Zhenzhu Zhang

Screening, Verification, and Optimization of Biomarkers for Early Prediction of Cardiotoxicity Based on Metabolomics

Drug-induced cardiotoxicity seriously affects human health and drug development. However, many conventional detection indicators of cardiotoxicity exhibit significant changes only after the occurrence of severe heart injuries. Therefore, we investigated more sensitive and reliable indicators for the evaluation and prediction of cardiotoxicity. We created rat cardiotoxicity models in which the toxicity was caused by doxorubicin (20 mg/kg), isoproterenol (5 mg/kg), and 5-fluorouracil (125 mg/kg). We collected data from rat plasma samples based on metabolomics using ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry. Following multivariate statistical and integration analyses, we selected 39 biomarker ions of cardiotoxicity that predict cardiotoxicity earlier than biochemical analysis and histopathological assessment. Because drugs with different toxicities may cause similar metabolic changes compared with other noncardiotoxic models (hepatotoxic and nephrotoxic models), we obtained 10 highly specific biomarkers of cardiotoxicity. We subsequently used a support vector machine (SVM) to develop a predictive model to verify and optimize the exclusive biomarkers. l-Carnitine, 19-hydroxydeoxycorticosterone, LPC (14:0), and LPC (20:2) exhibited the strongest specificities. The prediction rate of the SVM model is as high as 90.0%. This research provides a better understanding of drug-induced cardiotoxicity in drug safety evaluations and secondary development and demonstrates novel ideas for verification and optimization of biomarkers via metabolomics.

Plasma metabolic profiling analysis of nephrotoxicity induced by acyclovir using metabonomics coupled with multivariate data analysis

Acyclovir (ACV) is an antiviral agent. However, its use is limited by adverse side effect, particularly by its nephrotoxicity. Metabonomics technology can provide essential information on the metabolic profiles of biofluids and organs upon drug administration. Therefore, in this study, mass spectrometry-based metabonomics coupled with multivariate data analysis was used to identify the plasma metabolites and metabolic pathways related to nephrotoxicity caused by intraperitoneal injection of low (50mg/kg) and high (100mg/kg) doses of acyclovir. Sixteen biomarkers were identified by metabonomics and nephrotoxicity results revealed the dose-dependent effect of acyclovir on kidney tissues. The present study showed that the top four metabolic pathways interrupted by acyclovir included the metabolisms of arachidonic acid, tryptophan, arginine and proline, and glycerophospholipid. This research proves the established metabonomic approach can provide information on changes in metabolites and metabolic pathways, which can be applied to in-depth research on the mechanism of acyclovir-induced kidney injury.

A systematic strategy for screening and application of specific biomarkers in hepatotoxicity using metabolomics combined with ROC curves and SVMs

Current studies that evaluate toxicity based on metabolomics have primarily focused on the screening of biomarkers while largely neglecting further verification and biomarker applications. For this reason, we used drug-induced hepatotoxicity as an example to establish a systematic strategy for screening specific biomarkers and applied these biomarkers to evaluate whether the drugs have potential hepatotoxicity toxicity. Carbon tetrachloride (5 ml/kg), acetaminophen (1500 mg/kg), and atorvastatin (5 mg/kg) are established as rat hepatotoxicity models. Fifteen common biomarkers were screened by multivariate statistical analysis and integration analysis-based metabolomics data. The receiver operating characteristic curve was used to evaluate the sensitivity and specificity of the biomarkers. We obtained 10 specific biomarker candidates with an area under the curve greater than 0.7. Then, a support vector machine model was established by extracting specific biomarker candidate data from the hepatotoxic drugs and nonhepatotoxic drugs; the accuracy of the model was 94.90% (92.86% sensitivity and 92.59% specificity) and the results demonstrated that those ten biomarkers are specific. 6 drugs were used to predict the hepatotoxicity by the support vector machines model; the prediction results were consistent with the biochemical and histopathological results, demonstrating that the model was reliable. Thus, this support vector machine model can be applied to discriminate the between the hepatic or nonhepatic toxicity of drugs. This approach not only presents a new strategy for screening-specific biomarkers with greater diagnostic significance but also provides a new evaluation pattern for hepatotoxicity, and it will be a highly useful tool in toxicity estimation and disease diagnoses.

Rapid classification and identification of complex chemical compositions in traditional Chinese medicine based on UPLC-Q-TOF/MS coupled with data processing techniques using the KuDieZi injection as an example

The KuDieZi (KDZ) injection is prepared by extracting and processing Ixeris sonchifolia [Ixeris sonchifolia Hance] belonging to Lactuca genus of the Asteraceae family. The KDZ injection is a single-herb preparation in which the components are analysed by liquid chromatography-mass spectrometry. However, the chemical compositions of this preparation are complex and diverse. Hence, data processing is complicated and time consuming; furthermore, data processing cannot provide a systematic, accurate and repeatable method to rapidly classify and identify chemical constituents of the injection. In our study, the main components, particularly flavonoids, organic acids, amino acids and nucleosides in the KDZ injection, were rapidly classified and identified by data processing technology based on UPLC-Q-TOF/MS. After we reviewed lots of studies and collected information on the fragments, then compared them with the mass spectrometric analyses of the standards, the rules of diagnostic fragments (DFs) and neutral losses (NLs) of the four substances were found and summarised. A rapid classification and identification method of the chemical composition in the KDZ injection was then constructed using the DF filter (DFF) and the NL filter (NLF). This method was applied to analyse the KDZ injection. A total of 31 chemical components, which included 8 flavonoids, 13 organic acids, 6 amino acids and 4 nucleosides, were obtained. DFF and NLF were used to rapidly classify and identify chemical substances in the KDZ fingerprint. With this method, we effectively solved the technical difficulties in fingerprint resolution caused by complex components and low levels in traditional Chinese medicine (TCM). In addition, this study provided a novel approach for further studies on TCM.

A novel approach to the simultaneous extraction and non-targeted analysis of the small molecules metabolome and lipidome using 96-well solid phase extraction plates with column-switching technology

This study combines solid phase extraction (SPE) using 96-well plates with column-switching technology to construct a rapid and high-throughput method for the simultaneous extraction and non-targeted analysis of small molecules metabolome and lipidome based on ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry. This study first investigated the columns and analytical conditions for small molecules metabolome and lipidome, separated by an HSS T3 and BEH C18 columns, respectively. Next, the loading capacity and actuation duration of SPE were further optimized. Subsequently, SPE and column switching were used together to rapidly and comprehensively analyze the biological samples. The experimental results showed that the new analytical procedure had good precision and maintained sample stability (RSD<15%). The method was then satisfactorily applied to more widely analyze the small molecules metabolome and lipidome to test the throughput. The resulting method represents a new analytical approach for biological samples, and a highly useful tool for researches in metabolomics and lipidomics.

A rapid and integrated pyramid screening method to classify and identify complex endogenous substances with UPLC/Q-TOF MS-based metabolomics

Metabolomics plays a role in disease diagnosis, safety and efficacy of drug evaluation, and microbial research. Liquid chromatography-mass spectrometry (MS) is the main analysis tool in metabolomics studies. Given the existence of many different categories of endogenous metabolites is isomers of one another, some problems on classification and identification have emerged. These problems result in high false-positive results, as well as a complex and time-consuming substance identification process. Accordingly, this study reviewed literature and retrieved databases to identify endogenous substances in the same category accompanied by a certain mass range (MR) and mass defect range (MDR), as well as an identical or similar fragmentation pattern in the mass spectrum [i.e., characteristic and neutral loss (NL) fragments]. We conducted different MS/MS collision energies to analyze different categories of endogenous substances to discover and summarize their fragmentation patterns. We then used the MR and MDR of the parent ion, diagnostic fragments (together with their MDR), and NL as screening tools to establish a pyramid screening method (PSM) for the rapid classification and identification of metabolites. Finally, we compared the PSM with the conventional identification method through known compounds in the literature. PSM was found to solve the key problem in metabolomics to some extent, namely, the classification and identification of substances. This method also facilitated the further development of metabolomics and provides a new perspective on the screening and identification of target components in other complex samples.

Toxicity analysis of doxorubicin using plasma metabolomics technology based on rapid resolution liquid chromatography coupled with quadruple-time-of-flight mass spectrometry

Doxorubicin is a highly efficient antitumor drug, but it can induce toxicity, largely affecting peoples life. Metabolomics technology, a part of systems biology, can offer information on the changes in metabolic profiles of biofluids upon drug administration. Meanwhile, the study of plasma metabolomics of doxorubicin toxicity using liquid chromatography-mass spectrometry technology is not very clear. In this study, a plasma metabolomics approach using rapid resolution liquid chromatography coupled with quadruple-time-of-flight mass spectrometry technology was used to investigate the toxic mechanism of doxorubicin from a metabolic view. The biochemical analysis and histopathological examination results showed that a toxicity model can be built by intraperitoneal injection of doxorubicin with a dose of 15 mg kg−1 in male Wistar rats. Metabolomics results revealed that fifteen biomarkers were changed due to doxorubicin-induced toxicity. Besides, arachidonic acid metabolism, valine, leucine and isoleucine biosynthesis, sphingolipid metabolism, glycerophospholipid metabolism and primary bile acid biosynthesis were mainly responsible for the toxicity of doxorubicin. The changed metabolites and interrupted pathways found in this study are meaningful and the results can lay the foundation for further research on the toxicity mechanism of doxorubicin.

Screening and verification of linearly dependent biomarkers with acute toxicity induced by Aconiti Radix based on liquid chromatography-mass spectrometry-based metabolite profiling

Aconiti Radix, with its unique anti-inflammatory and analgesic effects, is a well-known form of traditional medication; however, improper use of the Aconiti Radix drug often leads to severe acute toxicity. Raw Aconiti Radix ethanol extraction is the most toxic ingredient, followed by raw Aconiti Radix water extraction; processed products have less toxic ingredients. Current clinical examinations primarily use biochemical tests and histopathological examination, but such approaches lack specificity, are time-consuming, and have low sensitivity, which can easily lead to false positive results. Therefore, a fast and accurate way to evaluate acute toxicity is needed. We have established a method that combines metabonomics with trend analysis of a gavage concentration series to find and validate acute toxicity biomarkers of Aconiti Radix. The purpose of this study is to identify Aconiti Radix acute toxicity biomarkers based on UPLC-Q-TOF-MS metabonomics technology. We use relative amounts of biomarkers with dosage and degree of toxicity to determine a dose-dependent trend; these substances may be exclusive Aconiti Radix acute toxicity biomarkers. These exclusive biomarkers were validated both in water extraction of Aconiti Radix and drug incompatibility with Aconiti Radix Cocta–Pinelliae rhizoma couple medicines; ultimately, the acute toxicity biomarkers (shikimic acid, L-acetylcarnitine, LysoPC (22:5), L-valine) were determined. This new method provides a better way to discover and validate specific metabonomics endogenous small molecule compounds.

Evaluation and optimization of biomarkers in a primary dysmenorrhea model using ultra performance liquid chromatography coupled with quadrupole-time-of-flight mass spectrometry combined with a support vector machine

A comprehensive plasma metabolic profiling analysis method was established to evaluate the potential biomarkers in a primary dysmenorrhea model combined with the method of a support vector machine which optimized the selected potential biomarkers. This is an original report to provide verification for the extensive application of the model both in a laboratory and clinic.

An integrated strategy for the rapid extraction and screening of phosphatidylcholines and lysophosphatidylcholines using semi-automatic solid phase extraction and data processing technology

This study attempts to establish a comprehensive strategy for the rapid extraction and screening of phosphatidylcholines (PCs) and lysophosphatidylcholines (LysoPCs) in biological samples using semi-automatic solid phase extraction (SPE) and data processing technology based on ultra-performance liquid chromatography-quadrupole-time of flight-mass spectrometry (UPLC-Q-TOF-MS). First, the Ostro sample preparation method (i.e., semi-automatic SPE) was compared with the Bligh-Dyer method in terms of substance coverage, reproducibility and sample preparation time. Meanwhile, the screening method for PCs and LysoPCs was built through mass range screening, mass defect filtering and diagnostic fragments filtering. Then, the Ostro sample preparation method and the aforementioned screening method were combined under optimal conditions to establish a rapid extraction and screening platform. Finally, this developed method was validated and applied to the preparation and data analysis of tissue samples. Through a systematic evaluation, this developed method was shown to provide reliable and high-throughput experimental results and was suitable for the preparation and analysis of tissue samples. Our method provides a novel strategy for the rapid extraction and analysis of functional phospholipids. In addition, this study will promote further study of phospholipids in disease research.