No abstract

Saikosaponins (SSs) are a class of triterpene saponins with a wide spectrum of bioactivities. A sensitive liquid chromatography tandem mass spectrometry (LC-MS/MS) method was developed for simultaneous determination of saikosaponin a, saikosaponin c, saikosaponin d and saikosaponin b₂ in rat plasma. Plasma samples were prepared by liquid-liquid extraction. The analytes and the internal standard (IS) digoxin were well separated on an octadecyl column using gradient elution and analyzed by monitoring the fragmentation transition pair of anionic adducts to deprotonated molecules in negative-mode electrospray. By neutral loss of HCOOH, the transition pairs of m/z 825 → 779 for SSa, SSd, SSb₂ and the IS, and m/z 971 → 925 for SSc were sensitive for MS/MS detection with the lower limits of quantification in the range of 0.20-0.40 ng/mL. Method validation experiments were performed, including selectivity, precision, accuracy, linearity, matrix effect, recovery and stability. The validated method was further applied to determine the pharmacokinetics parameters of SSa, c and d in rats following a single oral administration of the extract of chaihu (the dried roots of Bupleurum chinense DC).

Rationale:Discriminating between aglycone-substituted and saccharide-substituted saikosaponins by liquid chromatography/tandem mass spectrometry (LC/MS n ) is a long-standing issue that is still to be resolved. It is necessary to characterize the two types of substituted saikosaponins taking into consideration the potential significant difference in their bioactivity. Methods:Taking the substituents malonyl and acetyl as examples, we developed a MS strategy to discriminate between the aglycone-substituted and saccharidesubstituted saikosaponins through comparing their Y 0 − nH 2 O (n = 1-2) ions from the protonated molecules in the full-scan mass spectra and their B ions in the MS 2 spectra of sodium-adduct molecules in the positive mode. Results:The deprotonated molecules of the aglycone-substituted saikosaponins presented similar fragmentation patterns to those of saccharide-substituted ones in the negative mode, which could not discriminate whether the substitutes were located on the aglycone or the saccharide. In contrast, the Y 0 − nH 2 O (n = 1-2) ions containing or no substituent were observed respectively in the mass fragmentation of the protonated molecules of aglycone-substituted or saccharide-substituted saikosaponins in the positive mode. In addition, the B ions containing or no substituent were observed respectively in the mass fragmentation of the sodiumadduct molecules of the saccharide-substituted or aglycone-substituted saikosaponins in the positive mode. Two aglycone-malonylated saikosaponins were reported for the first time. Conclusions:Whether the substituents were located on the aglycone or the saccharide could be determined according to the Y 0 − nH 2 O (n = 1-2) ions from the protonated molecules in the full-scan mass spectra and the B ions in the MS 2 spectra of sodium-adduct molecules in the positive mode. Our results have updated the mass fragmentation patterns of substituted saikosaponins, which is helpful for the quality control of pharmaceutical preparations containing saikosaponins. More importantly, this MS strategy should be able to be extended to characterize other substituted saponins of bioactive significance in future studies.

An analytical method was developed for the determination of L-carnitine in milk and dairy products using hydrophilic interaction chromatography-tandem mass spectrometry (HILIC-MS/MS). The samples were extracted with 2% (v/v) acetic acid solution, and the protein was precipitated with acetonitrile subsequently. The separation of L-carnitine was carried out on an Acquity UPLC BEH HILIC column using ammonium acetate and acetonitrile as mobile phases. The quantification analysis of the target compound was performed under multiple reaction monitoring (MRM) mode by external standard method. A good linear relationship was obtained between the peak area and concentration of L-carnitine in the range of 1-100 µg/L with the correlation coefficient more than 0.99. The limit of quantification (LOQ ) of L-carnitine was 0. 01 mg/kg. The spiked recoveries were 96.0%-103.4%. The precisions (RSDs ) were 1.2%-4.3%. The sample preparation was simple and rapid, and the results were precise and sensitive. The developed method is suitable for the study of concentration of L-carnitine in milk and dairy products, and the technical support for the infant formula is provided.