ZMYND11-related syndromic intellectual disability: 16 patients delineating and expanding the phenotypic spectrum

1 INTRODUCTION

The chromosome 10p15.3 microdeletion syndrome is characterized by developmental delay (DD) and intellectual disability (ID), craniofacial dysmorphism, behavioral abnormalities, hypotonia, and seizures (DeScipio et al., 2013). Haploinsufficiency of ZMYND11 (NCBI Gene ID: 10771) is believed to account for many of the features associated with chromosome 10p15.3 microdeletion (Tumiene et al., 2017). ZMYND11 has been shown to act as a transcriptional repressor by inhibiting the elongation phase of RNA polymerase II by recognizing the histone modification present in transcribed regions, specifically H3K36 trimethylation (Wen et al., 2014).

In support of the critical role of ZMYND11 in the chromosome 10p15.3 microdeletion syndrome, patients with de novo truncating variants in ZMYND11 have a similar phenotype, including ID, seizures, and behavioral issues (Coe et al., 2014; Popp et al., 2017). In addition, missense variants in this gene have also been associated with ID and seizures, although there is a more severe phenotype in patients with specific variants, which may be related to a gain-of-function mechanism (Cobben et al., 2014; Moskowitz et al., 2016). A splice site variant has also been reported in a child with autism spectrum disorder (Iossifov et al., 2012). In total, 11 patients with pathogenic variants in ZMYND11 (MIM# 616083) have been reported to date (Aoi et al., 2019; Cobben et al., 2014; Coe et al., 2014; Iossifov et al., 2012; Moskowitz et al., 2016; Popp et al., 2017).

Here, we present 16 previously unreported individuals with pathogenic variants in ZMYND11, including four from the same family. We further delineate and expand the genotype–phenotype correlations and phenotypic spectrum of ZMYND11-related ID.

2 METHODS

All patients were ascertained after routine referral to their local Clinical Genetics service. Patients 1, 3, 5, and 8 were gathered through international collaboration using GeneMatcher (Sobreira, Schiettecatte, Valle, & Hamosh, 2016). Patients 2, 6, 7, 9, 11, and 12 were identified through the Wellcome Trust Deciphering Developmental Disorders study (Wright et al., 2015). Patients 13–15 were identified as affected relatives of Patient 12. Patients 4, 10, and 16 were identified through personal communication. Exome sequencing was performed on all probands, with a trio approach on Patients 1, 3, 5, 6, 9–12, and 16; and a duo approach on Patients 2, 4, 7, and 8, as DNA samples were only available from one parent. Sanger sequencing only was used to ascertain the presence of the familial variant in Patients 13–15, and all other patients had their ZMYND11 variant confirmed using this method. All sequence variants were described with reference to the ZMYND11 transcript NM_006624.5. All variants were classified according to the American College of Medical Genetics guidelines (Richards et al., 2015). Further information is available in the Supporting Information Data. Patient variants have been uploaded to either ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), Global Variome shared LOVD http://www.lovd.nl, or DECIPHER (https://decipher.sanger.ac.uk). Informed consent was obtained for all subjects for inclusion in this study.

3 RESULTS

3.2 Patient phenotypes

Phenotypic information for all patients is shown in Table 1. In-depth patient summaries are available in the Supporting Information Data (Supporting Information Patient Summaries). Prominent phenotypic features are detailed below. The denominators refer to the number of patients for whom the specific information is available.

Birth weight was at or above the 98th centile in three patients (3/14; 21%). Feeding problems (e.g., excess vomiting after feeds, bottle feeding requiring more than 1 hour), were present in 6/13 patients (46%). Most patients had normal growth parameters and head circumference.

Development was delayed in all patients (14/14; 100%). The median age at independent walking was 24 months (with an age range of 17 months to 4 years). Three patients remained unable to walk at the ages of two-and-a-half (for two individuals) and 4 years, respectively. Speech delay was prominent, with 14/14 (100%) affected. First words were achieved at a median age of two-and-a-half years (with age range of 2–4 years). Two patients were nonambulatory and had not achieved speech at two-and-a-half and 4 years age, respectively (2/14; 14%). All patients had mild to moderate ID (13/13; 100%).

Almost all patients had behavioral issues (14/16; 88%). These include attention deficit, hyperactivity, and impulsivity (8/16; 50%), aggressive behavior (8/16; 50%), and autism spectrum disorder or autistic traits (3/15; 20%). Neurological abnormalities were detected in 10/16 (63%); mostly hypotonia (5/16; 31%) and epilepsy (5/16; 31%).

Photographs of patients in this series are shown in Figure 1. Dysmorphic facial features were judged to be present in 11/16 (69%). There were a number of shared facial features, including thick eyebrows, prominent eyelashes, a depressed nasal bridge with a bulbous nose, anteverted nares, and thin vermilion of the upper lip and wide mouth.

Patients 12–14 in this series inherited their ZMYND11 variant from their mother (Patient 15). All individuals in this family had special educational needs; two of the siblings are now in employment. The ZMYND11 variant found in Patient 9 was paternally inherited. Detailed phenotypic information is not available for the father.

4 DISCUSSION

Here, we present 16 new individuals with predicted pathogenic variants in ZMYND11. Comparison with all previously published patients allows further delineation of the phenotypic spectrum associated with mutations in this gene (Tables 2 and S1; Aoi et al., 2019; Cobben et al., 2014; Coe et al., 2014; Iossifov et al., 2012; Moskowitz et al., 2016; Popp et al., 2017).

All patients (including our series) had DD, particularly affecting speech and ID. The severity of ID in this series is mild to moderate, but four patients have previously been described with severe ID (Cobben et al., 2014; Coe et al., 2014; Moskowitz et al., 2016; Popp et al., 2017). Behavioral issues are also a prominent feature both in our series and in those previously published (Coe et al., 2014; Popp et al., 2017), including aggression, attention deficit/hyperactivity, and autism/autistic traits. Therefore, this series provides further evidence that behavioral abnormalities are a significant part of the ZMYND11-associated phenotype. These behavioral problems may pose a substantial psychosocial burden, especially if the ID is mild. Hypotonia and epilepsy affect 48% and 39% of all patients, respectively (including our series). This enables us to indisputably establish hypotonia and epilepsy as part of the phenotype associated with this syndrome.

Dysmorphic features, particularly thick eyebrows, prominent eyelashes, and a bulbous nose, are present in the majority of patients (Figure 1). These are in line with the patients reported by Coe et al. (2014). These dysmorphisms may prove useful with regard to reverse phenotyping. Feeding difficulties were present in 59% of all patients (including our series), although only three patients required supplementary feeding.

Brachydactyly, seen in two patients in our series, is a possible novel feature. Interestingly, tooth enamel hypoplasia, present in one patient in our series, has previously been reported in another patient with a ZMYND11 variant (Coe et al., 2014), indicating this may be a rare and/or overlooked phenotypic feature, although formal dental assessment has not been documented for most patients.

In this series, three individuals inherited a predicted pathogenic ZMYND11 variant from their affected mother; another patient inherited the pathogenic variant from his father on whom detailed phenotypic information was lacking. Inheritance of a pathogenic ZMYND11 variant from an affected parent has been previously reported (Coe et al., 2014). Familial inheritance should, therefore, be considered in variant filtering and interpretation and reproductive counseling.

The majority of patients, including those in our series, have truncating variants, which are likely subject to nonsense-mediated decay, and hence, result in haploinsufficiency (Figure 2). Of note, the p.(Gln586*) variant in our series is located in the last exon and therefore may escape nonsense-mediated decay. The p.(Cys574Arg) variant is similarly located in the last exon. These variants may be expected to have a deleterious effect through disruption of the MNYD-type zinc-finger motif. This motif interacts with a number of intracellular partners, for example, the NCoR transcriptional corepressor (Masselink & Bernards, 2000), and amino acid variation within this motif has been shown to disrupt binding of these partners, resulting in reduced efficacy of ZMYND11-mediated transcriptional repression (Kateb et al., 2013; Masselink & Bernards, 2000). We suspect, therefore, that the two variants affecting the MYND-type zinc-finger motif domain in our series will at least result in a reduced function of the protein. The phenotype of these patients and a previously reported individual (Coe et al., 2014) with a p.(Gln587del) variant in this motif is not notably different to those patients harboring variants causing haploinsufficiency, supporting a loss-of-function mechanism. The p.(Glu416Lys) variant in this series is not in a functional domain. It has been classified as likely pathogenic given that it is de novo and not present in the gnomAD database; however further research is required to determine the effect of this variant.

In contrast, two missense pathogenic variants have been reported in patients with notably different phenotypes to those in this series. The p.(Ser421Asn) variant resulted in a severe Angelman-like phenotype and the p.(Arg600Trp) variant caused distinct facial dysmorphism, moderate to severe ID, and short stature (Cobben et al., 2014; Moskowitz et al., 2016). Given these distinct phenotypes, it is possible that other mechanisms, including a gain-of-function, may be at play, but further research is required to characterize the effects of these specific variants.

5 CONCLUSIONS

We present a series of 16 patients with predicted pathogenic ZMYND11 variants, predicted to result in haploinsufficiency or reduced protein function, together with a review of the published literature, allowing further delineation of the associated phenotype. DD and ID, usually mild to moderate, are universally present. Behavioral issues are frequent, and hypotonia and seizures are common. Feeding difficulties occur but are usually mild. Subtle dysmorphism includes prominent eyelashes and eyebrows. Novel features include brachydactyly and tooth enamel hypoplasia. Our data will contribute to successful reverse phenotyping following genomic sequencing.

CONFLICT OF INTERESTS

The authors declare that there are no conflict of interests.

DATA AVAILABILITY STATEMENT

Patient variants have been uploaded to either ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/), Global Variome shared LOVD http://www.lovd.nl, or DECIPHER (https://decipher.sanger.ac.uk). The accession numbers are: RCV000627377.1 (Clinvar), SGS 306759, SGS 307296, CAR 279594, SMB 307553, BWH 264849, GSH 282655 (DECIPHER), and via https://databases.lovd.nl/shared/genes/ZMYND11 (LOVD).