Novel ANKRD11 gene mutation in an individual with a mild phenotype of KBG syndrome associated to a GEFS+ phenotypic spectrum: a case report

Whole Exome Sequencing (WES) technology introduces a remarkable revolution in the identification of disease-causing genes and a powerful tool for genetic diagnosis, mainly relevant for rare diseases, which is remarkably difficult for clinicians to be diagnosed. WES can substantially reduce the number of cases remaining undiagnosed for many years and has become the current standard for the diagnosis of highly heterogeneous rare disorders with suspected Mendelian inheritance. This approach is leading to better dissecting the clinical phenotype of patients, in particular, those related to phenotypic progression in association with a single locus and those derived from mixed phenotypes developing from multilocus genomic variants. Recent studies reported the presence of multiple genomic diagnoses in a single individual in 3.2–7.2% of cases [1‐3].

Molecular variants in the SCN9A gene (MIM#603415) are responsible for a range of seizure disorders, which are characterized by early-onset isolated febrile seizures to generalized epilepsy with febrile seizures plus, type 7 (GEFS+), which identifies the most severe phenotype, as well as primary erythermalgia, callousness to pain linked to channelopathy and extreme paroxysmal pain disorder. Patients with isolated febrile seizures frequently showed an early onset between ages 5 months to 4 years and exhibit natural remission by age 6 years, while patients with GEFS+ persist in having various types of febrile and afebrile crises afterwards in life [4]. The SCN9A gene encodes for NaV1.7, a voltage-gated sodium channel mainly expressed in the hippocampus during the embryonic phase, suggesting a key function in the central nervous system [5] and in nociception signaling. SCN9A has been proposed suggested as a genetic modifier in SCN1A mutation linked with GEFS+ and as a potential susceptibility gene for Dravet syndrome [6, 7].

In 1975, KBG syndrome was identified and characterized by specific clinical findings: macrodontia of the upper central incisors, distinctive craniofacial signs, short stature, skeletal abnormalities and neurological involvement that encompasses developmental delay, convulsions and intellectual disability [8‐11].

The initial description of the KBG syndrome, named KBG based on the initials of first affected families’ surnames, referred to 7 patients from 3 unrelated families with a putative autosomal dominant inheritance [8, 12, 13]. To properly diagnose KBG syndrome, 4 or more of these 8 major criteria should be satisfied: (1) macrodontia of the upper central incisors, observed as a distinctive trait of KBG syndrome and reported in more than 95% of cases; (2) distinctive facial features (presence of at least three findings of six categories of craniofacial shape, hair/eyebrow, eyes, ears, nose and mouth); 3) hand abnormalities, (fifth finger clinodactyly, clinical brachydactyly, or short tubular bones on radiographic exam); (4) neurological implication, with magnetic resonance, global developmental delay, and/or a seizure disorder; (5) bone age > 2 SD below the average; (6) costovertebral abnormalities, (abnormal curvature of the spine, cervical ribs, or vertebral/endplate defects); (7) postnatal short stature (as a height less than 3rd centile); and (8) occurrence of a first-degree relative affected by KBG syndrome.

In 2011, Sirmaci et al., [14] identified pathogenic heterozygous variants in the ANKRD11 gene (MIM# 611192). Single nucleotide mutations and small indels represent about 83% of the pathogenic variants identified within ANKRD11 and larger copy number variants (mostly deletions) represent about 17% [15‐17].

ANKRD11 represents one of the family members of ankyrin repeat-containing cofactors that relates with p160 nuclear receptor coactivators (NCOA1) by recruiting histone deacetylases to inhibit ligand-dependent transcriptional activation [14, 18, 19]. ANKRD11 was also found to localize inside the neurons nuclei and to accumulate in distinct inclusions after their depolarization. This finding suggests that ANKRD11 represents one of the major players in neural plasticity [14].

Here, we report two sisters from a non-consanguineous family, presenting generalized epilepsy with febrile seizures plus (GEFS+; MIM#613863) associated with an heterozygous mutation in the SCN9A gene, p.(Lys655Arg), inherited from the father that was asymptomatic for the crisis, and a novel de novo loss of function mutation in the exon 10 of the ANKRD11 gene, p.(Tyr1715*) in one sister with a clinical phenotype compatible with KBG syndrome (MIM#148050).

We report here a dual diagnosis in case II-1 presenting with ANKRD11 and SCN9A pathogenic variants found by WES. The variant in ANKRD11 has never been reported so far in literature. Considering that it is de novo, we supposed a possible mosaicism in the parents but we were not able to find any by WES analysis. The reassessment of phenotypic features in case II-1 confirmed that she fulfilled the proposed diagnostic criteria for KBG syndrome, complicated by early-onset isolated febrile seizures, although EEG abnormalities with or without seizures have been reported in some KBG cases [17]. In particular, she presented with macrodontia, hand anomalies, neurological involvement, costovertebral anomalies and post-natal short stature. This is the first Brazilian case reported so far.

The KBG syndrome is very rare, with about 100 individuals reported so far in literature [15, 16]. Probably it is underdiagnosed due to the clinical features that can be mild and common to other diseases. ANKRD11 mutations have been found in patients identified in a large number of subjects with characteristics consistent with Cornelia de Lange syndrome (see, for example, CDLS1 MIM#122470), thus showing phenotypic overlap between the two disorders. As previously reported [22, 23], some KBG patients could be recognized by gestalt, others may look like Cornelia de Lange syndrome (CDLS). CDLS and KBG represent two rare and distinct syndromes, but they have clinical aspects that overlap such as cognitive deficit, growth retardation and certain craniofacial abnormalities (brachycephaly, wide eyebrows and nostrils anti-vertite). Other common features are limb abnormalities such as small hands and feet, clinodactyly of the fifth finger and syndactyly of the second and third toes. Five different genes associated with the cohesin complex and its regulation (NIPBL, SMC1A, SMC3, HDAC8 and RAD21), showed heterozygous mutations identified in patients with CDLS. The cohesin complex regulate gene expression mediating transcriptional activation and repression [23]. The main function of ANKRD11 is to suppress transcriptional activation of the target genes of nuclear receptors by enrolling deacetylase at different promoters [18]. It is reasonable to assume that dysregulation of functionally correlated genes by cohesin complex deficiency or ANKRD11 may result in overlapping phenotypic characteristics [23].

The shared variant p.(Lys655Arg), occurring in SCN9A, has been previously found in several individuals: one with GEFS+ and two with Dravet syndrome. Additionally, one of these individual also possesses a de novo SCN1A mutation [6] and another affected by atypical benign partial epilepsy (ABPE) of childhood, harbored variants in in the CPA6 and SCNM1 genes associated with epilepsy [24]. Although already shown to play important pathogenic roles in epilepsy and predicted to significantly alter protein function, the p.(Lys655Arg) variant was detected in asymptomatic parents and has been found with extremely low frequency in control cohorts (NHLBI GO Exome Sequencing Project and ExAC Browser).The SCN9A p.Lys655Arg variant found in our sisters seems to be associated with an early-onset isolated febrile seizures. Other subjects showing febrile seizures or GEFS+, that possess SCN9A mutations were sporadic and cannot offer powerful evidence for a specific role of SCN9A in seizure disorders that is presently under debate. Therefore, this report would further support that SCN9A mutations are linked to a monogenic pattern.

The phenotype of patients with GEFS+ combine febrile seizures, absence seizures, partial seizures, myoclonic seizures, or atonic seizures, with a variable degree of severity [25]. In a recent paper [7], none of the GEFS+ families analyzed could be completely clarified by high penetrance of SCN9A mutations. Moreover, it is not surprising to identify a non-penetrant individual because they are commonly identified in autosomal dominant diseases and well documented in pedigrees of febrile crises (60–80% of penetrance [6]).

This study provides an example of how WES has been instrumental allowing us to dissect the clinical phenotype, which is a multilocus variation aggregating in one proband. The successful identification of causal variant in a gene may not be sufficient, making it necessary to identify other variants that can fully explain the clinical picture. In several series of studies, the presence of multiple molecular diagnoses in a single individual has been described in 3.2–7.2% of cases that underwent molecular analysis, but large cohorts of patients and its associated clinical studies [3] are lacking to clearly define this phenomenon. This report emphasizes the critical role of the clinician in diagnostic genomic analyses and highlights the advantages of WES technology in the genetic dissection of a heterogeneous phenotype.

In our study, we prove that apparent phenotypic expansion may represent blended phenotypes resulting from pathogenic variation at more than one locus, so allowing a dissection of genotype–phenotype relationships. Due to the lack of the prevalence of blended phenotypes from multiple monogenic disorders, a systematic re-analysis of WES data sets is needed to a proper diagnosis in daily practice.