Utilization of targeted array comparative genomic hybridization, MitoMet®, in prenatal diagnosis of metabolic disorders

Abstract

Metabolic disorders are inborn errors that often present in the neonatal period with a devastating clinical course. If not treated promptly, these diseases can result in severe, irreversible disease or death. Determining the molecular defects in metabolic diseases is important in providing a definitive diagnosis for patient management. Therefore, prenatal diagnosis for families with known mutations causing metabolic disorders is crucial for timely intervention. Here we present three families in which standard Sanger sequencing failed to provide a definitive diagnosis, but the detection of genomic deletions by array comparative genomic hybridization (CGH) specifically targeted to mitochondrial and metabolic disease genes, MitoMet®, was fundamental in providing accurate prenatal diagnosis. In addition, to our knowledge, two deletions are the smallest detected by oligonucleotide array CGH reported for their respective genes, OTC and ARG1. These data highlight the importance of targeted array CGH in patients with suspected metabolic disorders and incomplete or negative sequencing results, as well as its emerging role in prenatal diagnosis.

Introduction

Many inborn errors of metabolism, including defects in the urea cycle, fatty acid oxidation, and organic acidemias present early on in the neonatal period. Common symptoms often include poor feeding, lethargy, and vomiting. Hyperammonemia and metabolic acidosis are often observed in a number of these conditions. If not recognized and managed early, metabolic diseases can have severe consequences. Therefore, identifying the causative molecular lesions of a metabolic disorder allows for optimal management. Prenatal diagnosis of metabolic disorders is often requested for the identification of affected fetuses for elective termination purposes, or for disease management immediately after birth. Until recently, enzymatic or analyte testing was performed for prenatal diagnosis of some metabolic disorders such as arginase deficiency or ornithine transcarbamylase deficiency (OTCD) [1].

However, some of these methods are not commonly available, can be technically challenging, or require tissue types that are difficult to test prenatally. For example, measuring enzyme activity for disorders such as OTCD requires a liver biopsy. This procedure is extremely invasive and the specificity is unclear due to the variability of fetal expression of urea cycle enzymes [2]. Arginase is present in fetal erythrocytes at 16–20 weeks' gestation, yet is not expressed in chorionic villi samples (CVS) or amniotic fluid (AF). As a result, prenatal testing for arginase deficiency requires fetal blood and cannot be performed prior to 16 weeks gestation [2]. Therefore, if the molecular defect is known, genetic analysis of the affected gene is the method of choice [2]. Sequence analysis remains the gold standard for a majority of mutations in genes involved in metabolic disorders. However, sequence analysis does not detect large deletions, duplications, and rearrangements. In autosomal recessive (AR) disorders, if only one point mutation is identified the disease cannot be confirmed, and prenatal diagnosis genotyping is incomplete or inaccurate. For example, a fetus carrying a heterozygous point mutation and a deletion in different regions of a gene would appear to only be a heterozygous carrier of the point mutation. X-linked disorders such as OTCD can often be difficult to confirm using sequence analysis if a deletion is the only molecular defect. In these cases, array CGH analysis is usually recommended. Indeed, array CGH has been successfully used in the identification of deletions in metabolic genes for whom point mutations were not identified [3], [4], [5].

In addition, even if the deletion junction region is known, it is difficult to assess a fetus using PCR due to the risk posed by possible maternal cell contamination (MCC) and PCR amplification bias. For example, if a mother is the carrier of a deletion and the fetal sample shows 10% MCC or less, there is still a possibility of PCR amplification of the deletion from a small amount of the maternal sample. This would lead to a false-positive result for the fetus. While array CGH technology is capable of detecting low-level mosaicism at 10–20% [6], its resolution is not sensitive enough to yield a false-positive result for MCC at 10% or lower.

Here we report the clinical presentation and molecular characterization of deletions in both X-linked and AR metabolic genes detected by MitoMet®, a custom designed CGH array (version 2.5) targeted to 192 nuclear genes involved in mitochondrial biogenesis and metabolic pathways, plus the whole mitochondrial genome. The deletions presented here were identified following sequence analysis that was either negative, failed to amplify particular exons, or could not confirm the expectation of a heterozygous point mutation in a parent of a proband. Two of these deletions encompass only a single exon from their respective genes (OTC and ARG1), and at 666 bp and 240 bp respectively, appear to be the smallest microdeletions detected by array CGH ever reported for OTC and ARG1. After identification and confirmation of familial deletions, MitoMet® array CGH was successfully utilized for prenatal detection in fetal samples. These data demonstrate the utility of targeted oligonucleotide arrays CGH in the prenatal diagnosis of metabolic disorders.