Abstract
Patzke C, Han Y, Covy J, Yi F, Maxeiner S, Wernig M, Südhof TC. J Clin Invest. 2015;125:3560–3571.
Heterozygous mutations in the syntaxin-binding protein 1 (STXBP1) gene, which encodes Munc18-1, a core component of the presynaptic membrane-fusion machinery, cause infantile early epileptic encephalopathy (Ohtahara syndrome), but it is unclear how a partial loss of Munc18-1 produces this severe clinical presentation. Here, we generated human ES cells designed to conditionally express heterozygous and homozygous STXBP1 loss-of-function mutations and studied isogenic WT and STXBP1-mutant human neurons derived from these conditionally mutant ES cells. We demonstrated that heterozygous STXBP1 mutations lower the levels of Munc18-1 protein and its binding partner, the t-SNARE-protein Syntaxin-1, by approximately 30% and decrease spontaneous and evoked neurotransmitter release by nearly 50%. Thus, our results confirm that using engineered human embryonic stem (ES) cells is a viable approach to studying disease-associated mutations in human neurons on a controlled genetic background, demonstrate that partial STXBP1 loss of function robustly impairs neurotransmitter release in human neurons, and suggest that heterozygous STXBP1 mutations cause early epileptic encephalopathy specifically through a presynaptic impairment.
Commentary
Epilepsy is a chronic disorder characterized by unprovoked seizures. In severe cases of childhood epilepsies, the cause of epilepsy may be related to genetic predisposition (1, 2), but how specific gene mutations lead to epilepsy is generally unknown. Recent advances in genomic technologies such as microarray-based comparative genomic hybridisation (array CGH) or DNA sequencing have revealed hundreds of heterozygous mutations in patients with neurodevelopment disorders including epilepsy (3). While various epilepsy genetics initiatives continue to reveal a treasure trove of potentially interesting gene mutations, a major bottleneck will be to distinguish disease-causing mutations from natural polymorphisms. Traditionally, animal models (e.g., mouse or zebrafish) have served as useful tools to uncover mutant phenotypes representing human epilepsy genes (4). However, to generate a mouse mutant for every human epilepsy gene mutation would be too time consuming and costly. While zebrafish models offer the possibility of high throughput mutational analysis, there may still be limitations in translating zebrafish to humans due to inherent differences in their nervous systems (5).
Fortunately, Takahashi and Yamanaka's landmark (6) discovery that adult somatic cells can be reprogrammed into induced pluripotent stem cells (iPSCs) with a cocktail of transcription factors has opened up the possibility to generate patient-specific cell lines to model disease. Zhang and colleagues (7) took it a step further by showing that human embryonic stem (ES) or iPSCs can be induced to differentiate into neurons (induced neuron or iN) more rapidly by forced expression of a defined transcription factor. Combined with the revolution of gene editing using CRISPR-Cas9 or other tools, this allows the generation of human ES or iPSC lines with specific mutations in the same background of the control cells (8).
Heterozygous mutations in the syntaxin-binding protein 1 (STXBP1) gene, which encodes a core component of the pre-synaptic membrane-fusion machinery called Munc18-1, has been associated with severe forms of early epileptic encephalopathy in patients with Ohtahara syndrome. How STXBP1 mutations, particularly a partial reduction in Munc18-1 cause human neuronal dysfunction is not known. To analyze the phenotypic effect of STXBP1 mutations in human neurons in a controlled genetic background, Patzke and colleagues produced hetero- and homozygous conditional knockout (cKO) cells by homologous recombination. They infected human H1 ES cells with a recombinant adeno-associated virus (AAV) that contained wild-type (WT) human STXBP1 sequences from the region encoding exon 2 flanked by loxP sites and two different drug selection resistance cassettes that are surrounded by frt sites, and converted them into iN cells by expression of the transcription factor Neurogenin-2 (Ngn2). First, they analyzed the protein levels of Munc18-1 by immunoblotting in hetero- and homozygous cKO cells, finding a lack of Munc18-1 protein in homozygous STXBP1-mutant neurons and a 30% decrease in the heterozygous neurons. In addition, they found a decrease in Syntaxin-1 levels, whereas other synaptic proteins showed no change. These results suggest that Munc18-1 and Syntaxin-1 are subunits of a complex and stabilize each other, as has been described in mice previously. Then, they found that the homozygous STXBP1-mutant neurons died during the first 3 weeks in culture, whereas the heterozygous mutant cells survived at a level similar to control neurons. This result could explain why heterozygous mutations in the STXBP1 gene have been observed in patients, and for this reason, they limited the study to heterozygous STXBP1-mutant neurons. Next, they analyzed the morphology of WT and STXBP1-mutant neurons, but no differences were found in the number and the total length of dendrites or in the size and number of synapses per dendritic segment. Consistently, they did not detect changes in intrinsic electrical properties. However, the STXBP1-mutant neurons showed a decrease in miniature excitatory postsynaptic current (mEPSC) frequency without a change in the amplitude and in action potential-evoked EPSCs. Then, the authors rejected the explanation that the phenotype observed could be due to a decrease in synapse numbers or a decrease in the size of the readily releasable pool (RRP) of vesicles per synapse, as the STXBP1-mutant neurons showed no differences either in the synapse numbers or the size and kinetic release of the RRP after hyper-tonic sucrose stimulation.
To further analyze the phenotype of STXBP1-mutant neurons in single connections, Patzke and colleagues used bacterially encoded opsins (called channelrhodopsin) to enable light-dependent control of neuronal activity in two different approaches. In the first one, they transfected the iNs after 21 days in vitro with a plasmid expressing a fluorescent reporter gene called tdTomato and codon-optimized channel-rhodopsin variant tdTomato-CHiEF either alone or together with a Munc18-1 rescue construct. In this way, they were able to evaluate the response to single-action potentials induced in a small proportion of neurons that establish synapses with nontransfected neurons by a brief optical stimulation and ask whether WT Munc18-1 protein rescues the phenotype. They found that optogenetic stimulation of presynaptic neurons promotes a decrease in EPSC amplitude in STXBP1-mutant neurons compared with control neurons, whereas STXBP1-mutant neurons expressing WT Munc18-1 gene showed a similar level in EPSC amplitude compared with control neurons. In the second approach, ES cells were infected at the same time with a tdTomato-CHiEF lentivirus, along with lentiviruses to induce neuronal differentiation from human ES cells and that generate control WT or heterozygous STXBP1-mutant neurons. With this approach, all of the iN cells express td-Tomato-ChiEF. After 7-days in vitro, the iNs were mixed with mouse newborn primary cortical neurons so that a sparse population of CHiEF-expressing iN cells were cocultured with an abundance of WT mouse neurons. After day 21, light was turned on to stimulate presynaptic iN cells and mouse neurons were patched. This experiment showed the same reduction in EPSC amplitude in STXBP1-mutant neurons found previously. Taken together, both approaches revealed that heterozygous STXBP1-mutant neurons were impaired in presynaptic neurotransmitter release, consistent with the data obtained in spontaneous and evoked neurotransmitter release. These results also suggest that a relatively small decrease in Munc18-1 together with a concordant decrease in Syntaxin-1 levels may lead to a large decrease in neurotransmitter release, thus explaining how a heterozygous mutation might cause a large phenotype.
Altogether, the results presented by Patzke and colleagues show that the heterozygous STXBP1 mutation impairs the synaptic transmission in neurons, which may explain the symptoms of Ohtahara syndrome patients. This study also suggests that synaptic phenotypes associated with STXBP1 mutations may be therapeutic targets. However, there are still many open questions:
Do STXBP1 mutations produce the same phenotype in other human pluripotent cell lines with different genetic backgrounds? Do neurons differentiated from Ohtahara syndrome patient-derived iPSCs show the same block in synaptic transmission? Could STXBP1-mutant neurons be used for drug testing to identify more efficacious drugs that work in Ohtahara syndrome patients?
In addition, we have to keep in mind that the human brain is a complex tridimensional structure, so can a neuronal two-dimensional network in vitro reproduce the brain's features? Recently, the development of three-dimensional cerebral organoids or “mini-brains” in vitro represents a promising approach to study brain development and disease in physiological and pathological conditions (9, 10). Despite these remaining questions, this study provides valuable information on the etiology of Ohtahara syndrome toward the development of future therapeutics.