental animals to discover the next heterogeneous molecule. According to this plan, a series of new heterogeneous molecules were discovered in succession in the next study of this case, and the corresponding inhibitors or agonists were selected according to the accessibility and price of the heterogeneous molecules to carry out sequential intervention verification. In this case, gastric cancer metastasis was targeted, and then intervention verification was performed with the heterogeneous molecules USP9X and PFN1. Therefore, in this case, a total of five heterogeneous molecules were sequentially intervened. In a sequential intervention validation study to verify the regulation of gastric cancer metastasis by heterogeneous molecules, the efficacy of each round of sequential intervention was determined by the number and size of lung metastatic nodules.
[0082] It should be noted that during the implementation of this validation model, in addition to JWA and MDM2, which are known important mutually independent molecules and were directly jointly intervened, FASN, USP9X, and PFN1, which were selected based on the availability of targeted molecular intervention drugs, are independent heterogeneous molecules, and the interaction strength scores predicted by different software were 0 points (Figure 16). EXAMPLES
[0083] This example is a validation of a system algorithm and sequential intervention to identify the heterogeneous molecules that drive and metastasize melanoma.
[0084] The main purpose of this example is to verify whether the heterogeneous molecules identified by the algorithm of the present invention have the independent function of driving melanoma to metastasis to the lung by using the theory of predicting and identifying the heterogeneous molecules that drive tumor metastasis according to the present invention and the rat lung metastasis model constructed with melanoma cell B16F10, and at the same time, combine with the conclusion of 1.6 of Example 1 in the method of identifying the heterogeneous molecules that drive gastric cancer metastasis and performing sequential intervention (the heterogeneous molecules that drive tumor metastasis are independent of each other in function, and even if one of the heterogeneous target molecules is intervened, the expression level of other heterogeneous molecules remains stable). During the implementation of this model, the selection plan for identifying the heterogeneous molecules that drive tumor metastasis was further optimized, and the model was used to verify the effectiveness of the optimization plan.
[0085] 2.1 Inhibition of melanoma metastasis in vivo by JWA mimetic peptide JP1. A passive transfer model was established by intravenous injection of melanoma cells B16F10 into rat tails. The inoculated cells were intraperitoneally injected into the rats on the second day, and Ctrl-R or JP1 polypeptide intervention was performed in each group. As shown in Figure 27, compared with the Ctrl-R group, the JP1 intervention group significantly inhibited melanoma lung metastasis, and the difference was statistically significant (Note: For Ctrl-R peptide and JP1 peptide, see Cui J, et al. Theranostics. 2020, 10(18):8036-8050).
[0086] To verify the effect of JP1 intervention on rat survival, a survival model of lung metastasis following tail vein injection of B16F10 cells was established. On the second day after tail vein injection of B16F10 cells, rats were randomly divided into groups according to body weight and administered Ctrl-R or JP1 intervention daily via intraperitoneal injection. As shown in Figure 28, compared with the Ctrl-R group, JP1 intervention significantly prolonged the survival time of lung metastasis rats, and the average survival days of the rats in the Ctrl-R group and JP1 intervention group were 25 days and 32 days, respectively, with a statistically significant difference.
[0087] Conclusion: Intervention of JWA mimetic peptide JP1 can effectively inhibit melanoma metastasis in vivo.
[0088] 2.2 Inhibition of melanoma metastasis in vivo using MDM2 inhibitor SP141. A rat model of passive metastasis was established using B16F10 cells. Different doses of SP141 were administered starting from the second day of cell inoculation. The number of lung metastasis nodules was recorded after the model experiment was completed. As shown in Figure 29, SP141 can inhibit the metastasis of rat melanoma lungs, and there is a relationship between the dose and the effect. Compared with the control group, the difference in the 80mg / kg group was statistically significant.
[0089] Conclusion: SP141 can effectively reduce melanoma metastasis in rats by inhibiting MDM2.
[0090] 2.3 To develop a model in which JWA and MDM2 cooperate to inhibit melanoma metastasis in vivo and in vivo based on their characteristics as functionally independent heterogeneous molecules. The relationship between the expression levels of JWA and MDM2 genes and the prognosis of melanoma groups was analyzed using the TCGA database. The results showed that patients with high JWA expression had a significantly longer survival time than patients with low expression, and patients with low MDM2 expression had a significantly longer survival time than patients with high expression. To analyze the relationship between the expression levels of the above two molecular proteins and the malignant phenotype of B16F10 cells, B16F10 cells were transfected with different amounts of Flag-JWA plasmid, and the JWA protein level was gradually increased, but the MDM2 protein expression level did not change significantly. At the same time, melanoma cells B16F1024h were treated with 0, 0.25 and 0.5 μM SP141 to detect the expression levels of MDM2 and JWA proteins, and found that the expression level of MDM2 protein decreased with increasing dose of SP141, but the expression level of JWA protein did not change significantly. Thus, as shown in FIG. 30, in melanoma, JWA and MDM2 are functionally independent heterogeneous molecules with no mutual regulatory relationship at the protein level.
[0091] After verifying the cell model, we further verified the animal model in rats. We established a lung metastasis model by tail vein injection of B16F10 cells, and on the second day, the rats were randomly divided into groups according to their body weight before drug intervention. As shown in Figure 31, the joint intervention of JWA and MDM2 can cooperate to inhibit the metastasis of melanoma in the body.
[0092] Conclusion: JWA and MDM2, which are heterologous molecules that drive melanoma metastasis, can be jointly intervened to cooperate in inhibiting melanoma metastasis to the lung.
[0093] 2.4 After JWA and MDM2 co-treatment of rats, melanoma lung metastasis was significantly inhibited, but the rats eventually died of lung metastasis. In response, other heterogeneous molecules that further drive melanoma metastasis in the residual metastatic foci of co-treatment of lung melanoma were selected and identified. Although joint intervention of JWA and MDM2 enhanced the inhibition of melanoma metastasis, melanoma cells have heterogeneous molecules that are not regulated by JWA and MDM2, so there are molecules that drive melanoma progression even with joint intervention. Protein profile analysis was performed on the residual lung metastatic lesions of rats jointly treated with JP1 and SP141, and protein molecules with more stable expression levels (expression levels between 0.95 and 1.05) than the control group were selected from the quantitative analysis of a total of more than 7,000 proteins, and initial selection of heterogeneous molecules was performed as follows. (1) There are ≥2 specific peptides, 2651 molecules. (2) Protein molecules predicted by the mass spectrometry results are removed. (3) Molecules with a coverage of identified peptide fragments in proteins of 10% or less are removed. (4) Molecules with more than two protein quantifications are retained. (5) Molecules with a total score of 200 or more are retained. (6) Identify protein molecules involved in tumor progression (54 in total, see Figure 32).
[0094] FIG. 33 and FIG. 42 show information on the relationships between each candidate protein molecule.
[0095] A series of bioinformatics relationship analyses were conducted on these 54 molecules, and the process was similar to that of 1.4 of Example 1 above. Based on experimental data and the rules of protein interaction, an initial interaction network was constructed, and it was found that the heterogeneous molecules that can be involved in melanoma are SMAD3, IDH1 / IDH2, mTOR, KHSRP / SRSF3, PGD, POLB, TOP1, L2HGDH, MOB2, PDIA4, CD2AP, FADS1, PGP, GSTP1, MAPAK3, MAPAK4, MEAF6, RNF181, D54, ITGA4, TYR, UFC1, NDRG1 and OAT (Figures 34-38).
[0096] The HR value of each molecule was calculated, and the order in which heterogeneous molecules can be involved in the sequential treatment of melanoma was determined as follows: RNF181 (0.47 (0.29-0.77), p = 0.003), mTOR (1.75 (1.12-2.74), p = 0.015), SRSF3 (1.68 (0.86-3.27), p = 0.127), TOP1 (0.62 (0.40-0.97), p = 0.035), L2HGDH (0.63 (0.43-0.92), p = 0.017), MOB2 (0.68 (0.45-1.0 5), p=0.079), UFC1(1.45(1.01-2.08), p=0.043), IDH1 / 2(1.42(1.07-1.89), p=0.014), PGP(1.41(0.8) 7-2.29), p=0.16), CD2AP(0.73(0.54-0.99), p=0.04), PGD(0.75(0.55-1.01), p=0.06), ITGA4(0.78, p= 0.066), FADS1(0.81(0.66-1.00), p=0.047), POLB(1.23(0.93-1.62), p=0.138), SMAD3(1.22(0.94-1. 60), p=0.139), MEAF6(1.13(0.66-1.93), p=0.652), OAT(1.11(0.83-1.47), p=0.486), GSTP1(1.10(0.9) 5-1.27), p=0.188), NDRG1 (0.96 (0.85-1.09), p=0.503), TYR (0.98 (0.90-1.07), p=0.706), PDIA4 (1.00 (0.70-1.43), p=0.991), MAP2K3 (1.00, p=0.74), MAP2K4 (1.00, p=0.87) and D54 (1.00, p=0.95) (Figures 34-38).
[0097] However, it is unknown whether the selected heterogeneous molecules have antagonist or agonist drugs that can be used in intervention models. In this validation model, we comprehensively examined the possibility that these heterogeneous molecules can be used as targets for intervention, and in addition to JWA and MDM2, we also selected SMAD3, IDH1 / 2, and mTOR as heterogeneous molecules that can intervene.
[0098] It should be noted that the importance of heterogeneous molecules is weighted according to the magnitude of the HR value in the present invention. Although the measure of intervention according to the gradual decrease of HR value may be desirable in theory, the actual contribution of molecules with different R values to the progression of H tumors may correspond to dynamic changes in big data, and the HR value of the same molecule may change with the continuous accumulation of data. Therefore, the HR value is a relative reference value. Since the molecules that can be intervened obtained by this prediction system are independent of each other, it is believed that any molecule (molecules with an interactive relationship or independent molecules) will be effective when intervened among them. In actual practice, depending on the availability and price of heterogeneous molecular drugs, it may be comprehensively considered and determined in combination with the HR value order. For example, in the model of this verification, the target molecules that sequentially intervene in melanoma metastasis are JWA, MDM2, SMAD3, IDH1 / 2 and mTOR in succession, and although they do not completely follow the order of decrease of HR value, the intervention results are effective.
[0099] Next, we performed pathway analysis on these five molecules. As shown in Figure 39, these five molecules were indeed significantly different among the signaling networks operating in tumors, and were relatively independent heterogeneous molecules in function.
[0100] 2.5 Effectively inhibiting melanoma metastasis through sequential intervention in JWA, MDM2, SMAD3, IDH1 / 2 and mTOR. According to the five target molecules selected, a four-stage sequential intervention plan was established for melanoma metastasis model rats. As shown in Figure 40, in the first stage, JWA and MDM2 (JP1 and SP141) were administered from the 1st to the 14th day, in the second stage, SMAD3 (SIS3) was administered from the 15th to the 21st day, in the third stage, IDH1 / 2 (AG881) was administered from the 22nd to the 28th day, and in the fourth stage, mTOR (Rapamycin) was administered from the 29th day until the completion of the model experiment. The effects of each stage of intervention were evaluated by administering the model rats to the previous stage (Note: JP1 and SP141 are the same as above, and SIS3, AG881 and Rapamycin are products of conventional technology).
[0101] A survival model of rats with lung metastasis after tail vein injection of B16F10 cells was established, and after cells were injected into the tail vein, the rats were randomly divided into groups according to their weight, and intervention was started from day 2. According to the results (see Figure 41), the average survival times of rats in the control group, JP1+SP141 sequential intervention group, JP1+SP141+SIS3 sequential intervention group, JP1+SP141+SIS3+AG881 sequential intervention group, and JP1+SP141+SIS3+AG881+Rapamycin sequential intervention group were 22.5, 25.55, 27.05, 29.35, and 30.7 days, respectively (Figure 41-D). When comparing the effects of the third intervention of JP1+SP141+SIS3, the fourth intervention of JP1+SP141+SIS3+AG881, the fourth intervention of JP1+SP141+SIS3+AG881, and the fifth intervention of JP1+SP141+SIS3+AG881+Rapamycin on the survival time of rats individually, as shown in Figures 41-E and 41-F, with each additional round of intervention, the survival time of rats increased significantly compared to the control group, and this difference was statistically significant.
[0102] Summary of Example 2 In order to improve the feasibility of the results of the present invention in the development of new drugs and clinical application, the selection strategy of heterogeneous molecules was further optimized in the study of sequential intervention verification to inhibit melanoma metastasis, that is, in the gastric cancer verification study, after each intervention, the prediction of heterogeneous molecules was optimized by resequencing, and after the first round of intervention, the remaining metastatic lesions were sequenced to predict all heterogeneous molecules and use them as the basis for the next intervention. If this method can be realized, the procedure of identifying and predicting heterogeneous molecules will be greatly simplified, and the feasibility of promoting application will inevitably be improved. Regarding melanoma, according to previous studies, the expression level of JWA is significantly lower than that of paracancerous tissue, while the expression level of MDM2 is significantly higher than that of paracancerous tissue, and JWA and MDM2 are mutually independent heterogeneous molecules. In the experimental model of verification, the heterogeneous molecules JWA and MDM2 were jointly intervened, and then the remaining metastatic lesions in the rat lung were extracted and subjected to high-throughput sequencing of proteins. The heterogeneous molecules SMAD3, IDH1 / 2 and mTOR were comprehensively selected for the next three rounds of sequential intervention.
[0103] In sequential interventions with heterologous molecules that drive melanoma metastasis, the efficacy of each round of sequential intervention was determined based on survival from pulmonary metastasis.
[0104] It should be noted that in the validation model for melanoma metastasis inhibition, in addition to JWA and MDM2 being known to be important mutually independent molecules that directly carry out joint intervention, the interaction strengths of SMAD3, IDH1 / 2 and mTOR were predicted by different software, and they were shown to be molecules with high interaction strength scores in the mutually independent interaction network (Figure 35).
[0105] In addition to the above examples, the present invention may have other implementations. All technical solutions by equivalent replacement or equivalent transformation are within the scope of the claims of the present invention.
Claims
1. A prediction system for driving important heterologous molecules in tumor metastasis, comprising an input module, an analysis module and an output module. The input module is for inputting the results of quantitative analysis of proteins before and after the intervention, the results of quantitative analysis of proteins before the intervention are a collection of quantitative analysis results of the expression quantity of each protein in the tumor metastasis lesions of the subject patient before drug intervention, and the results of quantitative analysis of proteins after the intervention are a collection of quantitative analysis results of the expression quantity of each protein in the residual tumor metastasis lesions of the subject patient after drug intervention. The analysis module includes a primary analysis sub-module, a secondary analysis sub-module and a calculation and ordering sub-module. The primary analysis submodule is used for initial selection and analysis, which includes comparing the quantitative analysis results of proteins before and after the intervention and entering proteins whose expression levels change within a predetermined range into a candidate protein set. The secondary analysis submodule is used for reselection analysis, and the reselection analysis includes: using a protein interaction network analysis tool to construct a protein interaction network based on experimental verification by the candidate protein set, linking with the signal pathways related to the candidate protein set in the signal pathway database to organize the protein interaction network; putting the most important proteins in each protein interaction network into a collection of independent molecules as independent heterogeneous molecules; using a protein interaction network prediction tool to construct a protein interaction prediction network based on the protein interaction discipline by the candidate protein set; and listing the proteins in the candidate protein set that have not participated in any protein interaction prediction network as independent heterogeneous molecules in the collection of independent molecules. The calculation and ordering module is used for calculation and ordering, and the calculation and ordering includes calculating an HR value for each heterogeneous molecule in the collection of independent molecules, ordering each heterogeneous molecule according to its HR value, and forming an ordered list of each heterogeneous molecule. The output module is used to output an ordered list of each heterogeneous molecule.
2. The prediction system for identifying important heterologous molecules that drive tumors and metastasis, as described in claim 1, characterized in that the quantitative analysis results of proteins before and after intervention are obtained by high-throughput sequencing of proteins, and the high-throughput sequencing of proteins includes protein profile analysis.
3. The prediction system for identifying important heterologous molecules that drive tumors and cause metastasis, as described in claim 1, characterized in that in the initial selection and analysis, the expression amount of proteins that are within a predetermined range is 0.95 to 1.05 times the expression amount before protein intervention, after protein intervention.
4. The prediction system for identifying important heterologous molecules that drive tumors and metastasis as described in claim 1, characterized in that in the initial selection and analysis, proteins whose expression levels change within a predetermined range are gradually selected and then added to the candidate protein set. The stepwise selection is then to select a protein that meets the following conditions: the number of specific peptides is ≥ 2, it is not a predicted protein, the coverage of the identified peptide amino acid in the current protein is ≥ 10%, the number of times of the current protein quantification is ≥ 2, the total score of the sequence quality of the current protein is ≥ 200 points, and there is evidence in the database that proves that it is involved in tumor progression; the specific peptide is a specific peptide fragment for identifying the current protein; the predicted protein is a predicted protein with the reliability properties of the current protein defined in Uniport or NCBI database; the identified peptide fragment is a specific peptide fragment used for indication and detected in the current protein; and the number of times of the protein quantification is the overlapping number of times that the current protein is quantified.
5. The prediction system for identifying important heterogeneous molecules that drive tumors and metastasis, as described in claim 1, characterized in that in the reselection analysis, the protein interaction network analysis tool includes the STRING network tool, the signal pathway database includes KEGG, SMPDB, WikiPathways, Cell Signaling Technology, BioCarta Pathway, Pathway Commons and PID, and the protein interaction network prediction tool includes the Genemania network tool, the Unihi network tool and the Hitpredict network tool.
6. The prediction system for identifying important heterologous molecules that drive tumors and metastasis, as described in 5, is characterized in that a protein interaction network is constructed and organized using Cytoscape tools, and a prediction network of protein interactions is constructed.
7. The prediction system for identifying important heterologous molecules that drive tumor metastasis as described in claim 1, characterized in that in the reselection and analysis, the most important protein in each protein interaction network is identified by quality characteristic analysis, and the said characteristic analysis includes an analysis of the importance and radial degree of each protein in the protein interaction network, among which the protein with the highest importance and radial degree is the most important protein in this protein interaction network.
8. The prediction system for identifying important heterogeneous molecules that drive tumor metastasis as described in claim 1, characterized in that in the calculation and ordering, the HR value of the heterogeneous molecule is calculated as follows: obtain sequence data related to the tumor from a public database, obtain a normalized count matrix of genes corresponding to the heterogeneous molecule from the sequence data, perform batch single-factor COX regression on the genes, and calculate the HR value.
9. The prediction system for identifying important heterologous molecules that drive tumor metastasis according to claim 8, characterized in that the public databases include TCGA, ICGC and GEO.
10. The prediction system for identifying important heterogeneous molecules that drive tumor metastasis as described in claim 1, characterized in that in the calculation and ordering, the ordering rule is that heterogeneous molecules with HR value ≧1 are directly ordered in reverse order according to their HR value, and heterogeneous molecules with HR value < 1 are ordered in reverse order of 1 / HR value.
11. A sequential intervention system for executing on heterogeneous molecules that drive tumor metastasis, comprising a prediction system for identifying important heterogeneous molecules that drive tumor metastasis according to any one of claims 1 to 10, and an execution module. The execution module is for determining the order of sequential interventions according to the ordered list of each heterogeneous molecule outputted to the prediction system and the determined rules.
Citation Information
Patent Citations
Tumor metastasis and recurrence prediction method and system based on TCGA database
CN109801680A
High-dimension low-sample gene data screening and protein network analysis method and system
CN111192638A
Recognized cell-based assays and their use
JP2014512520A
Pathway recognition algorithm using data integration from genome models (PARADIGM)
JP2014532859A
Multiplexed proteomics and predictive drug candidate assessment
US20200225242A1
