Pharmacokinetics, methionine depletion, antigenicity, and toxicity of recombinant methioninase (rMETase), which has shown efficacy in achieving cell kill in a broad range of human tumor models, were examined in macaque monkeys. Dose-ranging studies at 1000, 2000, and 4000 units/kg i.v. identified the 4000 units/kg dose as able to reduce plasma methionine to an undetectable level (less than 0.5 microM) by 30 min, and the level so remained for 8 h. Pharmacokinetic analysis showed that rMETase was eliminated with a T(1/2) of 2.49 h. A 2-week i.v. administration of 4000 units/kg every 8 h/day for 2 weeks resulted in a steady-state depletion of plasma methionine to less than 2 microM. The only manifest toxicity was decreased food intake and slight weight loss. Serum albumin and red cell values declined transiently during treatment, which may be related to extensive blood sampling. Re-challenge on day 28 resulted in anaphylactic shock and death in one animal. Subsequent pretreatment with hydrocortisone prevented the anaphylactic reaction, although vomiting was frequently observed. Re-challenge was carried out at days 66, 86, and 116. Anti-rMETase antibodies (at 10(-3)) were found after the first challenge, and these increased to 10(-6) after the fourth challenge and decreased to 10(-2) by 2 months post therapy. The main rMETase antibody was IgG, and although it has some in vitro features of being a neutralizing antibody, each challenge dose was effective in depleting plasma methionine levels. Thus, rMETase was able to effectively deplete plasma methionine levels with minimal toxicity in a primate model. These data provide the bases for alteration by polyethyleneglycol conjugation (PEGylation) of the enzyme to increase its duration of effect and reduce its immunogenicity.