A patient with van Maldergem syndrome with endocrine abnormalities, hypogonadotropic hypogonadism, and breast aplasia/hypoplasia

Features of VMS were originally described by Lionel Van Maldergem et al., in 1992 [2]. Van Maldergem et al., described an 11-year-old girl with intellectual disability (IQ of about 50), peculiar facial appearance including a large forehead, broad nasal bridge, with broad bulbous nose (pear shaped nose), hypertelorism, bilateral epicanthus, narrow palpebral fissures, small ears and helix malformation, hearing loss, dental malocclusion, camptodactyly of the 3rd and 5th fingers, clinodactyly, short 5th metacarpals, genu recurvatum, and hyperlaxity of the major joints. Echocardiogram, chest x-rays, and EKG, were normal. Radiographs showed maxillary hypoplasia, and malformed external ears. The karyotype was 46, XX. Zampino et al. described a 5 year-old girl with a similar phenotype [3]. An MRI of her brain showed enlarged lateral ventricles, and poor outline of the infundibular recess of the 3rd ventricle, and the optic nerves were abnormally thin. Additionally, there was marked reduction of the thickness of the white matter, a thin corpus callosum, abnormally low inserted tentorium with reduced volume of the pons and hypoplastic superior vermis and cerebellar tonsils. They suggested the name of Cerebro-facial-articular syndrome.

In 2012, Mansour et al.. described 6 additional patients as well as a follow up of the original patient [4]. The skeletal dysplasia and anomalous neuronal migration (periventricular neuronal heterotopia) were also shown to constitute common, but variable components of this phenotype. Radiological features included osteopenia, and skeletal abnormalities. The original patient when she was 32 years old had no breast or pubertal development. In 2012, Neuhann et al. [5] described a 4-year-old boy from 1st cousin parents, with severe developmental delay, with similar craniofacial, dysmorphic features and skeletal abnormalities. He, in addition, had genital abnormalities (microphallus, bifid scrotum, and cryptorchidism) and hypoplastic mammilae. Capello et al. in 2013 [6] conducted studies in 9 individuals from 7 families, 5 previously reported and 2 of their own. In 4 patients, the authors identified new non-synonymous homozygous sequence variants in DCHS1: Van Maldergem syndrome 1.

DCHS1 gene encodes a transmembrane calcium-dependent cell-cell adhesion molecule that belongs to the protocadherin superfamily, located on chromosome 11p15.4. Cell-cell adhesion is fundamental for multicellular architecture. Several classes of adhesion molecules are used to achieve this and cadherins represent a major family of such molecules. DCHS1 protein is the ligand for the FAT4 receptor. Additional human studies in 5 patients by Capello et al. [6] from 4 unrelated families (2 siblings in one family) led to identification of biallelic missense and nonsense mutations in FAT4 (FAT tumor suppressor, Drosophila, Homolog of, 4) in affected individuals, implicating mutations in this gene as a cause of the condition Van Maldergem syndrome 2.

FAT-4 encodes a protein that is a member of a large family of protocadherins (OMIM #612411, 2016). Dchs1-Fat4 signaling is required during mammalian development in multiple organs, including the brain, ear, cochlea, kidney, intestine, heart, lungs, and skeleton.

The phenotypes of Dchs1 ^−/− and Fat4 ^−/− single mutants and Dchs1^−/− and Fat4 ^−/− double mutant mice are highly similar [7]. Loss of Fat4 resulted in death at birth, while mice without DCHS1 could survive for a couple weeks at most, therefore did not reproduce. Mao et al. 2015 [7] did not report any hypothalamic-pituitary or other endocrine deficiencies in these mice.

Some pups failed to grow, remain in similar size to newborns pups. The cochlea were about 20% shorter than in the wild type cochlea; they had small cystic kidneys; the sternum was both wider and shorter and also exhibited an abnormal pattern of ossification. They show a modest increase in the width of the vertebral column but only in the lumbar and posterior thoracic region. These vertebra were also narrower along the anterior/posterior axis. The lungs were decreased in size. The intestines were also shorter than their wild type siblings, about 72% in length from the stomach to the rectum. The reduction in the size of internal organs (intestine, lung, and kidney) were not simply reflections of the decreased body size, because newborn pups did not differ significantly in overall size and other organs (e.g. skeletal, liver, heart) were not significantly smaller. The hearts were not smaller, but morphogenesis was affected; they exhibited defects in atrial septation.

Additional studies in mice are quite informative regarding the importance of Dchs1 Fat4 in brain development. Cadherins play an important role in the architecture of the brain, in neural development and cortical patterning.

The regulated proliferation and differentiation of neural stem cells before the generation and migration of neurons in the cerebral cortex are central aspects of mammalian development. Periventricular neuronal heterotopia, a specific form of mislocalization of cortical neurons, can arise from neuronal progenitors that fail to negotiate aspects of these developmental processes. Capello et al., 2013 [6] show that mutations in genes encoding the receptor-ligand cadherin pair DCHS1 and FAT4 lead to a recessive syndrome in humans (VMS) that includes periventricular neuronal heterotopia. Reducing the expression of Dchs1 or Fat4 within mouse embryonic neuroepithelium increased progenitor cell numbers, reduced their differentiation into neurons, and impaired their migratory capabilities to the cortical gray matter. This resulted in the heterotopic accumulation of cells below the neuronal layers in the neocortex, reminiscent of the human phenotype, manifested as a double cortex mainly in the occipital and parietal lobes. The studies of Zakaria et al. [8], provide further knowledge of problems with neuronal migration. Dchs1-Fat4 influence planar cell polarity (PCP), the polarization of cell structures and behaviors (direction of movement) within the plane of a tissue. PCP is essential for the generation of tissue architecture during embryogenesis and for postnatal growth and tissue repair.

Mutations in human FAT4 and DCHS1 cause VMS, characterized by severe neuronal abnormalities indicative of altered neuronal migration.

Studies of neuronal migration using the murine facial branchiomotor neurons (FBM) found that neurons, normally, undergo caudal and lateral migrations tangentially, within the plane of the neuroepithelium and finally migrate radially to form a condensed nucleus within the pia layer.

Tangential migrations contribute extensively to the architecture of the brain. Fat4 and Dchs1, key components of Fat 4-PCP signaling, are expressed in complementary gradients and are required for the collective tangential migration of FBM and for their PCP. Neuronal migration in Fat^−/− and Dchs^−/− mouse mutants disrupted migration laterally, delayed migration along the rostro-caudal axis and slowed radial migration [8]. Their studies identify Fat-PCP as a novel neuronal guidance system, and are in support of the disruptions observed in migration in mice and patients as periventricular neuronal heterotopia.

The Frizzel-Flamingo (Fz-PCP) pathway plays fundamental roles in neural development, including tangential migration of FBM and olfactory neurons. Whether the disruption of lateral migration in Fat4 ^−/− and Dchs1 ^−/− mouse affected migration of olfactory neurons to the hypothalamus was not stated.

Bested et al. [9] addressed the question of functional cerebral asymmetries in a patient with Van Maldergem Syndrome. Mammalian cadherins DCHS1-FAT4 affect functional cerebral architecture. Cadherins play a central role in structural left-right differentiation during brain and body development. Using neurophysiological (EEG) data, they show that when key regulators during mammalian cerebral cortical development are disrupted due to DCHS1-FAT4 mutations, functional cerebral asymmetries are stronger. This suggests that the strength and emergence of functional cerebral asymmetries is a direct function of proliferation and differentiation of neuronal stem cells. These results support the recent assumption that the molecular mechanisms establishing early left-right differentiation are an important factor in the ontogenesis of functional lateralization.

Two additional patients with VMS were reported for surgical rehabilitation of hearing impairments [10, 11].

The molecular genetic basis for endocrine anomalies observed in some VMS patients, including ours, remains unexplained. Here, we report detailed clinical and molecular genetic studies on a female VMS patient with hypogonadotropic hypogonadism and amazia. Whole exome sequencing identified compound heterozygous pathogenic missense variants in DCHS1, revealing a diagnosis of VMS.

Our patient, from normal parents (unaware of any consanguinity) of European ancestry, was seen when she was 13 years 2 months for short stature. The birth weight was 2.8 kg and length 53.3 cm after 9-month normal pregnancy. The Apgar scores were 3 and 8 and she had feeding problems early in life. She was evaluated by a clinical geneticist when she was 4 years of age because of dysmorphic features, but no syndrome was identified, and the chromosomes were normal, 46,XX.

Growth & development

She had been mildly developmentally delayed since birth and had dysmorphic features: prominent forehead, epicanthal folds, hypertelorism, narrow palpebral features, neural vision impairment, maxillary hypoplasia, flat face, flat broad root of the nose, pear shaped nose, small low-set ears, flat helix, sensorineural hearing loss, wide flat palate, dental malocclusion, narrow hands and fingers, hypothenar hypoplasia, small hands and feet, cubitus valgus, genu varum, genu recurvatum, asymmetry of the legs with the left shorter than the right (2.5 cm), scoliosis and osteopenia.

She had been growing in height along the 25th percentile, but her growth slowed down at 10 years of age. At 13 years of age she was on the 3rd percentile line (−1.88 SD). In weight, she grew for years along the 10th percentile line (Fig. 1). She had had pubic hair for 3 years and also axillary hair, but no breast development. The adrenal activity as judged by the levels of dehydroepiandrosterone sulfate (100.4 mcg/dl) was normal, within the range seen in adrenarche (20–535 mcg/dl). The bone age when she was 13 years 2 months was 9 years. Serum growth hormone (GH) after stimulation with arginine and L-dopa showed low values of 4 and 4.2 ng/ml (normal >10). Because of the short stature, lack of gonadal activity and possible, but not certain, growth hormone deficiency, she was treated with GH for 2 years from 13 ½ to 15 ½ and responded well. With the GH treatment, she went from the 3rd percentile (−1.88 SD) to the 25th percentile (−0.67 SD) and she had an adult height of 160 cm, the same as the mother (Fig. 1). Serum GH levels after stimulation tests, when she was 26 years and 7 months old were normal, 9.4 ng/ml (normal >5 ng/ml), indicating normal GH secretion.

MRI of the brain with axial and coronal images was obtained when she was 13 1/12 years. There was interference in the axial views on the frontal lobes from patient’s dental braces. No definite abnormalities were seen on the visualized portions of the brain.

Skeletal dysplasia

The patient has also exhibited some other skeletal changes: a lumbar levoscoliosis of 40 degrees, a thoracic dextroscoliosis of 10 degrees and osteopenia. A DXA was performed when she was 30 years of age, after she had been on estrogen for 15 years and was having menses. The manufacturer’s Z-Score for the lumbar bone mineral density was −1 8, −2.3 for the right total hip, and −1.1 for total bone mineral content. No adjustment was made for the different bone volume of the femoral dysplasia. Other radiographical changes have been: dysplasia of the femoral heads, with no dislocation; abnormalities of the left hand with thin 4th and 5th metacarpals; short 5th metacarpal; short middle phalanges of the 2nd and 5th digits; unequal leg length (2.5 cm); and the margins of the carpal bones more angular than smooth (Fig. 2).

Genetic evaluation

Genetic methods

Our molecular genetic studies were prospectively reviewed and approved by the Institutional Review Board (IRB) of The Research Institute at Nationwide Children’s Hospital. Consent was obtained from the patient and parents for whole exome sequencing (WES). All genomic DNA samples in the study were processed using Agilent SureSelectQXT Target Enrichment System for Illumina Paired End Sequencing Protocol (Agilent Technologies, CA). Genomic DNA was sheared and target exonic regions were captured with the Agilent SureSelect Clinical Research Exome kit. Paired-end 151 base pair reads were generated for exome-enriched libraries sequenced on the Illumina HiSeq 4000 (Illumina, CA). All samples were sequenced to a depth of 100X.

Secondary analysis was performed using Churchill, a pipeline developed in house for the discovery of human genetic variation [12]. Churchill utilizes the Burrows-Wheeler Aligner (BWA) to align sequence data to the reference genome (build GRCh37). Further refinement steps were performed on the aligned sequence data following the Broad Institutes guidelines for best practices (https://www.broadinstitute.org/gatk/guide/best-practices). Duplicate sequence reads were removed using PicardTools (v1.117). Local realignment was performed on the aligned sequence data using the Genome Analysis Toolkit (v3.4–0). Churchill’s own deterministic implementation of base quality score recalibration was used. The GATK’s HaplotypeCaller (HC) was used to call variants. To maximize sensitivity, variant calling was performed across all samples in the study. Use of the GATK’s variant quality score recalibration (VQSR) was excluded in favor of using Churchill’s own quality-based variant filtering algorithm. Comparison of variant calling performance with the NIST Genome in a Bottle (GIAB) gold reference standard (version 3.22) utilizing whole genome sequencing data for NA12878 (30X) demonstrated that Churchill is 99.9985% accurate, with a sensitivity (recall) of 99.6061%.

Tertiary analysis was performed using ANNOVAR and custom in-house scripts to annotate the variant call set with mutation and gene information, protein functional predictions, and population allele frequencies [13]. Common variation occurring at greater than 1% mean allele frequency (MAF) in the population was excluded, and variants outside of coding regions (defined as greater than 4 base pairs from an exon splice site) as well as exonic variants coding for synonymous single nucleotide polymorphisms were also dropped. Variants were further filtered based on the pattern of inheritance expected from examination of the pedigree.

Genetic results

Previously, karyotypes from the blood and the skin were normal (46, XX). A microarray comparative genomic hybridization (aCGH) showed an approximate 62.1 kb gain within chromosome 13, band q34. This region does not contain any genes with the UCSC genome browser (genome.ucsc.edu) and is in a region with limited copy number variation [14, 15]. Consequently, the laboratory report indicated that this finding is likely a benign copy number variation.

Clinically significant results from whole exome sequencing were validated by Sanger dideoxy sequencing of targeted DNA fragments, confirming that the patient had compound heterozygous pathogenic variants (rs145099391;P197L & rs753548138;T2334 M) in the DCHS1 gene (RefSeq NM_003737.3) (Table 2). Each parent was confirmed to be a heterozgyous carrier of one of the mutations (Fig. 3). Both of these DCHS1 variants have been reported in a heterozygous state in individuals of European descent, albeit very rarely (Table 2), but neither of them have been observed in a homozygous state in an individual, according to the genome Aggregation Database (gnomAD) of >15,000 healthy individuals of European ancestry [16]. We did not identify any known pathogenic variants in our patient in genes that have been reported to be associated with hypogonadotropic hypogonadism [17].

Chr	Gene	Positiona	dbSNP ID	SNP	Residue change	European population frequency (gnomAD database)	Type	Protein prediction CADDb Score
11p15.4	DCHS1	6,662,255	rs145099391	G > A	p.P197L	8.89 × 10−4	Missense Exon 2	22.30
11p15.4	DCHS1	6,646,574	rs753548138	G > A	p.T2334 M	9.02 × 10−6	Missense Exon 19	23.30

^aRefSeq NM_003737.3

^bCADD scores are derived from several different functional annotation tools. A score of 20 means that a variant is amongst the top 1% of deleterious variants in the human genome. The higher the score, the more likely that variant is predicted to be deleterious to the protein [33]

Our patient did not have growth hormone deficiency. The GH level after stimulation tests as an adult was normal indicating normal GH secretion [18, 19]. The low levels of GH after stimulation tests, at prepubertal age, of <10 ng/ml do not certainly indicate GH deficiency. Overlap exists in peak GH concentrations between normal children and those with GH deficiency. As many 30% of normal children may have peak values of <10 ng/ml in provocative tests [20]. There was no evidence of other pituitary deficiencies (prolactin, TSH, ACTH). The slow growth beginning at 10 years of age, lack of pubertal spurt, delayed bone age, and osteopenia are consistent with estrogen deficiency. She is clearly affected with Van Maldergem syndrome 1, sharing the clinical features described in previously reported patients and confirmed by our finding of compound heterozygous pathogenic variants in DCHS1. She also had idiopathic hypogonadotropic hypogonadism and after estrogen treatment, amazia (aplasia/hypoplasia of the breast, but normal nipples and areolas). Amastia is characterized by no breasts and no nipples, and athelia, by normal breasts and no nipples, terminology used to describe congenital abnormalities of the breasts (OMIM %113,700).

No endocrine abnormalities were evaluated in previously published patients with VMS 1 or 2 (Table 1). However, by the changes in two VMS patients described previously, some may have been affected also by hypogonadotropic hypogonadism. Six of the 13 patients reported previously were too young to make such conclusions, 4 to 9 years of age. In a 9-year-old female and a 14-year-old male there was no mention of the pubertal changes. A follow up of the original 11 year-old female patient of Van Maldergem, at the age of 32 years did not have any breast or pubertal development [4] suggesting the possibility of hypogonadotropic hypogonadism, and, with no adrenarche, possibly ACTH deficiency. Another patient [6] had microphallus, bifid scrotum, and bilateral cryptorchidism at birth, which are findings that may be present in patients with congenital hypogonadotropic hypogonadism. Hypoplasia of the mammillae was also a finding. Whether eleven of the previously reported patients also had hypogonadotropic hypogonadism and breast and/or nipple aplasia/hypoplasia is unknown. Present evaluation of the patients previously reported would be helpful.

This patient highlights an individual with VMS, characterized by compound heterozygous variants in DCHS1, a rare syndrome reported in only 13 patients, based on our literature search. Our observations provide additional information on the phenotypic spectrum of VMS, including hypogonadotropic hypogonadism and amazia (aplasia/hypoplasia of the breasts but normal nipples and areolas), that was not previously reported with VMS. Congenital breast aplasia/hypoplasia, in the presence of estrogen and progesterone is a rare anomaly of breast development, which pathogenesis is not well established.