Novel MLPA procedure using self-designed probes allows comprehensive analysis for CNVs of the genes involved in Hirschsprung disease

Hirschsprung disease (HSCR, OMIM 142623) is a congenital malformation characterized by the absence of intramural ganglion cells in the myenteric and submucosal plexuses along a variable portion of the distal intestine, due to a defect of craniocaudal migration of neuroblasts originated from the neural crest [1, 2]. HSCR presents an estimated incidence of 1/5000 live births, and has a non mendelian inheritance with reduced penetrance, variable expression and male predominance. Although familial forms exist, the vast majority of cases are sporadic. In addition, the disease can present as an isolated trait, although in a 30% of the cases it is associated with chromosomal abnormalities, neurodevelopment disorders and a variety of additional isolated anomalies and syndromes [2].

HSCR has a complex genetic aetiology with several genes being described as associated with isolated or syndromic forms. RET proto-oncogene is considered the major causal gene in HSCR and has been extensively studied in different HSCR series worldwide. Both traditional RET coding mutations and a common non-coding RET variant within a conserved enhancer-like sequence in intron 1, have been reported to be associated with a great proportion of HSCR cases [2‐4]. Other genes associated with HSCR encode for receptors, ligands (especially those participating in the RET and EDNRB signaling transduction pathways), and transcriptional factors, such as SOX10 and PHOX2B, among others, that are usually involved in the neural crest cell development and migration [2].

Interestingly, many recent reports point out the implications of altered gene dosage in diagnosis, prognosis and therapy in different human diseases [5]. Nonetheless, it does not seem to be apparently the case of HSCR, with the current data supporting a predominance of missense/nonsense mutations, although small deletions/insertions have been occasionally observed (Human Gene Mutation Database of the Institute of Medical Genetics in Cardiff, http://​www.​hgmd.​cf.​ac.​uk/​ac/​index.​php). In fact, no duplications and only one gross deletion affecting the entire sequence of RET have been reported [6, 7]. To date, only 2 studies have been reported investigating gene dosages anomalies in HSCR patients based on MLPA technique (Multiple Ligation-dependent Probe Amplification) [8, 9], which has an optimal performance to detect alterations of gene dosages [10]. Both of them used MLPA MRC-Holland commercial kit for HSCR, that analyses a limited number of genes (RET, ZEB2, EDN3 and GDNF), and revealed no CNVs associated to HSCR in those genes [8, 9]. In additions we have performed a SOX10 deletion screening on our HSCR patients [11] based on a previously reported QMF-PCR method (Quantitative Multiplex Fluorescent PCR), obtaining negative results [12]. Nevertheless, studies in other "HSCR genes" are necessary to rule out the potential implication CNVs in the pathogenesis of HSCR.

With the aim to analyse anomalies in the gene dosage of several genes described as associated to HSCR (EDNRB, GFRA1, NRTN and PHOX2B), but never previously analyzed by MLPA, we designed specific synthetic MLPA D-probes, following MRC-Holland recommendations. The hybridization, ligation and amplification of the MLPA probes were performed in 4 different probemixes of 8-10 probes each, together with the 3 control probes. Signal peaks height of the amplified products observed after electrophoresis, were as homogeneous as expected for self-designed probes, and peak normalization was successfully fulfilled between the patient samples and controls in all the probemixes tested.

After the validation of our probes, we screened a total of 208 HSCR patients and found a deletion in GFRA1 gene (c.(?-555)_431+?del; Figure 1) that affects exons 1a, 2a, 3 and 4 in isoform NM_005264 and exons 2b, 3 and 4 in isoform NM_145793. This deletion was detected in a heterozygous state, in a sporadic and isolated male HSCR patient presenting with short-segment HSCR (patient HSCR-115), and was not found in 100 control individuals tested. This deletion was inherited by his unaffected father, and was found to be absent in other healthy members of the family. There are 2 CNVs described for GFRA1 gene annotated in the Database of Genomics Variants http://​projects.​tcag.​ca, Variant_48418 and Variant_48004. The first variant is a 2.5Kb deletion located in the 3' untranslated region of the gene, while the other consisted on a 36 Kb deletion the genomic region containing exons 7, 8 and 9 of GFRA1. Therefore, the available data support that we are describing a novel deletion. Interestingly, this is the second time this deletion has been found in our HSCR patient series, in a patient not related with the one previously reported [Figure 1; 15].

In order to preliminarily examine the potential damaging effect of this deletion on GFRA1 expression and functionality, we used InterProScan and AlternativeSplicing tools from EBI and Transec from EMBOSS. We verified that the critical translation initiation signal in the gene was abolished; subsequently no wild-type (WT) protein was expected to be expressed from the deleted copy of the gene. In adition, we checked in silico whether the deleted allele could produce any protein with similar functional capacity as GFRα1. We found that an alternative peptide could be translated from deleted isoform NM_005264 with the same carboxyl-terminus aminoacidic sequence. This putative protein would maintain one of the 3 GDNF/GAS1 domains in the WT protein, but would also lack the localization signal in the N-terminal region. Although this deletion is well refined in its 3' end, we failed to establish the boundaries in the 5' end where all transcription and translation signals are located. Therefore it seems unlikely that the aberrant protein could be expressed, and in that case it would have a very limited, or even null, functionality.

HSCR has a complex genetic aetiology and point mutations in several genes have been reported to be implicated in a portion of isolated and syndromic HSCR forms [2]. It is tempting to speculate that other genetic events different from point mutation, such as CNVs, have a functional role in the pathogenesis of HSCR. Very little is known in this field for HSCR since typical screening methods based in conventional PCR are only able to detect small deletions/duplications (a few base pairs), and cytogenetic techniques can exclusively detect alterations in the order of megabases. Those techniques are neither powerful nor adequate to detect CNVs [10], so that those types of rearrangements would be missed. In this way, it would be possible that such mid-size deletions/duplications in several HSCR genes have been underreported. In addition, traditional techniques used to detect mid-size deletions/duplications, such as southern blot, are expensive, time consuming and not suitable for high-throughput results. For this reason we planned to perform CNVs screening in a large series of HSCR patients using MLPA technology, which can be performed in a large number of individuals within a short period time, in order to determine if it is a reliable technique suitable for a routine CNVs screening. Despite the negative results previously reported for HSCR MLPA commercial kit [8, 9], we have obtained positive results with the finding of a deletion affecting the 4 first exons in GFRA1. This deletion was previously identified in a sporadic HSCR patient, but its actual implication in the pathogenesis of this disease remained unknown [15]. The finding of the same deletion in an independent patient with the same phenotype and its absence in the control population, support that this deletion at the GFRA1 locus is a mutational event potentially related to HSCR. In addition, the implementation of MLPA technique for midsize deletion detection leads us to refine the deleted region at GFRA1 locus. The protein GFRα1 is one of the four co-receptors of the RET tyrosine kinase receptor. The binding of RET to GFRα1 is required for the specific recruitment of GDNF and the subsequent phosphorylation of RET. Therefore, the presence of such a deletion in GFRα1 would avoid the expression of the protein, presumably preventing RET phosphorylation and affecting the correct development of the ENS. The presence of this mutation in unaffected members of the family suggest that it could be necessary but not sufficient to produce the phenotype, and additional unidentified genetic events might be acting in this HSCR patient. In this sense, no point coding mutations were detected in this patient, or in the previously described patient harbouring the same deletion, in other HSCR-related genes tested such as RET, GDNF, NRTN, PSPN, ARTN, EDNRB, EDN3, NTF3, NTRK3, SOX10 or PHOX2B. The present results indicate that CNVs are not a common molecular cause of HSCR, although they should be taken into account for further studies.

One of our goals was to provide a simple, reliable, economic and fast method for CNVs screening in HSCR related genes, and the present study has successfully validated the self-designed MPLA probes for CNVs analysis. The design and validation of MLPA probes for additional genes represent an implementation for a technique that was restricted to the commercial production. In this sense, the present design, together with the commercial MLPA kit for HSCR, allows the complete analysis of CNVs in the coding region of the most prevalent genes for HSCR. In addition, the presence of a GFRA1 deletion that seems to impair protein function, in an unrelated HSCR patient supports and confirms the idea that this specific deletion might participate in the development of HSCR. Despite the fact that CNVs seems to be an uncommon susceptibility factor leading to this disease, our results point out the importance of taking into account those molecular events in HSCR studies from now on, at least in GFRA1 gene. Further screening of CNVs in additional series of patients would be necessary in order to completely address its real implications in the pathogenesis of HSCR.