X marks the spot

Mental retardation has been linked to mutations on the X-chromosome – however the genes on this chromosome are not fully understood, making screening a real challenge  

MENTAL RETARDATION affects a larger number of males in the population than females, and this has been attributed, in part, to the increased proportion of mutations in genes on the X-chromosome.

The number of causative mutations leading to X-linked mental retardation (XLMR) is unknown, and will remain so, until all of the genes on the X chromosome have been fully investigated. While this research is ongoing, screening for XLMR remains a challenge.

XLMR is estimated to affect approximately 1 in 1,000 males, and can be subdivided into two groups: syndromic X-linked mental retardation (MRXS), characterised by external features and/or neuromuscular and metabolic disorders; and non-syndromic X-linked mental retardation (MRX), where intellectual disability and the X-linked mode of inheritance is the only consistent indication. To date, there have been approximately 90 X-linked genes implicated in XLMR, with the majority of these genes associated with MRXS, while a subset of genes on the X chromosome have been associated with MRX. Significant overlap between these two sets of genes indicates that both syndromic and non-syndromic cases can result from alterations in many of the XLMR genes. MRX is thought to account for two thirds of XLMR cases, but the identification of males with MRX is often difficult without a positive family history.

To screen for MRX, the molecular diagnostic laboratory at the Greenwood Genetic Center (Greenwood, South Carolina) has developed a panel of sequencing primers for the interrogation of nine genes that have been most commonly associated with MRX. The panel provides sequence coverage of 95 amplicons encoding the exons and intron junctions for the following genes: acyl-CoA synthetase long-chain family member 4 (ACSL4); aristaless related homeobox (ARX), FtsJ homolog 1 (FTSJ1), GDP dissociation inhibitor 1 (GDI1), interleukin 1 receptor accessory protein-like 1 (IL1RAPL1), jumonji, AT rich interactive domain 1C (JARID1C), oligophrenin 1 (OPHN1), polyglutamine binding protein 1 (PQBP1), and zinc finger protein 41 (ZNF41).

For this study, extracted genomic DNA from 19 patients suspected of having MRX (obtained from the Greenwood Genetic Center) was directly sequenced on the Applied Biosystems 3500xL Genetic Analyser using the Greenwood MRX resequencing panel. Resequencing (also known as comparative, direct, medical, or PCR sequencing) is commonly used in medical research for the identification of causal or associated disease variants in specific genes or multigene disorders, such as MRX. Capillary electrophoresis-based platforms are commonly employed in these cases because it allows direct detection of DNA variants, including single nucleotide polymorphisms (SNPs) as well as small insertions and deletions. The Applied Biosystems fast resequencing workflow was used for the analysis of the 19 samples.

Sequencing data were analysed for missense variants (non-synonymous SNPs) using Applied Biosystems variant reporter software. In the 19 samples, a total of 12 coding variants were identified including five missense, six silent, and one in-frame indel (Table 1). The missense variants were subsequently assessed using PolyPhen (Polymorphism Phenotyping) (http://genetics.bwh.harvard.edu/pph/), a web-based tool that uses structural, phylogenetic, alignment, and annotation information to estimate the impact of an amino acid substitution on the structure/function of a protein, by assigning a score and predicting whether the change is benign, possibly damaging, probably damaging, or unknown.

One non-recurrent in-frame indel was identified in the ARX gene, which encodes a transcription factor required for normal brain development. In specimen A, the p.A112del variant (Exon 2a amplicon, c.333-335delGGC) was identified in the first poly-alanine tract of ARX (Figure 1c). The ARX gene contains four GCG trinucleotide repeat regions encoding the amino acid alanine, where expansion of the first poly-alanine tract from 10 residues to 17 has been found in association with infantile epileptic encephalopathy type 1 (EIEE1). In addition to EIEE1 and MRX, mutations in the ARX gene are responsible for lissencephaly X-linked type 2 (LISX2), myoclonic epilepsy X-linked with intellectual disability and spasticity (XMEIDS) and Partington syndrome (PRTS). The p.A112del variant seems unlikely to impact the function of ARX, since expansion of the poly-alanine tract has been associated with altering protein function.

One non-recurrent missense variant was identified in the gene JARID1C. The JARID1C gene encodes a histone demethylase that specifically demethylates the lysine-4 residue of histone H3, and is involved in the transcriptional repression of neuronal genes. In specimen P, the p.R1546Q variant (Exon 26 amplicon, c.4637G>A) was identified. The amino acid change resulted in a PolyPhen score of 1.283 and was predicted to be a benign change with minimal effect on protein function.

Two non-recurrent missense variants and one recurrent missense variant were identified in the gene ZNF41, which encodes a Krueppel-type, zinc-finger domain containing protein with a putative role in transcriptional regulation. In specimen E, the p.I125R variant (Exon 4a amplicon, c.374T>G) was identified (Figure 1a). This polymorphism has been previously identified (dbSNP: rs17147624) in a variant sequence region (141–176) of the gene that is missing in isoform 7 and isoform 8 of ZNF41 (Shoichet, et al., 2003). Despite the physicochemical property change from a medium-size, hydrophobic amino acid (isoleucine) to a large, basic amino acid (arginine), the variant was given a PolyPhen score of 1.043 and was predicted to have a benign effect on protein function. In specimen C, the p.D397E variant (Exon 4b amplicon, c.1107C>A) was identified. This polymorphism is in a region (425–447) of the ZNF41 gene that encodes a degenerate type 5 zinc-finger domain (C2H2). Both residues (aspartic acid and glutamic acid) are medium-size, acidic amino acids and the change was given a PolyPhen score of 0.488 and was predicted to be a benign change to protein function. In specimens D and N, the p.D315E variant (Exon 4b amplicon, c.945T>G) was identified (Figure 1b). This polymorphism has been previously identified (dbSNP: rs2498170) in a region of the ZNF41 gene that encodes a degenerate type 2 zinc-finger domain (C2H2) (Shoichet, et al., 2003). Both residues (aspartic acid and glutamic acid) are medium size, acidic amino acids, and the change was given a PolyPhen score of 0.191 and was predicted to be a benign change, with no effect on protein function.

A recurrent variant, p.V39I variant (Exon 2 amplicon, c.115G>A), was identified in specimens E, I, and N in OPHN1. The OPHN1 gene is expressed in the brain and encodes a rhoGAP protein. The change was given a PolyPhen score of 0.025, and was predicted to be a benign change to protein function.

A component of this study was to compare the sequencing results of these samples when generated on the new 3500xL Genetic Analyzer (performed at Applied Biosystems, Foster City, Ca), and confirm concordance with the sequencing results of the same samples analysed on the 3730 Genetic Analyzer. The resequencing data generated using fast workflow on the new 3500xL Genetic Analyzer corroborated the results observed by Dr Michael Friez and Dr Frank Bartel (Greenwood Genetic Center, personal communication). When coupled with the optimised fast resequencing workflow and secondary analysis with variant reporter software, the 3500xL genetic analyser provided a rapid, reliable, and easy-to-use platform for mutation detection across multiple genes and large numbers of samples.

Applied Biosystems would like to acknowledge the generous contribution of genomic DNA samples and primer sequences from Dr Mike Friez and Dr Frank Bartel, Molecular Diagnostic Laboratory, Greenwood Genetic Center, SC, which were used to generate the data shown.