A multi-round scanning screening method of molecular combination drugs with step-by-step superposition of molecules and step-by-step increase of therapeutic effect

Abstract

The application discloses a kind of molecular gradually superimposed curative effect gradually increasing molecular combination drug multi-round scanning screening method.In the first round single molecule scanning, n molecules are respectively used in disease model, and toxicity and curative effect are tested, and the molecule of toxicity / curative effect comprehensive better is selected, and enters second round double molecule combination scanning;In the second round double molecule combination scanning, the molecule selected in the first round and the remaining n-1 molecules are combined respectively, and the double molecule combination of toxicity / curative effect comprehensive better is selected, and enters next round……Such as experiencing multiple rounds, the molecule contained in molecule combination gradually increases, and curative effect also gradually increases, and a molecule combination containing m molecules is obtained, and the curative effect of the molecule combination on disease model reaches or exceeds the set curative effect threshold, while toxicity remains at a low level.The molecular combination drug screening method of the application is a kind of brand-new drug development paradigm.

Description

Technical Field

[0001] This invention relates to the field of drug development, specifically to a multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular cascading. Background Technology

[0002] Diseases threaten human health, and many currently lack effective treatments. The traditional mainstream drug development approach is a single-molecule drug development method, requiring knowledge of the disease's molecular mechanisms / targets. Without understanding the disease's mechanisms, drug development is extremely difficult. Understanding the disease's molecular mechanisms / targets allows for the design or screening of a new single compound that interacts with those mechanisms / targets, thereby inhibiting or curing the disease. However, designing or screening such a single compound that effectively interacts with a key disease mechanism / target is generally a very difficult and time-consuming process; some key mechanisms / targets are even undruggable. Furthermore, the single compound that specifically targets a disease, painstakingly identified, often fails in the final clinical trial stage, either due to insufficient clinical efficacy or inadequate safety profiles.

[0003] The inventors believe that drug development can still proceed even when the molecular mechanisms / targets of a disease are unknown, or when drug development is challenging despite known mechanisms / targets. It is still possible to develop low-toxicity and effective drugs. Based on the concept of molecular interaction networks in living organisms, the inventors propose a novel preclinical multi-molecule drug screening method: a multi-round scanning screening method for molecular combination drugs with progressively increasing efficacy through cascading molecular combinations. According to this method, for a specific (mechanism unknown) disease, theoretically, one or more highly effective and low-toxicity multi-molecule candidate drugs (i.e., molecular combination candidate drugs) can be rapidly identified through preclinical disease models for clinical testing. The inventors provided the first prototype proof of the feasibility of this method using a mouse Lewis lung cancer model. Summary of the Invention

[0004] The purpose of this invention is to provide a completely new drug development theory / method / paradigm / system, in addition to existing drug development theories / methods / paradigms / systems, thereby providing more options for drug development and providing a multi-round scanning screening method for molecular combination drugs with progressively increasing efficacy through molecular cascading.

[0005] The various aspects of the present invention will now be described in detail. Unless otherwise specified, all raw materials required for the present invention are commercially available or prepared according to conventional methods in the art. Unless otherwise defined or stated, technical terms and scientific terms used herein have the same meanings as those familiar to those skilled in the art. Furthermore, any methods and materials similar to or equivalent to those described herein may be applied to the methods of the present invention.

[0006] (I) Theoretical basis of the method described in the invention

[0007] From a chemical perspective, a living organism can be viewed as a massive self-organizing molecular aggregate, including DNA / RNA, proteins, amino acids, lipids, vitamins, minerals, and many other molecules. Within this massive self-organizing molecular aggregate, a highly complex network of interactions exists, encompassing various physical and chemical interactions. This network of molecular interactions underpins the existence of life. In a healthy organism, this network is in a normal state. In a diseased organism, it is in an abnormal state. When a molecule or combination of molecules enters the organism, it interacts with the massive molecular aggregate, causing a disturbance to the network. Positive disturbances can cause an abnormal network to return to a normal state, thus controlling or even curing the disease; negative disturbances can cause a normal network to shift towards an abnormal state, leading to disease.

[0008] From the perspective of molecular interaction networks, traditional drug development targets a key interaction (i.e., a critical target) within the molecular interaction network of a diseased organism. The goal is to design or screen a molecule to efficiently inhibit or disrupt this interaction, thereby effectively alleviating or even eliminating the abnormal state of the molecular interaction network. However, as mentioned earlier, such molecules are often very difficult to find. Even if a molecule that can efficiently inhibit / disrupt a critical interaction / target is found, it often has significant toxic side effects because this interaction / target is also crucial for the normal functioning of the organism.

[0009] The inventors believe that drug development does not necessarily require highly efficient targeting of a single key interaction / target; it can also target the entire molecular interaction network in a disease state. Within this network, there are often multiple disease-related interactions / targets, not just one. For example, in cancer, the proliferation of gene mutations causes many interactions to be in an abnormal state, and these abnormal interactions play a crucial role in the occurrence, development, and metastasis of cancer.

[0010] The inventors believe that diseases can be controlled / cured either by strongly interfering with a single key interaction / target (i.e., the traditional single-molecule drug development method) or by weakly interfering with multiple disease-related interactions / targets (i.e., the molecular combination drug screening method proposed in this invention). In other words, drug development can begin with weak perturbations of the molecular interaction network of organisms in a disease state by external molecules.

[0011] Many molecules possess diverse biological activities and can act on a variety of biological targets, thereby perturbing the molecular interaction network of living organisms to some extent. The inventors believe that there may exist a group of molecules (i.e., a molecular combination) where each molecule can exert a weak perturbation on the molecular interaction network in a disease state. Although the weak perturbation of each molecule is insufficient to significantly control / cure the disease, when all the weak perturbations of this group of molecules are accumulated / synergistically combined, they will produce a positive strong perturbation on the molecular interaction network, thereby effectively alleviating the abnormal state of the molecular interaction network and controlling or even curing the disease.

[0012] (II) Specific details of the method described in the invention

[0013] The molecular interaction networks of living organisms are extremely complex, and it is unlikely that humans will have a comprehensive understanding of these networks in the foreseeable future. Therefore, it is difficult to rationally design molecular combinations that can treat specific diseases. However, the inventors believe that it is possible to experimentally and relatively quickly find molecular combinations that can drive the molecular interaction networks of diseased organisms back to a normal state.

[0014] The inventors have proposed a multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular cascading, comprising the following steps:

[0015] 1) Apply n molecules from a molecular set to a disease model, test the toxicity and efficacy of the n molecules, take into account both toxicity and efficacy, prioritize toxicity over efficacy, select the molecule with the best overall toxicity / efficacy, and select one molecule from the selected molecules to enter the second round of bimolecular combination scanning.

[0016] 2) In the second round of bimolecular combination scanning, the molecules selected in the first round and the remaining n-1 molecules are combined to obtain n-1 bimolecular combinations. The toxicity and efficacy of the n-1 bimolecular combinations in the disease model are tested. The bimolecular combinations with better overall toxicity / efficacy are selected. One of the selected bimolecular combinations is selected to enter the next round of molecular combination scanning, i.e., the m-th round m-m-molecule combination scanning.

[0017] 3) In the m-th round of m-molecule combination scanning, the m-1 molecule combination selected in the (m-1)-th round is combined with the remaining n-m+1 molecules to obtain n-m+1 m-molecule combinations. The toxicity and efficacy of the n-m+1 m-molecule combinations in the disease model are tested, and the m-molecule combinations with better overall toxicity / efficacy are selected.

[0018] 4) Select 3 to n in sequence and repeat step 3). When the selected m molecular combination has an efficacy in the disease model that reaches or exceeds the set efficacy threshold, and the toxicity is equal to or lower than the set toxicity threshold, the molecular combination drug screening is completed.

[0019] If the molecular combination drug screening is not completed, replace the molecules in the molecular set in step 1), return to step 1), and scan and screen again. Repeat this process until the molecular combination drug screening is completed.

[0020] The disease model described is an in vitro disease model, which can be a two-dimensional cell planar model, a three-dimensional cell spheroid model, an organoid model, an organ / tumor microarray model, or a tumor slice culture.

[0021] In step 1), considering both toxicity and efficacy, toxicity is prioritized over efficacy, and molecules with a better overall toxicity / efficacy ratio are selected. Specifically, these include:

[0022] First, sort the molecules tested in step 1) from highest to lowest toxicity, and select the molecules with the lowest toxicity in the bottom 60%. Then, sort the molecules with the lowest toxicity in the bottom 60% from highest to lowest efficacy, and select the molecules with the highest efficacy in the top 50%. This gives you the molecules with the best overall toxicity / efficacy.

[0023] In step 2), bimolecular combinations with a good balance of toxicity and efficacy are selected, specifically including:

[0024] First, sort the bimolecular combinations tested in step 2) from highest to lowest toxicity, and select the bimolecular combinations with the lowest toxicity in the bottom 60%. Then, sort the bimolecular combinations with the lowest toxicity in the bottom 60% from highest to lowest efficacy, and select the bimolecular combinations with the highest efficacy in the top 50%. This gives you the bimolecular combinations with the best overall toxicity / efficacy.

[0025] In step 3), combinations of m molecules with a good balance of toxicity and therapeutic effect are selected, specifically including:

[0026] First, sort the n-m+1 m-molecule combinations tested in step 3) according to their toxicity from highest to lowest, and select the m-molecule combinations with the lowest toxicity in the bottom 60%; then sort the m-molecule combinations with the lowest toxicity in the bottom 60% according to their efficacy from highest to lowest, and select the m-molecule combinations with the highest efficacy in the top 50% to obtain the m-molecule combinations with the best overall toxicity / efficacy.

[0027] The inventors believe that, theoretically, for a given molecular combination that can effectively treat a certain disease, there should be an additive therapeutic chain between the molecules in that combination. That is, assuming there is a molecule A, it can act on one or more interaction sites in the molecular interaction network of a certain disease, thereby exerting a certain therapeutic effect on the disease (E). A Therefore, there exists another molecule B that can act on one or more sites in the disease molecular interaction network. When A and B are used simultaneously, they will act on all the aforementioned sites simultaneously, and the therapeutic effect of these effects on the disease will be at least as good as the therapeutic effect of using A alone, which is E. A+B ≥E A Then, there exists a third molecule C that can also act on different sites in the disease molecular interaction network, such that the efficacy of using A, B, and C simultaneously is at least no less than the efficacy of A+B, i.e., E. A+B+C ≥E A+B There exists a fourth molecule D such that the therapeutic effect of A+B+C+D is at least no less than that of A+B+C, i.e., E. A+B+C+D ≥E A+B+C ...and so on; ultimately, a molecular combination containing A, B, C, D, ... M is obtained, whose therapeutic effect reaches or exceeds the pre-set therapeutic threshold E. threshold (meaning that the disease can be effectively controlled or even cured), i.e., E A+B+C+D+…+M ≥E threshold .

[0028] Therefore, for a disease (whose molecular mechanism is unknown but for which an effective preclinical model has been established), if there is a molecular set (M1, M2, ... M...) containing n active molecules... i ,…M n-1 M n Suppose that within this molecular set, there exists one or more groups of molecules (i.e., one or more molecular combinations) that can collectively and effectively act on the molecular interaction network of the diseased organism, thereby effectively controlling or even curing the disease. Then, a multi-round scanning screening method for finding these molecular combinations, characterized by a step-by-step synergistic effect and progressively increasing therapeutic efficacy, is described below:

[0029] The first round of single-molecule scanning involves applying each of the n molecules from the aforementioned molecule set to the disease model to test their toxicity and efficacy. Considering both toxicity and efficacy, toxicity is prioritized over efficacy, with molecules exhibiting low toxicity but some therapeutic effect being given priority. The molecule with the optimal overall toxicity / efficacy is selected, or a molecule with a good overall toxicity / efficacy is selected from the list of molecules. This molecule is hypothesized to be M. A Its therapeutic effect is E A Meanwhile, its toxicity is within an acceptable range.

[0030] Second round of bimolecular combination scanning: M A Combine the remaining n-1 molecules to obtain n-1 bimolecular combinations. Test the toxicity and efficacy of each bimolecular combination against the disease model, and select the bimolecular combination with the best overall toxicity and efficacy, or one from the bimolecular combinations with a good overall toxicity / efficacy ratio. This bimolecular combination is assumed to be M. A +M B Generally speaking, M A +M B The therapeutic effect should be better than M. A The therapeutic effect, or at least not less than M A The therapeutic effect, namely E A+B ≥E A Meanwhile, its toxicity is within an acceptable range.

[0031] Third round of trimolecular combination scanning: M A +M B Combined with the remaining n-2 molecules, n-2 trimolecular combinations are obtained. The toxicity and efficacy of each of these trimolecular combinations are tested, and the trimolecular combination with the best overall toxicity and efficacy is selected, or one is selected from the trimolecular combinations with a good overall toxicity / efficacy. This trimolecular combination is assumed to be M. A +M B +M C Generally speaking, M A +M B +M C The therapeutic effect should be better than M. A +M B The therapeutic effect, or at least not less than M A +M B The therapeutic effect, namely E A+B+C ≥E A+B Meanwhile, the toxicity is within acceptable limits.

[0032] ...

[0033] The m-th round of m-molecule combination scanning will include M A +M B +…+M m-1Combined with the remaining n-m+1 molecules, n-m+1 m-molecule combinations are obtained. The toxicity and efficacy of each of these m-molecule combinations are tested. The m-molecule combination with the best overall toxicity and efficacy is selected, or one is selected from the m-molecule combinations with a good overall toxicity / efficacy. This m-molecule combination is assumed to be M. A +M B +…+M m-1 +M m Generally speaking, M A +M B +…+M m-1 +M m The therapeutic effect should be better than M. A +M B +…+M m-1 The therapeutic effect, or at least not less than M A +M B +…+M m-1 The therapeutic effect, namely E A+B+C+…+(m-1)+m ≥E A+B+C+…+(m-1) Meanwhile, the toxicity is within an acceptable range. At this point, the therapeutic effect of the optimal m-molecule combination reaches or exceeds the pre-set therapeutic threshold E. thresh old, i.e. E A+B+C+…+m ≥E thresh old.

[0034] (III) Disease models to which the method described in the invention is applicable

[0035] The multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic efficacy, as described in this invention, is applicable to any preclinical disease model. The closer the disease model is to human physiological conditions, the better. Two-dimensional cell models are far removed from human physiological conditions, and the molecular combinations screened may not have high reliability. Animal models are closer to human physiological conditions than two-dimensional cell models, and therefore have higher reliability. Furthermore, the method can also be used to screen molecular combination drugs on three-dimensional cell spheroids, organoids, organ / tumor microarrays, tumor slice cultures, and other models. If a certain number of promising molecular combinations can be screened through effective preclinical disease models, the probability of one or more of them passing clinical trials will increase, potentially improving the success rate of drug development.

[0036] In theory, if we can accurately simulate the molecular interaction network of living organisms—ultra-large-scale self-organizing molecular aggregates—we can first conduct virtual screening of molecular combination drugs on computer virtual models of human diseases, and then conduct preclinical and even clinical tests on real disease models. This may be the most ideal situation, which would greatly increase the probability of successful drug development and thus achieve precision medicine.

[0037] (iv) Selection criteria for each molecule in the starting molecule set of the method described in the invention

[0038] The selection criteria for each molecule in the starting molecule set are as follows:

[0039] (1) The differences in chemical structure and physicochemical properties among the molecules should be as large as possible. The differences in chemical structure and physicochemical properties mean that these molecules act on different sites in the molecular interaction network, so that after combination, these molecules can cover as many sites in the disease molecular interaction network as possible, and thus may produce the desired positive perturbation to the disease;

[0040] (2) Each molecule should possess as much pleiotropic biological activity as possible, acting on multiple targets in the body to produce a variety of physiological / pharmacological activities. A typical example of this is the antidiabetic drug metformin, for which numerous studies have reported that the compound possesses a variety of physiological / pharmacological activities beyond antidiabetic effects.

[0041] (3) The toxicity of the selected molecules should be as low as possible. If the molecules in the molecular set are all low-toxic or non-toxic active molecules (for example, FDA-approved drug molecules that have been proven to have low toxicity through long-term use, and many non-toxic or low-toxic natural product molecules, etc.), then the effective molecular combination found by the above method is more likely to remain low in toxicity.

[0042] (4) Bioavailability should be as high as possible. For example, molecules that are well absorbed in the human gastrointestinal tract after oral administration and are widely distributed in the body, or molecules that have good water / lipid solubility and are suitable for administration by subcutaneous / intraperitoneal / intravenous injection. In addition, the in vivo half-life of these molecules should be as long as possible.

[0043] (5) They do not react chemically with each other.

[0044] (6) The selected molecules should be readily available / synthesized. The inventors hope that the effective molecular combinations obtained through the molecular combination screening method will contain molecules that are readily available or readily synthesized. Otherwise, if the molecules are difficult to obtain / synthesize or are costly, the application of the molecular combinations will be greatly limited.

[0045] (v) Diversity of the starting molecular set of the method described in the invention

[0046] Since the various interactions in the molecular interaction network of living organisms always occur within a certain spatial shape / conformation, and the number of such spatial shapes / conformations is finite, and these spatial shapes / conformations may also have a certain degree of similarity / overlap, when the initial molecule set is large and contains a certain number of molecules with diverse structures, then for any disease, it is highly likely that some molecules will be found in this initial molecule set. Each of these molecules will fit to a certain extent with the spatial shape / conformation of one or more disease-related interactions in the disease molecular interaction network, thereby exerting a certain degree of influence (weak perturbation) on these interactions. After a certain combination, these weak perturbations of molecules may produce an additive / synergistic effect, generating a positive strong perturbation on the disease molecular interaction network, thereby achieving the effect of controlling or curing the disease. Researchers can set an initial molecule set containing a large number of molecules based on their own experience and the physical, chemical, and pharmacological properties of various molecules, and determine experimental conditions such as the dosage, administration method, administration frequency, and administration sequence of each molecule in the molecule set. Then, for any disease, the same initially set of molecules can be used to perform multiple rounds of stepwise superposition scanning and screening of the molecules to find potentially effective molecular combinations.

[0047] Researchers can also target a specific disease and, based on their experience, literature data on the effects of drugs / molecules on that disease, and the physical, chemical, and pharmacological properties of relevant drugs / molecules, define an initial set of molecules for that disease. They can then determine experimental conditions such as the dosage, route of administration, frequency of administration, and order of administration for each molecule in this set. Researchers can use this initial set of molecules designed for that disease to perform multiple rounds of progressively layered molecular scanning to identify potentially effective molecular combinations.

[0048] Whether targeting any disease or a specific one, different researchers, based on their varying research experience, are likely to design different molecular combinations and varying dosages / methods / frequencies / sequences, and then conduct multi-round superimposed scanning screenings of these molecular combinations. Within a scanning cycle comprising m rounds of scanning, the dosage / frequency / sequence of each molecule, once determined, should generally not be changed in any round. After setting the initial dose for each molecule, the dose in subsequent rounds can be adjusted based on the toxicity and efficacy observed in each round. Especially in later-stage scanning rounds, when some molecular combinations show significant efficacy, based on the general principle of a positive correlation between efficacy and dose, it may be considered to appropriately increase the dose of each molecule in that combination to see if the efficacy can be amplified while maintaining low toxicity.

[0049] Furthermore, even when using the same starting set of molecules, different researchers may have different initial settings for the administration method, dosage, frequency, and order of each molecule in the starting set. As a result, the pharmacological effects / pharmacokinetics of these molecules after combination may also be different, which may lead to different effective molecular combinations obtained after multiple rounds of molecular stacking and screening.

[0050] (vi) Diversity of molecular combination drug candidates obtained from screening

[0051] Due to the complexity of molecular interaction networks in living organisms and the diversity of molecular structures in the initial molecular set, the inventors hypothesized that in any round of scanning, there might not be just one molecular combination (or a single molecule, for the first round of scanning) with a good overall toxicity / efficacy, but rather multiple combinations. By starting from these molecules / molecule combinations and conducting subsequent rounds of scanning, it would be possible to obtain multiple molecular combinations with therapeutic effects but different compositions.

[0052] In fact, the inventors believe that, considering the finiteness and overlap of the spatial shapes / conformations of interactions mentioned earlier, when the initial molecule set contains a sufficient number of molecules with sufficiently large structural differences, it is highly likely that some single molecules with weak therapeutic effects and low toxicity can be found during the first round of single-molecule scanning. This is because, for those key interactions in the molecular interaction network of organisms in a disease state, molecules that perfectly match the specific spatial shape / conformation of these key interactions while also having low toxicity are not easy to find (which is a major reason why the development of specific single-molecule drugs is so difficult), but molecules that partially match such specific spatial shapes / conformations and have low toxicity can still be found in large numbers. For example, for a key enzyme that causes a disease, it is often very difficult to find a molecule that can have a more perfect match with its active site than the substrate while not causing excessive toxicity. However, compounds that partially match the active site, that is, compounds that have a weak inhibitory effect on enzyme activity and also have low toxicity, can still be found in considerable numbers. Such compounds are eliminated in the traditional drug development model and will not be further considered. However, under the drug development theory / method / model described in this invention, these molecules are all worthy of consideration. As mentioned earlier, if we start from these molecules with weak therapeutic effects and sequentially perform subsequent rounds of molecular combination scanning, it is possible to obtain multiple different therapeutic molecular combinations.

[0053] In summary, the inventors believe that for a large set of starting molecules containing many structurally diverse molecules, for any given disease, it is almost certain that there exist molecules that can introduce various weak perturbations into the molecular interaction network of the disease. These weak perturbations, after certain combinations, may produce a superposition / synergistic effect, becoming a strong perturbation while maintaining a low level of toxicity. Therefore, for a given disease model and a given set of starting molecules, through a limited number of rounds of progressively superimposed molecular scanning and screening as described in this invention, it is possible to find one or more therapeutic multi-molecule candidate drugs. If no therapeutic, low-toxicity molecular combination is found, the molecular set can be changed (i.e., a new batch of molecules can be used), and the scanning and screening can be repeated, in a cyclical manner, until a therapeutic molecular combination is found.

[0054] (vii) Why not directly combine the weakly effective molecules found in the first round of scanning?

[0055] Another issue is that if some low-toxicity molecules with weak therapeutic effects are found in the first round of single-molecule scanning, why not simply combine these molecules directly to obtain a molecular combination? This is because, due to the complexity of the molecular interaction network in living organisms, these molecules, after combination, may not produce additive / synergistic effects. Some molecules may exhibit antagonistic effects; some molecules that did not show activity in the previous round of combination scanning may show activity in the next round; and the low toxicity of some molecules may be cumulatively enhanced after combination. Therefore, the above simple combination cannot guarantee good efficacy, nor can it guarantee that the toxicity will remain low after combination. Therefore, the aforementioned multi-round, step-by-step molecular screening method is needed to find molecular combinations with synergistic effects but not synergistic toxicity. In summary, the multi-round, step-by-step molecular screening method proposed by the inventors can effectively avoid the problems of antagonistic efficacy / cumulative toxicity often encountered in the traditional molecular combination drug development process based on mechanism and rational design concepts.

[0056] (viii) The difference between the method described in this invention and traditional drug combination development: no need to know the molecular mechanism of the disease

[0057] The molecular combination drug development method described in this invention differs significantly from traditional drug combination therapies. Drug combination is a commonly used treatment approach for diseases, targeting certain known key interactions / targets in the molecular interaction network of abnormal disease states. Existing drugs that can act on these key interactions / targets are combined and administered to control / treat the disease. Because the drugs used in the combination are developed through traditional single-molecule drug development models, they often possess toxicity, resulting in stronger toxic side effects when used in combination. Furthermore, the molecular mechanisms of many diseases are not yet fully understood, or even if key interactions / targets are known, there are still no effective drugs to target them. In these cases, there are no readily available drugs for combination therapy. In contrast, the method proposed by the inventors does not combine existing clinical drugs with clearly defined targets for a specific disease, but rather represents a novel drug development method that does not require knowledge of the disease's molecular mechanisms. This method is based on the idea of ​​weak perturbations of the molecular interaction network of various low-toxicity active molecules on disease molecules. Starting from a molecular set containing various low-toxicity active molecules, it uses multiple rounds of molecularly superimposed scanning and screening experiments based on preclinical disease models to discover one or more low-toxicity molecular combination drugs that can effectively control or even cure a disease. Using this method, researchers do not need to understand the molecular mechanisms of the disease; they only need to consider efficacy and toxicity as indicators to quickly discover one or more preclinical multi-molecule drug candidates for a specific disease. Any disease that is currently incurable (such as cancer) can be screened for promising molecular combination drug candidates using the method described in this invention, provided that an effective preclinical disease model is established.

[0058] (ix) Similarities and differences between the method described in this invention and single-molecule drug development methods

[0059] The multi-round scanning efficacy-layering molecular combination drug screening theory / method / technology described in this invention strictly follows modern drug development standards. It first conducts preclinical disease model screening tests to identify promising molecular combination candidate drugs, and then these molecular combinations can be clinically tested. In this respect, it is consistent with mechanism / target-based single-molecule drug development. However, while targeted drugs focus on a specific mechanism / target to find a single molecule with specific efficacy, the method described in this invention focuses on the specific molecular interaction network of the disease to find efficacy-specific molecular combinations. It seeks the cumulative superposition of multiple weak physiological activities resulting from the synergistic action of multiple molecules on multiple disease-related interactions / targets in the disease molecular interaction network, ultimately achieving higher therapeutic efficacy and lower toxicity.

[0060] (x) Universality of the drug development method described in this invention

[0061] While not all diseases can necessarily find effective molecular combinations through the aforementioned molecular stacking and multi-round scanning screening method, the inventors believe that due to the widespread existence of weak perturbations of the molecular interaction network of various molecules in living organisms, at least a considerable number of diseases may be able to find effective multi-molecule clinical drug candidates through this method. Therefore, the molecular multi-round stacking and multi-round scanning screening method described in this invention can serve as a powerful alternative to existing drug development methods, and is particularly suitable for drug development work for diseases with unknown molecular mechanisms / targets. Attached Figure Description

[0062] Figure 1 The diagram shows the tumor anatomy of the five-molecule combination QD+BL+AQ+SL+EJ in two experiments. Note: Circles indicate that no tumor was found in the mouse, and crosses indicate that the mouse died of peritonitis caused by intraperitoneal injection during the experiment. Detailed Implementation

[0063] To enhance understanding of the present invention, an embodiment (prototype embodiment) is provided below to further illustrate the technical solution / details of the present invention.

[0064] The inventors selected a mouse Lewis lung cancer model as the research subject, and the Lewis lung cancer cell line used was provided by the Animal Experiment Center of Zhejiang University of Traditional Chinese Medicine. Lewis lung cancer is a mouse-derived lung cancer that can form tumors in mice with intact immune systems, thereby validating the above-mentioned drug development theories / methods / models in a physiological environment as close as possible to that of humans.

[0065] The inventors aimed to select molecules that had been proven non-toxic or low-toxic through long-term practice. These molecules were expected to maintain low toxicity even after combination. Ultimately, based on their research experience and literature review, the inventors selected 16 low-toxic molecules with broad biological activity to form a small set of starting molecules: fisetin (FS), luteolin (MX), genistein (RL), hesperidin (JP), isoliquiritigenin (YG), resveratrol (BL), EGCG (EG), curcumin (JH), perillyl alcohol (ZS), emodin (DH), oleanolic acid (QD), azithromycin (AQ), metformin (EJ), celecoxib (SL), hydroxychloroquine (QL), and trimetazidine (QM).

[0066] The selected molecules are all readily available or synthesized and have very low toxicity. Some of these are even found in everyday fruits and vegetables, such as genistein, hesperidin, and resveratrol. Others are established drugs with well-understood pharmacology and toxicology, such as metformin, celecoxib, and hydroxychloroquine, which have been used clinically for many years. Most of the selected molecules possess pleiotropic biological activities, acting on multiple targets in the body to produce diverse physiological and pharmacological activities. Furthermore, the significant differences in their chemical structures and physicochemical properties mean that they act on different sites within the molecular interaction network of living organisms. This combination allows them to potentially act on as many interaction sites / targets as possible within the molecular interaction network of tumor-bearing mice, thereby generating the desired positive perturbation of the disease. While some of the selected molecules have been shown to possess certain broad-spectrum anticancer activities in vitro and in vivo, these activities are insufficient for their sole use in the clinical treatment of cancer.

[0067] Oleanolic acid is very poorly soluble, but its solubility can be improved by adding alkali to convert it into a salt. Therefore, in terms of administration, oleanolic acid is dissolved in water / PEG400 solvent system with alkali and then administered alone, while the remaining 15 molecules are combined and dissolved together in water / PEG400 solvent system before administration. The dosage of each molecule is generally determined with reference to relevant pharmacological research literature, especially animal model pharmacology literature, including toxicity data, pharmacokinetic data, and dosage data. The dosage is generally set below, or even far below, the maximum safe dosage found in the literature. In addition, if it is an approved drug, the clinical dosage of the drug is also referenced, and the clinical dosage is converted to the dosage used in mouse models according to the human-mouse dosage conversion relationship. Many of the selected molecules have low absorption in the digestive system, so intraperitoneal injection is used to increase the bioavailability of these molecules in mice.

[0068] After selecting the aforementioned 16 molecules, the inventors initially considered them as a single 16-molecule combination and tested their anticancer activity in a Lewis lung cancer mouse model. They found that this combination achieved a tumor inhibition rate of 45-70% with no significant toxicity. However, this combination likely contained molecules that were ineffective or had antagonistic activities. Furthermore, due to the large number of molecules, this 16-molecule combination suffered from poor solubility, resulting in a suspension formulation that impacted clinical application. Therefore, the inventors used these 16 molecules as a starting set and employed the proposed multi-round scanning screening method, where molecules with the best overall anticancer activity / toxicity ratio were selected in each round of scanning experiments to proceed to the next round. Ultimately, through multiple rounds of scanning screening experiments, a five-molecule combination with good inhibitory effects on the Lewis lung cancer tumor model was discovered. The specific process of each round of molecular combination scanning screening is described below:

[0069] In previous studies, the inventors discovered that oleanolic acid (QD) exhibits certain anticancer activity against Lewis lung cancer in mice when used alone, with a tumor inhibition rate of approximately 0-45% in the Lewis tumor model within a safe dosage range. Therefore, the inventors skipped the first round of single-molecule scanning and instead started the second round of bimolecular combination anticancer activity scanning, using oleanolic acid (QD) as the starting point.

[0070] In the second round of bimolecular combination anticancer activity scanning experiments, oleanolic acid was combined with 15 other molecules to obtain 15 bimolecular combinations. The oleanolic acid monotherapy group served as the control group. The results are shown in Table 1. None of the 15 molecules produced significant toxicity after being combined with oleanolic acid, as evidenced by a general increase in mouse body weight during the experiment (body mass index > 1). Body mass index is the ratio of final body weight to initial body weight. Simultaneously, the average tumor weight after combining the 15 molecules with oleanolic acid was not significantly different from that of the oleanolic acid monotherapy group, indicating that these molecules, after being combined with oleanolic acid, either did not produce an additive / synergistic effect or only had a weak additive / synergistic effect.

[0071] The inventors believe that the lack of a significant additive / synergistic anticancer effect between these molecules and oleanolic acid may be due to the following reasons: 1) The sites on the molecular interaction network of tumor-bearing mice that these molecules act on are located in different regions of the molecular interaction network, and are not adjacent to each other. Therefore, no direct connection is established between them, resulting in no significant synergistic anticancer effect; 2) The physiological activity produced by these molecules after acting on the molecular interaction network itself does not have a significant anticancer effect, thus failing to produce a significant additive effect with the anticancer effect of oleanolic acid. The inventors further speculate that at least some of these molecular action sites that do not show a significant additive / synergistic effect under bimolecular combination conditions may be connected through some kind of "bridging" effect of the action sites of a third, fourth, or even more molecules, potentially producing an additive / synergistic anticancer effect.

[0072] Table 1: Results of the second round of bimolecular combinatorial scanning

[0073] QD 5 0 20.0±0.7 22.1±1.4 1.11 0.70±0.28 100.0±40.0 QD + RL 5 1 19.2±1.2 20.1±1.8 1.04 0.50±0.12 71.4±16.5 QD + FS 5 0 19.9±0.5 20.4±1.2 1.02 0.67±0.21 95.1±29.4 QD + MX 5 0 20.1±0.5 22.3±1.1 1.11 0.92±0.22 132.1±31.7 QD + JP 5 0 18.7±1.2 21.0±1.5 1.12 0.89±0.26 127.6±36.9 QD + YG 5 0 19.7±1.0 21.4±0.6 1.08 0.89±0.28 127.3±39.6 QD + EG 5 0 19.1±0.5 21.1±0.8 1.10 0.61±0.19 86.8±26.9 QD + BL 5 0 20.6±0.5 22.1±0.4 1.07 0.53±0.11 75.8±15.1 QD + DH 5 0 19.1±1.3 21.4±1.6 1.12 0.95±0.25 135.0±35.5 QD + JH 5 0 19.7±1.0 21.4±1.0 1.09 0.69±0.15 98.4±21.5 QD + ZS 5 0 19.3±0.5 20.4±0.9 1.06 0.70±0.16 100.4±22.3 QD + EJ 5 0 19.4±0.8 21.1±1.6 1.09 0.75±0.37 106.5±53.2 QD + QM 5 0 20.1±1.0 21.8±1.2 1.08 0.66±0.14 94.7±20.0 QD + AQ 5 0 19.9±0.4 20.9±1.2 1.05 0.61±0.23 87.6±32.8 QD + QL 5 0 19.3±0.7 20.5±0.5 1.06 0.64±0.15 91.9±22.1 QD + SL 5 0 19.6±0.6 21.4±0.6 1.09 0.60±0.28 86.3±39.7

[0074] Table 2: Results of the third round of trimolecular combination scanning

[0075] QD + BL 5 0 22.6±1.4 24.4±1.1 1.08 1.00±0.25 100.0±24.6 QD + BL + RL 5 0 19.8±0.5 21.6±0.6 1.09 0.90±0.20 90.2±19.5 QD + BL + FS 5 0 21.5±0.9 22.8±1.3 1.06 0.96±0.43 96.3±43.1 QD + BL + MX 5 1 21.6±1.6 24.8±1.8 1.14 1.20±0.40 119.9±39.5 QD + BL + JP 5 0 22.4±0.8 25.0±1.6 1.12 0.96±0.17 96.2±17.3 QD + BL + YG 5 0 21.4±1.6 23.4±2.2 1.09 0.76±0.29 76.0±29.3 QD + BL + EG 5 0 22.0±1.2 24.2±1.5 1.10 0.88±0.12 87.9±11.9 QD + BL + DH 5 0 21.9±0.9 24.7±1.6 1.13 1.00±0.24 100.4±24.0 QD + BL + JH 5 0 21.6±0.7 22.8±2.4 1.06 0.86±0.22 85.6±21.9 QD + BL + ZS 5 1 22.4±1.0 23.6±1.2 1.05 1.02±0.37 101.6±36.7 QD + BL + EJ 5 1 21.8±0.7 23.6±1.4 1.08 0.90±0.26 89.9±25.7 QD + BL + QM 5 0 22.0±0.8 24.0±1.5 1.09 0.98±0.43 97.8±43.1 <![CDATA[ Q D+BL+AQ]]> 5 0 21.5±1.3 22.8±1.7 1.06 0.65±0.37 65.4±36.8 QD + BL + QL 5 0 22.0±0.9 24.6±1.2 1.12 1.08±0.26 107.6±26.1 QD + BL + SL 5 0 21.7±1.4 23.9±0.7 1.11 1.28±0.35 127.5±34.6

[0076] Based on the above considerations, the inventors deemed it necessary to select one of the 15 bimolecular combinations for a third round of trimolecular combination scanning. Two bimolecular combinations were considered: one with genistein (RL) and the other with resveratrol (BL), each combined with oleanolic acid (QD). These two bimolecular combinations showed slightly better tumor inhibition rates than oleanolic acid alone, with relative tumor weights of 71.4% and 75.8% respectively. Considering mouse weight and mortality, the QD+BL group had a higher body mass index than the QD+RL group, and no mouse deaths were observed, indicating that QD+BL had lower toxicity than QD+RL. After comprehensive consideration, the inventors selected "oleanolic acid + resveratrol" as the starting point for the third round of trimolecular combination anticancer activity scanning.

[0077] In the third round of three-molecule combination anticancer activity scanning experiments, the "oleanolic acid + resveratrol" (QD+BL) bimolecule combination was used as the starting point, and combined with the remaining 14 molecules to obtain 14 three-molecule combinations. The QD+BL bimolecule combination administration group served as the control group. The anticancer activity and toxicity of these three-molecule combinations were tested, and the specific experimental results are shown in Table 2. None of the 14 three-molecule combinations showed significant toxicity, as evidenced by a general increase in mouse body weight during the experiment (body mass index > 1). The average tumor weight of most three-molecule combinations did not change significantly compared with the QD+BL group. However, the three-molecule combinations QD+BL+YG and QD+BL+AQ showed certain additive / synergistic effects compared with the QD+BL bimolecule combination, with the QD+BL+AQ group showing the most significant effect, with a relative tumor weight of 65.4%. Taking all factors into consideration, "oleanolic acid + resveratrol + azithromycin" (QD+BL+AQ) was selected as the starting point for the fourth round of four-molecule combination anticancer activity scanning.

[0078] Table 3: Results of the fourth round of tetramolecular combination scanning

[0079] Negative group 8 0 24.4±1.2 25.6±2.0 1.05 1.30±0.43 100.0±32.8 QD + BL + AQ 8 0 25.3±1.4 24.9±2.5 0.98 0.95±0.36 73.3±27.9 QD + BL + AQ + RL 7 0 24.9±1.2 25.3±1.7 1.02 1.00±0.48 76.7±37.2 QD + BL + AQ + FS 7 0 24.9±2.8 25.2±3.3 1.01 1.28±0.53 98.8±40.9 QD + BL + AQ + MX 7 0 25.3±1.1 25.7±1.3 1.02 0.95±0.19 73.0±14.5 QD + BL + AQ + JP 8 0 24.5±1.5 24.6±2.0 1.00 0.50±0.32 38.5±24.6 QD + BL + AQ + YG 8 0 24.7±1.4 24.2±1.5 0.98 0.33±0.32 25.7±24.8 QD + BL + AQ + EG 7 0 25.0±1.6 24.4±1.5 0.98 0.46±0.33 35.5±25.2 QD + BL + AQ + DH 8 0 24.1±1.7 23.5±1.7 0.98 0.89±0.55 68.5±42.6 QD + BL + AQ + JH 8 0 25.2±1.6 24.7±1.9 0.98 0.75±0.47 57.7±35.8 QD + BL + AQ + ZS 8 3 24.1±1.8 24.3±2.6 1.01 0.36±0.20 27.9±15.5 QD + BL + AQ + EJ 8 0 24.2±1.5 22.8±2.2 0.94 0.48±0.29 36.6±22.0 QD + BL + AQ + QM 8 0 24.4±1.0 23.7±1.8 0.97 0.55±0.27 42.3±20.7 QD + BL + AQ + QL 8 0 24.0±0.8 22.9±0.9 0.95 0.25±0.32 19.6±24.8 QD + BL + AQ + SL 8 0 23.9±0.8 23.2±0.9 0.97 0.16±0.23 12.6±17.7

[0080] In the fourth round of four-molecule combination anticancer activity scanning experiments, the three-molecule combination of "oleanolic acid + resveratrol + azithromycin" (QD+BL+AQ) was used as the starting point, and was combined with the remaining 13 molecules to obtain 13 four-molecule combinations. The QD+BL+AQ administration group served as the positive control group, and a solvent injection group was added as the negative control group. The anticancer activity and toxicity of these four-molecule combinations were tested, and the experimental results are shown in Table 3. Except for QD+BL+AQ+ZS, none of the tested combinations showed significant toxicity, with body mass index ranging from 0.94 to 1.05, and no mouse deaths. Among the 13 molecules, hesperidin (JP), isoliquiritigenin (YG), EGCG (EG), perillyl alcohol (ZS), metformin (EJ), hydroxychloroquine (QL), and celecoxib (SL) showed significantly enhanced anticancer activity after being combined with "oleanolic acid + resveratrol + azithromycin" (relative tumor weight <40% compared to the negative group). Hydroxychloroquine (QL) and celecoxib (SL) showed the most significant effects (relative tumor weight reduction of 19.6% and 12.6%, respectively), with minimal weight loss in both groups (not exceeding 5%). Subsequently, the anticancer activity of the "oleanolic acid + resveratrol + azithromycin + celecoxib" (QD+BL+AQ+SL) combination was tested twice, achieving relative tumor weight reductions of 138% and 58.6% relative to the negative control group, respectively—indicating that the anticancer effect of this molecular combination is unstable due to fluctuations in the tumor model's state. Tumors are complex systems with chaotic characteristics. Even for the same tumor model, the state of tumor modeling and its corresponding molecular interaction network often differ with random fluctuations in modeling experimental conditions, leading to varying responses to the same drug. Sometimes these differences can be very large—a phenomenon frequently reported in cancer research literature. To address this issue, the inventors believe that by building upon the discovered molecular combinations with unstable anti-cancer effects and adding more molecules, it may be possible to overcome or cover the random variability of tumor state and molecular interaction networks, thereby obtaining molecular combinations with more stable anti-cancer effects.

[0081] Therefore, given the unstable anticancer effect of the aforementioned QD+BL+AQ+SL combination, the inventors used this four-molecule combination as a starting point and combined it with the remaining 12 molecules to obtain 12 five-molecule combinations, conducting a fifth round of five-molecule combination anticancer activity scanning experiments. The experiment was conducted twice, with significant differences in the tumor model status between the two experiments. It was found that the five-molecule combination "oleanolic acid + resveratrol + azithromycin + celecoxib + metformin" (QD+BL+AQ+SL+EJ) exhibited relatively significant anticancer activity and low toxicity in both experiments, recording relative tumor weight increases of 31.7% and 48.2% compared to the negative control group, corresponding to tumor inhibition rates of 68.3% and 51.8%, respectively. Figure 1 ).

[0082] In summary, in this (prototype) embodiment, the molecular combination screening method described above, through four rounds of stepwise molecular superposition scanning, revealed a five-molecule combination "oleanolic acid + resveratrol + azithromycin + celecoxib + metformin" that stably demonstrated a high tumor inhibition rate in a mouse Lewis lung cancer model. This demonstrates the effectiveness and feasibility of the molecular combination screening method described in this invention. Theoretically, based on this five-molecule combination, further rounds of stepwise molecular superposition scanning can be performed to obtain molecular combinations with even higher anticancer activity.

[0083] This embodiment also demonstrates that the inventors did not consider the molecular mutation mechanism of Lewis lung cancer during the screening process of molecular combination drugs. This empirically proves that the method described in this invention does not require knowledge of the characteristics of the disease's molecular mechanism, and is therefore particularly suitable for drug development work for diseases with unclear mechanisms. For example, for recurrent tumors, the mechanism is often unclear and there are often no cures. However, if a suitable preclinical disease model can be established, this method can also be used to find effective molecular combination drug candidates.

[0084] The starting molecule set used in this embodiment contains only 16 molecules, a rather small number. However, even with such a small starting molecule set, a molecular combination exhibiting certain anticancer activity against Lewis lung cancer in mice was still found using the molecular combination screening method described in this invention. The inventors speculate that if the number of structurally diverse, low-toxicity active molecules in the starting molecule set is expanded to 50-100 (or even 200), such a starting molecule set might be used to screen various cancers, and there is a high probability of finding more effective molecular combination drug candidates.

[0085] The above description is merely a (prototype) embodiment of the drug development theory / method / mode described in this invention, and is only used to explain and illustrate this invention and related details, and should not be used to limit the spirit and scope of the essential technology of this invention. The essential technical spirit and content of this invention have been broadly defined within the scope of the claims. Any completed technical entity or method that is the same as defined in the scope of the claims of this invention, or that constitutes an equivalent modification, is considered to be covered within the scope of those claims. By reading the text of this invention, those skilled in the art can make various modifications to this invention, and all equivalent modifications or alterations made according to the spirit and essence of this invention also fall within the scope defined by the appended claims.

Claims

1. A multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular cascading, characterized in that, Includes the following steps: 1) Apply n molecules from a molecular set to a disease model, test the toxicity and efficacy of the n molecules, take into account both toxicity and efficacy, prioritize toxicity over efficacy, select the molecule with the best overall toxicity / efficacy, and select one molecule from the selected molecules to enter the second round of bimolecular combination scanning. The disease model described is an in vitro disease model, which may be a two-dimensional cell planar model, a three-dimensional cell spheroid model, an organoid model, an organ / tumor microarray model, or a tumor slice culture. 2) In the second round of bimolecular combination scanning, the molecules selected in the first round and the remaining n-1 molecules are combined to obtain n-1 bimolecular combinations. The toxicity and efficacy of the n-1 bimolecular combinations in the disease model are tested. The bimolecular combination with the best overall toxicity / efficacy is selected. One of the selected bimolecular combinations is selected to enter the next round of molecular combination scanning, i.e. the m-th round m-m-molecule combination scanning. 3) In the m-th round of m-molecule combination scanning, the m-1 molecule combination selected in the (m-1)-th round is combined with the remaining n-m+1 molecules to obtain n-m+1 m-molecule combinations. The toxicity and efficacy of the n-m+1 m-molecule combinations in the disease model are tested, and the m-molecule combinations with better overall toxicity / efficacy are selected. 4) Select 3 to n in sequence and repeat step 3). When the selected m molecular combination has an efficacy in the disease model that reaches or exceeds the set efficacy threshold, and the toxicity is equal to or lower than the set toxicity threshold, the molecular combination drug screening is completed. If the molecular combination drug screening is not completed, replace the molecules in the molecular set in step 1), return to step 1), and scan and screen again. Repeat this process until the molecular combination drug screening is completed.

2. The multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular-level superposition as described in claim 1, characterized in that, In step 1), considering both toxicity and efficacy, toxicity is prioritized over efficacy, and molecules with a better overall toxicity / efficacy ratio are selected. Specifically, these include: First, sort the molecules tested in step 1) from highest to lowest toxicity, and select the molecules with the lowest toxicity in the bottom 60%. Then sort the molecules with the lowest toxicity in the bottom 60% from highest to lowest efficacy, and select the molecules with the highest efficacy in the top 50%. This gives you the molecules with the best overall toxicity / efficacy.

3. The multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular-level superposition as described in claim 1, characterized in that, In step 2), bimolecular combinations with a good balance of toxicity and efficacy are selected, specifically including: First, sort the bimolecular combinations tested in step 2) from highest to lowest toxicity, and select the bimolecular combinations with the lowest toxicity in the bottom 60%. Then, sort the bimolecular combinations with the lowest toxicity in the bottom 60% from highest to lowest efficacy, and select the bimolecular combinations with the highest efficacy in the top 50%. This gives you the bimolecular combinations with the best overall toxicity / efficacy.

4. The multi-round scanning screening method for molecular combination drugs with progressively increasing therapeutic effects through molecular-level superposition as described in claim 1, characterized in that, In step 3), combinations of m molecules with a good balance of toxicity and therapeutic effect are selected, specifically including: First, sort the n-m+1 m-molecule combinations tested in step 3) according to their toxicity from highest to lowest, and select the m-molecule combinations with the lowest toxicity in the bottom 60%. Then, sort the m-molecule combinations with the lowest toxicity in the bottom 60% according to their efficacy from highest to lowest, and select the m-molecule combinations with the highest efficacy in the top 50%. This gives you the m-molecule combinations with the best overall toxicity / efficacy.

Images