Short Communication Calvo, Ramos, Fontecha 927 María V. Calvo 1 Lourdes Ramos 2 Javier Fontecha 1 1 Dpto. de Tecnología de los Productos Lµcteos, Instituto del Frío (CSIC), C/JosØ Antonio Novais, 10, Ciudad Universitaria, 28040 Madrid, Spain 2 Dpto. de Anµlisis Instrumental y Química Ambiental, Instituto de Química Orgµnica General (CSIC), C/Juan de la Cierva, 3, 28006 Madrid, Spain Determination of cholesterol oxides content in milk products by solid phase extraction and gas chromatography-mass spectrometry A method for quick and reliable analysis of cholesterol oxidation products (COPs) in dairy products has been developed. After lipid extraction, fat was transesterified under mild conditions to avoid degradation of the target compounds. Isolation of the COPs studied from other components in the lipid fraction was carried out on an ami- nopropyl SPE cartridge. Finally, analytes were analysed by GC-MS without derivati- sation. The method developed provides high specificity and good sensitivity, allowing the direct and unambiguous determination of the underivatized COPs investigated. Application of the method to dairy food analysis revealed the presence of COPs in powder milks and milk-based infant formulas commercially available in Spain, show- ing that attention should be focused on reduction of cholesterol oxidation levels during both the processing and the storage of these types of foodstuffs. Key Words: Cholesterol oxidation products (COPs); SPE; GC-MS; Milk products; Received: April 6, 2002; revised: September 19, 2002; accepted: January 3, 2003 DOI 10.1002/jssc.200301249 1 Introduction Cholesterol is an unsaturated lipid susceptible to oxidation under a variety of conditions to form so-called oxysterols. Foodstuffs containing cholesterol, particularly those that have been exposed to heat and air during processing or shelf-life, could contain cholesterol oxidation products (COPs) that display a variety of adverse biological effects [1 – 4]. Determination of COPs is still considered to be an analytical challenge due to their low concentrations (a lg/g lipid) in lipid extracts, which necessitates their separation from the main components of the fat to ensure proper determination. In addition, the great structural simi- larity of the different COPs mandates the use of highly selective methods for their unambiguous isolation and determination. Also special attention should be focused on avoiding artefact formation during sample processing. Some alternatives to the tedious and time-consuming pro- cedures [5 – 7] typically used for the determination of COPs have been described. Despite their satisfactory sensitivity and reproducibility, none of them can be con- sidered as really adequate for routine analysis [8 – 9]. Methods focusing on the determination of a single COP [10], which may limit their practicability when analys- ing real-life samples, and an expensive approach using labelled cholesterol standard [11] to prevent interference from artefacts, have also been reported. This variety of methods has sometimes yielded discrepancies among the COP levels determined by different laboratories. The aim of this work was to develop an analytical method which is sufficiently straightforward, fast, and accurate to be routinely used for the isolation and determination of all major COPs in milk products. 2 Materials and methods 2.1 Chemicals All solvents were of analytical grade (Panreac, Barcelona, Spain). 5a-Cholestane, cholesterol, 25-hydroxycholes- terol, 19-hydroxycholesterol, 6-ketocholestanol, 5a,6a- epoxycholesterol, 7-ketocholesterol, and cholestanetriol (Sigma Chemical, St. Louis MO, USA) were used as stan- dards (A 99% purity). Stock solutions (2 lg/lL) of each standard in chloroform and mixtures containing 100 lL of each individual standard were used for method optimisa- tion. A butter fat sample spiked with COPs and several milk products from local markets in Madrid (Spain) were analysed. 2.2 Sample preparation Lipids were extracted according to Hara and Radin [12] but with slight modifications. Briefly, 1 g of milk powder dissolved in 10 mL of deionised water (or 10 mL of milk for liquid samples) was vigorously mixed with 12 mL of iso- propanol for 3 min. Afterwards, 9 mL of hexane was added and the tube was shaken for 3 min. The mixture was centrifuged at 2 000 g for 10 min and the hexane Correspondence: María V. Calvo, Dpto. de Tecnología de los Productos Lµcteos, Instituto del Frío (CSIC), C/JosØ Antonio No- vais, 10, Ciudad Universitaria, 28040 Madrid, Spain. Phone: +34 915 445 607/915 492 300. Fax: +34 915 493 627. E-mail: mcalvo@if.csic.es. J. Sep. Sci. 2003, 26, 927–931 i 2003 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim 1615-9306/2003/0907–0927$17.50+.50/0