The goal of this study was to quantify in the dog the error that is made in assessing drug tissue concentrations when no correction for blood contamination is performed and hence to determine in which organs such a correction should be made. The organs investigated were the heart, the brain, the liver and the skeletal muscle, and the test drug used was the H1-antihistamine, cetirizine (0.1 or 0.6 mg/kg/day for 3 days, orally, n = 6 dogs). Radiolabelled serum albumin was used to quantitate blood trapped in the tissues. Blood and tissue samplings were performed 2 h after the last drug administration. Mean (+/-SEM) volumes of blood trapped in the liver, heart, muscle and brain were 263 +/- 12, 91 +/- 3, 27 +/- 1 and 20 +/- 2 microL/g, respectively. Apparent tissue/blood concentration ratios of cetirizine were 2.39 +/- 0.33, 1.11 +/- 0.09, 0.77 +/- 0.07 and 0.37 +/- 0.05 in the four organs. When correction for residual blood is not performed, cetirizine concentrations are underestimated (-13.6 +/- 3.2%) in the liver, slightly overestimated (+4.7 +/- 1.5 to +6.3 +/- 2.8%) in the brain, and neither over nor underestimated in the heart and muscle. Simulation data over a wide range of theoretical drug tissue/blood concentration ratios indicate that in the dog: (a) for the liver, correction of apparent tissue concentration for residual blood should be performed when the drug tissue/blood concentration ratio achieved is <0.8 or >4, (b) for the heart, correction should be made when this ratio is < or =0.4 and (c) for the brain and muscle, no correction is necessary unless the ratio is < or =0.1.

Therapeutic angiogenesis via administration of recombinant growth factors or vector-mediated overexpression has been largely unsuccessful in part due to the inadequacy of a single growth factor to induce all aspects of vascular growth. We have previously shown that human tissue kallikrein (hK1) promotes angiogenesis by kinin-mediated activation of the Akt-eNOS pathway, independently of VEGF. To further characterise the cellular mechanisms of hK1 induced neovascularisation, we delivered adenovirus encoding the wild type kallikrein gene ( Ad.KLK1 , 5x10 6 or 5x10 5 pfu) or a polymorphic variant ( Ad.R53H-KLK1 ), which is 50% less potent as a kinin generating enzyme, or control virus ( Ad.eGFP ) in the rat mesenteric assay, a non-ischemic model of angiogenesis. In separate experiments, Ad.KLK1 was injected into the mesentery of rats given the bradykinin-2-receptor (B2R) antagonist, Icatibant (Subc 300nmol/kg/day). Intravital and confocal imaging was used to examine vessel morphology and histology. Ad.KLK1 increased the functional vessel area (445±76 vs 32±4% in Ad.eGFP, p<0.01), conduit vessel density (193±12 vs 63±7mm −2 , p<0.001) and average vessel diameter (15±1 vs 12.6±1μm, p<0.001). Ad.KLK1 -induced neovasculature showed an increased pericyte (58±5% vs 31±5, p<0.05) and smooth muscle cell coverage (4±1 vs 0%, p<0.001). In addition, we found that B2R co-localises with pericytes in the mesentery. Following Icatibant, Ad.KLK1 stimulated the formation of a network of small-size vessels, without pericyte or smooth muscle cell coverage and surrounded by spots of micro-hemorrhaging. Ad.R53H-KLK1 was less efficient than Ad.KLK1 in promoting angiogenesis and recruiting vascular smooth muscle cells. In a model of limb ischemia, disruption of the kallikrein gene resulted in dysfunctional reparative arteriogenesis and delayed hemodynamic recovery, which were both corrected by local Ad.KLK1 but not by Ad.R53H-KLK1. These data provide a mechanistic explanation for the robust arteriogenic effect of Ad.KLK1 , indicating a key role for the kinin B2R. We also demonstrate for the first time that a single nucleotide polymorphism in the KLK1 gene, present in 5–7% of the Caucasian population, leads to a reduced neovascularisation in-vivo . All data mean±SEM, n=6