Early onset or syndromic epilepsy

Gene: BAIAP2

Green List (high evidence)

Comment on list classification: There are 7 unrelated individuals reported in literature with de novo heterozygous missense variants in BAIAP2. 6 individuals presented with developmental and epileptic encephalopathy (PMID: 41133935), while 1 individual was diagnosed with classic lissencephaly (PMID: 38149472). All individuals presented with early onset epilepsy. Baiap2 knockdown in mouse cause abnormalities in neuronal migration, morphogenesis, and differentiation (PMID: 38149472). Additional functional evidence from coimmunoprecipitation studies shows that missense variants around aa 340-366 in BAIAP2 are likely to disrupt binding of 14-3-3, necessary for BAIAP2 inhibition (PMID: 30696821). Based on the available evidence, BAIAP2 should be promoted to Green for Early onset or syndromic epilepsy.

Created: 17 Nov 2025, 1:02 p.m. | Last Modified: 17 Nov 2025, 1:06 p.m.

Panel Version: 8.66

PMID: 38149472 Tsai et al., 2024
6yo individual from Taiwan with a de novo c.85C>T, p.Arg29Trp variant in BAIAP2 with lissencephaly. Method: WES +Sanger.
Phenotype: severe global developmental delay since infancy, lack of speech development, poor motor milestone achievement; refractory epilepsy with multiple seizure patterns, including focal, myoclonic, tonic, absence and tonic-clonic seizures, and epileptic spasms since infancy. Classical lissencephaly features noted on brain MRI at 5yrs.
Functional evidence: Baiap2 knockdown in mouse cause abnormalities in neuronal migration, morphogenesis, and differentiation. Expression of p.Arg29Trp failed to rescue the defect - suggests LoF effect for this variant.

PMID: 41133935 Zhang et al., 2025
6 patients reported with de novo missense variants in BAIAP2. All six patients exhibited severe infantile or early childhood onset epilepsy - refractory seizures in four individuals. Age of onset of epilepsy: 5 mo - 1 yr 7 mo. Additionally, 6/6 individuals presented with global developmental delay; 4/6 are nonverbal and 2/6 only know a few words; ID severity was variable: mild (1/6), moderate-severe (3/6), and severe-profound (2/6). Diverse ancestry: Chinese, British, Turkish, White Kosovar, Austrian. Patients were heterozygous for the following missense variants: c.1088G>A, p.Arg363His; c.1019C>A, p.Thr340Lys; c.1018A>G, p.Thr340Ala; c.566A>T, p.Glu189Val; c.1019C>T, p.Thr340Ile; c.1024C>G, p.Pro342Ala. The variants are not present in gnomAD v4.1.0.
Authors pose that 5/6 variants reported are likely to destabilise binding of 14-3-3 to BAIAP2.

Functional evidence: PMID: 30696821 Kast & Dominguez, 2019
Protein 14-3-3 binds to two pairs of sites in the linker between the CRIB-PR and SH3 domains of BAIAP2 (pT340/pT360 or pT340/pS366) - consistent with the hotspot region seen in patients (5/6 variants reported in PMID: 41133935 are in the 340-366 region). The binding of 14-3-3 is crucial for inhibition of BAIAP2 action - particularly regulating actin and filopodia formation.

This gene is not yet associated with a phenotype in OMIM (accessed 17th Nov 2025).

Created: 17 Nov 2025, 12:59 p.m. | Last Modified: 17 Nov 2025, 1:03 p.m.

Panel Version: 8.66

Mode of inheritance
MONOALLELIC, autosomal or pseudoautosomal, imprinted status unknown

Phenotypes
developmental and epileptic encephalopathy, MONDO:0100620; classic lissencephaly, MONDO:0015146

Publications

• 30696821
• 38149472
• 41133935

I don't know

De novo missense variants detected in 6 unrelated individuals with DEE. One of which has also been seen on DECIPHER and deposited as a research variant (also de novo).
Seizure type is highly variable.
ID ranges from mild to profound.
Animal models (zebrafish) provide some evidence of pathogenicity through abnormal neurite growth (lack or absence of filopodia) and increased neuronal excitability (reduced rheobase & increased AP firing patterns in whole cell recordings). Ion channels within membrane remain unaffected - no significant difference between WT and mutants in specific AP characteristics such as amplitude, voltage threshold, afterhyperpolarisation, and half-width values.
The above evidence led the authors hypothesis that GoF is the mechanism of disease for the variants in this study, however, they propose that LoF variants may also be pathogenic (no evidence to confirm this).
Further studies needed to verify findings and confirm this proposed biphasic disease mechanism model (LoF & GoF variants cause disease).
Currently only evidence for GoF in for a very specific repertoire of residues that are clustered within the phosphorylation site region that is critical to 14-3-3 binding for autoinhibited conformation: T340, Pro342, Arg363 and one variant within the I-BAR domain which is crucial for maintaining the overall positive charge of the dimer: Glu189Val.
Sources: Literature

Created: 7 Nov 2025, 10:47 a.m.

Mode of inheritance
MONOALLELIC, autosomal or pseudoautosomal, NOT imprinted

Phenotypes
DEE

Publications

• PMID: 41133935

Mode of pathogenicity
Other