The patient was a male infant, first born after a 37 weeks gestation to second-cousin, asymptomatic parents of Turkish origin. Both parents had unremarkable medical histories, a normal head size and a normal intellect. They reported no paresis, and showed no spasticity nor hyperreflexia. The pregnancy was unremarkable, with no report of alcohol use or substance abuse, nor infection. Severe microcephaly was noted at birth, with a head circumference (HC) of 29 cm (−4.3SD), a length of 45 cm and a weight of 2.31 kg. Microcephaly progressed, with a HC of 41.5 cm at 1.1 year (−4.1SD), 43 cm at 3yo (−4.5SD), and 44 cm at 5.2 years (−4.7SD). The patient also had short stature, with a length at −3.5SD and a weight at −2.6SD. No dysmorphia was present except for clinodactily of the fifth fingers. Global hypotonia was present in the first months of life. Hypertonia was first noted at age 9 months, and progressed, with a bilateral Babinski sign. Partial complex epilepsy appeared at age 2 years and was treated successfully with valproate and lamotrigine. Motor development was severely impaired. The patient sat without support at age 1 year, crawled at 2 years, and stood with support at age 2.5 years. He walked with support after age 3 years. He had mild clubfeet deformities with flattened arches. Intellectual development was severely impaired. Testing at the age of 4 months revealed a developmental stage corresponding to the age of 3 months, and at the age of 9.5 months corresponding to 5 months (Bayley scales of infant and toddler development second edition). At the age of 8 years he had no words, but showed emotions. MRI at age 2y8m showed a normal cortex, temporo-parietal subcortical atrophy and hippocampal atrophy bilaterally, enlarged lateral ventricles, a thin corpus callosum predominantly in its caudal portion, a normal cerebellum and a normal white matter in FLAIR sequence acquisition (Fig. 1). A standard karyotype was normal 46,XY. Plasma amino acids and urine organic acids chromatograms showed normal patterns. The parents divorced and the mother had two other healthy children with a new partner.
After paediatric neurology evaluation and follow up, the patient was referred to the genetics clinic for testing of primary microcephaly genes. In a first round of analysis, we used a 250 K SNP GeneChip® microarray (AffymetrixTM, Inc.) and found heterozygosity of alleles at loci MCPH1,2,3,4,5,7 and 9, making the corresponding genes unlikely to cause microcephaly in the proband. Direct sequencing of the coding exons and exon-intron junctions of genes ASPM and WDR62 showed no abnormality.
For whole-exome sequencing, the proband’s genomic DNA was sheared and exonic sequences enriched using Roche SeqCap EZ Human Exome v3.0 (64 Mb) DNA capture. Sequencing was performed on a HiSeq1500 Illumina sequencer at the BRIGHTcore BRussels Interuniversity Genomics High Throughput core [8]. Raw sequences were aligned to the reference genome GRCh37 using BWA algorithm version 0.7.10 [9], duplicated reads were then marked using Picard version 1.97 [10], alignment quality was improved using the GATK [11] realigner and base recalibrator version 2.7, and finally, variants were called using GATK Haplotype Caller version 2.7. The resulting variant set was annotated and filtered using the Highlander software [12]. Variants were filtered for quality criteria (pass GATK standard filter, read depth > 5, variant confidence by depth ≥ 10), allelic frequency < 0.5% (based on the maximum minor allele frequency found in ExAC, 1000G, ESP6500, gonl, ARIC5606 and our in-house database), nonsynonymous or splice junction effect in protein coding genes (using biotype from Ensembl [13] and snpeff_effect from SnpEff [14]) and genotype (homozygous or heterozygous variants in a subset of 68 primary microcephaly genes that we extracted from the literature (Additional file 1), and homozygous or biallelic variants from all other genes of the exome). Variants were then sorted by decreasing Combined Annotation Dependent Depletion (CADD) score [15]. A CADD score between 0 and 10 is associated with non-deleterious variants, and scores greater or equal to 20 are associated with the 1% most deleterious substitutions possible. The variant of interest was confirmed by Sanger sequencing of exon 13 of AP4M1. DNA was amplified using a standard Polymerase Chain Reaction (forward primer: AGTACAGCCCACACCCACAC, reverse primer: CACCTTCTTGAGGCAGACCC). The PCR product was purified with Exosap-IT (Affymetrix), and sequenced by the company Beckman Coulter Genomics.
The affected child’s exome sequence data were first analyzed for rare (allele frequency <0.5%) variants in 68 primary microcephaly-related genes. This showed heterozygous missense variants in three genes: ATR (c.6109A > G p.Met1996Thr), MCPH1 (c.85C > T p.Ala3Ala) and BLM (c.3967G > A p.Gln1283Gln). Analysis of the rest of the exome data revealed 5 hemizygous, 9 compound heterozygous, and 26 homozygous variants, two of which being encompassed by a 9 Mb homozygous stretch (chr7: 91599825–100494222), consistent with homozygosity by descent (autozygosity) for this chromosomal segment. One of these three variants, located at Chr7:99,703,901, consisted of a homozygous truncating mutation in exon 13 of the AP4M1 gene, c.1170C > T (transcript_uniprot_id O00189) changing the Arginine at position 338 of the polypeptide into a stop codon, p.Arg338X (Fig. 2). The variant frequency was 3.31×10^(−5) in the Exome Aggregation Consortium [16] with only four alleles reported, all heterozygous, in 1 Latino and 3 European subjects. The variant was absent from 1000G, GoNL, ESP, and our in-house database. The mutation was located in the middle of the predicted Mu homology domain of AP4M1 which spans amino acid residues 176 through 453 (Fig. 2, left panel). Sanger sequencing confirmed homozygosity of the mutation in the proband and heterozygosity in both parents, as shown in Fig. 2 right panel (father not shown). Among all variants observed in the patient, this was the only variant predicted to cause a premature termination codon. It yielded a CADD score = 39, the highest of all observed variants.