J. Sep. Sci. 2007, 30 S. S. Jacob et al. Shana Sunny Jacob Norman William Smith Cristina Legido-Quigley Pharmaceutical Sciences Research Division, King's College London, UK Original Paper Assessment of Chinese medicinal herb metabolite profiles by UPLC-MS-based methodology for the detection of aristolochic acids In this study, aristolochic acid in different herbal medicines containing a mixture of varying herb species was identified through fingerprint pattern similarities. Ari- stolochic acid I and II are nephrotoxic compounds naturally present in the Aristolo- chia plant species that are commonly used in Chinese herbal medicines. Twenty- four commercially available herbal formulations were extracted into an aqueous solution and injected into a UPLC-MS system. All the samples were analysed by mul- tiple reaction monitoring (MRM) to check for the presence of aristolochic acids I and II. The same samples were then fingerprinted using two different gradient methods and the chromatograms deconvoluted into retention time (RT) and masses for the chemicals present taking concentration into account. Statistical analysis of this data revealed that samples were highly heterogeneous, and that the main diffe- rences between the preparations were concentrations of polar compounds. A model was constructed where the samples could be separated into two groups differentia- ted by the presence of the two forms of aristolochic acid. Keywords: Chemometrics / Fingerprinting / Herbal medicines / UPLC-MS / Received: December 13, 2006; revised: January 29, 2007; accepted: January 30, 2007 DOI 10.1002/jssc.200600517 1 Introduction Chromatographic fingerprinting is a way to represent the chemical information from a sample, in this case Chi- nese herbal medicines. It is an integrated method that uses techniques such as LC-MS in combination with che- mometrics, which is the application of statistical tech- niques to extract the chemical information out of com- plex datasets. The analytical technique that we used was ultra-per- formance LC (UPLC). This is a relatively new technique for fingerprinting, and to the authors' knowledge, this is the first time that the profile assessment of herbal pre- parations has been attempted with UPLC-MS. UPLC, as the name states, generates higher chromatographic per- formance, by using smaller stationary phase particle size (1.7 lm) columns. This decreases band-broadening, thereby giving high efficiency to the separation, which concurrently increases resolution and sensitivity. When UPLC is hyphenated to a mass spectrometer (MS), super- ior resolution and a higher S/N is achieved. LC-MS and UPLC-MS produce complex 3-D chromatograms as we deal with mass, retention time (RT) and concentration for all the separated metabolites. Thus, peak assignment and identification of compounds is difficult, the data files are large and treatment using statistical methods is necessary for the elucidation of the key metabolites. This approach was applied to a group of 24 diverse her- bal preparations. Plants are a fabulously rich source of diverse functional biochemicals, and provide value in this applied context. By creating unique opportunities to explore plant systems and characterise their biochemical composition, fingerprint profiling will greatly assist in identifying and defining much of the still unexploited biodiversity available today [1]. Metabolic profiling tech- niques can simultaneously track changes in many plant metabolites and have found a number of applications in both biology and biosafety [2]. Together with multivari- ate statistical analyses, NMR spectroscopy has already been successfully applied to the characterisation of vari- ous herbs and plant products for quality control, authen- tication, determining geographical origin and for detect- ing adulteration of products [2]. It was recently stated that this approach is considered to have the potential to Correspondence: Dr. Cristina Legido-Quigley, Franklin-Wilkins Building, 150 Stamford Street, London SE1 9NH, UK E-mail: cristina.legido_quigley@kcl.ac.uk Fax: +44-20-78484800 Abbreviations: MRM, multiple reaction monitoring; PCA, prin- cipal component analysis; PLS-DA, partial least square discrimi- nant analysis; RT, retention time i 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.jss-journal.com