Arginine deprivation causes death of up to 80% of cancer cell lines in vitro, but in the body, citrulline would be available as a convertible source of this amino acid in vivo. Some tumour cell lines, notably the vast majority of melanomas and hepatocellular carcinomas, tend to be deficient in argininosuccinate synthetase (EC 6.5.4.3.), and therefore cannot recycle citrulline to arginine. Argininosuccinate synthetase is present at levels that convert enough citrulline to arginine to allow limited growth in about half of a modest range of malignant cell types analysed in this study. Attempts to rescue cells that are unable to utilise citrulline with the immediate downstream product, argininosuccinate, had very limited success in a few tumour cell lines. Particularly noteworthy is the demonstration that argininosuccinate was totally incapable of rescuing cells that utilise citrulline efficiently, consistent with tight channelling (coupling) of argininosuccinate synthetase and argininosuccinate lyase in the urea cycle. The findings suggest that an excellent opportunity exists for further exploitation of arginine deprivation in the selective killing of tumour cells. Arginine is a vital amino acid for many metabolic processes other than protein synthesis (e.g. creatine production, polyamine synthesis and nitric oxide (NO) generation). Its removal from culture medium by medium formulation or arginase treatment quickly leads to death in B80% of tumour cell lines (Scott et al, 2000). The addition of citrulline can often circumvent this deficiency, but few cells are capable of its biosynthesis in culture. Arginine deficiency in vivo has been induced by various means, but is less effectively achieved because body homeostasis is too robust, and citrulline generation is difficult to control. Having successfully reduced L1210 cells in culture to negligible viability in 3-5 days with arginase (EC 3.5.3.1), citrulline supplementation in the continued presence of the enzyme partially circumvented the arginine deficiency. The rate of conversion became a rate-limiting factor, most noticeably in such fast-growing leukaemia as L1210 (Philip et al, 2003). In a corresponding in vivo study, arginase treatment of DBA/2 mice injected intraperitoneally with 1210 cells increased neither their survival time nor decreased tumour burden, despite plasma arginine falling to B1-2 mM level. Plasma citrulline remained normal, and therefore the tumour cells must have converted citrulline to arginine fast enough to sustain relatively normal tumour growth rate (Wu and Morris, 1998;Philip et al, 2003). This makes it difficult to understand how previous in vivo work with arginase and arginine deiminase could have achieved significant inhibition of tumour growth under similar conditions (Bach and Swaine, 1965;Miyazaki et al, 1990), and begs further analysis of the ability of tumour cells to use citrulline in lieu of arginine (Wheatley and Campbell, 2002). Sugimura et al (1992) found that four out of five human melanoma cell lines deprived of arginine w...