Abstract
Wan Y, Morikawa M, Morikawa M, Iwata S, Naseer MI, Chaudhary AGA, Tanaka Y, Hirokawa N. J Cell Biol. 2023;222(2):e202208108. doi:10.1083/jcb.202208108
Epilepsy is a common neurological disease worldwide, and one of its causes is genetic abnormalities. Here, we identified a point mutation in KIF4A, a member of kinesin superfamily molecular motors, in patients with neurological disorders such as epilepsy, developmental delay, and intellectual disability. KIF4 is involved in the poly (ADP-ribose) polymerase (PARP) signaling pathway, and the mutation (R728Q) strengthened its affinity with PARP1 through elongation of the KIF4 coiled-coil domain. Behavioral tests showed that KIF4-mutant mice exhibited mild developmental delay with lower seizure threshold. Further experiments revealed that the KIF4 mutation caused aberrant morphology in dendrites and spines of hippocampal pyramidal neurons through PARP1-TrkB-KCC2 pathway. Furthermore, supplementing NAD, which activates PARP1, could modulate the TrkB-KCC2 pathway and rescue the seizure susceptibility phenotype of the mutant mice. Therefore, these findings indicate that KIF4 is engaged in a fundamental mechanism regulating seizure susceptibility and could be a potential target for epilepsy treatment.
Commentary
The present study by Wan et al identified a KIF4A gene variant (human exon 20, R728Q) in a family with affected individuals presenting with intractable epilepsy, developmental delay, and severe intellectual disability. 1 The mother, 2 daughters, and 1 son carried a mutant allele, and only the son presented with severe symptoms, indicating sex-linked inheritance. None had abnormalities on MRI. KIF4A is an X-chromosome-linked kinesin-4 kinesin family gene on Xq13.1 and its paralog KIF4B is an intron-less version of KIF4A on 5q33.1. 2 Human KIF4 belongs to the protein kinesin superfamily (KIF) that have fundamental roles in brain functioning, development, survival, and plasticity by regulating the transport of cargo along microtubules. 3,4 A previous study had also identified mutations in KIF4A that were responsible for intellectual disability and epilepsy and altered the balance between synaptic excitation and inhibition (in vitro). 5 Here, following the KIF4A variant identification, the authors performed an exquisite and thorough study that identified the 5 predicted impact of the mutation on KIF4 function, generated transgenic mice that exhibited many symptoms reported in humans, and identified a clear molecular and cellular mechanism contributing to increased anxiety and seizure susceptibility by building on their previous work on KIF4 and its binding partner, poly (ADP-ribose) polymerase 1 (PARP1). 6 The authors even reported a potential treatment for seizures that can be readily implemented in patients.
To address the mechanisms linking KIF4A mutation to behavioral phenotype, the authors generated Kif4 R728Q mutant transgenic mice using the CRISPR/Cas9 system. Mouse Kif4 is on chromosome X and has 82% sequence homology with human KIF4A. 2 Most analyses were performed using Kif4a mut/y mice (male) and Kif4a y/y (control) mice that showed smaller brain (especially the hippocampus) and global delay of developmental milestones. Behavioral characterization of the mice illustrated increased anxiety, impaired learning and memory (using hippocampal based tasks), increased susceptibility to seizures using the pentylenetetrazole (PTZ) test, and epileptic EEG patterns in the primary motor cortex and hippocampus (using deep recording electrodes) following PTZ injections at 3 weeks of age.
Based on the well-known structure of KIF4 (containing an N-terminal motor region, coiled-coil region, and C-terminal loading region), and the fact that the mutations occurred in the coiled-coil domain, the mutation was predicted to result in elongation of the coiled-coil domain that could affect KIF4 binding with several of its partners. 3 The authors pursued the PARP1 binding partner based on their previous work and the importance of PARP signaling pathway in DNA repair, regulation of gene transcription, and neuronal survival. 6 Poly (ADP-ribose) polymerase 1 catalyzes a post-translational modification, poly(ADP-ribosyl)ation, by promoting the transfer of ADP-ribose moieties from NAD+ to specific amino acids and the elongation of a protein-mono (ADP-ribose) into a PAR branched chain. 7 Using carefully crafted biochemistry experiments, they found that mutant KIF4 displayed increased binding affinity to PARP1 resulting in suppression of PARP1 activity, but no change in PARP1 levels. As PARP1 is a critical epigenetic factor that can regulate several key genes, the authors hypothesized that decreased PARP activity mediates the behavioral and seizure phenotypes through dysregulation of specific downstream signaling pathways.
First, the authors examined CA3 hippocampal neuron dendrite morphology and identified a hyper-branched phenotype in vivo and elongated dendritic spines in vitro. They then followed a hypothesis-driven path and examined the expression levels of epilepsy-related genes in these mice and in cultured mutant versus control hippocampal neurons. They identified increased full-length and truncated TrkB expression, increased AKT and ERK/1/2 activity, and reduced levels of PAR and KCC2, which is well-known to be regulated by TrkB activity. As KCC2 is a potassium-chloride cotransporter, this resulted in higher intracellular chloride (iCl) concentration in mutant hippocampal neurons (assessed using chloride imaging). Using an in vitro combination of pharmacological blockers for PARP1 and TrkB, and a PARP1 activator, NAD, they validated altered KIF4A-PARP1-TrkB-KCC2-iCl signaling in Kif4a mut/y mice and neurons. Further, using these pharmacological approaches, as well as PARP1 overexpression to manipulate PARP1 or TrkB activity, they identified a direct role of these signaling pathways in the hyper-branched dendrite and elongated spine phenotypes of Kif4a mutant neurons. Finally, NAD injection in Kif4a mut/y mice inhibited the anxiety and epileptic phenotype, suggesting that increasing PARP1 activity can reverse these neurological symptoms. Intriguingly, using a PARP1 blocker in Kif4a y/y mice to decrease PARP1 activity as seen in Kif4a mut/y mice was sufficient to trigger the seizure susceptibility phenotype observed in Kif4a mut/y mice. Collectively, the authors identified a critical role of the KIF4A-PARP1-TrkB signaling pathway in KCC2 expression and iCl homeostasis, dendrite and spine abnormalities, and behavioral deficits in Kif4a mut/y mice. Importantly, they identified a potential treatment option using NAD injections.
One of the most important findings of this study is the identification of a key intracellular signaling—PARP—that is directly involved in the anxiety and seizure phenotype of Kif4a mut/y mice. This may be viewed as a surprising finding considering that KIF4 has multiple binding partners, the affinity and activity of which may be altered and contribute to the behavioral and seizure phenotypes. However, even if different molecules and signaling pathways were to be altered in Kif4a mut/y mice in addition to PARP1 signaling, it is conceivable that all of these alterations combined are necessary to trigger behavioral alterations while blocking only one is sufficient to prevent abnormal cellular and behavioral phenotypes as reported for other epilepsy disorders. 8 Intriguingly, blocking PARP1 in control mice phenocopied the anxiety and seizure susceptibility observed in mutant mice. However, the degree of PARP1 inhibition was not assessed and may be more pronounced than in Kif4a mut/y mice. Considering the critical importance of PARP signaling, it is nevertheless not too surprising that inhibition of such signaling led to altered behavior and increased seizure susceptibility. With respect to the PARP1 downstream players, TrkB and KCC2, their role in neuron dysmorphogenesis was supported using the TrkB blocker, however, whether they contribute to the behavioral and epilepsy phenotype was not directly tested. It is likely that PARP1 has multiple downstream effects in addition to TrkB-KKC2 that contribute to the disease pathology and phenotype. Finally, additional approaches using shRNA strategies would be beneficial in future studies as well as identifying additional KIF4 binding partners and analyses of neuronal firing and synaptic integration. Independent of these unknowns, the present study identified a potential treatment option using NAD injections for individuals with KIF4A mutations. Whether this can be applied for other types of epilepsy remains to be investigated.
