Gene therapy decreases seizures in a model of Incontinentia pigmenti
IP is caused by pathogenic variants in the X‐chromosomal Nemo (IKBKG) gene that inactivate the protein. A genomic rearrangement deleting exons 4 to 10 is found in 75% of patients.3 IP presents mainly in females and occasionally in males with XXY karyotype (Klinefelter's syndrome) or mosaicism attributed to somatic mutations because the complete absence of NEMO is lethal. Studies in a mouse model of IP have shown that NEMO is essential for the function and survival of brain endothelial cells, whereas neurons and glial cells seem to cope without NEMO.4 In the absence of endothelial NEMO, the blood–brain barrier (BBB) is leaky. Moreover, brain endothelial cells die, resulting in a rarefaction of capillaries and hypoperfusion of the brain. Attributed to the disrupted BBB and the cerebral ischemia, mice suffer from epileptic seizures that closely mimic the neurological manifestations of IP.4 The concept that IP is a cerebrovascular disease is further supported by clinical case reports showing a vascular pattern of lesions on magnetic resonance imaging.5 So far, there is no causal therapy.
Gene therapy of the severe neurological manifestations of IP would require a vector to specifically deliver the Nemo gene to brain endothelial cells. This aim came within reach when a novel strategy of in vivo screening led to a suitable adeno‐associated virus (AAV) 2–based vector with mutated capsid.6 The new AAV‐BR1 vector shows unprecedented specificity for the brain endothelium and low transduction of the liver or other easily accessible peripheral organs after intravenous injection.6 Using this vector, we found that transduction of brain endothelial cells with the Nemo gene indeed prevented the disruption of the BBB and the death of brain endothelial cells. However, whether this gene therapy would also improve epileptic seizures was unclear, because, in contrast to humans, mice with a heterozygous Nemo deficiency die of skin manifestations at an early age before the presence of epileptic seizures could be evaluated.7 To circumvent this problem, we used a conditional model of Nemo knockout in brain endothelial cells in which mice develop cerebral symptoms of IP, that is, a disrupted BBB, loss of brain endothelial cells, and epileptic seizures, without the lethal skin manifestations.4 Here, we show that gene therapy with a brain endothelial‐specific AAV vector reverses albumin extravasation and ameliorates the seizure phenotype, suggesting a high translational potential of the new approach. Importantly, the low transduction of the liver and other peripheral organs by AAV‐BR1 was associated with a favorable safety profile after 11 months, especially with no evidence of hepatocellular carcinoma (HCC).