[en] Introduction
: Metabonomics studies based on spectroscopic anal-
ysis of urine samples is a valuable tool in the discovery of new
biomarkers of drug-induced toxicity. However, urine components
may originate from many different organs, making speculative the
designation of the source of a biomarker detected in urine. In this
work, we have used an isolated and perfused organ approach to
validate the liver origin of some urine biomarkers of hepatotoxicity.
Materials and methods
: In vivo studies: Wistar male rats
(
n
= 5/group) placed in metabolic cages received either a single dose
of hydrazine (100 mg/kg, i.p.) or 2 doses of valproic acid (750 mg/kg,
s.c.). Urine samples collected before and after dosing were analyzed
by
1
H NMR spectroscopy at 400 MHz. After processing and binning,
multivariate data analysis was applied to the NMR data (Umetrics,
SIMCA-P+).
Perfused livers: Livers isolated from male Wistar rats were
perfused with a recycling Krebs-Henseleit solution. Hydrazine or
valproic acid were added to the fluid (same doses as used in vivo),
which was sampled over time for
1
H NMR spectroscopy. NMR spec-
tra of acid extracts of livers were also obtained.
Results
: As already reported by others, our in vivo investiga-
tions confirmed the increased excretions of taurine and/or creatine
in urine of rats treated with either hepatotoxicants. Conflicting
findings were obtained when acid extracts of perfused livers and
perfusion fluids were analyzed, especially in the intracellular levels
of the proposed biomarkers, taurine and creatine, after exposure to
hydrazine. In addition, perfused livers seemed more resistant to
hepatotoxicants as compared to the in vivo situation, a finding that
could not be explained by a lower liver exposure or by differences
in drug metabolism. Our hypothesis is that the stress induced to
liver cells during the isolation procedure in some way protects the
hepatocytes against the toxic assault
