Greek study describes two new genetic mutations of EHMT1 and highlights the broad spectrum of Kleefstra syndrome

Two children with the same diagnosis can have a very different course. In Kleefstra syndrome, this is something families know well in practice, but modern genetic research is now beginning to help us better understand the molecular reasons that may lie behind this great variability.

One Greek study, published on April 29, 2025 in the scientific journal Genes, describes two boys of Greek descent with Kleefstra syndrome type 1 (KLEFS1), in whom were identified two novel, de novo pathogenic variants in the EHMT1 gene. One child carried a novel frameshift variant, while an extremely small partial deletion of EHMT1, measuring just 11 kb, was identified in the second. The two cases showed noticeable differences in the severity of the neurodevelopmental presentation.

The work was carried out by researchers and clinicians of National and Kapodistrian University of Athens, University of Thessaly, University Hospital of Larissa, Agia Sofia Children's Hospital, and the GeneTech Analytics laboratory. First author is Maria Tseti and last/corresponding author is Aspasia Tsezou.

Two children, the same diagnosis, but a different clinical picture

The first child was 7 years old during the description of the study. He was born prematurely at 35 weeks and from the very first years of his life presented a delay in psychomotor development. He achieved head control at about 6 months, sat without support at 12 months, and walked independently at the age of 2 years. His first words also appeared around age 2 and simple sentences around 4.

Along the way, he was diagnosed with moderate intellectual disability, ADHD, axial hypotonia, sleep disorders and frequent temper tantrums. At the age of 6 years, she presented with focal epileptic seizures. The EEG recorded epileptiform activity in the left temporo-occipital region with generalization, and oxcarbazepine was administered, with a good response up to the publication of the study. A small ventricular septal defect had also been present in the neonatal period, which subsequently closed completely.

To this child, the exome sequencing revealed a new heterozygous frameshift variant in EHMT1:

c.2075-2097del, p.(Val692Glyfs*64).

The variant is located in exon 13 and alters the reading frame of the gene, leading to premature protein termination. This pathological protein is predicted to lack important regions of the C-terminal end, including the regions ANKR and SET, which play a central role in the function of EHMT1. Sanger sequencing analysis of the parents showed that the variant was again, meaning it was not detected in either parent.

The smallest partial deletion of EHMT1 described up to that time

The second case presents particular genetic interest.

The second child was 9 years old. He also had a developmental delay, but his overall picture was noticeably milder. He achieved head control at 6 months, sat at 10 months, walked independently at 22 months, and spoke his first words around the age of 3.

At the age of 5, he had been diagnosed with mild intellectual disability, ADHD and axial hypotonia. Despite his/her learning and especially language difficulties, he/she followed the general school curriculum with additional educational support.

The initial karyotype and Fragile X testing were normal. The diagnosis was made when performed chromosomal microarray analysis (CMA), which revealed a new deletion of just 11 kb within EHMT1 itself, which included exons 19 to 25:

arr[GRCh37] 9q34.3(140703393-140714454)x1.

The parents were also tested, and the deletion proved to be de novo.

According to the authors, this was the smallest partial deletion of EHMT1 that had been identified up to that time by chromosomal microarray. And this is interesting not only as a new genetic finding, but also because the deletion did not remove the entire gene. It affected a specific part of it and, among other things, the catalytic SET domain.

Why it matters which part of EHMT1 is affected

The EHMT1 it doesn't function like a simple switch that is either «on» or «off.» The protein it produces has different functional regions that perform different roles.

Two of the most important are the region ANKR (ankyrin repeat) and the area SET. Very simply put, ANKR participates in the recognition of specific epigenetic marks on histones (a function often described in the literature as reader function) while SET is the catalytic domain involved in methyltransferase activity, the so-called writer function.

In the first Greek case, the frameshift variant is expected to lead to the loss of both the ANKR and SET domains and to functional insufficiency of EHMT1. The child exhibited the more typical and severe KLEFS1 presentation of the two cases.

In the second incident, the small intragenic deletion mainly affects the final part of the gene and the SET region. The authors hypothesize that, provided that the mutant RNA escapes the nonsense-mediated decay mechanism and a partially functional protein is produced, this could potentially contribute to a milder clinical picture. However, this is a biologically plausible hypothesis and not something functionally proven in the specific study.

This distinction is important.

We cannot simply say that «deletions cause milder Kleefstra» or that a specific position in the gene reliably predicts a child's path.

What the large international EHMT1 study has already shown us

The Greek study was published at a time when the understanding of genotype-phenotype correlations in Kleefstra syndrome is changing significantly.

In 2024, a major international study analyzed 209 individuals with rare EHMT1 variants and molecularly confirmed KLEFS1 in 191 of them. The work showed that the spectrum of the syndrome is much broader than previously thought and even includes people with a relatively mild presentation, normal cognitive function in some cases, and rare familial forms.

Even more important was that the researchers showed that different categories of alterations in EHMT1 can affect protein function in different ways.

Alterations affecting the ANKR region and the «reading» function of epigenetic marks can cause KLEFS1, while certain alterations that exclusively disrupt the enzymatic function of the SET region are not necessarily accompanied by the typical molecular signature or the classical clinical picture of the syndrome.

This provides important context for understanding the second Greek incident as well. It does not prove that this specific 11 kb deletion always causes a mild form, but it supports the broader idea that it doesn't just matter whether EHMT1 is affected, it also matters how exactly its function is affected.

What the two Greek cases teach us

Perhaps the most interesting part of the work for families may not be the exact genetic coordinates, but the different picture of the two children.

Both had molecularly confirmed KLEFS1.

Both exhibited developmental delay, hypotonia, and facial features compatible with the syndrome.

However, the first child exhibited moderate intellectual disability, epilepsy, major behavioral difficulties and sleep disorders, while the second had mild intellectual disability and could attend a general school program with additional support.

This difference is yet another example of the great clinical heterogeneity of Kleefstra syndrome.

For a parent, this is of particular importance: genetic diagnosis is extremely important, but it is not in itself an accurate prediction of a child's future.

Even two children with pathogenic variants in the same gene can differ significantly in language, learning, behavior, epilepsy, associated organic manifestations, and degree of autonomy.

Nor does the absence of features rule out Kleefstra

Neither of the two children in the Greek study presented with obesity or had a severe multisystem burden at the age at which they were evaluated. The first child had a history of a small cardiac defect that had resolved spontaneously, as well as epilepsy, while the second child had no recorded significant cardiac, renal, gastrointestinal, or sensory complications.

This brings to mind an important point: no person needs to exhibit all the characteristics described for Kleefstra in order to have the syndrome.

Earlier descriptions of KLEFS1 relied heavily on individuals with a more prominent and severe clinical presentation. As genetic diagnosis becomes more widespread, people with milder or less «typical» characteristics are now also being identified. The large 2024 international study clearly confirmed this broadening of the phenotypic spectrum.

Why different genetic tests were needed

The two cases also show very nicely why in modern genetic diagnosis there is not a single test that always detects every type of damage.

In the first child, the diagnosis was made with exome sequencing, because it was a small change in the sequence of EHMT1.

In the second, the diagnosis was made with chromosomal microarray, because it involved the loss of a small segment of DNA, a copy-number variant, which included several exons of the gene.

In both cases, older tests such as conventional karyotyping had not provided the diagnosis.

This is particularly important in children with unexplained developmental delay: a «normal» result on an older genetic test it does not necessarily mean that a molecular genetic cause has been ruled out. The choice of the appropriate test depends on the clinical presentation and should be made by a clinical geneticist or the responsible medical team.