Chemiluminescence Arising From the Oxidation of Bilirubin in Aqueous Media Leonidas P. Palilis 1 , Antony C. Calokerinos 1 and Nikos Grekas 2 1 University of Athens, Laboratory of Analytical Chemistry, Panepistimiopolis, Zogfrafou, 157 71 Athens, Greece 2 FARMALEX S. A., Research and Development Laboratory, Tsocha 15–17, 115 10 Athens, Greece Biomed. Chromatogr. 11, 71–72, (1997) No. of Figures: 2. No. of Tables: 1. No. of Refs: 2. INTRODUCTION Bilirubin (BIL) is a compound of great biological interest since it is found in human plasma as metabolic product of haemoglobin and haeme proteins. Analytical methods currently used for the determination of bilirubin are, usually, spectrophotometric (direct or after diazo reaction), enzy- matic and high-performance liquid chromatography (HPLC) methods (Westwood, 1991). The knowledge of the chemical properties of BIL is limited mainly due to the lack of stability and solubility of the compound in water. Nevertheless, it was decided to investigate the possibility of developing a chemilumines- cence or bioluminescence analytical procedure for bilirubin. Our initial target was to find chemiluminogenic reactions of BIL in aqueous solutions. It was found that bilirubin generates chemiluminescence during the action of various oxidizing agents, such as N-bromosuccinimide, hypo- chlorite and ferricyanide, in alkaline solutions (Palilis et al., 1996). The severe interference from albumin diminishes the value of this procedure, which can only be applied after a separation step. Another idea for reducing interference was to use the enzymatic reaction of bilirubin with bilirubin oxidase and investigate whether this reaction is bioluminogenic. Although a weak radiation was generated after mixing the two reactants, the signal was not analytically useful. Our attempts were, therefore, focused on the enhance- ment of this weak radiation, using several molecules as fluorophores. It was found that luminol generated radiation during the action of bilirubin oxidase. The intensity of the radiation is relatively constant for several minutes and is reduced when BIL is injected into the solution. The height of the ‘negative’ peak allows the measurement of BIL with excellent sensitivity and acceptable selectivity. The proce- dure was found to suffer from less interferences and was further investigated. This communication describes the use of the enzymatic reaction of BIL with bilirubin oxidase in order to investigate whether this reaction is bioluminogenic with less inter- ference from other species. RESULTS AND DISCUSSION Bilirubin oxidase is an enzyme specific for the oxidation of BIL through the following process: Bilirubin + O 2 ________ → Bilirubin oxidase Biliverdin + H 2 O Biliverdin + O 2 ________ → Purple compund(s) Since radiation is emitted immediately after mixing bilirubin oxidase with luminol and since hydrogen peroxide is not included in the reacting mixture, the reaction pathway is different from the well known chemiluminescence reaction of luminol with hydrogen peroxide. Enzymes such as cholesterol oxidase, glucose oxidase and horseradish peroxidase, all of which were used, did not show similar behaviour, under the same experimental conditions, for the measurement of BIL. Hence the reduction of the emission intensity was used as the analytical signal. Measurement procedure A buffer solution of luminol was transferred to the reaction cell of a batch chemiluminometer, followed by a portion of bilirubin oxidase. After the signal reached the maximum level, an aliquot of BIL was injected (Fig. 1). Effect of pH value Luminol is soluble in alkaline solution and its chemilumino- genic properties appear at pH 9–13, with the best response at a pH around 11. On the other hand, bilirubin oxidase catalyses the oxidation of both conjugated and unconjugated bilirubin at less basic pH values (optimum pH 8.2), while in an acidic medium it catalyses only conjugated bilirubin. Hence the effect of pH on the decrease of the emission intensity was examined as a compromise between reactivity of luminol and action of the enzyme (Fig. 2). The optimum  Correspondence to: L. P. Palilis Figure 1. Schematic diagram of recorded negative peaks. CCC 0269–3879/97/020071–02 $17.50 Received 30 April 1996 © 1997 by John Wiley & Sons, Ltd. Accepted 21 May 1996 BIOMEDICAL CHROMATOGRAPHY, VOL. 11, 71–72 (1997)