This article is mentioned in:

Introduction

Next-generation sequencing (NGS) has introduced a novel era, in which it has become possible to rapidly and cost-efficiently obtain large amounts of sequence data for patients. NGS analysis is therefore currently regarded as a first-choice diagnostic tool, particularly in cases in which phenotypical and clinical features are unclear or atypical, leaving Sanger sequencing for the confirmation of detected variants (1). The correct placement of a preliminary clinical diagnosis and subsequent selection of candidate genes for molecular analysis are challenging when patients present with signs and symptoms that do not match any known clinical condition. In these cases, the use of Sanger sequencing to produce a molecular diagnosis is unreasonable due to the high cost and effort required.

In these cases, whole exome sequencing (WES) can be an appropriate tool to detect pathogenic mutations; however, this requires an effective strategy for filtering results and in-depth analysis to meaningfully interpret incidental findings (2,3). Furthermore, in silico and functional studies may be necessary to attribute the correct value to unexpected results.

The present study reports on a patient who was selected, given the clinical phenotype complexity, for proband-parents (trios) analysis using WES with the aim of detecting all the possible genetic variants that may have generated the patient's phenotype.

The present study confirmed that it is essential to filter and confirm each step of the NGS based on the decision of a team that has the necessary expertise to assess the impact of the detected variants.

Materials and methods

Ethical approval

The technical and scientific review board of the Institute for Maternal and Child Health, Istituto di Ricovero e Cura a Carattere Scientifico (IRCCS) 'Burlo Garofolo' (Trieste, Italy) (no. 185/08, 19/08/2008) approved the present study. For enrolment of the child and the parents in the present study, informed consent was obtained from the parents.

Patient

A six-year-old boy with non-consanguineous parents was admitted to the Paediatric Department of the IRCCS 'Burlo Garofolo' (Trieste, Italy) due to neurogenic diabetes insipidus, mild hepatomegaly and recurrent episodes of fever with skin rash and cervical adenitis.

DNA extraction

Genomic DNA (gDNA) was extracted from 1–2 ml EDTA-anticoagulated blood obtained from the proband and his parents using an EZ1 DNA Blood kit (Qiagen, Hilden, Germany) according to manufacturer's instructions.

WES analysis

A total of 3 μg gDNA was used for library preparation using TruSeq™ Exome Enrichment 62 Mb kit (Illumina Inc., San Diego, CA, USA). A paired-end sequencing of each library was performed on an Illumina HiSeq1000 (Illumina Inc.). Raw sequencing data were collected as unmapped 100-bp reads in fastQ format. CLC Genomics Workbench version 6.5 software (CLC Bio; Qiagen) was employed to check the quality of the reads, to map reads back to the human reference genome, version hg19, to calculate the overall coverage, to perform local realignment and base quality recalibration and, finally, to call single nucleotide variants (SNVs) and small insertion/deletions (INDELs). All of these data were collected into a standardized variant call format, version 4.1 (4). SNVs/INDELs were annotated using ANNOVAR software (http://www.openbioinformatics.org/annovar/; version 2013Aug23) referring to several public databases, including National Center of Biotechnology Information dbSNP build138 (http://www.ncbi.nlm.nih.gov/SNP/), the 1000 Genomes Project (http://www.1000genomes.org/) and the National Heart, Lung, and Blood Institute Exome Sequencing Project Exome Variant Server (http://evs.gs.washington.edu/EVS/) (5).

Bioinformatic and statistical analyses

SN Vs/INDELs were first analyzed with the aim of evaluating all known mutations already associated with the clinical manifestations of the patient. Subsequently, SNVs/INDELs were filtered with the purpose of selecting variants which are poorly represented within the general population and potentially 'pathogenic'; to this end, variants were selected according to the following inclusion criteria: A) Minor allelic frequency <0.03 in residents with ancestry from northern and western Europe, referring to the 1000 Genomes Project (http://www.1000genomes.rg/) and NHLBI Exome Sequencing Project (ESP) Exome Variant Server (http://evs.gs.washington.du/EVS/); B) homozygous variants carried by the proband; and C) variants predicted as 'pathogenic' based on Polyphen-2 (6), Mutation Taster and LRT scores, as recorded in dbNSFP v2.0 (7,8).

Polymerase chain reaction (PCR) and Sanger analyses

PCR amplification was performed for the specific genecoding sequences of the MVK and RAB40AL genes. 50 ng DNA was amplified using KAPA 2G Fast Hot Start Readymix (initial concentration, 2X; final concentration, 1X; respective primer initial concentration, 10 μM; final concentration, 0.2 μM; final volume, 25 μl) (Resnova, Roma, Italy) and a 2720 Thermal Cycler (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA). Amplification products were purified with ExoSAP-IT enzyme (Affymetrix, Inc., Santa Clara, CA, USA) at 37°C for 20 sec, subsequently inactivated at 85°C for 10 sec, and directly sequenced by the Sanger method using ABI Prism 3.1 Big Dye terminator chemistry (Invitrogen Life Technologies, Carlsbad, CA, USA) and BigDye XTerminator Purification kit (Applied Biosystems). Runs were performed on the ABI Prism 3130XL automated DNA sequencer (Applied Biosystems). Sequences were analyzed with Seqman II Software (DNASTAR I Lasergene 7.0; DNASTAR, Madison, WI, USA). PCR amplification and Sanger sequencing primers to confirm the MVK (NM_000431) mutations (c.1129G>A; p.V377I) and (c.404C>T; p.S135L) were: MVK5 forward, 5′-GTTGAGAAAACTGGACCAGATGC-3′ and MVK5 reverse, 5′-CTCAGCTTCCTCATGTTAAAATG-3′; and MVK11 forward, 5′-GGCTTTTGCCTTGAATATGATGA-3′ and MVK11 reverse, 5′-GGGCCTCTCCAGCAGTGTC-3′. PCR amplification and Sanger sequencing primers to confirm the RAB40AL (NM_001031834) known dinucleotide missense changes c.A176G and c.C177A, leading top.D59G were: RAB40AL forward, 5′-CCTCTGCGCACAACCTTGC-3′ and RAB40AL reverse, 5′-CTGGAGCGATTCCAGCTTG-3′. Primers were obtained from Integrated DNA Technologies (Coralville, IA, USA). PCR amplification cycle: Initial denaturation (95°C for 1 sec), 30 cycles of denaturation (95°C for 10 sec), annealing (60°C for 10 sec) and elongation (72°C for 5 sec), followed by a final elongation (72°C for 30 sec). Sequence reaction cycle: Initial denaturation (96°C for 3 sec), 26 cycles of denaturation (96°C for 30 sec), annealing (53°C for 15 sec) and elongation (60°C for 4 sec).

RAB40AL inactivation analysis

300 ng genomic DNA of the patient was digested with 40 U HpaII methylation-sensitive restriction enzyme and 20 U RsaI methylation-insensitive restriction enzyme (New England Biolabs, Ipswich, MA, USA) in a total volume of 20 μl for 24 h at 37°C. After digestion, the mixtures were incubated at 95°C for 10 min. 50 ng digested and undigested DNA were amplified with RAB40AL forward and reverse primers as previously described to verify the inactivation of the gene.

X-chromosome inactivation (XCI) analysis via androgen receptor

300 ng genomic DNA of the patient's mother was digested as described above and incubated at 95°C for 10 min. 50 ng of digested and undigested DNA was amplified with primers 5′-GCTGTGAAGGTTGCTGTTCCTCAT-3′ and 5′–56-FAM/TCCAGAATCTGTTCCAGAGCGTGC-3′ according to the protocol of a previous study (9) for amplification of a region of the human androgen receptor (HAR) gene, localized on Xq12 and characterized by a polymorphic number of CAG repeats. Since this fragment has two HpaII restriction enzyme sites, only methylated DNA was amplified.

XCI analysis via the polymorphic marker DXS1214

300 ng genomic DNA of the patient's mother was digested as described above and incubated at 95°C for 10 min. 50 ng digested and undigested DNA was a mplified with specific DXS1214-labeled primers.

The labeled PCR products were denatured and run on the automated sequencer (ABI 3730XL Platform; Applied Biosystems) using the POP7 polymer and the ROX size standard as size markers (Invitrogen Life Technologies, Inc.). GeneMapper software (v. 4.0; Applied Biosystems) was used for fragment analysis.

Biochemical analysis

Urinary mevalonic acid was determined by means of gas chromatography/mass spectrometry. Spectra were obtained using a Hewlett Packard gas chromatograph 6890 series equipped with a Hewlett Packard 5973 quadrupole operating in electron-impact mode at 70 eV (Hewlett Packard, Palo Alto, CA, USA) (10).

Results and Discussion

WES revealed two missense mutations of the MVK gene (heterozygous): NM_000431, c.1129G>A; p.V377I (rs28934897); and NM-000431, c.404C>T; p.S135L (rs104895297). In addition, the known dinucleotide missense mutations of the RAB40AL (NM_001031834) gene were detected: c. A176G (rs145606134) and c. C177A (rs138133927), leading to p.D59G. Sanger sequencing analysis confirmed all mutations. No mutations were detected in genes associated with diabetes insipidus.

The MVK gene encodes mevalonate kinase, the third enzyme on the pathway leading to the synthesis of cholesterol from acetyl-CoA (11). Mutations of the MVK gene cause a recessive inherited disease called mevalonate kinase deficiency (MKD), which manifests as a clinical and biochemical continuum with mevalonic aciduria at the most severe end [MA; online mendelian of inheritance in man (OMIM) #610377] and hyper immunoglobulin D and periodic fever syndrome at the mildest end (HIDS; OMIM #260920) (12).

The D59G variant of RAB40AL has already been reported as causative of the Martin-Probst syndrome (MPS; OMIM #300519), a recessive X-linked disease characterized by congenital sensorineural hearing loss, mental retardation, short stature, congenital umbilical hernia, facial dysmorphism, abnormal teeth, widely spaced nipples and abnormal dermatoglyphics (13,14). In the present study, a multidisciplinary team discussed the relevance of the genetic results in order to compare the patient's clinical features with those pertaining to MKD and MPS, based on clinical, genetic and functional evaluations.

The diagnosis of MKD was supported not only by the fact that the two mutations detected have a well known role in the disease, but also by a biochemical analysis that demonstrated the accumulation of mevalonic acid [5,496 μg/ml; the normal range of mevalonic acid levels is 34–323 μg/ml (internal laboratory reference)] in urine during febrile crisis, as a proof of MK enzyme deficiency (10).

In the case of the present study, the diagnosis of MKD adequately explained the proband's condition, apart from the presence of neurogenic diabetes insipidus (Table I).

Table I Main symptoms for MP syndrome and MK deficiency compared with the symptoms of patient of the present study.

Table I

Main symptoms for MP syndrome and MK deficiency compared with the symptoms of patient of the present study.

Feature	MP syndrome	MK deficiency	Patient
Growth	Short stature	From failure to thrivea to normal; hyper immunoglobulin D; periodic fever syndrome	Normal
Head	Microcephaly	Microcephalya	Normal
Ears	Hearing loss	Normal	Normal
Eye	Normal	Central cataractsa	Normal
Dysmorphisms	Low-set ears Telecanthus Hypertelorism Epicanthal folds Broad nasal root Broad mouth Full lower lip Abnormal teeth	Triangular facea, Hypoplastic alae nasia	Normal
Genitourinary	Bifid scrotum Small phallus Cryptorchidism Hypoplastic kidneys Renal insufficiency	Normal	Normal
Skin	Telangiectasias	Skin rashes	Skin rashes
Central nervous system	Mental retardation, mild to severe	Psychomotor retardationa, Cerebella ataxiaa	Neurogenic diabetes insipidus
Osteoarticular system	Normal	Arthralgia/arthritis	Normal
Endocrine features	Hypothyroidism	Normal	Normal
Hematology	Pancytopenia	Anemia	Mild anemia
Lymphoid organs	Leukopenia	Lymphadenopathy, hepato-splenomegaly	Lymphadenopathy, hepato-splenomegaly
Molecular Basis	RAB40AL	MVK	MVK V377I/S135L RAB40AL D59G

a Phenotypic features of mevalonic aciduria. MV, mevalonate kinase; MP, Martin Probst.

In fact, most of the clinical features of the patient are characteristics of MKD, although a large number of MKD-associated symptoms were not present in the patient. In particular, diabetes insipidus, which was one of the main complaints of the patient, is not characteristic of MKD. By contrast, no sign or symptom of MPS was present in the patient. Since MPS is described as an X-recessive disease with complete penetrance, it was unexpected that the patient had not developed any of the severe impairments, which are typical of the disease. It was thus hypothesized that either the defect in MK has a protective effect against the development of MPS or that the variant detected in RAB40AL is not the actual cause of MPS. Of note, the variant identified in RAB40AL was predicted to affect protein farnesylation; however, MKD is also associated with an abnormal pattern of protein prenylation due to a shortage of mevalonate-derived isoprenoids (15,16). However, considering that the two defects can reduce protein farnesylation, the combined effect of the two defects would have been expected to be even more severe. On the contrary, the National Heart, Lung and Blood Institute-exome sequencing project (ESP)6500 database (http://evs.gs.washington.edu/EVS/) lists two healthy males carrying the D59G variant of RAB40AL, suggesting that it represents an irrelevant albeit rare polymorphism. In addition, recent studies described the same RAB40AL variant in a patient with a distinctly different phenotype from that reported in MPS patients (17,18).

To ascertain the possibility that RAB40AL is not the causative gene of MPS, the present study exploited the opportunity of studying the mother of the patient, who is a heterozygous carrier of the same gene variant. Bedoyan et al (13) demonstrated that, while males affected by MPS develop progressive loss of white blood cells, female carriers have a normal leukocyte count and display a skewed XCI in peripheral blood cells. The present study then attempted to analyze the inactivation profile of RAB40AL by amplifying the gene after digestion with a methylation-sensitive enzyme (19,20). However, it was not possible to amplify RAB40AL after HpaII digestion, suggesting that this gene escapes inactivation (Fig. 1A). In order to verify the hypothesis of a skewed XCI, the present study then used polymorphic X-chromosome markers known to undergo methylation and inactivation. As the most widely used androgen receptor humara locus was uninformative in the patient's family, the polymorphic marker DXS1214 was then analyzed. After digestion with HpaII, which is selective for active unmethylated DNA, the two DXS1214 polymorphic alleles remained present to similar extents, confirming a balanced XCI (Fig. 1B). These results suggested that the pattern of skewed XCI in the family examined by Bedoyan et al (13) is not due to the D59G variant in RAB40AL, but to another unidentified X-chromosome gene mutation. In conclusion, these results suggested that mutations of RAB40AL are not the cause of MPS.

Figure 1 (A) Humara PCR results from digested genomic DNA (HpaII) and uncut DNA. The RAB40AL PCR product could only be amplified from undigested DNA. (B) Analysis of the DXS1214 polymorphic marker in genomic DNA digested with HpaII and in DNA digested with RsaI. The graphs suggest a balanced X inactivation. PCR, polymerase chain reaction.

The case of the present study confirmed the potential of the NGS technique to identify pathogenic and causative mutations, but also highlighted the crucial role of an expert team performing the critical revision of all detected WES variants.

Acknowledgments

The authors would like to thank the patient and his family for participating in this study. This work was supported by an Intramural grant of the Institute for Maternal and Child Health (IRCCS 'Burlo Garofolo', Trieste, Italy). The present study was also supported by the 'Associazione Azzurra Malattie Rare' and 'Beneficentia Stiftung in Vaduz, Liechtenstein'.

References