A metabonomic approach to the investigation of drug-induced phospholipidosis: an NMR spectroscopy and pattern recognition study.

(1)H NMR spectroscopy of urine and pattern recognition analysis have been used to study the metabolic perturbations caused following dosing of five novel drug candidates, two of which (GWA, GWB) caused mild lung and liver phospholipidosis, whilst the rest (GWC-GWE) did not cause any detectable toxicity. Urine samples were collected predose, 0-8 h, 8-16 h, 16-24 h and 24-32 h after single, oral dosing with each compound to Han Wistar rats (n = 3 per group), and liver and lung samples for were taken at 48 h for histology. (1)H NMR spectra of whole urine were acquired, processed and subsequently analysed using principal component analysis. All animals administered the drug candidates showed a significant reduction in serum triglycerides and those animals administered either GWA or GWB were observed to have foamy alveolar macrophages and the presence of multilamellar bodies in hepatocytes by electron microscopy. In the plot of the first two principal components, urinary spectra of those animals dosed with GWA or GWB mapped separately to controls, all pre-dose samples and animals dosed with GWC-GWE. Inspection of the principal components loadings indicated an increase in urinary phenylacetylglycine with a concomitant decrease in urinary citrate and 2-oxoglutarate, possibly constituting a novel urinary biomarker set for phospholipidosis. This work exemplifies the use of NMR spectroscopy and pattern recognition methods for the detection of novel biomarker combinations for poorly understood toxicity types and the potential in screening novel drugs for toxicity.