PHACTRing in actin: actin deregulation in genetic epilepsies

This scientific commentary refers to ‘De novo PHACTR1 mutations in West syndrome and their pathophysiological effects’, by Hamada et al. (doi:10.1093/brain/awy246). West syndrome is a severe epileptic encephalopathy (EE) characterized by early-onset epileptic spasms, developmental regression and a prototypical electroencephalographic pattern termed hypsarrhythmia. Although close to 50% of cases are attributable to brain injuries (e.g. hypoxic-ischaemic encephalopathy, infections, neurometabolic disorders) or malformations (e.g. lissencephaly), many others remained of unknown aetiology until recently. The advent of Next Generation Sequencing has helped identify genetic causes for a large proportion of the remaining cases, revealing a preponderance of rare monogenic disorders caused by de novo variants in a variety of genes, and demonstrating the genetic heterogeneity of this clinical syndrome (Epi4K Consortium et al., 2013; Michaud et al., 2014; Hamdan et al., 2017). In this issue of Brain, Hamada and co-workers provide compelling genetic and functional evidence suggesting that mutations in the PHACTR1 gene are a novel genetic cause of West syndrome (Hamada et al., 2018). By investigating a large cohort of 800 children with early-onset EE using whole exome sequencing (WES), the authors identified two de novo variants in the PHACTR1 gene [c.1449T>C/p.(Leu500Pro) and c.1436A>T/p.(Asn479Ile)] in children with West syndrome, suggesting that this might be a novel candidate gene for the disorder. Moreover, by mining earlier WES datasets, they catalogued three additional de novo variants in PHACTR1 [c.1148C>T/p.(Ser383Leu); c.1553T>A/p.(Ile518Asn) and c.1561C>T/ p.(Arg521Cys)] in patients with a range of neurodevelopmental disorders, supporting a likely causality. All five de novo mutations were rare variants (absent or highly infrequent in control populations) and were predicted to be deleterious using four different bioinformatic tools, suggesting their probable pathogenicity. While PHACTR1 polymorphisms predispose to cardiovascular diseases (Kathiresan et al., 2009), this is the first direct evidence of an association between the PHACTR1 gene and neurological disorders. PHACTR1 encodes phosphatase and actin regulator protein-1 (PHACTR1), which comprises an N-terminus nuclear localization signal, four RPEL repeats and a C-terminus protein phosphatase 1 (PP1)-binding domain. RPEL repeats bind actin and are considered key players in regulating cell motility and morphology during neuronal development. Further, PP1 proteins are serine/threonine phosphatases that regulate the phosphorylation status of a variety of target proteins involved in neuronal migration and cell signalling (Shmueli et al., 2006). PHACTR1 is expressed in cortical neurons throughout brain development and its levels are particularly high in pre- and postsynaptic densities (Ito et al., 2018), suggesting a role in synaptic activity. However, the functions of PHACTR1 during brain development remain largely unknown. Using RNAi-mediated gene repression and in utero electroporation in the ventricular zone of embryonic mouse brains [at embryonic Day (E) 14.5], Hamada et al. demonstrate that reduced Phactr1 expression impairs neuronal migration. Indeed, instead of reaching the more superficial cortical layers (II–IV), mutant pyramidal cells become trapped in the intermediate ventricular zone and lower cortical plate. Further, Hamada et al. document significant alterations in neuronal morphology and polarity of the misplaced neurons. These deficits in migration and morphology were persistent at postnatal days (P)0 and P7, consistent with a blockade rather than a delay of neuronal migration. Importantly, this migration deficit was not attributable to a reduction in neuronal proliferation, as revealed by preserved numbers of Ki67-positive cells. Further, by using a lower concentration of RNAi plasmids, presumably resulting in a lesser reduction of Phactr1 expression, the authors show that mutant neurons that migrate properly to their correct position in the cortical plate still present significant morphological deficits, with striking reductions of dendritic length and complexity at P21. Thus, PHACTR1 function is essential for both embryonic development and postnatal maturation of pyramidal cells (Fig. 1A). The deficits observed were rescued using co-electroporation of an RNAi-resistant cDNA, supporting the specificity of the findings to the repression of Phactr1 rather than non-specific off-target effects. Schematic representation of the effect of Phactr1 repression during neuronal development, as demonstrated by Hamada et al. (2018). (A) In a healthy developing brain (left), neurons exit the ventricular zone and migrate along the radial glia towards the superficial layers of the cortical plate (shown here from E14.5 to P0). By P21, pyramidal neurons ramify their dendritic arborization, enabling them to process incoming inputs from other neurons. Mutations in PHACTR1 (right) induce severe morphological and migration defects, trapping the neurons in the intermediate ventricular zone and lower cortical plate during their migration phase. By P21, they display a simplified dendritic arborization and reduced excitability (modified from Kwan et al., 2012). (B) Under normal conditions (wild-type), PHACTR1 binds actin filaments and PP1, and interacts with the Slack potassium channel (KNa1.1). However, patient-derived mutations prevent these molecular interactions. (C) Illustrations of (i) a normal gene; (ii) haploinsufficiency; and (iii) a dominant-negative effect of a genetic mutation. (iv) Silencing Phactr1 by electroporating either 2 μg of RNAi alone or 2 μg of RNAi together with 1 μg of mutant Phactr1 results in inactivation of the corresponding protein. (v) By contrast, protein function of PHACTR1 is not altered when silencing the gene with half the amount of either RNAi or mutated cDNAs. (vi) A dominant-negative effect is observed when the same reduced amounts of RNAi and Phactr1 mutated cDNAs are co-transfected, resulting in protein inactivation. R521C = c.1561C>T/ p.(Arg521Cys); L500PS = c.1449T>C/p.(Leu500Pro); N479I = c.1436A>T/p.(Asn479Ile); I518NS = c.1553T>A/p.(Ile518Asn); F-actin = filamentous actin; IZ = intermediate zone; PP1 = protein phosphatase 1; RNAi = RNA interfering molecule; SP = subplate; SVZ = subventricular zone. The observed deficits in cell migration and morphology are interesting given the essential roles of actin remodelling in neuronal development. Indeed, cortical neurons have their origins in the ventricular zone and migrate radially towards the subventricular zone, where they adopt a transient multipolar morphology. They then adopt a bipolar morphology to reinitiate migration towards the cortical surface by attaching to and migrating along the radial glia. To migrate efficiently, cortical neurons must continuously extend their leading process and retract their trailing process, events intimately dependant on the dynamic remodelling of the actin cytoskeleton (Marín et al., 2006). Notably, a large proportion of the genes associated with EE, autism and intellectual disability encode regulators of actin dynamics involved in cytoskeleton remodelling (Michaud et al., 2014; Zamboni et al., 2018). Thus, the observation that Phactr1 repression causes severe neuronal migration deficits along with abnormal cell morphology is consistent with the predicted impact of PHACTR1 on actin remodelling and adds to a growing body of evidence implicating this biological mechanism in genetic neurodevelopmental disorders. To support the association of the PHACTR1 gene with neurodevelopmental disorders, Hamada et al. sought to investigate the functional impact of patient-derived mutations to confirm their predicted pathogenicity. First, using expression of mutant PHACTR1 cDNA in a heterologous system (COS7 cells) and immunoprecipitation assays, the authors show a reduction of actin binding by three variants (p.Leu500Pro, p.Asn479Ile, and p.Ile518Asn) and a loss of binding to PP1 by the other variant (p.Arg521Lys). Further, the authors show that the four mutant Phactr1 cDNAs fail to rescue the cellular deficits observed after gene repression of Phactr1 in E14.5 neurons. Thus, all mutations tested appear to have significant functional impact on Phactr1 function and on neuronal development, confirming their pathogenicity. Heterozygous missense mutations (i.e. variants changing one amino acid on one copy of the gene) may exert their effect either through a loss-of-function of one copy of the gene (so-called haploinsufficiency), where 50% reduction in gene expression is sufficient to cause a given disorder; or through a dominant-negative effect, in which the mutant allele prevents proper function of the remaining wild-type allele, for instance by trapping the wild-type protein in the endoplasmic reticulum, directing it for degradation, or by interacting with critical binding partners and blocking essential interaction sites for the wild-type protein (Fig. 1B). Clarifying the molecular mechanisms of novel mutations thus carries significant therapeutic implications since replacement therapy would be effective in haploinsufficiency, whereas chaperone therapies may be required to salvage the wild-type allele in cases of dominant-negative mutations. Hamada et al. now provide strong evidence for a dominant-negative effect of the four mutations studied. Indeed, not only did the mutant cDNA fail to rescue the migratory deficit after co-electroporation with the RNAi-expressing plasmids in e14.5 neurons, it actually exacerbated the deficits. Further, mere overexpression of the mutant cDNA was sufficient to impair neuronal migration, whereas expression of the wild-type cDNA did not affect neuronal migration, supporting an intrinsic negative impact of the mutant proteins on the wild-type proteins. Notably, although partial repression of Phactr1 with a lower concentration of the RNAi did not impair migration, the additional expression of the mutant cDNA did lead to migration deficits, further suggesting an impact of the mutant proteins on the residual wild-type protein (Fig. 1C). Similar dominant-negative effects have recently been reported for other EE-associated genes (He et al., 2018). The discovery that most PHACTR1 mutations are dominant-negative expands the group of genetic EEs that may benefit from chaperone therapies and could thus have major repercussions for future therapeutic strategies for patients with PHACTR1-associated neurodevelopmental disorders. Despite the strong evidence presented by Hamada et al. for the critical role of PHACTR1 in neuronal development and for the association of de novo PHACTR1 dominant-negative mutations with a spectrum of neurodevelopmental disorders, a number of issues remain to be resolved. First, as the authors point out, in utero electroporation and gene repression approaches induce variable levels of gene silencing in the target cells. Thus, further validation of these findings in heterozygous and homozygous knockout mice will help assess the precise impact of Phactr1 haploinsufficiency compared to homozygous loss-of-function on neuronal development and network function. Second, it is not entirely clear how the observed cellular phenotypes relate to the human disease. In particular, although the knockdown experiments impaired neuronal migration, none of the patients described had cortical malformations that would have signalled a defect in pyramidal cell migration (i.e. lissencephaly or neuronal heterotopia). Further, a reduction in dendritic complexity or in neuronal excitability (as demonstrated by reduced input resistance and decreased firing rates) is unlikely to result in epilepsy, although these may well contribute to the observed cognitive deficits. However, the additional involvement of other neuronal populations, for instance inhibitory GABAergic interneurons, probably contributes to the disease pathophysiology. Indeed, mutations causing delays in interneuron migration or altered interneuron morphology have been shown to result in epilepsy and cognitive comorbidities in a variety of animal models of EE and autism (Rossignol, 2011; Katsarou et al., 2017). Thus, the impact of targeted Phactr1 deletions on different neuronal populations and the sufficiency for epilepsy onset will need to be investigated, perhaps with conditional knockout models in the future. Finally, the cellular and molecular mechanisms by which Phactr1 mutations affect neuronal migration and dendritic development remain uncertain. The authors report a dramatic decrease in the actin-binding activity of three mutated PHACTR1 cDNAs, and a reduced affinity for PP1 for the fourth mutated cDNA. However, the net impact of the loss of these interactions on actin remodelling is unknown. Determining the specific function of PHACTR1-PP1 interaction on actin dynamics and on the processes of neuronal locomotion and neurite remodelling will help shed light on the precise mechanisms underlying PHACTR1-associated neurodevelopmental disorders. Other unanswered questions include whether other PHACTR1 binding partners play a role in the pathophysiology of the disorder and whether the phenotype may be rescued using other actin modulators. Answering these questions will help further clarify the disease mechanisms and identify potential therapeutic targets. In summary, Hamada and colleagues report two novel and three published de novo mutations in the PHACTR1 gene in patients with a spectrum of neurodevelopmental disorders. They provide functional evidence for the pathogenicity of these variants, both in terms of neuronal development and of disrupted molecular interactions with actin or PP1. Importantly, they show that these mutations have a dominant negative effect, a finding with major therapeutic implications. Altogether, this study demonstrates the power of functional studies to decipher the role of novel candidate genes identified through WES in patients with genetically heterogeneous disorders, such as EE. Glossary Dominant negative effect: A dominant negative effect occurs when a heterozygous mutation, affecting only one allele, results in an aberrant product that represses the healthy wild-type allele, for instance by trapping the wild-type protein in cytoplasmic compartments away from its normal site of action (i.e. endoplasmic reticulum or lysosomes) or by impairing interactions of the wild-type protein with necessary binding partners. Thus, heterozygous dominant negative mutations exert a molecular effect akin to a homozygous loss-of-function. Dominant negative mutations usually result in more severe phenotypes than those causing haploinsufficiency. Haploinsufficiency: Haploinsufficiency refers to a heterozygous mutation in a gene that will cause one of the two alleles to be non-functional. This results in a 50% decrease in the amount of protein generated from this particular locus and will lead to pathological findings in circumstances when this reduced protein expression is insufficient to maintain cellular function. In utero electroporation: A transfection method that delivers an experimental plasmid (i.e. expressing an RNAi for genetic repression or a mutant cDNA for over-expression or rescue experiments) to dividing cells of the embryonic mouse brain by the application of electrical pulses that drive negatively charged DNA molecules into cells on the beam path. The plasmid does not integrate into the cell genome but is rather stably expressed as an independent plasmid DNA throughout the lifetime of the cell. Modified cells can thus be tracked and assessed at later embryonic stages, or at postnatal time points. RNA interference (RNAi): An RNAi is a small interfering RNA fragment designed to recognize a particular sequence of a target mRNA. It can be expressed in cells in the form of a short-hairpin RNA (shRNA), in which complementary sequences are cloned on each side of the RNAi sequence, thus leading to the stable expression of a folded hairpin-like structure once transcribed. ShRNAs are processed by the endogenous cellular RNA interference machinery into double-stranded RNA, which is then cleaved by the endogenous Dicer cleavage enzyme into small RNA fragments, the RNAi. RNAi then bind the target mRNA and RNAi-mRNA dimers are diverted for degradation by the intracellular machinery, thus preventing mRNA translation and effectively resulting in gene repression (silencing). We thank Mathieu Lachance for designing the figure. L.E. receives post-doctoral training awards from the Fond de Recherche du Québec en Santé (FRQS) and from Brain Canada-Kids Brain Health Network through the Canada Brain Research Fund, with the financial support of Health Canada. E.R. receives a Young Investigator Salary Award from the Canadian Institutes for Health Research (CIHR). The authors report no competing interests.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.