The 3rd European CDKL5 Research Conference was held at The St Johns Hotel in Solihull, Birmingham on the 2ndand 3rd of June 2016. The conference consisted of 6 sessions over the 2 days
Session 1 – Where we are with CDKL5
After an introduction from Carol-Anne Partridge, Chair of CDKL5 UK, there were short presentations from parents Sarah Skillicorn and Shelley Ashton and sibling Nathan Partridge on living with CDKL5. Sunny Philip, consultant paediatric neurologist at the Birmingham Children’s Hospital, then gave an overview of his practice and of our current understanding of the clinical aspects of the CDKL5 disorder. Finally, Majid Jafar, CDKL5 dad and co-founder of the LouLou Foundation reviewed the work that the Foundation has been supporting in terms of research both in the UK and overseas and highlighted the CDKL5 Forum which interested researchers can access through the Foundation’s website portal.
Session 2 – An update on the biology of CDKL5
Ralph Hector from Glasgow, presented his work on re-characterisation of the CDKL5 gene. Previous studies have shown that CDKL5 is a complex gene, but there are still several gaps in our knowledge. Improvements in molecular diagnosis and more extensive research into CDKL5 disorders necessitated an updated analysis of the CDKL5 gene. Before any protein is produced the gene is first read (transcribed) to produce RNA and this is known as its transcript. There is often more than one transcript (isoform) and in this way one gene can produce slightly different versions of the same protein. This is the same for CDKL5. Using bioinformatics, a computer science and statistics approach, 5 different transcript isoforms for CDKL5 were identified and these findings confirmed using molecular biology tools. The CDKL5 transcript isoforms for a range of tissues were identified and by doing this the major isoform in the brain was identified as a 9.7 kb transcript which was named hCDKL5_1. This study has enabled the description of an updated gene model in humans and in mice as a result of which a standardised nomenclature system for CDKL5 transcripts has been suggested. These findings provide an essential backdrop for the design of gene-based and molecular therapies, and provide an important update for the molecular diagnosis of mutations in the CDKL5 gene.
Charlotte Kilstrup-Nielsen presented work from her laboratory where they are interested in identifying specific molecular networks that are directly controlled by CDKL5 and using this knowledge to identify novel therapeutic strategies to correct defects caused by dysfunctional CDKL5. The morphological defects characterizing neurons with reduced CDKL5 levels can be linked to a general defect in the neuronal cytoskeleton. The cytoskeleton is a network of proteins involved in shaping all kind of cells. The regulation of cytoskeleton dynamics is conserved between different cell types and can easily be studied in proliferating cells. Using this simple and easy-to-handle model system they have identified a novel role of CDKL5 in regulating a specific protein complex involved in controlling cytoskeleton functions. Interestingly, by searching existing literature they found a compound known to act on this pathway and which is currently being tested in clinical trials against autism. To start understanding its therapeutic potential for CDKL5 they treated CDKL5-deficient neurons and analysed its capacity to rescue morphological and molecular defects. Importantly, they found that specific defects in both young and mature neurons can revert to the normal situation. They therefore are going to test the full therapeutic potential of this compound in Cdkl5-null mice.
Mike Cousin presented work being done in Edinburgh. Therapies for CDKL5 can be either primary (direct replacement of the CDKL5 gene or protein) or secondary (pharmaceutical intervention to correct altered cell signalling). The research team at University of Edinburgh focuses on the latter and is integrating a multi-disciplinary approach (examining molecules, cells, circuits and behaviour) to determine the biological role of CDKL5. Central to this are a series of what are called unbiased screens, which examine changes in gene expression, synaptic / circuit function and animal behaviour during altered CDKL5 function. Preliminary studies have indicated that synaptic dysfunction (both presynaptic and postsynaptic) may contribute to altered brain function in CDKL5. In addition to these unbiased screens, the team at Edinburgh will also examine similarities between CDKL5 and other models of epilepsy / intellectual disability. This is because similar alterations in cell signalling pathways have been observed in a number of these models, with drug trails already ongoing. It is hoped that therapeutic strategies which are already established for other neurodevelopmental disorders can be quickly translated for CDKL5 if similar pathways are affected. Central to this approach will be a new rat model of CDKL5 which will allow more subtle and flexible behavioural tasks to be performed and will establish a screening platform for potential therapeutics that target convergent signalling pathways.
Ivan Munoz from Dundee discussed the role of kinases and in particular Serine/Threonine protein kinases which regulate many important processes inside the cell, such as cellular proliferation, differentiation and apoptosis. Up to one third of the human proteins are modified by phosphorylation. The human genome encodes for more than 500 different protein kinases, which are usually kept in an inactive form until activated by certain specific stimuli. Every kinase acts on a specific subset of substrates, which phosphorylates upon stimulus activation in order to respond to both external and internal signals.
CDKL5 is a large Serine/Threonine protein that contains a catalytic domain in its N-terminal half and a large C-terminal side that might play a role in the control of the catalytic activity. Despite several studies showing the involvement of CDKL5 in neuronal development, bona fide in vivo targets (substrates) of CDKL5 remain largely unknown. A few proteins have been suggested to be substrates of CDKL5 in vitro, but whether they are in vivo substrates of the kinase is unclear. In order to identify substrates of CDKL5 the team at Dundee propose a series of unbiased approaches, based on mass spectrometric identification of peptides phosphorylated by CDKL5. Because no stimulus/stimuli that activate CDKL5 have been described yet, they predict that CDKL5 substrate phosphorylation will be scarce. They therefore propose an alternative strategy to identify possible substrates through the generation of both human and mouse CDKL5 knockout cells complemented with wild type CDKL5 or a constitutive hyperactive CDKL5 mutant. Finally, the quality of the available tools to study CDKL5 is, in some cases quite poor, in particular the commercially available polyclonal antibodies. Having a good antibody against CDKL5 is essential to study the function of this protein. A range of different CDKL5 recombinant fragments expressed both in bacteria and in insect cells in order to raise monoclonal antibodies have been tried. They have recently been able to generate a fragment of CDKl5 (1-823) that is expressed reasonably well and they are now planning to develop monoclonal antibodies with it. These protein preps will be sent to Nanotools, a German company specialized in raising monoclonal antibodies, in order to get a good quality antibody for CDKL5.
The team from Dundee would also like to remind all the CDKL5 research community that they are producing a list of all the CDKL5 reagents that they generate, especially cDNA clones and antibodies, that will be available for the whole CDKL5 research community, and that it will be available at http://mrcppureagents.dundee.ac.uk/
Tommaso Pizzorusso presented work on the consequences of CDKL5 mutation at the neuronal level, about which little is still known. Knowledge acquired on neuronal alterations can be used to understand the defective mechanisms and possibly to correct them, and to obtain good biomarkers to monitor the state of neuronal circuits during development or after treatment. They first analysed dendritic spines, the sites at which the majority of synapses, the structures allowing communication between neurons, are located. This was motivated by the lack of neuronal degeneration in CDKL5 disorder suggesting that neurons are there, but they might not communicate well. Using an imaging method (2-photon imaging) that allows repeated imaging of the same neurons, they were able to observe that dendritic spines form normally but are then excessively pruned. This occurs largely during development but also in adults. By applying IGF-1 in developing and adult CDKL5 mice dendritic spines became more resistant to pruning, consequently ameliorating their biochemical alterations. They then tested whether the dendritic spine impairment was accompanied by functional changes in neuronal communication. Using electrophysiological methods they found that neuronal communication is reduced and is not able to adapt to the changes in activity, a property called plasticity important for learning. Finally, to analyse the integrated behaviour of the circuits in the cerebral cortex of CDKL5 mutants they examined visual responses at different developmental stages. Visual responses can be monitored objectively and non-invasively in humans and in mice, and cortical visual deficits have been reported in CDKL5 patients. They found that visual responses are initially normal but fail to develop both in male mice with no CDKL5 and in females with mosaic expression of CDKL5. Many phenotypes at different levels (functional, morphological, and behavioural) are now available for preclinical tests of treatments
David Millar from Cardiff, presented the early results of some research looking at the relation of CDKL5 to what is known as the mTOR/ GSK3B signalling pathways. Firstly, neurons derived from patient iPS cells are a useful model for the CDKL5 disorder and show the same increased GSK3B activity as seen as the mouse model. Secondly, different types of mutation MAY have a different molecular phenotype i.e. both missense and frameshift mutations showed increased GSK3B activity whereas only the frame shift mutation showed reduced mTOR activity. This work needs to be repeated and ideally with further samples. Finally, neurons derived from patient iPS cells show similar deficits to those seen in the mouse models and so may be a useful model system to test re-purposed drugs. This research has been funded by CDKL5 UK.
Claudia Fuchs presented her work, also on the AKT/GSK3B signalling pathways, as possible treatment option for the CDKL5 disorder. CDKL5 is highly expressed in the developing brain, suggesting its importance for correct brain maturation and function. However, very little is known as to the role of CDKL5 in brain development and no therapeutic approaches are available so far for the CDKL5 disorder. In order to shed light on the role of CDKL5 on brain development, they took advantage of the newly generated Cdkl5 knockout (KO) mouse model from the Gross group (Amendola et al., 2014) and found that CDKL5 plays a fundamental role in postnatal hippocampal neurogenesis, by affecting neuronal survival and maturation (Fuchs et al., 2014). Looking at the molecular mechanisms underlying these neurodevelopmental defects, they found increased activity of GSK3B, a crucial inhibitory regulator of many neuronal functions.
More recently, they had found that treatment with the ATP-competitive GSK3B inhibitor, SB216763, recovers hippocampal defects in Cdkl5 KO mice, suggesting that GSK3B inhibitors may be exploited in order to improve the neurodevelopmental alterations due to CDKL5 loss-of-function (Fuchs et al., 2015). However, compounds that inhibit GSK3B kinase activity in an ATP-competitive manner may also affect other kinases, with harmful effects, while ATP-noncompetitive GSK3B inhibitors, such as Tideglusib (or NP-12) are more selective. Moreover, Tideglusib is already approved for use in humans and is being tested in Phase II clinical trials for other pathologies. Based on this evidence they wondered whether pharmacotherapy with Tideglusib could rescue hippocampal neurodevelopmental defects due to Cdkl5 loss. They found that treatment with Tideglusib completely restores neurodevelopmental defects and behavioural abnormalities in the hippocampus of Cdkl5 KO mice. These results suggest that treatment with Tideglusib may represent a therapy for CDKL5 disorder that is potentially translatable into clinical trials. This work was supported by the International Rett Syndrome Foundation and the International Foundation for CDKL5 research (Mentored Training Fellowship to Claudia Fuchs).
Maurizio Giustetto presented a study on the cellular and molecular determinants of cortical defects produced by CDKL5 mutation. They used the visual cortex as a model because CDKL5 mice show severe defects of central visual processing. They tested neuronal activity in this area by studying the expression of the gene cFos (a marker of neuronal activity) and found that cortical circuits are less active in KO mice compared to WT littermates (note: WT = Wild Type which means “normal”). Growing evidence indicates that disruption of excitation/inhibition (E/I) balance in the forebrain may represent one biological signature of neurological diseases. Although the disruption of E/I balance is likely to be the cellular determinant of circuit dysfunction associated with CDKL5 pathology, very little is known about the possible mechanism involved. To shed light on this issue, confocal microscopy and synaptic imaging was used to study the organization of cortical connections in the visual cortex of Mecp2-KO mice. They found that the number of both excitatory and inhibitory synapses was increased.
Next, because CDKL5 is known to be present at synapses, they investigated the molecular organization of the postsynaptic side of neuronal contacts and found that both PSD95 and the scaffold protein Homer were abnormally expressed. Thus, the increased number of excitatory synapses may not correspond to more active contacts but probably the opposite. Their study, indicates for the first time an involvement of the scaffold Homer, a protein implicated in several neurological disorders (e.g. ASDs and Fragile X) in CDKL5 pathology. Because of the elevated expression of CDKL5 in GABAergic interneurons (INs), they also investigated the impact of CDKL5 mutation on the assembly of INs in the cerebral cortex of Cdkl5-KO mice. Their data demonstrate for the first time that loss of CDKL5 leads to significant alterations of INs assembly and in a progressive enhancement of inhibitory connectivity/tone in the V1 cortex, in agreement with our cFos data.
Session 3 – The CDKL5 Associations
Presentations were given by the associations of Slovakia, Netherlands, Germany, Spain, Switzerland, Ireland and Italy.
Session 4 – An update on treatment at the molecular level
Paul Ross from Glasgow discussed his lab’s work in gene therapy for Rett syndrome and also an alternative approach for CDKL5. Rett syndrome and the CDKL5 disorder are both severe X-linked genetic disorders affecting mainly females, with males being much more severely affected. Since both disorders are caused by mutations in a single gene, MECP2 in the case of Rett, they are good candidates for gene replacement therapy. Promisingly, mouse studies have shown that Rett syndrome is reversible in mouse models, suggesting the disease may be treatable, even in adults. Ongoing studies are investigating whether this is also the case for CDKL5 disorder. His lab carried out the first ever gene-therapy study for Rett syndrome and showed that delivery of a normal copy of the MECP2 gene to severe male Rett mouse models, using viral vectors, led to improvement in lifespan and Rett symptoms. A further study by an American lab also showed improvement in female mouse models. However, follow-up studies have shown that the level of MeCP2 protein expressed in mice has to be very tightly controlled. If too little of the gene to the is delivered to mice using the viral vector, Rett symptoms don’t improve, but if too much is delivered, the animals get severe liver toxicity due to overexpression of the protein.
To try and overcome this we have developed a new approach targeting the disease at the level of RNA. Permanent copies of genes are stored as DNA in the cells. When the gene is needed a temporary RNA copy is made and translated into a protein, the molecule that carries out the function of the gene. Using a technique called RNA trans-splicing, we can bypass mutations in this RNA copy in order to make normal MeCP2 protein. Since the amount of protein made will be remain under the normal control of the cell, there is no danger of overexpression toxicity, giving this strategy an important advantage over normal gene replacement therapy. We have identified several promising candidate molecules capable of replacing mutated DNA in cells, and are now testing these molecules in Rett mouse models. We believe that the similarity between Rett and CDKL5 disorder means that this same approach could also be applied to CDKL5. For this reason, CDKL5 UK have funded a year-long pilot project to enable our lab to test trans-splicing therapy for CDKL5 disorder, to be carried out by Ralph Hector (session 2(1)).
Elisabetta Ciani presented her work on Protein Substitution Therapy. No therapies are presently available to improve the neurological phenotypes associated with the CDKL5 disorder. Since mutations in the CDKL5 gene lead to a lack of functional CDKL5 protein, delivery of a functional CDKL5 protein into the brain represents the best therapeutic approach. It has been discovered that certain proteins and peptides exhibit the unique property of efficient translocation across cell membranes. This unique translocation is usually due to the presence of a Protein Transduction Domain (PTD) in these molecules. The HIV-1 Transactivator of Transcription (TAT) protein is the best characterized viral PTD containing protein. Earlier experiments with the TAT-PTD protein domain demonstrated successful transduction of high molecular weight proteins into the mouse brain. Importantly, no toxic effects or immunogenicity problems of the TAT-PTD have been reported so far.
This study provides novel evidence that the innovative approach, named protein substitution therapy, aimed at compensating for the lack of CDKL5 function by targeting a functional recombinant CDKL5 protein into the brain, is feasible. To deliver an active CDKL5 into the nervous system and within brain cells, we constructed a TAT-CDKL5 fusion protein using a modified HIV protein transduction domain TAT as a delivering moiety. We demonstrated that TAT-CDKL5 fusion proteins can be delivered into cells and retain CDKL5 activity after internalization. When administrated in vivo, TAT-CDKL5 fusion protein was able to cross the blood brain barrier and diffuse into the brain. Finally, we treated CDKL5 knockout mice with TAT-CDKL5 protein and showed that neurobiological and neurobehavioral defects underwent an improvement, in several cases bringing brain development and behaviour up to wild-type levels. It is worth noting that all the presented results were obtained in adult mice, indicating that neurodevelopmental defects in the CDKL5 knockout condition may be corrected even after the early stages of brain development. Such promising results strengthen the idea that a protein substitution therapy with TAT-CDKL5 fusion protein may be successfully developed for CDKL5 patients.
Tariq Ali discussed the role of drug repurposing and how this might be used as a therapeutic option for CDKL5. Currently there is a proposed study looking at the role of a drug called ataluren/translarna. This may have a role to play where the genetic mutation is a nonsense mutation. A nonsenese mutation causes a premature “stop” codon in the gene resulting in a truncated or abnormally short protein being produced. Ataluren acts by preventing the “stop” codon from being read so that a normal length protein is produced.
Session 5 – Dealing with the everyday problems
Sunny Philip presented an overview of the disorder which can be seen here.
Patrick Forcelli presented some of his research that focused on firstly, the effect of anti-seizure medications on brain development and secondly, strategies to avoid these negative effects. He showed data indicating that several common anti-seizure drugs trigger cell death (apoptosis) in the baby rat brain; this included abnormal pruning of neurons as well as damage to white matter tracks (the fibres that connect different brain regions). He then showed data indicating that these same drugs impair the development of connections between neurons (synapses) in two important brain areas, the striatum and hippocampus. The damage to synapses was evident both functionally when recording the activity of these neurons and structurally when looking at dendritic spines. These drugs were also associated with changes in behaviour, for example, impaired learning and memory, altered sociability, and decreased behavioural flexibility. He also showed that treatments, like melatonin, can improve the therapeutic effect of phenobarbital, while simultaneously blocking some of the negative effects phenobarbital has on the developing brain. Finally, he presented some data indicating that drugs that act on cannabinoid receptors can suppress seizures in baby rats.
Eveline Hagebeuk presented an update on the sleep studies she has done on children with CDKL5. 90% of children with CDKL5 have sleep disturbance. Using polysomnography (PSG) to study these children, she identified various disturbances of sleep including a delay to sleep onset in 1 child, a relatively low percentage of REM sleep in all and relatively low sleep efficiency caused by frequent and long lasting awakening during the night. Of interest was that although EEG evidence of epileptiform activity was present these disturbances did not appear to be associated with seizures apart from in 1 child. She discussed treatment strategies using a combination of seizure treatment, behavioural strategies and bedtime reduction, occasionally combined with medication, and all tailored to individual needs.
Sam Amin from Bristol Children’s Hospital presented his findings from a small cohort of patients who completed a questionnaire. You can see a video of this here.
Thank you to Safespaces, Tobii Dynavox, Red Goat Productions, Two Viz Thinkers, Footsteps, Newlife Foundation and Livanova and our sponsors the LouLou Foundation, PTC Therapeutics, the National Lottery and St James Place Foundation.