1 In vitro adduct formation of phosgene with albumin and hemoglobin in human blood Daan Noort, † * Albert G. Hulst, ‡ Alex Fidder, † Ronald A. van Gurp, † Leo P.A. de Jong † and Hendrik P. Benschop † † Department of Chemical Toxicology and ‡ Department of Analysis of Toxic and Explosive Substances. TNO Prins Maurits Laboratory, P.O. Box 45, 2280 AA Rijswijk, The Netherlands phone +31 15 2843497; fax + 31 15 2843963; e-mail: noortd@pml.tno.nl Abstract The development of procedures for retrospective detection and quantitation of exposure to phosgene, based on adducts to hemoglobin and albumin, has been initiated. Upon incubation of human blood with [ 14 C]phosgene (0 - 750 μM) a significant part of radioactivity (0-13%) became associated with globin and albumin. Upon Pronase digestion of globin, one of the adducts was identified as the pentapeptide O=C-(V-L)-S-P-A, representing amino acid residues 1-5 of α-globin, with a hydantoin function between N-terminal valine and leucine. Micro-LC/tandem MS analyses of tryptic as well as V8 protease digests identified one of the adducts to albumin as a urea resulting from intramolecular bridging of lysine residues 195 and 199. The adducted tryptic fragment could be sensitively analyzed by means of micro- LC/tandem MS with multiple reaction monitoring (MRM), enabling the detection in human blood of an in vitro exposure level of ≥1 μM of phosgene. Introduction Phosgene is an important intermediate for industrial production of insecticides, isocyanates, plastics, aniline dyes, and resins with an estimated production of almost 1 billion pounds per year. As a result of this extensive usage, thousands of industrial workers are potentially at risk of exposure to phosgene. In addition, phosgene can be formed by the thermal decomposition of chlorinated hydrocarbons during fires, and thus be a potential hazard to firefighters. Reliable diagnosis of exposure to phosgene other than observation of the developing lung edema by means of chest roentgenology is not available. Consequently, precious time is lost before the severity of the intoxication can be assessed, i.e., the very period of time in which dose-related therapy is supposed to be most effective (1). Presently, only passive dosimeters are available for those at risk of accidental exposure to phosgene, which monitor the external dose of phosgene (2). It follows that development of simple, rapid methods of diagnosis and dosimetry of exposure to phosgene is highly worthwile. Our approach towards diagnosis is based on analysis of phosgene adducts with proteins present in the blood, which are presumably formed upon respiratory exposure. It is well known that phosgene is highly reactive towards amino, hydroxyl and thiol groups (3), not only in organic solvents but also in aqueous solution (4). Binding of phosgene, formed from chloroform by metabolic activation, to proteins and endogeneous compounds has already been observed by others. For example, Pohl et al. (5, 6) reported binding of [ 14 C]phosgene formed from [ 14 C]chloroform to microsomal protein. Evidence that inhaled phosgene can actually enter the bloodstream was presented by Sciuto et al. (7). Upon exposure of rodents to a high dose (Ct = 1740 mg.min/m 3 ) of phosgene, a change in absorption at 413 nm in the plasma was observed, indicating that the integrity of the erythrocytes had been affected. In view of the long life time of hemoglobin in erythrocytes (half life time ca.