A method, kit and application for evaluating the risk of heart toxicity of processed aconite based on NFKB1, PSMB1 and FGR markers

Abstract

This invention relates to a method, kit, and application for assessing the cardiotoxicity risk of processed Aconitum carmichaelii products based on NFKB1, PSMB1, and FGR biomarkers, belonging to the field of traditional Chinese medicine processing technology. The invention provides a biomarker specifically any one of the NFKB1, PSMB1, or FGR genes, or any combination thereof. This invention is the first to identify the key causal biomarkers NFKB1, PSMB1, and FGR that cause cardiotoxicity from Aconitum carmichaelii toxic components. A dynamic response model composed of these three genes can accurately quantify the degree of toxicity transformation during Aconitum carmichaelii processing: elevated NFKB1 expression indicates a risk of inflammatory activation, decreased PSMB1 expression reflects the critical point of protein homeostasis collapse, and persistent inhibition of FGR expression indicates the failure of endogenous cardioprotective mechanisms. The predictive model constructed using these three genes can accurately assess the cardiotoxicity of processed Aconitum carmichaelii products and evaluate the Aconitum carmichaelii processing technology.

Description

Technical Field

[0001] This invention relates to the field of traditional Chinese medicine processing technology, and in particular to a method, kit, and application for assessing the cardiotoxicity risk of processed Aconitum carmichaelii products based on NFKB1, PSMB1, and FGR markers. Background Technology

[0002] Aconite (ACONITI RADIX LATERALIS PRAEPARATA) is a processed root of the plant Aconitum carmichaelii Debx., belonging to the Ranunculaceae family. It is a key herb in Traditional Chinese Medicine (TCM) for treating heart failure (HF), possessing the function of "restoring yang and rescuing from collapse." However, aconite has a narrow therapeutic window, and its diester-type alkaloids (such as aconitine, mesonacetin, and 10-hydroxymesonacetin) have strong cardiotoxicity.

[0003] Currently, the toxicity assessment and quality control of Aconitum carmichaelii and its processed products mainly rely on quantitative analysis of chemical components. This involves using instruments such as high-performance liquid chromatography (HPLC) to detect the content of diester-type alkaloids (toxic components) in the sample. The current Chinese Pharmacopoeia and industry standards use this as the primary basis for determining whether Aconitum carmichaelii has passed the "toxicity reduction test." However, existing technologies neglect the differences in the metabolic responses of organisms to toxic components, making it difficult to accurately reflect their toxicological risks in vivo. Relying solely on static assessments of chemical component content cannot characterize the dynamic biological processes of toxic components, such as liver metabolism, intestinal flora transformation, and target organ responses, thus presenting the following limitations: (1) The "dosage-toxicity" relationship is not completely equal: Measuring only the content of chemical components cannot fully reflect the true toxicological effects of drugs in organisms. Even if the content of alkaloids decreases, the toxic effects (or molecular perturbations) of the drug at specific biological targets still exist.

[0004] (2) Lack of biomarker indicators: Existing chemical detection is a physicochemical indicator, lacking toxicity evaluation standards based on biological mechanisms (such as gene expression and signaling pathway perturbation), and cannot explain the "toxic mechanism" and "precision detoxification principle".

[0005] (3) Inability to assess individual genetic susceptibility: Existing technology does not take into account the heterogeneous influence of human genetic background on toxic reactions and cannot predict the risk of Aconitum carmichaelii toxicity in highly sensitive populations.

[0006] (4) Neglecting the disruption of “protective mechanisms”: Traditional toxicology only focuses on the direct damage of toxic components to the target site, ignoring the potential inhibition or loss of function of endogenous cardiac defense mechanisms (such as protective kinases) caused by toxic components.

[0007] Therefore, from the perspective of biological effects, screening combinations of biomarkers that exhibit dynamic changes in gene expression and constructing a "dynamic toxicity response model based on changes in gene expression levels" has become a key path to overcome existing evaluation bottlenecks. Summary of the Invention

[0008] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method, kit and application for assessing the cardiotoxicity risk of processed Aconitum carmichaelii products based on NFKB1, PSMB1 and FGR biomarkers.

[0009] To achieve the above objectives, the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides a biomarker or a combination thereof, wherein the biomarker is any one of the NFKB1 gene, PSMB1 gene, and FGR gene; and the combination of biomarkers is any combination of the NFKB1 gene, PSMB1 gene, and FGR gene.

[0010] This invention utilizes large-scale human genetic data (pQTL) and causal inference techniques to identify for the first time key causal biomarkers for cardiotoxicity caused by aconite toxic components (especially 10-hydroxyaconitine / 10-OH mesaconitine), namely the NFKB1, PSMB1, and FGR genes.

[0011] NFKB1 (nuclear factor κB subunit 1) is a positive risk factor for heart failure caused by the toxic components of Aconitum carmichaelii. The toxic components cause myocardial damage by activating the NFKB1-mediated inflammatory storm.

[0012] PSMB1 (proteasome 20S subunit β1) is a positive risk factor for heart failure caused by the toxic components of Aconitum carmichaelii. The toxic components may inhibit the expression of PSMB1 or interfere with its function, leading to an imbalance in the ubiquitin-proteasome system (UPS) homeostasis, thereby causing toxicity.

[0013] FGR (Gardner-Rasheed feline sarcoma virus oncogene homolog) is a protective factor against heart failure. This invention found that the toxic components of Aconitum carmichaelii can lead to a significant downregulation or loss of function of FGR, resulting in a "loss of cardioprotective mechanism".

[0014] The dynamic response model composed of the above three genes can accurately quantify the degree of toxicity-efficacy conversion during the processing of Aconitum carmichaelii: elevated NFKB1 expression indicates the risk of inflammatory activation, decreased PSMB1 expression reflects the critical point of protein homeostasis collapse, and persistent inhibition of FGR expression indicates the failure of endogenous myocardial protection mechanism.

[0015] As a preferred embodiment of the first aspect, the biomarker combination is a combination of the PSMB1 gene and the FGR gene.

[0016] This invention optimizes the "PSMB1+FGR" biomarker combination, which is highly specific and sensitive, based on the initial screening of NFKB1, PSMB1, and FGR. This combination achieved an AUC of 0.773 on the independent validation set, significantly improving predictive efficacy compared to NFKB1 (AUC=0.653), PSMB1 (AUC=0.0707), and FGR (AUC=0.716) alone.

[0017] In a second aspect, the present invention provides the application of the biomarkers or combinations thereof described in the first aspect in predicting the cardiotoxicity risk of processed Aconitum carmichaelii products.

[0018] The predictive model constructed based on the above-mentioned biomarkers or their combinations has AUC values ​​of NFKB1: AUC=0.653, PSMB1: AUC=0.0707, FGR: AUC=0.716, and PSMB1+FGR combination: AUC=0.773. It can be seen that the NFKB1 gene, PSMB1 gene, FGR gene or their combination can quantitatively assess the cardiotoxicity of Aconitum carmichaelii extracts with different processing methods, and accurately identify the residual toxicity risk caused by insufficient processing or the safety problems caused by improper processing.

[0019] Thirdly, the present invention provides the application of products that detect the biomarkers or combinations thereof described in the first aspect in predicting the cardiotoxicity risk of processed Aconitum carmichaelii.

[0020] As a preferred embodiment of the third aspect, the product is a primer for amplifying the PSMB1 gene and / or the FGR gene; or / and, the product is a probe for detecting the PSMB1 gene and / or the FGR gene; the nucleotide sequence of the primer is shown in SEQ ID NO:1 to SEQ ID NO:4.

[0021] This invention achieves toxicity grading and early warning by detecting the relative expression levels of the PSMB1 gene and / or FGR gene in test samples and combining this with threshold criteria: a simultaneous decrease in the expression levels of both PSMB1 and FGR genes indicates high risk; normal expression levels of both PSMB1 and FGR genes indicate low risk. Therefore, by designing specific primers to amplify the PSMB1 and FGR genes and constructing a real-time quantitative PCR detection kit, rapid on-site screening and interpretation of the cardiotoxicity risk of processed Aconitum carmichaelii products can be achieved.

[0022] Fourthly, the present invention provides a kit for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, the kit comprising primers for amplifying the PSMB1 gene and / or the FGR gene; and / or probes for detecting the PSMB1 gene and / or the FGR gene.

[0023] Fifthly, the present invention provides a scoring model for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, the model being calculated using the following formula: RiskScore =( β1 × Exp PSMB1 )+( β2 × Exp FGR )+C in, Exp The expression level of the corresponding gene after normalization or using 2 -ΔΔCt The relative expression level calculated by the method, or the expression level after Log2 transformation; β1 for PSMB1 The regression coefficient for the relative expression level of the gene is negative; β2 for FGR The regression coefficient for the relative expression level of the gene is negative; C The constant term represents the intercept; in, RiskScore The higher the value, the higher the risk of cardiotoxicity.

[0024] RiskScore For risk score, when Risk Score If the risk level is >0, then processed aconite products are considered high-risk. Risk Score If the value is ≤0, then processed aconite products are considered low-risk or safe.

[0025] The risk scoring model constructed based on PSMB1 and FGR genes in this invention has an AUC value of 0.773, which is significantly better than that of a single biomarker. Its 95% confidence interval is 0.595-0.951, indicating that the scoring model has robust discriminative ability.

[0026] As a preferred embodiment of the fifth aspect, the coefficient β1、β2 and constant term C The range of values ​​is as follows: β The value of 1 ranges from -3.0 to -1.0; β 2. The value range is from -3.0 to -1.0; C The value ranges from 3 to 35.

[0027] In this embodiment of the invention, based on different data processing methods, β1, β2, and C in the RiskScore model have different values: When the expression data (Exp) of PSMB1 and FGR are values ​​transformed by Log2, the PSMB1 coefficient (β1) in the RiskScore formula is -2.7809, the FGR coefficient (β2) is -2.4653, and the constant term intercept (C) is 29.273. The model formula is as follows: RiskScore =(-2.7809× Exp PSMB1 )+(-2.4653× Exp FGR )+29.273.

[0028] When the expression data (Exp) of PSMB1 and FGR are used as 2 -ΔΔCt When calculating the relative expression level, the PSMB1 coefficient (β1) in the RiskScore formula is -3.0, the FGR coefficient (β2) is -3.0, and the constant term intercept (C) is 4.0. The model formula is as follows: RiskScore =(-3.0× Exp PSMB1 )+(-3.0× Exp FGR )+4.0.

[0029] The parameter adaptations under the different data processing methods and detection platforms described above all maintained the core biological principle that target gene expression levels are negatively correlated with cardiotoxicity risk, and the center cut-off threshold of RiskScore remained constant at 0. That is, when RiskScore > 0, the processed Aconitum carmichaelii product to be tested was determined to be high-risk; when RiskScore ≤ 0, the processed Aconitum carmichaelii product to be tested was determined to be low-risk or safe.

[0030] Sixthly, the present invention provides a method for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, the method comprising the following steps: S1. Resuscitate and culture human cardiomyocytes. Set up a blank control group and a test group. Add the processed Aconitum carmichaelii extract to the cardiomyocyte culture system of the test group and incubate together. S2. Extract total RNA from cardiomyocytes from step S1; S3. Total RNA was reverse transcribed to synthesize cDNA; S4. Real-time quantitative PCR amplification was performed using the primers and internal reference gene primers described in the third aspect; S5. Using the blank control group as a benchmark, and using internal reference genes to normalize the data, calculate the relative expression levels of PSMB1 and FGR genes in the experimental group to be tested. S6. Substitute the relative expression level into the formula described in the fifth aspect to calculate. RiskScore ; S7. Determine the risk of cardiotoxicity based on preset thresholds.

[0031] As a preferred embodiment of the sixth aspect, the threshold is 0, when Risk Score If the risk level is >0, then processed aconite products are considered high-risk. Risk Score If the value is ≤0, then processed aconite products are considered low-risk.

[0032] In a preferred embodiment of the sixth aspect, the human cardiomyocytes are preferably the AC16 human cardiomyocyte line; the incubation time in step S1 is preferably 24 hours; the PCR amplification program in step S4 is preferably: 95℃ pre-denaturation for 5 min; 40 cycles (95℃ for 10 s, 60℃ for 30 s); melting curve analysis is used to verify amplification specificity. All experiments are performed in triplicate, and data are presented in duplicate. -ΔΔCt The relative expression level is calculated to ensure that the results are reproducible and traceable.

[0033] This invention constructs a new method to predict the cardiotoxicity risk of processed Aconitum carmichaelii products. By combining the detection of PSMB1 and FGR genes with quantitative model verification, the specificity and sensitivity of the discrimination are significantly improved, enabling the assessment of toxicity risk to move from traditional empirical judgment to molecular weight measurement.

[0034] In a seventh aspect, the present invention provides a computer-readable carrier, which is a computer program storing the prediction model described in the fifth aspect and the method described in the sixth aspect, which can be read and executed by a processor to achieve automated identification of the cardiotoxicity risk of processed Aconitum carmichaelii products.

[0035] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Overcoming species differences and establishing strong causal mechanisms: The biomarkers provided by this invention are not based on simple correlations derived from traditional animal models or in vitro experiments, but rather on large-scale human whole-proteome association data (deCODEpQTL). Through multidimensional causal inference analysis using Mendelian randomization (MR), semi-synthetic MR, and Bayesian colocalization, they precisely identify strong causal targets. This method effectively eliminates false positives and pleiotropic interference in traditional drug target screening, overcomes species immune heterogeneity, and can more realistically and accurately reflect the pathological responses of human target organs to toxic components.

[0036] 2. Constructing a multi-dimensional evaluation system with both offensive and defensive aspects: Existing toxicological evaluations are mostly limited to unidirectional "tissue damage" caused by toxic substances. This invention innovatively introduces FGR kinase as a novel assessment indicator of "loss of endogenous protective mechanisms" in the myocardium. By jointly detecting pathogenic factors (NFKB1 / PSMB1) and defensive factors (FGR) that mediate toxicological damage, a multi-dimensional molecular toxicity evaluation system covering "damage activation" and "protective failure" is constructed, significantly improving the comprehensiveness and specificity of toxicity risk prediction.

[0037] 3. Achieving a leap from "physicochemical quality control" to "quantification of biotoxicity": Compared to the static physicochemical testing standards in the current pharmacopoeia that rely solely on the content of diester alkaloids, the biological indicators provided by this invention can more fundamentally reflect whether the processing technology has truly eliminated the deep molecular toxicity of the drug. Using this model, the optimal processed products that "have qualified chemical component content and do not cause disturbances to key pathogenic targets of the myocardium" can be accurately screened, providing a scientific basis at the functional mechanism level for the process optimization of Aconitum carmichaelii to "reduce toxicity while retaining efficacy".

[0038] 4. Possesses extremely high industrial transformation and promotion value: The real-time fluorescence quantitative PCR detection method based on specific gene expression levels constructed in this invention has advantages such as simple operation, low cost, high throughput, and short detection cycle. This method is highly compatible with the needs of modern industrialized production of traditional Chinese medicine and is easily promoted and applied to various aspects of traditional Chinese medicine production enterprises, such as safety screening of processed medicinal materials, dynamic monitoring of processing techniques, and quality control throughout the entire chain. Attached Figure Description

[0039] Figure 1 This is a schematic diagram showing the results of differential expression analysis of LCT, PSMB1, FGR, and NFKB1 genes between heart failure samples (HF) and control samples (Control) in an independent clinical cohort. Figure 2 This diagram illustrates the significant positive correlation between the expression of PSMB1 and NFKB1 genes in an independent clinical cohort. Figure 3 ROC curves for FGR, NFKB1, and PSMB1 predicted separately; Figure 4 This is the ROC curve jointly predicted by FGR and PSMB1. Detailed Implementation

[0040] To better illustrate the purpose, technical solution, and advantages of the present invention, the present invention will be further described below in conjunction with specific embodiments.

[0041] Example 1: Genetic screening of key biomarkers for Aconitum carmichaelii toxicity 1. Prediction of potential toxicological targets of Aconitum carmichaelii Using pharmacophore models (such as SuperPred and SwissTargetPrediction), 173 potential biological targets of 10-OH mesaconitine, the core toxic component of Aconitum carmichaelii, were predicted.

[0042] Primary screening of biomarkers based on strong causal inference from genetics To eliminate false positives prevalent in traditional network pharmacology, a rigorous "strong causal" inference system was introduced. Based on ultra-deep proteomics pQTL data of over 35,000 individuals in the deCODE database, genetic variation was used as an instrumental variable. A combined analysis of two-sample Mendelian randomization (Two-Sample MR) and summative MR (SMR) was employed to preliminarily screen four candidate genes significantly associated with heart failure: LCT, PSMB1, FGR, and NFKB1.

[0043] Professional selection and co-location confirmation of landmarks Artificial interference removal: Further analysis by the inventors revealed that the genetic association signal of the candidate gene LCT (lactase) is highly susceptible to interference from population stratification factors related to dietary habits and geographical distribution, rather than from direct pathological and toxicological factors. To ensure target organ specificity of the detection, this invention excluded LCT from the core biomarker library, ultimately establishing NFKB1, PSMB1, and FGR as core toxicity biomarkers.

[0044] Colocalization analysis confirmed the presence of the core biomarkers. Further Bayesian colocalization analysis rigorously confirmed the presence of these biomarkers in the UKB-PPP and Irish cohorts of the deCODE database. The results showed that only protein FGR exhibited a directly shared causal variation with the risk of heart failure (HF) (posterior colocalization probability PP.H4.abf>0.5), demonstrating a very strong protective factor (OR<1, P<0.05). Simultaneously, NFKB1 and PSMB1 were also identified as significant pathogenic risk factors via MR and SMR (OR>1, P<0.05). This multidimensional analysis based on human genetic principles objectively established the core causal role of these three biomarkers in cardiotoxicology.

[0045] Example 2: Validation, efficacy evaluation, and combination optimization of biomarkers in clinical samples This embodiment builds upon the genetic screening results of Example 1, using the independent transcriptome dataset GSE16499 (containing 15 heart failure samples and 15 control samples) from the GEO database to simulate the endpoint state of cardiac injury caused by Aconitum carmichaelii toxicity, and to conduct clinical validation, predictive efficacy evaluation and combination optimization of the aforementioned initial screening biomarkers.

[0046] Gene expression and correlation validation in independent clinical cohorts The results of differential expression analysis between heart failure samples (HF) and control samples (Control) are as follows: Figure 1 As shown, the expression levels of LCT, PSMB1, FGR, and NFKB1 genes were all different. The FGR gene was statistically significantly downregulated in the heart failure samples (P<0.05), directly validating its "loss of protection" toxicological mechanism at the transcriptomic level. Although the absolute mean expression levels of PSMB1 and NFKB1 between the two groups did not reach the strict univariate statistical significance threshold (P>0.05), correlation analysis showed a significant positive correlation between PSMB1 and NFKB1 gene expression (Spearman correlation coefficient r = 0.49, P = 0.007). Figure 2 This high level of co-expression confirms the dynamic synergistic effect of the "proteasome-inflammatory axis" during toxicity, suggesting that they may have joint predictive value in disease diagnosis.

[0047] 2. Evaluation of the predictive power of individual biomarkers To further evaluate the diagnostic and classification value of each gene, PSMB1, FGR, and NFKB1 were used as individual detection indicators. A cardiotoxicity risk score (CRS) model was constructed using multivariate binary logistic regression, and ROC curves were plotted. Figure 3 ).

[0048] (1) PSMB1 prediction alone: ​​Although the mean expression levels do not differ significantly, the risk prediction formula constructed based on its distribution characteristics is: RiskScore=β1×Exp PSMB1 +C, its area under the curve (AUC) still reached 0.707, proving that it has independent classification prediction ability (Note: β1 is negative, indicating that the more severe the damage to the PSMB1 proteasome, the higher the risk).

[0049] (2) FGR prediction alone: ​​The prediction formula RiskScore=β2×Exp is used. FGR +C, with an AUC value of 0.716, is consistent with its significantly differential expression results, demonstrating good predictive performance (Note: β2 is negative, indicating that the more severe the loss of FGR protection mechanism, the higher the risk).

[0050] (3) NFKB1 prediction alone: ​​The prediction formula RiskScore=β3×Exp is used. NFKB1 +C, its AUC value is only 0.653.

[0051] In summary, both PSMB1 and FGR, when used alone, possess a predictive power >0.7, making them independent and effective diagnostic biomarkers. While NFKB1 is recognized as a core toxicological driver at the genetic level, its predictive and diagnostic efficacy as a single marker is relatively weak at the transcriptome (mRNA) level (AUC only 0.653). This suggests that NFKB1 activity is primarily regulated by post-translational modifications (such as protein phosphorylation), and its mRNA abundance fluctuations are insufficient to serve as a sensitive clinical detection target.

[0052] 3. Performance evaluation and model optimization of marker combinations Based on the above findings, the inventors eliminated redundant indicators (NFKB1) with weak predictive efficacy and established the PSMB1+FGR dual biomarker combination, consisting of downstream effector PSMB1 (proteasome subunit) and FGR (protective kinase), as the preferred embodiment of the present invention.

[0053] Based on this combination, a final predictive model for the cardiotoxicity risk of processed Aconitum carmichaelii products is constructed using Logistic regression. The calculation formula is as follows: RiskScore =( β1 × Exp PSMB1 )+( β2 × Exp FGR )+ C .

[0054] (Note: Exp The relative expression levels of genes detected by techniques such as qPCR; β1 , β2 The regression coefficients for the corresponding genes are denoted as . β1 , β2 All are negative; C is the intercept of the constant term. Model parameter construction: In this embodiment, the expression data (Exp) of PSMB1 and FGR after Log2 transformation in the GSE16499 training set are extracted and logistic regression analysis is performed. The model parameters obtained are as follows: PSMB1 coefficient (β1): -2.7809 FGR coefficient (β2): -2.4653 Intercept (C) of constant term: 29.273; The formula for this model is: RiskScore =(-2.7809× Exp PSMB1 )+(-2.4653× Exp FGR )+29.273.

[0055] (Note: In practical industrial applications, such as using qPCR technology 2) -ΔΔCt The coefficients β and constant C mentioned above can be linearly corrected based on the data baseline of the detection platform, but the core biological trend of negative values ​​for β1 and β2 remains unchanged. After calculating the Risk Score using the aforementioned PSMB1+FGR joint model, the ROC curve was plotted (see [link]). Figure 4 The results showed that the AUC value of the combined detection jumped to 0.773.

[0056] Depend on Figure 3 and Figure 4 It is evident that the AUC of the combined detection (0.773) is significantly superior to that of PSMB1 alone (0.707) or FGR alone (0.716). This demonstrates that the "dual detection strategy" proposed in this invention produces a significant synergistic effect. This model essentially monitors the "simultaneous collapse of the cleansing system (PSMB1) and the defense system (FGR)" in the cardiac microenvironment. When the expression levels of both significantly decrease, the Risk Score increases sharply, effectively reducing the risk of missed detection that may exist with single-index detection. Furthermore, removing NFKB1 and retaining only two primer pairs offers significant industrial advantages, including more consistent logic (both exhibit downregulation in the toxicological model, with consistent detection directions) and lower detection costs.

[0057] Risk assessment criteria (Cut-off Value) in industrial applications Based on the Youden index maximization principle and the logistic regression probability transformation principle (usually with a disease incidence probability of 0.5 as the boundary, the corresponding Logit transformation value is 0), this invention sets the RiskScore center cut-off threshold to 0. The specific determination method includes the following steps: Step 1 (Detection): Measure the relative expression levels (Exp) of PSMB1 and FGR in cardiomyocytes after treatment with the extract of the test sample. Step 2 (Calculation): Substitute into the formula to calculate the combined Risk Score; Step 3 (Judgment): High risk (unacceptable): If the Risk Score > cutoff value (Cut-off = 0), it means that the expression levels of PSMB1 and FGR are too low, indicating that the Aconitum carmichaelii sample has strong biological toxicity and that the processing to reduce toxicity was insufficient.

[0058] Low risk (qualified): If the Risk Score ≤ Cut-off value (Cut-off = 0), it means that the expression of PSMB1 and FGR is maintained at a level close to that of the normal control, without triggering the collapse of the "proteasome-defense system", proving that the processing technology effectively achieves "reduction of toxicity and preservation of efficacy".

[0059] Example 3: Cellular validation and parameter adaptation of the kit for the core toxic component of Aconitum carmichaelii (10-OH mesaconitine). This embodiment utilizes the AC16 human cardiomyocyte model to perform real-world cellular validation of the dual-marker kit provided by this invention, targeting the core cardiotoxic component of Aconitum carmichaelii, 10-OHmesaconitine monomer, and demonstrates a model parameter calibration method based on the qPCR platform.

[0060] I. Experimental Materials and Preparation 1. Cell model: AC16 human cardiomyocyte line.

[0061] 2. Sample to be tested: 10-OH mesaconitine standard (purity >98%, dissolved in DMSO).

[0062] 3. Reagents: AC16-specific culture medium (DMEM / F-12 1:1 mixture, with 12.5% ​​FBS and 1% penicillin antibody added), TRIzol total RNA extraction reagent, reverse transcription kit, and SYBR Green qPCR universal premix.

[0063] II. Primer Design Specific amplification primers were designed for PSMB1 and FGR. 1. PSMB1 (NM_002793.4) ​​primer pair: Forward (F): 5'-CAGCCATGTATTCGGCTCCT-3' (SEQ ID NO: 1) Reverse (R): 5'-TGCCAGTATAGTACCTCCGTT-3' (SEQ ID NO: 2) 2. FGR (NM_005248.3) primer pair: Forward (F): 5'-GGGCAGAATTGGGCTCCAG-3' (SEQ ID NO: 3) Reverse (R): 5'-ATTCCAGGTTCCCTGCCCTT-3' (SEQ ID NO: 4) 3. Primer pair for internal reference gene (GAPDH, NM_002046.7): Forward (F): 5'-GTCTCAGCAACTCTCGCCTT-3' (SEQ ID NO: 5) Reverse (R): 5'-GTGGAGCCATTGTTGGCAAG-3' (SEQ ID NO: 6) III. Detailed Experimental Procedure Step S1: Cell drug delivery treatment 1. AC16 cells were seeded in 6-well plates at a density of 5 × 10⁶ cells / well. 5 1 cell / well, cultured overnight until adhered to the wall.

[0064] 2. Group processing: Blank control group (Control): Add an equal volume of culture medium containing 0.1% DMSO.

[0065] Toxicity Model Group: 10-OH mesaconitine was added to the culture medium (the final concentration was set to a toxic concentration that resulted in a cell survival rate of about 60%-70%, for example, 100 μM).

[0066] 3. Incubate at 37℃ in a 5% CO2 incubator for 24 hours. Step S2: Total RNA extraction and reverse transcription After removing the culture medium, the cells were washed with PBS and lysed with 1 mL of TRIzol. Total RNA was extracted using the standard chloroform-isopropanol method. 1 μg of total RNA was used to synthesize cDNA (20 μL system) using a reverse transcription kit; the reaction conditions were: 37℃ for 15 min, 85℃ for 5 s.

[0067] Step S3: Real-time quantitative PCR (qPCR) detection Prepare a 20 μL reaction mixture: 10 μL SYBR Green Master Mix; 0.4 μL each of forward and reverse primers; 2 μL cDNA template; 7.2 μL nuclease-free water. Amplification program: 95℃ pre-denaturation for 30 s; 40 cycles (95℃ for 5 s, 60℃ for 30 s fluorescence collection). Set up 3 replicates for each sample.

[0068] IV. Data Calibration and Parameter Adaptation of Risk Prediction Model on the qPCR Platform 1. Analysis of differences in data volume: The model constructed in Example 2 based on transcriptome microarray data (with a constant term C value of 29.273) uses Log2 transformation expression values, whose data distribution typically ranges from 8.0 to 12.0. However, in industrially widespread qPCR assays, relative gene expression levels (Exp) are typically expressed using a 22... -ΔΔCt The expression level was calculated using a specific method. In this system, the theoretical baseline value of Exp for the normal physiological state (i.e., the blank control group) was constant at 1.0, while the expression level in the highly toxic state typically ranged between 0 and 0.5. Directly applying the constants of the chip model would lead to significant systematic errors, making effective judgment impossible.

[0069] 2. Parameter calibration specific to the qPCR platform: To make the model suitable for qPCR kits, this invention targets 2 -ΔΔCt The model parameters underwent equivalent mathematical adaptation based on the data distribution characteristics. While maintaining the core biological principle (i.e., the expression levels of PSMB1 and FGR are negatively correlated with toxicity), the regression coefficients and intercepts were adjusted to ensure the model could accurately distinguish between normal baseline and significantly downregulated states. The adaptation derivation is as follows: To maintain the strong negative correlation between the two genes with equal weights, calibration coefficients β1 = -3.0 and β2 = -3.0 were set. To ensure that the decision threshold center remains unchanged (i.e., with RiskScore = 0 as the boundary), the calculated score of normal cells (exp = 1.0) must be less than 0 (safe), while the calculated score of severely damaged cells (e.g., exp decreased to 0.4) must be greater than 0 (toxic). After calculation and balancing, the optimal constant term intercept C = 4.0 was established for the qPCR kit.

[0070] Thus, this invention establishes a standardized calculation formula directly applicable to industrial qPCR testing: RiskScore =(-3.0× Exp PSMB1 )+(-3.0× Exp FGR )+4.0 V. Cellular Verification Results and Judgment Criteria This invention is unified with RiskScore = 0 was used as the cut-off value for determining the risk of high or low cardiotoxicity. The representative relative expression level (Exp) detected by qPCR was substituted into the above calibration formula, and the verification results are shown in Table 1: Table 1 Experimental analysis of 10-OH mesaconitine, the core cardiotoxic component of Aconitum carmichaelii, showed that this monomer significantly inhibited the expression of PSMB1 and FGR in AC16 cardiomyocytes. After substituting the values ​​into the qPCR-specific calibration formula provided in this invention, the RiskScore of the toxicity model group jumped to 1.75 (above the threshold of 0), successfully achieving quantitative early warning of high toxicity risk. This indicates that this kit and model can not only accurately predict the cardiotoxic risk of 10-OH mesaconitine molecules, but can also be directly applied to the evaluation of toxicity reduction processes and routine quality control systems for complex traditional Chinese medicines containing this component (such as raw Aconitum carmichaelii and different processed products).

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the scope of protection of the present invention. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the essence and scope of the technical solutions of the present invention.

Claims

1. A biomarker or a combination thereof, characterized in that, The biomarker is any one of the NFKB1 gene, PSMB1 gene, and FGR gene; The biomarker combination is any combination of the NFKB1 gene, PSMB1 gene, and FGR gene.

2. The biomarker combination as described in claim 1, characterized in that, It is a combination of the PSMB1 gene and the FGR gene.

3. The use of the biomarkers or combinations thereof as described in claim 1 or 2 in predicting the cardiotoxicity risk of processed Aconitum carmichaelii products.

4. The use of products containing the biomarkers of claim 1 or 2 or combinations thereof in predicting the cardiotoxicity risk of processed Aconitum carmichaelii.

5. The application as described in claim 4, characterized in that, The product is a primer for amplifying the PSMB1 gene and / or the FGR gene; or / and the product is a probe for detecting the PSMB1 gene and / or the FGR gene, the nucleotide sequences of the primers being shown in SEQ ID NO:1 to SEQ ID NO:

4.

6. A kit for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, characterized in that, The kit includes primers for amplifying the PSMB1 gene and / or the FGR gene; and / or probes for detecting the PSMB1 gene and / or the FGR gene.

7. A scoring model for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, characterized in that, The model is based on logistic regression, and its calculation formula is as follows: RiskScore =( β 1× Exp PSMB1 )+( β 2× Exp FGR )+ C in, Exp The expression level of the corresponding gene after normalization or using 2 -ΔΔCt The relative expression level calculated by the method, or the expression level after Log2 transformation; β1 for PSMB1 The regression coefficient of the gene is negative; β2 for FGR The regression coefficient of the gene is negative; C The constant term represents the intercept; RiskScore For risk score, when Risk Score If the risk level is >0, then processed aconite products are considered high-risk. Risk Score If the value is ≤0, then processed aconite products are considered low-risk or safe.

8. The scoring model as described in claim 7, characterized in that, The coefficient β1、β2 and constant term C The range of values ​​for is as follows: β The value of 1 ranges from -3.0 to -1.0; β 2. The value range is from -3.0 to -1.0; C The value ranges from 3 to 35.

9. A method for predicting the cardiotoxicity risk of processed Aconitum carmichaelii products, characterized in that, The method includes the following steps: S1. Resuscitate and culture human cardiomyocytes. Set up a blank control group and a test group. Add the processed Aconitum carmichaelii extract to the cardiomyocyte culture system of the test group and incubate together. S2. Extract total RNA from cardiomyocytes from step S1; S3. Total RNA was reverse transcribed to synthesize cDNA; S4. Real-time quantitative PCR amplification was performed using the primers and internal reference gene primers described in claim 5; S5. Using the blank control group as a benchmark, and using the internal reference gene to normalize the data, calculate the relative expression levels of PSMB1 and FGR genes in the experimental group to be tested. S6. Substitute the relative expression level into the formula described in claim 7 to calculate. RiskScore ; S7. Determine the risk of cardiotoxicity based on preset thresholds.

10. A computer-readable medium, characterized in that, The carrier is a computer program storing the prediction model of claim 7 and the method of claim 9, which can be read and executed by a processor to achieve automated identification of the cardiotoxicity risk of processed Aconitum carmichaelii products.

Images