Vascular tissue contains kallikrein and kallikrein mRNA, suggesting a vascular kallikrein-kinin system. We questioned whether 1) kallikrein concentration varies among large and small vessels; 2) kallikrein is released by vascular tissue; and 3) blocking protein synthesis inhibits release, suggesting de novo synthesis. Using rat vascular rings and isolated-perfused hindquarters, we examined kallikrein in the bath and perfusate. Active kallikrein was higher in tail arteries than the aorta (P < 0.001); tail veins had six times more kininogenase than the vena cava (P < 0.001). Total kallikrein showed a similar pattern, being highest in tail vessels. Arterial rings released active and total kallikrein. After 1, 2, and 3 h incubation, cumulative release was as follows: active, 90 +/- 13, 201 +/- 25, and 311 +/- 41 pg.h-1 x mg tissue-1; total, 170 +/- 14, 366 +/- 24, and 537 +/- 40 pg.h-1 x mg tissue-1, indicating constant release up to > or = 3 h. In contrast, lactic dehydrogenase fell from 6.7 +/- 2.5 to 2.5 +/- 0.4 U.h-1 x mg tissue-1. Total kallikrein in the rings was 302 +/- 51 pg bradykinin/mg wt tissue before 3 h and 298 +/- 68 afterward. Kallikrein released by the hindquarters after 3 h was as follows: active, 6.2 +/- 2.8 ng bradykinin.min-1 x kg.body wt-1; total, 85.2 +/- 17 ng bradykinin.min-1 x kg body wt-1. Puromycin pretreatment (10 mg ip) reduced total perfusate kallikrein from 105 +/- 19 to 8.5 +/- 3.6 (P < 0.005).(ABSTRACT TRUNCATED AT 250 WORDS)