Method for evaluating neurotoxicity of quinolone antibiotic environmental residues based on network toxicology

Abstract

The present invention provides a method for evaluating the neurotoxicity of environmental residues of quinolone antibiotics based on network toxicology. The present invention collects the types of quinolone drugs remaining in the environment and their drug targets, and collects neuropsychiatric diseases and their disease genes; establishes a drug target network, a neuropsychiatric disease module, and an overlapping network of drug target-disease modules, calculates network parameters to obtain key genes, uses the DAVID database to perform enrichment analysis on the key genes, and then performs network ranking analysis. Compared with traditional methods, the network toxicology of the present invention using big data high-throughput integration analysis technology has a wide coverage range of research objects, considers the toxicity of drugs from the perspectives of drug targets and disease genes, and provides a new idea for evaluating the risk size of adverse reactions of compounds.

Description

Technical Field

[0001] The present invention belongs to the field of pharmacy, relates to network toxicology, and particularly relates to a method for evaluating the neurotoxicity of quinolone antibiotic environmental residues based on network toxicology. Background Art

[0002] Neurotoxicity is an etiological phenomenon that alters the structure and function of the central nervous system through biological, chemical, or physical agents. Air pollutants, pesticides, insecticides, heavy metals, industrial wastes, etc. in the environment all have neurotoxicity. Diseases caused by neurotoxicity include Alzheimer's disease, Parkinson's disease, cognitive impairment, neurovascular diseases such as cerebral aneurysms, and neurodevelopmental diseases such as attention deficit hyperactivity disorder, intellectual disability, and autism spectrum disorder.

[0003] Quinolone antibiotics have a broad antibacterial spectrum and are considered a priority and extremely important class of antibiotics. However, the problem of the residue of quinolone drugs in the environment is becoming increasingly serious. The residues of quinolone drugs have been detected in water and soil in multiple regions, and even the presence of quinolone drugs has been detected in the urine of children and the elderly in some areas. It is considered that quinolone drugs are the primary factor causing health risks. There are even reports suggesting that quinolone drugs in the environment may enter the human body through the food chain and drinking water and have an adverse impact on the nervous system. Neurotoxicity has been clearly identified as one of the main adverse reactions of quinolone drugs, and the neurotoxicity of quinolone environmental residues cannot be ignored. The traditional methods for evaluating the neurotoxicity of quinolones by establishing animal models or cell models require a large amount of time and cost, and the occurrence and development of diseases are the result of the combined action of multiple biological processes. There is an urgent need to establish a new method for evaluating the neurotoxicity of quinolone environmental residues. Summary of the Invention

[0004] In view of the deficiencies that the existing methods for evaluating the neurotoxic subtypes of quinolone environmental residues are biased and fail to comprehensively consider the biological processes of diseases, the problem to be solved by the present invention is to provide a method for more comprehensively evaluating the neurotoxicity of quinolone environmental residues, so as to more efficiently and accurately determine the neurotoxic subtypes that quinolone drugs may cause and the biological processes related to diseases. The present invention provides a method for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology. The method includes collecting the types of quinolone drugs remaining in the environment and their drug targets, collecting neuropsychiatric diseases and their disease genes, establishing a drug target network, a neuropsychiatric disease module, and an overlapping network of drug target-disease modules, calculating network parameters to obtain key genes, using the DAVID database to perform enrichment analysis on the key genes, evaluating the proximity between quinolone drugs and neuropsychiatric disease genes by the network proximity method, evaluating the functional similarity between quinolone drug targets and disease genes by the network diffusion method, comprehensively ranking the results of the two, evaluating the neurotoxicity risk of quinolone drugs, and selecting important disease genes in the overlapping network for enrichment analysis to analyze the potential neurotoxic mechanism of quinolone drugs. By revealing the neurotoxic subtypes and the mechanisms of adverse reactions that quinolone environmental residues may cause, the present invention provides a theoretical reference for evaluating the health risks of quinolone environmental residues.

[0005] To solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for evaluating the neurotoxicity of quinolone antibiotic environmental residues based on network toxicology, comprising the following steps:

[0007] (a) Obtaining of quinolone drug targets and establishment of a target network

[0008] Determine the types of quinolone drugs according to the literature, collect drug targets using an online database, and distribute the collected drug targets into the protein-protein interaction group (PPI) to establish a drug target network;

[0009] (b) Determination of neuropsychiatric diseases, obtaining of disease genes and establishment of a neuropsychiatric disease module

[0010] Determine neuropsychiatric diseases according to the literature, obtain the medical subject headings corresponding to neuropsychiatric diseases in the MeSh database, use the MeSh subject headings as search terms to collect the corresponding Entrez Gene IDs in the OMIM database, and obtain the corresponding gene names after conversion by the UniProt database. The disease genes of each neuropsychiatric disease collected are distributed into the PPI network to construct the corresponding neurotoxic disease module;

[0011] (c) Calculation of network proximity

[0012] First, calculate the shortest path length \(d_c(V, T)\) between the quinolone drug target group (T) and the neuropsychiatric disease genome (V) in the PPI network according to formula (1); as shown in formula (1):

[0013]

[0014] Generate a reference distance distribution by calculating the proximity between two random target groups and genomes. When converting the observed distance to a normalized distance using the mean \(\mu_{d_c(V, T)}\) and standard deviation \(\sigma_{d_c(V, T)}\) of the reference distribution, the relative average shortest distance \(Z\) between T and V dc Calculate according to formula (2):

[0015]

[0016] If \(Z\) dc < 0, it is considered that the quinolone drug target is adjacent to the neuropsychiatric disease gene;

[0017] (d) Calculation of network diffusion distance

[0018] First, calculate the diffusion state distance (DSD) between interacting genes in the PPI network, and then calculate the DSD(\(t, v\)) between the quinolone target (\(t\)) and the neuropsychiatric disease gene (\(v\)). The smaller the DSD(\(t, v\)), the higher the similarity between \(t\) and \(v\) in the human interactome, is the average minimum DSD between drug targets and disease genes in the PPI network, used to describe the influence of drug targets on disease genes; The smaller it is, the more significant the influence of quinolone drugs on neuropsychiatric diseases; The calculation formula of is as follows:

[0019]

[0020] (e) Comprehensive ranking

[0021] Take the arithmetic mean of the network proximity result and the network diffusion result as the comprehensive value Use the comprehensive value as the final measure to evaluate the risk of quinolone drugs on neuropsychiatric diseases;

[0022] (f) Network overlap

[0023] Overlap the constructed drug target network and the neuropsychiatric disease module based on the PPI network, and perform enrichment analysis on disease genes that completely overlap or are adjacent to drug targets.

[0024] The online prediction database of drug targets described in step (a) of the present invention is selected from SuperPred; the protein database is selected from Uniprot; the correction for matching relevant gene symbols is the Uniprot ID; the sorted target names are converted into gene names to obtain drug targets; the sorted drug targets are processed by the online tool jvenn for mapping intersection; the human interactome is compiled from 21 public databases, which compile experimentally derived protein-protein interaction (PPI) data; all proteins are mapped to the corresponding Entrez ID, and proteins that cannot be located are removed.

[0025] Further, the acquisition of quinolone drug targets and the establishment of the target network in step (a) above are as follows: The quinolone drugs are determined according to the literature, and the UniProtID of the predicted targets of quinolone drugs is obtained from the SuperPred database. After conversion by the UniProt database, the corresponding gene names are obtained. The collected drug targets are distributed into the PPI network to establish a drug target network. The nodes in the target network are drug targets, and the associations between two nodes are determined according to the known connections between proteins in the PPI network to form the edges of the drug target network.

[0026] The medical subject heading database described in step (b) of the present invention is selected from MeSh; the gene database is selected from OMIM; the correction for matching relevant gene symbols is the Entrez Gene ID, and the disease genes are obtained through transformation and sorting. The distance between disease genes in the PPI network is measured by the network parameter ds (shortest distance). The disease genes in the disease module are uploaded to the DAVID database for GO enrichment analysis to obtain the enrichment results of biological process (BP), cellular component (CC), molecular function (MF), and KEGG pathway analysis.

[0027] Further, for the determination of the above-mentioned neuropsychiatric diseases in step (b), the acquisition of disease genes and the establishment of the neuropsychiatric disease module, medical subject headings corresponding to the neuropsychiatric diseases were obtained from the MeSh database. Using the MeSh subject headings as retrieval entries, the corresponding Entrez Gene IDs were collected from the OMIM database. After conversion through the UniProt database, the corresponding gene names were obtained. The disease genes of each neuropsychiatric disease collected were distributed into the PPI network to construct the corresponding neurotoxic disease module; the shortest path connecting two genes in the PPI was calculated and represented by ds. At the same time, the largest connected component (LCC) was calculated, which consisted of S continuously connected disease genes with ds = 1 and was a subset of interconnected genes (i.e., genes with ds = 1). Enrichment analysis was performed on the disease genes in the disease module: (1) those located in the PPI, (2) those with ds = 1, and (3) those in the LCC.

[0028] The disease genes that completely overlap or are adjacent to the drug targets described in step (f) refer to the disease genes in the overlapping network with the shortest distance to the drug targets less than or equal to 1 (i.e., ds ≤ 1).

[0029] Further, the above-mentioned network overlap in step (f) is the overlap of the quinolone drug target network and the neuropsychiatric disease module based on the PPI. The network parameter ds is used to measure the distance between the disease genes and the nearest drug target in the PPI. ds = 0 indicates that the disease gene completely overlaps with the drug target, ds = 1 indicates that the disease gene is adjacent to the drug target, ds = 2 indicates that the shortest distance between the disease gene and the drug target is 2, ds > 2 indicates that the shortest distance between the disease gene and the drug target is more than 2 steps, and ds = NA indicates that there is no association between the disease gene and the drug target; genes with ds ≤ 1 in the overlapping network were selected for gene enrichment analysis using the DAVID database.

[0030] The above-mentioned quinolone drugs are ciprofloxacin, danofloxacin, difloxacin, enoxacin, enrofloxacin, flumequine, norfloxacin, ofloxacin, pefloxacin, and sarafloxacin.

[0031] The above-mentioned neuropsychiatric diseases are Alzheimer's disease (AD), Autistic Spectrum Disorder (ASD), Depression, Epilepsy, Parkinson's Disease (PD), and Schizophrenia.

[0032] The present invention has the following beneficial effects and advantages:

[0033] 1. The method for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology according to the present invention belongs to the category of systems bioinformatics. Through data mining and analysis, relying on the known protein-protein interactionome, a drug target network, a disease module, an overlapping network, and network ranking analysis are constructed. Compared with traditional methods, the network toxicology of the present invention using big data high-throughput integration analysis technology has a wide coverage of research objects, considers the toxicity of drugs from the perspectives of drug targets and disease genes, and provides a new idea for evaluating the risk of adverse reactions of compounds.

[0034] 2. The risk ranking method based on the PPI network proposed by the present invention not only considers the proximity of quinolone drug targets and neuropsychiatric disease genes in the PPI, but also considers the functional similarity or influence on each other of the targets and genes.

[0035] 3. Starting from disease genes and drug targets, the present invention considers the interconnections between disease genes and between disease genes and drug targets, rather than simply considering these genes in isolation, providing insights into the comprehensive nature of the human pathogenesis mechanism.

[0036] 4. The method for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology according to the present invention first uses the network toxicology method to evaluate the neurotoxicity of quinolones, reveals the neurotoxic subtypes more likely to be caused by quinolone residues in the environment, and explains the related mechanisms, providing a theoretical reference for evaluating the health risks of quinolone residues in the environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Flowchart of the method for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology according to the present invention;

[0038] Figure 2 Differential analysis of genes for 6 neuropsychiatric diseases;

[0039] Figure 3 Top 10 GO biological processes of the gene enrichment results of the depression-difloxacin overlapping network with ds≤1 (each 10 from top to bottom is a group, and the three groups are BP, CC, and MF respectively);

[0040] Figure 4 Top 10 pathways of the gene enrichment results of the depression-difloxacin overlapping network with ds≤1 (the signal pathway is on the left and the corresponding genes of the signal pathway are on the right);

[0041] Figure 5 Top 10 GO biological processes of the gene enrichment results of the Parkinson-ciprofloxacin overlapping network with ds≤1 (each 10 from top to bottom is a group, and the three groups are BP, CC, and MF respectively);

[0042] Figure 6 Top 10 pathways of gene enrichment results of Parkinson-ciprofloxacin overlapping network with ds≤1 (the signal pathway is on the left and the genes corresponding to the signal pathway are on the right);

[0043] Figure 7 Top 10 GO biological processes of gene enrichment results of schizophrenia-sarafloxacin overlapping network with ds≤1 (every 10 items from top to bottom form a group, and the three groups are BP, CC, and MF respectively)

[0044] Figure 8 Top 10 pathways of gene enrichment results of schizophrenia-sarafloxacin overlapping network with ds≤1 (the signal pathway is on the left and the genes corresponding to the signal pathway are on the right). Detailed implementation manners

[0045] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described and illustrated below through specific embodiments in combination with the attached drawings. The descriptions of the following embodiments are only used to help understand the method and its core idea of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements and modifications can still be made to the present invention, and these improvements and modifications also fall within the protection scope of the claims of the present invention. The method flow of the present invention for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology is as Figure 1 shown.

[0046] Embodiment 1 Method for evaluating the neurotoxicity of quinolone environmental residues based on network toxicology

[0047] 1. Method steps

[0048] 1.1 Determination of quinolone drugs, acquisition of targets and establishment of target network

[0049] According to the literature on the residues of quinolone antibiotics in the research environment, 10 quinolone drugs were determined. The UniProtIDs of the predicted targets of the 10 quinolone drugs were obtained from the SuperPred database, and the corresponding gene names were obtained after conversion through the UniProt database. The collected drug targets were distributed into the PPI network to establish a drug target network. The nodes in the target network are drug targets, and the associations between two nodes are determined according to the known connections between proteins in the PPI network to form the edges of the drug target network.

[0050] 1.2 Determination of neuropsychiatric diseases, acquisition of disease genes and establishment of neuropsychiatric disease modules

[0051] Six neuropsychiatric diseases were identified based on the literature on the neurotoxicity of environmental pollutants. The medical subject headings corresponding to the six neuropsychiatric diseases were obtained from the MeSh database. Using the MeSh subject headings as search entries, the corresponding Entrez Gene IDs were collected from the OMIM database. After conversion through the UniProt database, the corresponding gene names were obtained. The disease genes of each neuropsychiatric disease collected were distributed into the PPI network to construct the corresponding neurotoxicity disease module. The shortest path connecting two genes in the PPI was calculated and denoted as ds. At the same time, the largest connected component (LCC) was calculated, which consists of S continuously connected disease genes with ds = 1 and is a subset of the interconnected genes (i.e., genes with ds = 1). Enrichment analysis was performed on the disease genes (1) located in the PPI, (2) with ds = 1, and (3) in the LCC in the disease module.

[0052] 1.3 Calculation of network proximity

[0053] The network proximity method was used to calculate the proximity between neuropsychiatric diseases and quinolone drugs in the PPI network. First, the shortest path length dc(V,T) between the quinolone drug target group (T) and the neuropsychiatric disease genome (V) in the PPI network was calculated. As shown in formula (1):

[0054]

[0055] The reference distance distribution was generated by calculating the proximity between two random target groups and genomes, and this process was repeated 1000 times. To define the proximity metric and avoid repeatedly selecting the same highly connected nodes, when converting the observed distance to a normalized distance using the mean μdc(V,T) and standard deviation σdc(V,T) of the reference distribution, the relative average shortest distance Z between T and V dc was calculated according to formula 2:

[0056]

[0057] If Z dc < 0, it is considered that the quinolone drug target is adjacent to the neuropsychiatric disease gene.

[0058] 1.4 Calculation of network diffusion distance

[0059] The network diffusion method was used to calculate the diffusion state distance (DSD) to evaluate the functional similarity between drug targets and disease genes in the PPI network. First, the DSD values between interacting genes in the PPI network were calculated. On this premise, the DSD(t,v) between the quinolone target (t) and the neuropsychiatric disease gene (v) was calculated. The smaller the DSD(t,v), the higher the similarity between t and v in the human interactome. It is the average minimum DSD of drug targets and disease genes in the PPI network, used to describe the impact of drug targets on disease genes. The smaller it is, the more significant the impact of quinolone drugs on neuropsychiatric diseases. The calculation formula of it is as follows:

[0060]

[0061] 1.5 Comprehensive ranking

[0062] Take the arithmetic mean of the network neighborhood results and the network diffusion results as the comprehensive value Use the comprehensive value as the final metric to evaluate the risk of quinolone drugs on neuropsychiatric diseases.

[0063] 1.6 Network overlap

[0064] The quinolone drug target network and the neuropsychiatric disease module overlap based on PPI. The network parameter ds is used to measure the distance between the disease gene and the nearest drug target in the PPI. ds = 0 indicates that the disease gene and the drug target completely overlap, ds = 1 indicates that the disease gene and the drug target are adjacent, ds = 2 indicates that the shortest distance between the disease gene and the drug target is 2, ds > 2 indicates that the shortest distance between the disease gene and the drug target is more than 2 steps, and ds = NA indicates that there is no association between the disease gene and the drug target. Genes with ds ≤ 1 in the overlapping network are selected for gene enrichment analysis using the DAVID database.

[0065] 2. Results

[0066] 2.1 Quinolone drug targets and target networks

[0067] Ten quinolone drugs and the number of their targets are shown in Table 1. Through intersection mapping analysis, the ten quinolone drugs have a total of 53 targets. The collected drug targets are distributed into the PPI network to form a drug target network, and the number of targets contained in each target network is shown in Table 1.

[0068] Table 1 The number of ten quinolone drugs and their targets, and the number of targets in the target network

[0069]

[0070]

[0071] 2.2 Neuropsychiatric disease genes and disease modules

[0072] The inventors identified six neuropsychiatric diseases: Alzheimer's disease (AD), Autistic Spectrum Disorder (ASD), also known as autism, Depression, Epilepsy, Parkinson's Disease (PD), and Schizophrenia. The number of genes collected for each disease is shown in Table 2. Through intersection mapping analysis, it was found that two disease genes, APOE and BDNF, exist in five diseases except ASD, as Figure 2 shown. The disease genes collected were distributed in the PPI network to construct corresponding neuropsychiatric disease modules. The gene localization of each disease is shown in Table 2.

[0073] Table 2 Disease genes of six neuropsychiatric diseases and their localization in PPI

[0074]

[0075] After gene enrichment, in Alzheimer's disease, biological processes such as beta-amyloid formation (β-amyloid formation, BP) and tau protein binding (tau protein binding, MF) were obtained. In autism, biological processes such as regulation of NMDA receptor activity (NMDA receptor activity regulation, BP), neuroligin family protein binding (neuroligin family protein binding, MF), and glutamatergic synapse (glutamatergic synapse, Pathway) were obtained. In depression, biological processes such as ionotropic glutamate receptor signaling pathway (ionotropic glutamate receptor signaling pathway, BP), serotonin binding (serotonin binding, MF), and dopaminergic synapse (dopaminergic synapse, pathway) were obtained. In the epilepsy disease module, biological processes such as neuronal action potential (neuronal action potential, BP) and voltage-gated ion channel activity (voltage-gated ion channel activity, MF) were obtained. In the Parkinson's disease module, biological processes such as cellular response to oxidative stress (cellular response to oxidative stress, BP), ubiquitin protein ligase binding (ubiquitin protein ligase binding, MF), and dopaminergic synapse (dopaminergic synapse, Pathway) were obtained. In the schizophrenia disease module, biological processes such as G-protein coupled glutamate receptor signaling pathway (G-protein coupled glutamate receptor signaling pathway, BP), dopamine neurotransmitter receptor activity, coupled via Gi / Go (dopamine neurotransmitter receptor activity, coupled via Gi / Go, MF), and dopaminergic synapse (dopaminergic synapse, Pathway) were obtained. Differential analysis was performed on the enrichment results of the three parts of the disease module to obtain the gene enrichment results (specific enrichment results) that only exist in ds = 1 and LCC.

[0076] 2.3 Network proximity

[0077] In the proximity analysis based on the PPI network, the relative average proximity distance Z between quinolone drug targets and neuropsychiatric disease genes was calculated. dc , if Z dc is negative, it represents that the drug is close to the disease gene. The network proximity results of 10 drugs and 6 neuropsychiatric diseases are shown in Table 3 (rounded to 4 decimal places).

[0078] Table 3 Network proximity results of 6 neuropsychiatric diseases and 10 quinolone drugs

[0079]

[0080] The proximity degrees of 10 quinolone drugs to Alzheimer's disease, depression, epilepsy, Parkinson's disease, and schizophrenia are all negative, indicating that quinolone drugs are relatively close to the genes of these 5 types of neuropsychiatric diseases. However, the proximity degree to autism is around 0, suggesting a weak correlation between quinolone drugs and autism.

[0081] 2.4 Network diffusion

[0082] Network diffusion is used to calculate the functional similarity between quinolone drug targets and neuropsychiatric disease genes in the PPI network or the mutual influence between drug targets and disease genes. , the smaller the value, the higher the functional similarity between the target and the gene and the greater the impact on the rest of the network. The network diffusion results of 10 quinolone drugs and 6 neuropsychiatric diseases are shown in Table 4. The network diffusion value of sarafloxacin and autism is the smallest, which is 5.277, and the network diffusion value of Alzheimer's disease and difloxacin is the largest, which is 6.148.

[0083] Table 4 Network diffusion results of 6 neuropsychiatric diseases and 10 quinolone drugs

[0084]

[0085]

[0086] 2.5 Comprehensive ranking of the impact of quinolone drugs on neurotoxicity

[0087] Based on the arithmetic mean of the network proximity value and the network diffusion value as the comprehensive value, the comprehensive value is used as the final metric to evaluate the risk of quinolone drugs to different subtypes of neuropsychiatric diseases. The smaller the comprehensive value, the greater the risk. The comprehensive values of 10 quinolone drugs and 6 neuropsychiatric diseases are shown in Table 5.

[0088] Table 5 Comprehensive values of 6 neuropsychiatric diseases and 10 quinolone drugs

[0089]

[0090] Depression, Parkinson's disease, and schizophrenia always ranked among the top three in the ranking of 10 quinolone drugs. Therefore, based on the comprehensive value, it is considered that quinolone drugs have a relatively high risk for depression, Parkinson's disease, and schizophrenia. Among them, depression - difloxacin, Parkinson's disease - ciprofloxacin, and schizophrenia - sarafloxacin have the smallest comprehensive values among the three diseases.

[0091] 2.6 Network overlap

[0092] Based on the network ranking results, an overlapping network of depression - difloxacin, Parkinson's disease - ciprofloxacin, and schizophrenia - sarafloxacin was established. Gene enrichment analysis was performed on genes with ds ≤ 1 in the overlapping network. The top 10 results of the enrichment analysis are as follows Figures 3 - 7As shown. In the enrichment results of the depression-ciprofloxacin overlapping network, biological processes such as long-term synaptic potentiation, dopamine metabolic process (BP), glutamatergic synapse (CC), serotonin binding, dopamine neurotransmitter receptor activity, glutamate-gated calcium ion channel activity (MF), dopaminergic synapse, serotonergic synapse (Pathway), etc. were obtained. In the enrichment results of the Parkinson's-ciprofloxacin overlapping network, biological processes such as negative regulation of neuron death, response to oxidative stress (BP), protein kinase binding, tau protein binding (MF), JAK-ST-AT signaling pathway, PI3K-Akt signaling pathway (Pathway), etc. were obtained. In the enrichment results of the schizophrenia-sarafloxacin overlapping network, biological processes such as regulation of neuron projection development, dopamine metabolic process (BP), glutamatergic synapse (CC), serotonin binding, dopamine neurotransmitter receptor activity (MF), dopaminergic synapse, calcium signaling pathway (Pathway), etc. were obtained.

[0093] As can be seen from the network sorting results, the effects of 10 quinolone drugs on depression, Parkinson's disease, and schizophrenia are relatively obvious. The combined value of depression and 6 quinolone drugs is negative, and the combined value with danofloxacin ranks second and is negative. The combined value with enoxacin is the largest and positive; Parkinson's disease has a unique negative combined value with ciprofloxacin, the combined value with danofloxacin ranks second, and the combined value with enoxacin is the largest; the combined values of schizophrenia and quinolone drugs are all positive, the combined value with danofloxacin ranks second, and the combined values with norfloxacin, ofloxacin, enrofloxacin, flumequine, and enoxacin are all greater than 0.7. At the same time, the aggregation value rankings of ciprofloxacin with Alzheimer's disease, epilepsy, and schizophrenia are also relatively significant. To verify the sorting results, the inventors analyzed the differences in drug targets of danofloxacin and enoxacin, and ciprofloxacin and enoxacin located in the PPI. The specific drug targets of danofloxacin and enoxacin were respectively compared with the genes corresponding to the specific enrichment results of depression, the specific drug targets of ciprofloxacin and enoxacin were respectively compared with the genes corresponding to the specific enrichment results of Parkinson's disease, and the specific drug targets of danofloxacin and enoxacin were respectively compared with the genes corresponding to the specific gene enrichment results of schizophrenia.

[0094] After comparison, the genes coinciding between the specific target of danofloxacin and the genes corresponding to the specific enrichment results of depression are DRD2 and CYP3A4. DRD2 is the target of the other 9 quinolone drugs except enoxacin. There is already literature proving the relationship between DRD2 and depression and schizophrenia. There is no coincidence between the specific drug target of enoxacin and the genes corresponding to the specific enrichment results of depression.

[0095] SLC6A3, GABRA1, and SERPINE1 are the genes coinciding between the specific drug target of ciprofloxacin and the genes corresponding to the specific enrichment results of Parkinson's disease. GABRA1 is the specific target of both ciprofloxacin and danofloxacin. GABRA1 mediates GABA transmission and participates in the formation of GABAergic synapses. GABAergic dysfunction may be related to the pathogenesis of Parkinson's disease. The gene PTGS2, which coincides between the specific drug target of enoxacin and the specific gene enrichment results of Parkinson's disease, is the specific drug target of enoxacin and has no obvious association with Parkinson's disease.

[0096] DRD2, DRD3, SLC6A3, and GRIN1 are genes that overlap between the specific targets of danofloxacin and the genes corresponding to the specific enrichment results of schizophrenia. DRD3 is a drug target shared by the other 6 drugs except enoxacin, flumequine, norfloxacin, and sarafloxacin. SLC6A3 is a target shared by 5 drugs, namely ciprofloxacin, danofloxacin, difloxacin, enrofloxacin, and sarafloxacin. GRIN1 is a target of the other 9 drugs except enoxacin. DRD2, DRD3, and GRIN1 are related to schizophrenia. There is no overlap between the specific drug targets of enoxacin and the genes corresponding to the specific enrichment results of schizophrenia.

[0097] From the above analysis results, it can be seen that there is a negative comprehensive value between depression and Parkinson's disease and quinolone drugs. The specific drug targets obtained are consistent with the sorting results. However, for schizophrenia, since there is no negative value in the comprehensive value between it and quinolone drugs, the specific drug targets obtained deviate from the sorting results, but there is a certain degree of discrimination, indicating that the network sorting results have a certain degree of reliability.

[0098] According to the enrichment results of network overlap, dopamine, 5-hydroxytryptamine, and glutamate, as key neurotransmitters in the brain, play key roles in various mental disorders and neurodegenerative diseases. Anhedonia is a core symptom of depression and involves the downregulation of the dopamine system. The tight molecular interactions between 5-hydroxytryptamine and 5-hydroxytryptamine receptors, nerve growth factors, synaptic plasticity regulatory proteins, and neurogenesis may be the cause of impaired neural plasticity in depression. The pathological accumulation of glutamatergic neurotransmission leads to a decrease in the number of glutamate receptors, dendritic retraction, and even spinal cord injury. And neuronal atrophy is a potential anatomical pathophysiological change in the nervous system in depression.

[0099] Parkinson's disease is a progressive neurodegenerative movement disorder mainly caused by the death of dopaminergic neurons in the substantia nigra pars compacta. In addition, oxidative stress, impaired intracellular Ca 2+ homeostasis, mitochondrial dysfunction, and abnormal protein phosphorylation can also lead to the disease. The dysregulation of the dopamine system is an integral part of the pathophysiology of schizophrenia. The neuropathological changes in schizophrenia include an increase in striatal dopamine levels and an increase in the density of dopamine receptor 2. Drugs that increase dopamine can cause positive symptoms of schizophrenia, and antipsychotic drugs act by blocking dopamine receptor 2. The morphology of the dendrites of glutamatergic neurons in the cerebral cortex of schizophrenia patients has changed, and the level of the axonal synaptic protein marker synaptophysin has decreased.

[0100] Network toxicology can be used to evaluate the effects of quinolone environmental residues on different subtypes of neurotoxicity while exploring the disease-related biological processes affected by quinolone drugs. The present invention for the first time reveals that quinolone environmental residues may more easily lead to depression, Parkinson's disease, and schizophrenia by affecting biological processes such as dopamine, serotonin, and glutamate. Among them, the associations of difloxacin with depression, ciprofloxacin with Parkinson's disease, and sarafloxacin with schizophrenia are the most significant. The present invention provides new ideas for evaluating the risk of adverse reactions of compounds and also provides a theoretical reference for evaluating the neurotoxicity of quinolone environmental residues.

Claims

1. Method for evaluating the neurotoxicity of quinolone antibiotic environmental residues based on network toxicology, comprising the following steps: (a) Acquisition of quinolone drug targets and establishment of a target network Determine the types of quinolone drugs according to the literature, collect drug targets using an online database, and distribute the collected drug targets into the protein-protein interaction group (PPI) to establish a drug target network; (b) Determination of neuropsychiatric diseases, acquisition of disease genes, and establishment of neuropsychiatric disease modules Determine neuropsychiatric diseases according to the literature, obtain the medical subject headings corresponding to neuropsychiatric diseases in the MeSh database, use the MeSh subject headings as search terms to collect the corresponding Entrez Gene IDs in the OMIM database, and obtain the corresponding gene names after conversion by the UniProt database. The disease genes of each neuropsychiatric disease collected are distributed into the PPI network to construct the corresponding neurotoxicity disease module; (c) Calculation of network proximity First, calculate the shortest path length dc(V,T) between the quinolone drug target group (T) and the neuropsychiatric disease genome (V) in the PPI network according to formula (1); as shown in formula (1): Generate a reference distance distribution by calculating the proximity between two random target groups and the genome. When converting the observed distances to normalized distances using the mean μdc(V,T) and standard deviation σdc(V,T) of the reference distribution, the relative average shortest distance Z between T and V dc is calculated according to formula (2): If Z dc < 0, it is considered that the quinolone drug target is adjacent to the neuropsychiatric disease gene; (d) Calculation of network diffusion distance First, the DSD values between interacting genes in the PPI network were calculated, and then the diffusion distance DSD(t, v) between the quinolone target (t) and the neuropsychiatric disease gene (v) was calculated. The smaller the DSD(t, v), the higher the similarity between t and v in the human interactome. is the average minimum DSD of drug targets and disease genes in the PPI network, which is used to describe the impact of drug targets on disease genes; The smaller it is, the more significant the impact of quinolone drugs on neuropsychiatric diseases; The calculation formula of is as follows: (e) Comprehensive ranking Take the arithmetic mean of the network proximity and the network diffusion distance as the comprehensive value, and use the comprehensive value as the final measure to evaluate the risk of quinolone drugs to neuropsychiatric diseases; (f) Network overlap The quinolone drug target network and the neuropsychiatric disease module overlap based on the PPI. The network parameter ds is used to measure the distance between the disease gene and the nearest drug target in the PPI. ds = 0 indicates that the disease gene and the drug target completely overlap, ds = 1 indicates that the disease gene and the drug target are adjacent, ds = 2 indicates that the shortest distance between the disease gene and the drug target is 2, ds > 2 indicates that the shortest distance between the disease gene and the drug target is more than 2 steps, and ds = NA indicates that there is no association between the disease gene and the drug target; Select genes with ds ≤ 1 in the overlapping network and perform gene enrichment analysis using the DAVID database.

2. The method according to claim 1, wherein: The step (a) of acquiring quinolone drug targets and establishing a target network is to determine quinolone drugs according to the literature, obtain the UniProtID of the predicted targets of quinolone drugs in the SuperPred database, obtain the corresponding gene names after conversion by the UniProt database, distribute the collected drug targets into the PPI network to establish a drug target network, the nodes in the target network are drug targets, and the association between two nodes is determined according to the known connections between proteins in the PPI network to form the edge of the drug target network.

3. The method according to claim 1, characterized in that: The determination of the neuropsychiatric disease in step (b), the acquisition of disease genes, and the establishment of the neuropsychiatric disease module are as follows: Obtain the medical subject headings corresponding to the neuropsychiatric disease in the MeSh database, use the MeSh subject headings as retrieval entries to collect the corresponding Entrez Gene IDs in the OMIM database, and obtain the corresponding gene names after conversion through the UniProt database. The disease genes of each neuropsychiatric disease collected are distributed into the PPI network to construct the corresponding neurotoxic disease module; calculate the shortest path connecting two genes in the PPI and denote it as ds, and at the same time calculate the largest connected component LCC, which consists of S continuously connected disease genes and is a subset of interconnected genes, that is, the genes with ds = 1. Perform enrichment analysis on the disease genes in the disease module: (1) disease genes located in the PPI, (2) disease genes with ds = 1, and (3) disease genes in the LCC.

4. The method according to any one of claims 1 to 3, characterized in that: The quinolone drugs are ciprofloxacin, danofloxacin, difloxacin, enoxacin, enrofloxacin, flumequine, norfloxacin, ofloxacin, pefloxacin, and sarafloxacin.

5. The method according to claim 4, wherein: The neuropsychiatric diseases are Alzheimer's disease, autism spectrum disorder, depression, epilepsy, Parkinson's disease, and schizophrenia.

Images