Understanding EEF1A2

1 · The protein and its job

Delivering aminoacyl-tRNA to the ribosome

eEF1A2's molecular function is compactly stated: it promotes the GTP-dependent binding of aminoacyl-tRNA to the A site of the ribosome during protein biosynthesis. Established

Unpacking that: as the ribosome works along a messenger RNA, each new amino acid must arrive already attached to its matching transfer RNA, and it must arrive at the right moment in the right orientation. eEF1A2 is the carrier. In its GTP-bound state it holds a charged tRNA; it docks that cargo into the ribosome's A site; if the tRNA matches the codon being read, GTP is hydrolysed to GDP, the factor's grip relaxes, and it leaves — now in the GDP state, and unable to carry anything until it is reloaded.

The reloading is not spontaneous in practice. Exchanging GDP for GTP is accelerated by a dedicated guanine nucleotide exchange factor; for eEF1A the cognate exchange factor is eEF1B. Established

Why the nucleotide matters so much: the two switches

Translational GTPases of this family work as conformational switches rather than as simple enzymes. The identity of the bound nucleotide — GTP or GDP — is read out by two mobile structural elements, conventionally called switch I and switch II, which change position depending on what is in the nucleotide pocket. Those movements are what propagate a chemical event in one small pocket into a change in how the whole protein grips its cargo. Established

This matters for variant interpretation, and the published structural analysis of patient variants makes the point directly. Of the variants modelled in one cohort, p.G70 is the only one placed in the switch I loop, positioned towards the eEF1B/tRNA pocket formed by domains I, II and III. p.F98C and p.M102V sit buried within the switch II helix, while p.D91N, p.G384R and p.R423C are placed directly adjacent to it. EEF1A2 evidence

And the largest genotype–phenotype study to date reports that the variants associated with a severe phenotype, and the majority of those associated with a moderate phenotype, cluster within the switch II region. EEF1A2 evidence

Three domains

eEF1A2 is built from three domains. Boundaries differ slightly between annotation sources, so treat these as approximate.

The variant database plots published variants along this domain map.

A second job, and a second reason neurons care

Beyond translation, eEF1A proteins interact with the actin cytoskeleton. This is relevant here because functional work on recurrent patient variants has reported altered actin bundling and altered neuronal morphology alongside the translation defect. EEF1A2 evidence A factor that touches both protein synthesis and cytoskeletal organisation has two routes by which it could shape how a neuron grows and connects.

Expression: why this gene and not its twin

Humans have two closely related genes for this factor. EEF1A1 is expressed broadly across tissues; eEF1A2 protein expression is concentrated in brain, heart and skeletal muscle. Established

Crucially, the two are not simply alternatives — there is a developmental handover. eEF1A2 replaces the ubiquitously expressed eEF1A1 as neurons develop, becoming the predominant factor. EEF1A2 evidence

Two consequences follow, and both shape this project:

• Timing. A variant in a gene that takes over during neuronal maturation is a variant whose consequences are tied to a developmental window, not a constant background.
• Experimental access. Cell models that still express eEF1A1 can mask what a variant does. The E122K mouse authors note explicitly that transformed, immortalised, iPSC-derived and primary cultures studied so far all express eEF1A1, either exclusively or alongside eEF1A2, confounding genotype–phenotype work. Anyone designing experiments on this gene has to solve that problem first.

2 · EEF1A2-associated neurological disorders

The reported clinical spectrum

Across cohorts, the features reported in association with pathogenic EEF1A2 variants include developmental delay — particularly speech delay — intellectual disability ranging from mild to severe, autism spectrum features, attention deficit, early-onset epilepsy, hypotonia, ataxia and sleep disorder.

In a cohort selected for epilepsy and/or a dyskinetic movement disorder, the described picture was more severe: global developmental delay in infancy usually progressing to moderate-to-profound intellectual disability, with most individuals non-verbal and non-ambulatory; near-universal severe hypotonia; movement disorders — mainly dystonia and choreoathetosis — in most patients; and behavioural features, dysphagia and respiratory problems in a subset.

Epilepsy

Where epilepsy is present, reported onset is early. In the epilepsy-ascertained cohort the median age of seizure onset was four months, with a range from the first day of life to eight years; myoclonic seizures were the most common presenting type, seizures were often refractory, and no single anticonvulsant was consistently effective. Late onset is documented too: one individual carrying D252H did not develop epilepsy until the age of eight.

Neuroimaging and course

Early MRI is often normal or shows delayed myelination and a thin corpus callosum; longitudinal imaging showed cortical and cerebellar atrophy in some patients, a subset had developmental regression, and two patients in that cohort died by age four.

Other organ systems

Reports beyond the nervous system exist but are individually sparse, and we flag them as single observations rather than established features:

• Respiratory. An infant presenting at three months with hypotonia, microcephaly, failure to thrive and respiratory insufficiency with central apneas requiring support.
• Cardiac. A nine-year-old with a de novo heterozygous p.V16L and dilated cardiomyopathy, described as the first heterozygous patient with this cardiac phenotype. One case. Not an established feature.
• Overlapping diagnoses. A de novo EEF1A2 variant identified in a patient presenting as Rett-syndrome-like, after the common Rett genes were excluded.

How much the picture widened — and why that matters

The most important development in the clinical literature is not a new symptom but a correction of severity. The largest cohort to date, 26 individuals recruited through the European Reference Network ERN-ITHACA and compared against 42 previously published cases, found a significantly milder phenotype than the literature had suggested: 83% walking versus 29%, 84% with language skills versus 15%, epilepsy in 63% versus 93%, hypotonia in 58% versus 96%, pyramidal signs in 24% versus 68%. Three individuals in that cohort had no intellectual disability at all.

The authors attribute this to ascertainment: early reports came from cohorts selected for severe epileptic encephalopathy, so the gene's description inherited that selection. They also raise the possibility of incomplete penetrance for certain variants, not previously described for this gene.

A published Correction to that cohort study exists and must be read alongside it; resolving it is an open item in our bibliography, and figures quoted above should be treated as provisional until we have done so.

Phenotypic variability is a finding, not noise

Several observations make clear that genotype does not determine outcome in any simple way:

• The same variant has produced overlapping but non-identical presentations in different individuals.
• Two individuals with de novo variant mosaicism did not have milder phenotypes — so mosaicism cannot be assumed to predict severity.
• p.M102V impaired growth in a yeast functional assay yet was inherited from an unaffected father. A functional assay result is not a diagnosis.

Browse published variants How the mechanism is being investigated

References