Chemometrics strategy coupled with high resolution mass spectrometry for analyzing and interpreting comprehensive metabolomic characterization of hyperlipemia

Abstract

Hyperlipidemia (HLP) is a metabolic disorder which is characterized by a disturbance in lipid metabolism and is a primary risk factor for cardiovascular disease and atherosclerosis. Patients who were diagnosed with hyperlipidemia are usually with a serious dyslipidemia, which can be evaluated by current clinical markers such as triglyceride, total cholesterol, low density lipoprotein cholesterol, and high-density lipoprotein. Metabolic profiling has great potential to help the diagnosis and prognosis of hyperlipidemia patients in an earlier stage. To explore the metabolic profiling, we used ultraperformance liquid chromatography mass spectrometry, monitored the metabolites' alterations and impaired pathways in HLP patients. Of note, a total of 37 serum metabolites from HLP patients were tentatively identified. Furthermore, they were involved in 11 pathways such as linoleic acid metabolism, glycerophospholipid metabolism, terpenoid backbone biosynthesis, sphingolipid metabolism, glycosylphosphatidylinositol-anchor biosynthesis, steroid hormone biosynthesis, arachidonic acid metabolism, pentose and glucuronate interconversions, starch and sucrose metabolism, glycerolipid metabolism and purine metabolism. These results provide potential biomarkers for early risk assessment of HLP and offer further insights into the complex metabolic pathway changes in HLP. They also suggest the distinction of utilizing UPLC-Q-TOF-MS for exploring metabolic characterization of HLP patients.

---

Introduction

Hyperlipidemia (HLP) results from defects of genes or metabolic disorder. It is characterized by abnormally increased or decreased serum levels of triglycerides (TG), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), very low-density lipoprotein cholesterol (VLDL-C) and high-density lipoprotein cholesterol (HDL-C).¹ The disordered lipid levels are risk factors of many diseases, such as coronary heart disease (CHD), atherosclerosis, pancreatitis, liver disease and so on.^2–4 The role of hyperlipidemia as a risk factor continues especially in the older age group,^5,6 while the morbidity of HLP in young adults is also rising with the improvement of living standards, which could increase the subsequent risk of CHD.^7,8 According to a survey data from the “National Health and Nutrition Examination”, only some patients who get early HLP diagnosis have received treatment.⁹ The high cost of the treatment and the followed complications have made the hyperlipidemia became a burden to the society.^10,11 More importantly, the current parameters such as TG, TC, LDL-C, VLDL-C, and HDL-C do not address the mechanisms and what really happened in the lipid metabolism, with these puzzles, more effective methods for specific diagnosis of hyperlipidemia need to be developed.

Fortunately, metabolomics, a study focusing on all metabolites which are the last products of cellular adjustment processes in the body, can describe the portrayal of the physical condition and is considered as closely correlative with a patient's phenotype.^12,13 Metabolomics identifies and quantifies the disordered low molecular compounds which resulted from the endogenous or exogenous perturbations, represents the comprehensive assessment of biological systems, and builds a bridge between the disturbed metabolites of the body and the complicated clinical symptoms.^14–16 As an analysis method, urine, serum, saliva and other biological fluids and tissues are the main research objects, and GC/MS and UPLC/MS are usually applied as the analysis tools.^17–19 In this study, serum samples from HLP patients were collected and the high sensitive UPLC-G2-Si-MS system combined with pattern recognition approaches was used to analyze and verify the specific biomarkers influenced by the HLP. Through this study we showed that metabolomics is not only a powerful tool for the better understanding of the hyperlipidemia pathophysiologic mechanism but also helps clinical specific diagnosis.

Materials and methods

Ethical statement

Written informed consents were obtained from all subjects. The experimental protocol were reviewed and approved by the Ethical Committee of Heilongjiang University of Chinese Medicine and was conducted according to the principles expressed in the Declaration of Helsinki.

Subjects

With the aim of estimating the serum metabolic profile of the hyperlipidemia and assess its precise molecular mechanism, 71 consecutive hyperlipidemia patients and 100 healthy participants of both sex and age between 35 and 80, who fulfilled the inclusion and exclusion criteria and gave informed consent, were enrolled from four different outpatient departments between February 2014 and March 2015, all consenting patients and healthy participants were clinically diagnosed to have dyslipidemia or normal lipidemia and the detailed information of the patients and healthy participants has been uploaded to the Clinical Trial Management Public Platform. The threshold of hyperlipidemia obeys to the China Adult Dyslipidemia Prevention Guide formulate joint committee and American National Cholesterol Education Program (Table S1†), partial information of the HLP patients is shown in the following Table S2.† After collecting the samples, the serum of the participants were transported in a low-temperature environment to the National TCM Key Laboratory of Serum Pharmacochemistry, Heilongjiang University of Chinese Medicine. In addition, all participants have been notified that before the prior week of the experiment, smoking, alcohol, fast food containing cheese and preservative, fruit juice, fasting coffee, tea, chocolate, cola and other caffeinated foods or beverages are forbidden. The exclusion criteria is that the patients who had been diagnosed with autoimmune diseases, cancer, infectious diseases, tuberculosis, AIDS and chronic active hepatitis, neuromuscular disease, endocrine disease would be excluded.

Chemicals and reagents

Acetonitrile and methanol (HPLC grade) was purchased from Merck (Germany); distilled water was purchased from Watson's Food &Beverage Co., Ltd. (Guangzhou, China); formic acid (HPLC grade, FA) was purchased from Aladdin reagent Co., Ltd. (Shanghai, China); leucine enkephalin was purchased from Sigma-Aldrich (St. Louis, MO, USA). All other reagents were of analytical grade.

Sample preparation

The fasting blood samples were collected from the cubital vein at morning after one week's light diet. The blood samples were collected in vacuum tubes and the serum were separated immediately by centrifuging at 4000 rpm for 30 min at 4 °C. The samples were stored at −80 °C until analysis, in order to minimize sample degradation. Before UPLC/MS analysis, the serum samples were thawed at room temperature and every sample was normatively prepared. 800 μl of methanol was added to each serum sample (200 μl) for one-step protein precipitation. The samples were vortex-mixed for 15 min and centrifuged at 13 000 rpm for 15 min at 4 °C. The supernatant (800 μl) was transferred to a 1.5 ml centrifuge tube and evaporated to dryness under vacuum in a SpeedVac concentrator. Subsequently, 100 μl of methanol was added to dissolve each dried residue by vigorously vortexing for 10 min, the sample was added to each autosampler vial (Waters Ltd, Elstree, UK) and then injected into the UPLC/MS system for analysis. For another, a pooled sample of all 171 samples was prepared as a QC sample and was dissolved in the same way. The QC sample provides a representative analyte containing all the samples that would be encountered during the entire sample sequence in an average length of one QC every 10 injections.

Chromatography conditions

The mass spectrometry detection was performed by the Waters Synapt-G2-Si High Definition Mass Spectrometer (Manchester, UK). The chromatographic analysis was performed in a Waters ACQUITY UPLC system controlled with Masslynx (V4.1, Waters Corporation, Milford, USA). An aliquot of 1 μl of samples solution was injected into an ACQUITY UPLC BEH C18 column (50 mm × 2.1 mm, 1.7 μm, Waters Corporation, Milford, USA) at 50 °C, the flow rate was 0.5 ml min⁻¹. The optimal mobile phase consisted of a linear gradient system of (A) 0.1% formic acid in acetonitrile and (B) 0.1% formic acid in water, 0–0.5 min, 5–20% A; 0.5–2.5 min, 20–60% A; 2.5–4.0 min, 60–66% A; 4.0–5.0 min, 66% A; 5.0–7.5 min, 66–86% A; 7.5–8.0 min, 86–99% A; 8.0–10.0 min, 99% A; 10.0–10.5 min, 99–5% A; 10.5–12.5 min, 5% A. After every sample injection, a needle wash cycle was done to remove the remnants and prepare for the next sample. In addition, the eluent was transferred to the mass spectrometer directly without a split.

Mass spectrometry conditions

The optimal conditions of analysis were as follows: the source temperature was set at 110 °C, desolvation gas temperature was 350 °C, cone gas flow was 50 L h⁻¹ and desolvation gas flow was 1000 L h⁻¹. For positive ion mode, the capillary voltage was 3.0 kV and for negative ion mode, the capillary voltage was 2.5 kV. The data acquisition rate was set to 0.2 s per scan, with a 0.1 s inter scan delay. Data were collected in centroid mode from 50 to 1200 Da. For accurate mass acquisition, a lock-mass of leucine enkephalin at a concentration of 4 ng μl⁻¹ was used via a lock spray interface at a flow rate of 10 μl min⁻¹ monitoring for both ion mode ([M + H]⁺ = 556.2771, [M − H]⁻ = 554.2615) to ensure accuracy during MS analysis.

Data processing

To ensure the repeatability and reproducibility of the data which generated from multiple batches, six peaks were randomly picked out from all QC samples and the coefficient of variation (CV%) were calculated, the paired retention time and m/z of these peaks were 0.55–229.1542 (ESI⁺), 3.71–223.1342 (ESI⁺), 7.88–447.3448 (ESI⁺), 0.88–365.1153 (ESI⁻), 3.37–295.1944 (ESI⁻), 6.54–391.2846 (ESI⁻). The results proved that the coefficient of variation (CV%) were all below 10%, suggesting that the reproducibility is acceptable. The method of metabolomics coupled with pattern recognition and pathway analysis on potential biomarkers in disease had been performed reproducibly and precisely in our previous publications.²⁰ MS data were processed by the Progenesis-QI manager (Waters Corp.) to peak detection and auto-alignment. All of the data were normalized to the summed total ion intensity per chromatogram, and the resultant data matrices were imported into the EZinfo 2.0 software for OPLS-DA. The MassFragment™ application manager (Waters corp, Milford, USA) was used to facilitate the MS/MS fragment ion analysis process by way of chemically intelligent peak-matching algorithms. The reconstruction, interaction, and pathway analysis of potential biomarkers were performed with MetPA software based on the above database sources to identify the metabolic pathways. All statistical analyses were performed by the Student's t-test. Differences with a P value of 0.05 or less were considered significant, assays were performed in triplicate, and the results are expressed as mean ± SD.

Results and discussion

The data of serum samples were extracted by the UPLC-MS system (Fig. 1). All potential metabolites including the retention time, the exact mass, and the MS/MS data were supplied by the chromatographic peaks in the BPI chromatograms. All detect ion peaks were processed for uniformization and alignment by Progenesis QI, and then the data of HLP and healthy groups were investigated by OPLS-DA analysis, through this analysis, the metabolite profile of HLP was gathered within the group and was significantly separated from healthy group. The result of S-plot based on serum profiling data offered the significantly different ions between the HLP and healthy groups which we selected as potential biomarkers (Fig. 2). Potential markers were inferred by peak elemental composition and the MassFragment™ application manager (Waters corp., Milford, USA), for example: azelaic acid's chemical structure and mass fragment information are illustrated in Fig. 3. The precise molecular mass and its mass fragments were detected by mass spectrometer (SYNAPT-G2-Si) and determined within a reasonable degree of measurement error (<5 mDa). Then they were matched to the network database such as Human Metabolome Database (http://www.hmdb.ca/) and MassBank (http://www.massbank.jp/). After all, 37 endogenous metabolites were tentatively identified and summarized in Table S3.† They are glycerophospholipids (GPs), phosphatidylcholine (PC), phosphatidyl ethanolamine (PE), testosterone sulfate, sphingomyelin (SM), azelaic acid, 1-[sn-glycero-3-phospho]-1D-myo-inositol, gamma-glutamyl-beta-cyanoalanine, acetyl-N-formyl-5-methoxykynurenamine, p-cresol, 4-hydroxybenzaldehyde, N-[[3a,5b,7a]-3-hydroxy-24-oxo-7-[sulfooxy]cholan-24-yl]-glycine, 2-phenylethanol glucuronide, uric acid, muricholic acid, sphingosine 1-phosphate, 14,15-epoxy-5,8,11-eicosatrienoic acid, linoleic acid, maslinic acid, oleic acid and chenodeoxycholic acid. To investigate the magnitude of change in the tentatively identified markers, partial relative intensity of the markers were compared between the HLP and healthy participants (Fig. 4). Correlation analysis was analyzed for the further understanding of the disturbed metabolites in hyperlipidemia group and as these can be regarded as potential biomarkers (Fig. 5A). Clustering heatmap (Fig. 5B) manifested the differences of relative value between hyperlipidemia group and healthy group. Referring to the previous experiments, our tentatively identified results were proved valuable, Xu Q. et al.²¹ found the same biomarker “Linoleic acid” as a potential marker for hyperlipidemia in their high-lipid diet rats' plasma and in another experiment, H. Miao et al. found a great number of disturbed lipid compounds such as LysoPC(18:0) and LysoPC(16:0)²² in the hyperlipidemia rats' plasma, which were also tentatively identified in our experiment.

Fig. 1 Base peak intensity chromatograms obtained from the positive ion UPLC-MS analyses of HLP group [A] and health group [B].

Fig. 2 OPLS-DA score plot (A) based on the serum profiling of hyperlipidemia patients [red color] and healthy participants [black color]; S-plot of the biomarkers selection (B), the variables marked [box] are the markers selected as potential biomarkers.

Fig. 3 Chemical structure and major mass fragment information proposed for azelaic acid in the negative mode.

Fig. 4 Comparison of the relative intensity of partial potential biomarkers in hyperlipidemia patients and healthy participants.

Fig. 5 Correlation analysis (A) of the potential metabolites in hyperlipidemia patients and healthy participants. Heat map (B) for identified metabolites in healthy participants and hyperlipidemia patients. The heatmaps were constructed based on the potential candidates of importance, which were extracted with OPLS-DA analysis. The color of each section is proportional to the significance of change of metabolites (red: up-regulated; blue: down-regulated).

To investigate the relative pathways of the biomarkers, KEGG and HMDB were used efficiently. KEGG is a database resource for understanding high-level functions and utilities of the biological system (http://www.kegg.jp/) and the HMDB is a freely available electronic database containing detailed information about small molecule metabolites found in the human body (http://www.hmdb.ca/). Through the KEGG research, a Sankey diagram was sketched for the pathways associated with the potential HLP biomarkers we found in this experiment (as shown in Fig. 6). The result showed that the lipid metabolism is the main disturbed metabolism in HLP patients, such as fatty acids metabolism, linoleic acid metabolism and glycerophospholipid metabolism. At the same time, according to the literature of Z. Y. Li et al., they found 35 biomarkers as the potential markers for hyperlipidemia by analysis of the metabolites of high-fat diet mices, they finally found six pathways were involved which also included the fatty acid metabolism.²³ In our experiment, MetPA software was used to identify the network pathways, the result of MetPA revealed that the potential biomarkers are mainly involved in the pathway of linoleic acid metabolism, glycerophospholipid metabolism, glycosylphosphatidylinositol (GPI)-anchor biosynthesis, sphingolipid metabolism, arachidonic acid metabolism, in which glycerophospholipid metabolism and linoleic acid metabolism were the most important metabolic pathways in HLP patients (Fig. 7). Just like the results showed in our experiment, the experiment by C. Zhou et al. found the same disturbed pathway “glycerophospholipid metabolism” as a key pathway in the hyperlipidemia rats.²⁴ In addition, a potential relationship between all metabolites was illustrated in Fig. 8.

Fig. 6 Sankey diagram for the pathways associated with the potential biomarkers that were found in this experiment.

Fig. 7 Summary of pathway analysis with MetPA tool. (1) Linoleic acid metabolism, (2) glycerophospholipid metabolism, (3) glycosylphosphatidylinositol (GPI)-anchor biosynthesis, (4) sphingolipid metabolism, (5) arachidonic acid metabolism.

Fig. 8 A systemic view of metabolic pathways that associate with hyperlipidemia in this study, providing a disease specific picture of human physiology.

Lipid is the fundamental components of biological membranes, previous reports had shown that the perturbation of lipid metabolism played a critical role in the initiation and progression of HLP and HLP-related diseases.^25–27 In classical biochemical assays, the main lipid changes of HLP involve TC, TG, HDL-C, VLDL-C and LDL-C and in our experiment the metabolic perturbations of glycerophospholipids (GPs), phosphatidylcholine (PC), phosphatidyl ethanolamine (PE), sphingomyelin (SM) were detected in HLP patients' fasting serum. Glycerophospholipids were derived from the hydrolysis of phosphatidylcholine by the regulation of phospholipase and they can play a major role in cell signaling and lipid metabolism, in our experiment we found increased glycerophospholipids such as LysoPC(15:0), LysoPC(18:3(6Z,9Z,12Z)), LysoPC(20:5(5Z,8Z,11Z,14Z,17Z)), LysoPC(18:3(9Z,12Z,15Z)), LysoPC(16:1(9Z)), LysoPC(22:6(4Z,7Z,10Z,13Z,16Z,19Z)), LysoPC(18:0), LysoPC(20:3(5Z,8Z,11Z)), LysoPC(22:5(7Z,10Z,13Z,16Z,19Z)), LysoPC(P-18:0), LysoPC(22:4(7Z,10Z,13Z,16Z)) and decreased glycerophospholipids such as LysoPC(14:0) LysoPC(18:1(11Z)), LysoPC(P-16:0) in HLP patients' serum. Phosphatidylcholine is the major phospholipid component of all plasma lipoprotein classes,²⁸ it is a glycerophospholipid in which a phosphorylcholine moiety occupies a glycerol substitution site. As is the case with diacylglycerols, glycerophosphocholines can have many different combinations of fatty acids of varying lengths and saturation attached at the C-1 and C-2 positions. The disordered of phosphatidylcholine could directly influence the levels of low-density lipoprotein in the blood which could cause other diseases such as chronic kidney disease.^29,30 After all, the disturbed phosphatidylcholines play an important role in hyperlipidemia, atherosclerosis and acute cardiovascular diseases.^31,32 Sphingomyelin (SM) is a type of sphingolipid found in animal cell membranes, especially in the membranous myelin sheath which surrounds some nerve cell axons. It usually consists of phosphorylcholine and ceramide. Serum sphingolipids have close relation with hepatic injury such as chronic hepatitis B (CHB) and HBV-related cirrhosis, what's more, liver is an important part of lipid metabolism.^33–35 The observed changes in the lysoPC, PC and SM appeared to be important feature markers of HLP.

Bile acids are the end products of cholesterol catabolism and are physiological detergents that facilitate excretion, absorption, and transport of fats and sterols in the intestine and liver. Bile acids are also steroidal amphipathic molecules derived from the catabolism of cholesterol. They modulate bile flow and lipid secretion, are essential for the absorption of dietary fats and vitamins, and have been implicated in the regulation of all the key enzymes involved in cholesterol homeostasis. They mainly exist in the enterohepatic circulation and the bile acids from the liver play an important role in nutrient absorption and distribution, metabolic regulation and homeostasis. They are secreted into the bile then promote the lipid emulsification and absorption. The disorders in bile acid metabolism may cause cholestatic liver diseases, dyslipidemia, fatty liver diseases, cardiovascular diseases, and diabetes.^36,37 In this experiment, we found the bile acids such as muricholic acid and chenodeoxycholic acid at an abnormal level, which suggested the imbalance in lipid absorption and metabolism. The level of uric acid in our hyperlipidemia patients was found higher than that of healthy participants. Various epidemiological evidences have shown the increased incidence of hyperuricemia in the subjects with hyperlipidemia and obesity³⁸ and the high uric acid level could cause cardiovascular diseases.^39,40

We also found linolenic acid was higher in hyperlipidemia patients and it is an essential intermediate chain of N-3 polyunsaturated fatty acid which exists in human's tissue cells, it plays an important role in the vivo synthesis and metabolism and it is necessary for life activity factors such as eicosapentaenoic acid (DHA) and docosahexoenoic acid (EPA). The abnormal level of linolenic acid could cause a disturbing state in lipid metabolism then following with a series of complications.^41,42 Arachidonic acid is a polyunsaturated, essential fatty acid. It mediates inflammation and the functioning of several organs and systems either directly or upon its conversion into eicosanoids. Arachidonic acid in cell membrane phospholipids is the substrate for the synthesis of a range of biologically active compounds (eicosanoids) including prostaglandins, thromboxanes, and leukotrienes. These compounds can act as mediators in their own right and can also act as regulators of other processes, such as platelet aggregation, blood clotting, smooth muscle contraction, leukocyte chemotaxis, inflammatory cytokine production, and immune function. According to surveys, these biologically active substances have a very important role in both the human lipid metabolism and cardiovascular system.^43,44 The disordered arachidonic acid metabolites could lead to the lipid metabolism disorders then turn into the higher morbidity of HLP. 14,15-Epoxy-5,8,11-eicosatrienoic acid is a biomarker which we found in the HLP patients' serum and it is a metabolite of arachidonic acid in the human body. The disorder of it could indirectly respond the arachidonic acid is at a disturbed level.

In this study, some other biomarkers were also found in the HLP patients' serum compared to healthy participants. They are p-cresol, beta-D-galactose, acetyl-N-formyl-5-methoxykynurenamine, testosterone sulfate, azelaic acid, 4-hydroxybenzaldehyde, N-((3a,5b,7a)-3-hydroxy-24-oxo-7-(sulfooxy)cholan-24-yl)-glycine, beta-D-galactose, sphingosine 1-phosphate and maslinic acid, 1-(sn-glycero-3-phospho)-1D-myo-inositol. These markers could be as a reference for the clinical diagnosis and mechanism study of HLP.

Conclusion

In this study, a metabolomics approach based on UPLC-MS detection has been successfully established for the metabolites study of hyperlipidemia, 37 metabolites and 11 disordered pathways were measured to distinguish the characterized information between diseased and health status. Among them, large amounts of lipids and fatty acids were found metabolic abnormalities and they had been widely investigated to be closely associated with cardiovascular and steroid diseases. These metabolites and pathways we found may provide a more perspective way to explore the mechanism of HLP. As an incentive of cardiovascular and other disease, it's a fact that the earlier HLP or its associated symptoms are identified, the more likely the treatment is successful, also supports the need of novel diagnostic approaches.

Competing financial interests

The authors declare no competing financial interests.

Acknowledgements

This work was supported by grants from the Key Program of Natural Science Foundation of State (Grant No. 81302905, 81373930, 81430093, 81673586), National Key Subject of Drug Innovation (Grant No. 2015ZX09101043-005, 2015ZX09101043-011), TCM State Administration Subject of Public Welfare of (Grant No. 2015468004), Natural Science Foundation of Heilongjiang Province of China (H2015038), University Nursing Program for Young Scholars with Creative Talents in Heilongjiang Province (UNPYSCT-2015118).

References

1. E. Esteve, W. Ricart and J. M. Fernández-Real, Dyslipidemia and inflammation: an evolutionary conserved mechanism, Clin. Nutr., 2005, 24(1), 16–31.
2. N. B. Ann Marie, E. D. Peterson and R. B. D'Agostino, et al. Hyperlipidemia in early adulthood increases long-term risk of coronary heart disease, Circulation, 2015, 131(5), 451–458.
3. I. A. Sobenin, A. V. Zhelankin and E. V. Chernova, et al. Conventional and novel cardiovascular risk factors and predisposition to the development of atherosclerosis, Patol. Fiziol. Eksp. Ter., 2015, 4–11.
4. F. J. Bosques-Padilla, V. E. Genaro and G. S. Omar, et al. Hypertriglyceridemia-Induced Pancreatitis and Risk of Persistent Systemic Inflammatory Response Syndrome, Am. J. Med. Sci., 2015, 349(3), 206–211.
5. N. Kagansky, S. Levy and Y. Berner, et al. Cholesterol lowering in the older population: time for reassessment?, QJM, 2001, 94(9), 457–463.
6. W. M. Freitas, L. A. Quaglia and S. N. Santos, et al. Low HDL cholesterol but not high LDL cholesterol is independently associated with subclinical coronary atherosclerosis in healthy octogenarians, Aging: Clin. Exp. Res., 2015, 27(1), 61–67.
7. N. B. Ann Marie, E. D. Peterson and R. B. D'Agostino, et al. Hyperlipidemia in early adulthood increases long-term risk of coronary heart disease, Circulation, 2015, 131(5), 451–458.
8. S. S. Martin and E. D. Michos, Mapping hyperlipidemia in young adulthood to coronary risk: importance of cumulative exposure and how to stay young, Circulation, 2015, 131(5), 445–447.
9. M. J. Pencina, N. B. Ann Marie and R. B. D'Agostino, et al. Application of New Cholesterol Guidelines to a Population-Based Sample, N. Engl. J. Med., 2014, 370(15), 1422–1431.
10. K. A. Schulman, S. Balu and S. D. Reed, Specialty Pharmaceuticals for Hyperlipidemia – Impact on Insurance Premiums, N. Engl. J. Med., 2015, 373(17), 1591–1593.
11. S. Hallberg, S. R. Gandra and K. M. Fox, et al. Healthcare costs associated with cardiovascular events in patients with hyperlipidemia or prior cardiovascular events: estimates from Swedish population-based register data, The European Journal of Health Economics, 2015, 1–11.
12. N. Mackinnon, B. S. Somashekar and P. Tripathi, et al. MetaboID: a graphical user interface package for assignment of ¹H NMR spectra of bodyfluids and tissues, J. Magn. Reson., 2012, 226, 93–99.
13. Z. Aihua, S. Hui and Y. Guangli, et al. Metabolomics in diagnosis and biomarker discovery of colorectal cancer, Cancer Lett., 2014, 345(1), 17–20.
14. A. Zhang, H. Sun and X. Wang, Urinary metabolic profiling of rat models revealed protective function of scoparone against alcohol induced hepatotoxicity, Sci. Rep., 2014, 4, 6768.
15. A. Zhang, S. Hui and X. Wu, et al. Urine metabolomics, Clin. Chim. Acta, 2012, 414(4), 65–69.
16. X. Wang, A. Zhang and H. Sun, Future perspectives of Chinese medical formulae: chinmedomics as an effector, OMICS: J. Integr. Biol., 2012, 16(16), 414–421.
17. W. Ping, Y. Quan-Wei and Z. Ai-Hua, et al. Preliminary identification of the absorbed bioactive components and metabolites in rat plasma after oral administration of Shaoyao-Gancao decoction by ultra-performance liquid chromatography with electrospray ionization tandem mass spectrometry, Pharmacogn. Mag., 2014, 10(40), 497–502.
18. J. J. W. Mikkonen, S. P. Singh and M. Herrala, et al. Salivary metabolomics in the diagnosis of oral cancer and periodontal diseases, J. Periodontal Res., 2016, 51(4), 431–437.
19. H. Sun, S. Zhang and A. Zhang, et al. Metabolomic Analysis of Diet-Induced Type 2 Diabetes Using UPLC/MS Integrated with Pattern Recognition Approach, PLoS One, 2014, 9(3), e93384.
20. S. Hui, Z. Ai-Hua and Z. Di-Xing, et al. Metabolomics coupled with pattern recognition and pathway analysis on potential biomarkers in liver injury and hepatoprotective effects of Yinchenhao, Appl. Biochem. Biotechnol., 2014, 173(4), 857–869.
21. Q. Xu, Y. Liu and Q. Zhang, et al. Metabolomic analysis of simvastatin and fenofibrate intervention in high-lipid diet-induced hyperlipidemia rats, Acta Pharmacol. Sin., 2014, 35(10), 1265–1273.
22. H. Miao, H. Chen and S. Pei, et al. Plasma lipidomics reveal profound perturbation of glycerophospholipids, fatty acids, and sphingolipids in diet-induced hyperlipidemia, Chem.–Biol. Interact., 2015, 228, 79–87.
23. Z. Y. Li, L. L. Ding and J. M. Li, et al. ¹H-NMR and MS Based Metabolomics Study of the Intervention Effect of Curcumin on Hyperlipidemia Mice Induced by High-Fat Diet, PLoS One, 2015, 10(3), e0120950.
24. C. Zhou, G. Li and Y. Li, et al. A high-throughput metabolomic approach to explore the regulatory effect of mangiferin on metabolic network disturbances of hyperlipidemia rats, Mol. BioSyst., 2015, 11(2), 418–433.
25. S. Shankar, F. Eoin and G. Shakti, et al. Bioinformatics and systems biology of the lipidome, Chem. Rev., 2011, 111(10), 6452–6490.
26. B. Engelmann, S. Streich and U. M. Schönthier, et al. Changes of membrane phospholipid composition of human erythrocytes in hyperlipidemias. I: Increased phosphatidylcholine and reduced sphingomyelin in patients with elevated levels of triacylglycerol-rich lipoproteins, Biochim. Biophys. Acta, Lipids Lipid Metab., 1992, 1165(1), 32–37.
27. B. Engelmann, U. M. Schönthier and W. O. Richter, et al. Changes of membrane phospholipid composition of human erythrocytes in hyperlipidemias. II: Increases in distinct molecular species of phosphatidylethanolamine and phosphatidylcholine containing arachidonic acid, Biochim. Biophys. Acta, Gen. Subj., 1992, 1165(1), 38–44.
28. L. K. Cole, J. E. Vance and D. E. Vance, Phosphatidylcholine biosynthesis and lipoprotein metabolism, Biochim. Biophys. Acta, Gen. Subj., 2012, 1821(5), 754–761.
29. T. Sangsawang and A. Sriwijitkamol, Type of dyslipidemia and achievement of the LDL-cholesterol goal in chronic kidney disease patients at the University Hospital, Vasc. Health Risk Manage., 2015, 11, 563–567.
30. K. S. Jelena, P. A. Amira and S. Slavica, et al. Hyperlipidemia, oxidative stress, and intima media thickness in children with chronic kidney disease, Pediatr. Nephrol., 2013, 28(2), 295–303.
31. S. Gerd and R. Katharina, Metabolism and atherogenic disease association of lysophosphatidylcholine, Atherosclerosis, 2010, 208(1), 10–18.
32. M. Takayuki, K. Tsuneo and K. Katsuo, Role of lysophosphatidylcholine [LPC] in atherosclerosis, Curr. Med. Chem., 2007, 14(30), 3209–3220.
33. S. J. Zheng, Q. Feng and J. F. Li, et al. Serum sphingomyelin has potential to reflect hepatic injury in chronic hepatitis B virus infection, Int. J. Infect. Dis., 2015, 33, 149–155.
34. P. Nguyen, V. Leray and M. Diez, et al. Liver lipid metabolism, J. Anim. Physiol. Anim. Nutr., 2008, 92(3), 272–283.
35. C. J. Green, D. Johnson and H. D. Amin, et al. Characterization of lipid metabolism in a novel immortalized human hepatocyte cell line, Am. J. Physiol.: Endocrinol. Metab., 2015, 309(6), E511–E522.
36. J. Y. L. Chiang, Bile Acid Metabolism and Signaling, Comprehensive Physiology, 2013, 3(3), 1191–1212.
37. Y. Qi, C. Jiang and C. Jie, et al. Bile acid signaling in lipid metabolism: metabolomic and lipidomic analysis of lipid and bile acid markers linked to anti-obesity and anti-diabetes in mice, Biochim. Biophys. Acta, Mol. Cell Biol. Lipids, 2014, 1851(1), 19–29.
38. H. Nakamura, Association of hyperuricemia with hyperlipidemia and obesity, Nippon Rinsho, 1996, 54(12), 3289–3292.
39. L. A. Hebert and B. Rovin, Uric Acid and Cardiovascular Risk, N. Engl. J. Med., 2009, 360(17), 1811–1821.
40. P. Su, L. Hong and Y. Zhao, et al. Relationship Between Hyperuricemia and Cardiovascular Disease Risk Factors in a Chinese Population: A Cross-Sectional Study, Med. Sci. Monit., 2015, 21, 2707–2717.
41. J. A. Simon, Y. H. Chen and S. Bent, et al. The relation of alpha-linolenic acid to the risk of prostate cancer: a systematic review and meta-analysis, Am. J. Clin. Nutr., 2009, 89(5), 1558S–1564S.
42. A. Pan, M. Chen and R. Chowdhury, et al. α-Linolenic acid and risk of cardiovascular disease: a systematic review and meta-analysis, Am. J. Clin. Nutr., 2012, 96(6), 1262–1273.
43. A. T. Høstmark and A. Haug, Percentages of oleic acid and arachidonic acid are inversely related in phospholipids of human sera, Lipids Health Dis., 2013, 12(2), 106.
44. E. Demetz, A. Schroll and K. Auer, et al. The arachidonic acid metabolome serves as a conserved regulator of cholesterol metabolism, Cell Metab., 2014, 20(5), 787–798.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra24267g

---