New biomarkers for early diagnosis of Lesch-Nyhan disease revealed by metabolic analysis on a large cohort of patients

Irène Ceballos-Picot, Aurélia Le Dantec, Anaïs Brassier, Jean-Philippe Jaïs, Morgan Ledroit, Julie Cahu, Hang-Korng Ea, Bertrand Daignan-Fornier, Benoît Pinson

BackgroundLesch-Nyhan disease is a rare X-linked neurodevelopemental metabolic disorder caused by a wide variety of mutations in the HPRT1 gene leading to a deficiency of the purine recycling enzyme hypoxanthine-guanine phosphoribosyltransferase (HGprt). The residual HGprt activity correlates with the various phenotypes of Lesch-Nyhan (LN) patients and in particular with the different degree of neurobehavioral disturbances. The prevalence of this disease is considered to be underestimated due to large heterogeneity of its clinical symptoms and the difficulty of diagnosing of the less severe forms of the disease. We therefore searched for metabolic changes that would facilitate an early diagnosis and give potential clues on the disease pathogenesis and potential therapeutic approaches.MethodsLesch-Nyhan patients were diagnosed using HGprt enzymatic assay in red blood cells and identification of the causal HPRT1 gene mutations. These patients were subsequently classified into the three main phenotypic subgroups ranging from patients with only hyperuricemia to individuals presenting motor dysfunction, cognitive disability and self-injurious behavior. Metabolites from the three classes of patients were analyzed and quantified by High Performance Ionic Chromatography and biomarkers of HGprt deficiency were then validated by statistical analyses.ResultsA cohort of 139 patients, from 112 families, diagnosed using HGprt enzymatic assay in red blood cells, was studied. 98 displayed LN full phenotype (86 families) and 41 (26 families) had attenuated clinical phenotypes. Genotype/phenotype correlations show that LN full phenotype was correlated to genetic alterations resulting in null enzyme function, while variant phenotypes are often associated with missense mutations allowing some residual HGprt activity. Analysis of metabolites extracted from red blood cells from 56 LN patients revealed strong variations specific to HGprt deficiency for six metabolites (AICAR mono- and tri-phosphate, nicotinamide, nicotinic acid, ATP and Succinyl-AMP) as compared to controls including hyperuricemic patients without HGprt deficiency.ConclusionsA highly significant correlation between six metabolites and the HGprt deficiency was established, each of them providing an easily measurable marker of the disease. Their combination strongly increases the probability of an early and reliable diagnosis for HGprt deficiency.