Research Articles

34. Synaptic synthesis, dephosphorylation and degradation: a novel paradigm for an activity dependent neuronal control of CDKL5.
Journal of Biological Chemistry 2015.    Link to text                                                   

This is a laboratory study from Italy looking at mechanisms that regulate or control the levels of the CDKL5 protein in the neuron. Before reading this, it might be worth your while revising some of the basic neurophysiology that I have put on The CDKL5 Protein page, particularly in relation to the “structure of a neuron” and “dendrites and spines”. Much of the research that the authors have produced in this study is in relation to the regulation of the CDKL5 protein in the dendrite.

In this study, the authors have examined how levels of the CDKL5 protein change in relation to stimulation of the neuron. From their results it would appear that CDKL5 is regulated in different ways depending on the maturity of the neuron. In both mature and immature neurons, stimulation produces a rapid rise in CDKL5 levels over about a 5 minute period. However, whereas CDKL5 levels in mature neurons return relatively quickly back to their base level over the next 10 to 30 minutes depending on the maturity of the neuron, levels in immature neurons remain elevated for over an hour or so.

In both cases, the initial rise in CDKL5 levels appears to be due to local production of the protein – by translation of localized mRNA (see Genetics Page). The subsequent fall in CDKL5 levels then occurs due to the protein first being dephosphorylated (removal of a phosphate group) and then broken down by protein splitting enzymes. The stimulus for the change in activity of the CDKL5 protein appears to be a localized depolarisation – where the electrical charge in the neuron flips from negative to positive. Various mechanisms are involved - again depending on the maturity of the neuron - but it appears that the same neuronal stimulation initiates both the initial increase in CDKL5 and its subsequent degradation

Note - This study isn’t telling us what the CDKL5 protein does but rather is telling us something about the way in which it is regulated or controlled. What is apparent is that whatever it is that the CDKL5 protein does, in order for it to function properly its levels have to be tightly regulated. As I have mentioned before in other reviews this may well have implications for certain therapies – particularly protein replacement or even gene therapy. It may not be enough just to try and replace the CDKL5 protein – it may also be necessary to ensure that it is being regulated correctly.

33. Catastrophic Childhood Epilepsy: A Recent Convergence of Basic and Clinical Science.
w   2014.   Link

This is an excellent review of what are termed the “Catastrophic” epilepsies, their causes and the development of future research and therapeutic strategies. The paper is not about CDKL5, but we are in there somewhere. The authors make suggestions about recommendations and proposed future directions.

1: Create new experimental models to understand seizure mechanisms and develop therapies  
2: Apply new technologies for probing distributed neuronal networks
3: Use a top-down approach 
4: Test currently approved drugs for other clinical indications
5: Engage pharma and the biotech industries
6: Expand resource sharing
7: Create data warehouses

I think points 6 and 7 are particularly relevant as they refer to the “sharing” of research resources such as animal models and other tools such as iPS cells, so that any interested research group could utilize resources that have already been developed. They also advocate the development of “data warehouses” which would include common data elements so that individuals from different institutions could add to the database. Patient data would then be widely available to researchers and clinicians – now there’s an idea!!!

Note – The point I personally take away from this is that CDKL5 is but a very small fish in a small pond in a very big world. If we don’t all stick together and pool our resources both in terms of supporting research and the development of therapeutic strategies then we may be inadvertently shooting ourselves (or perhaps our children) in the proverbial foot. I don't personally believe that there is any place for academic or any other territorial self-interests which ultimately can only hinder the important and necessary work that needs to be done in order to progress the understanding of CDKL5 and hopefully the development of future therapies.

This excellent article can now be accessed in full via the link above.

32. Sleep, Plasticity and the Pathophysiology of Neurodevelopmental Disorders: The Potential Roles of Protein Synthesis and Other Cellular Processes.   Brain Sciences 2014.          Link to article                                                   

This is an interesting review from the US about sleep and its relationship to the development of memory and learning. The “laying down” of memory processes can be categorized into three stages - encoding, consolidation and retrieval, and sleep is clearly more than just having a few zzzzz’s after a good night out!  In normal development, sleep appears to be important for the consolidation of various types of memory and therefore can affect the development of learning. Sleep may be particularly important for memory consolidation during the periods of intense learning that occur throughout early development.

The drive for sleep at any point in time is determined by two processes: a circadian process that mainly controls REM sleep and a homeostatic process that is mainly confined to the description of non-REM sleep. There are a number of substances implicated as being sleep regulators, including adenosine, prostaglandins, nitric oxide and cytokines. There are also a number of cellular processes involved and one hypothesis suggests that sleep is required for synaptic pruning – NB) disorders of synaptic pruning have been implicated in neurodevelopmental disorders. 

Having reviewed the normal processes of sleep, the authors discuss the disturbance of sleep in neurodevelopmental disorders, pointing out that recent data has revealed that sleep abnormalities are among the most prevalent and common symptoms and may contribute to the progression of these disorders. The authors review a number of neurodevelopmental disorders including Autism, Fragile X Syndrome, Rett syndrome, Prader-Willi and Angelman’s Syndrome, where sleep disorders have been reported. They then go on to discuss the cellular consequences of normal and prolonged wakefulness and their effects on plasticity of the brain. The authors also point out that sleep has been thought of as a symptom of a wide variety of psychiatric and neurological disorders and that if the “primary” disorder was treated then the sleep disorder would resolve. However, this view is being replaced with the idea that the disordered sleep is not merely a symptom but that it can also play a role in the progression of the primary disorder. As the authors again point out, cell signaling and cortical connections are often altered in neurodevelopmental disorders so that the effects of any “perturbations” on neurons will be unpredictable. The combination of abnormalities of sleep and the altered brain conditions that are present might therefore lead to “synergistic and deleterious” effects. Finally, the authors note that there are very few sleep-dependent memory studies in patients with neurodevelopmental disorders or studies in any subject group that test whether social skills-learning is sleep-dependent.

Note – Theories around sleep and the disturbance of sleep in relation to memory and learning remain contentious as there appears to be conflicting experimental evidence. However, this is a good review around the subject and although CDKL5 isn’t specifically mentioned, we all know how relevant this is to us. A difficult subject to study as there are many forms of memory and learning. The authors point out that more research into sleep disturbance in individuals with neurodevelopmental disorders is needed. Although future therapeutic strategies are discussed, this paper isn’t giving us any practical answers yet into how we might deal with sleep disturbance. It is, however, making a very good case that managing and trying to improve sleep patterns should be a priority. There is quite a lot more technical stuff in the review so certainly worth a read.

31. Neurodevelopmental and neurobehavioral characteristics in males and females with CDKL5 duplications.
 Journal of Human Genetics 2014.  Link to abstract.

This is a report from the US on a group of children who were found to have mutations involving duplications of the CDKL5 gene. This has previously been reported in 1 individual and here, the authors present clinical and genetic details of a further 7 children (4 males and 3 females) and 4 of the mothers. 

Clinical - All of the 7 children had an element of intellectual impairment and developmental delay.There were 3 males and 2 females who had an autism spectrum disorder while the same 3 males and 1 of the females also had an attention deficit hyperactivity disorder. These same 3 males also had an obsessive-compulsive disorder. Speech impairment or language delay was present in all the children while behavioural abnormalities were present in all females and all but 1 of the males. All the children seemed to be mobile although difficulties with various motor skills were present. Interestingly, none of the 7 had epileptic seizures and 3 of the 7 had macrocephaly as opposed to the microcephaly we can see in other CDKL5 mutations. Parental testing showed that the duplications were inherited from the mother in 4 cases and the father in 1 case. Testing was unavailable for the remaining 2 cases. Of the 4 mothers who were carriers, 2 reported having mild learning difficulties while 2 were said to be healthy. 

Genetic - The CDKL5 duplications were associated with duplications of other genes in the region of the CDKL5 gene. Furthermore, 3 of the children were also found to have duplications affecting other chromosomes (chromosomes 9, 16 and 17). The authors did not feel that these other duplications contributed to the observed phenotypes. Blood testing revealed random X-inactivation in all the females.

Note – A very interesting report which is telling us that having too much CDKL5 protein does not seem to be as bad as having too little or none at all. The assumption here is that through the duplication of the CDKL5 gene, the brain is producing twice as much CDKL5 protein as would normally be present. This is something that is seen in other conditions including Rett syndrome. Whilst having too much is not as good as having normal levels, these individuals are clearly functioning relatively well compared to our children with mutations causing little of no CDKL5 protein activity. The fact that 2 of the mothers who were carriers were said to be healthy is also an interesting point in this report. The notion that we might be able cope better with too much than too little protein may have implications for therapeutic options in the future, particularly in relation to replacement therapies

30. Loss of CDKL5 impairs survival and dendritic growth of newborn neurons by altering AKT/GSK-3β signaling.   
Neurobiology of Diseases 2014.   Link to text.                                                                                                                  

This is a paper from the University of Bologna and the European Molecular Biology Laboratory (EMBL), Monterotondo, Italy. The authors describe studies of the role of the Cdkl5 gene in brain development using knockout mice - that is, mice which have had the Cdkl5 gene removed. Their histological studies focused mainly on a part of the brain called the hippocampal dentate gyrus,  because this part of the brain undergoes most of its development after birth which therefore allows for the study of the early development of neurons. The hippocampus also plays a key role in learning and memory.

The development of neurons (nerve cells) in the hippocampus goes through a specific pathway. Cells called primary progenitor cells (PPC’s) produce intermediate progenitor cells (IPCs) that are undifferentiated (ie - not yet committed to becoming nerve cells). Undifferentiated IPCs divide and differentiate (specialize) rapidly to form IPCs that are now destined to become nerve cells. These differentiated IPCs will then generate immature neurons that integrate into the neuronal circuits of the hippocampus as mature cells. Most of the immature neurons are subjected to a “selection process”, during which they are either recruited or eliminated.

In this study, the authors found that loss of the Cdkl5 gene caused an increase in cell death of immature neurons but with no effect on their preceding IPCs. This suggests that CDKL5 plays an important role in the survival of mature neurons. They also found that KO mice had less neurons, and that dendritic arborization was significantly reduced. This all suggests that the CDKL5 gene also has a role in dendritic development and/or stabilization.

In association with these neuro-anatomical defects, the authors noted that Cdkl5 KO mice exhibited impairment in hippocampus-dependent memory. They also point out that as cortical areas of the brain share these same defects this might explain the severe cognitive impairment of individuals with CDKL5 mutations. Finally, the authors propose that the loss of Cdkl5 impairs neurogenesis and dendritic development through disruption of the AKT/GSK-3β signalling pathway. 

Note - again, it has taken me a while to digest this paper and I have omitted a lot of detail from my review. We are getting into the nitty-gritty of some of the molecular biology of the CDKL5 protein and I think it is only going to get more complex as further studies emerge. Some of the results in this study looking at the changes in development and appearances of nerve cells confirm other reports. What is new here is that CDKL5 appears to act through the AKT/GSK-3β signalling pathway - although the signalling factor Akt has already been implicated in the role of CDKL5 (see review 20). I have briefly discussed signalling pathways under “Future Research” at the bottom of The CDKL5 Protein. So, we are still only seeing glimpses of the bigger picture which can make it difficult to put into context the results of individual studies such as this one - but still a great study.

29. GluD1 is a common altered player in neuronal differentiation from both MECP2-mutated and CDKL5-mutated iPS cells.
European Journal of Human Genetics 2014.   
 Link to abstract.                                                              

This is a joint paper from Italy and the US about research into the molecular targets of CDKL5 and MeCP2.  The study arises from the observation that children with a CDKL5 disorder share some characteristics with those who have Rett syndrome. A previous study has shown that CDKL5 and MeCP2 belong to the same molecular pathway. The authors of this study further hypothesise that amongst the genes that both CDKL5 and MeCP2 each influence, some may be common to both, and this might therefore explain some of the similarities in phenotype between the two conditions. The aim of their study therefore, was to identify whether there were any common genes.

The authors studied gene expression in iPS cells derived from 2 patients (1 male and 1 female) who had CDKL5 mutations and 1 patient with a MeCP2 mutation. They identified one particular gene - the GRID1 gene - that was downregulated in both CDKL5 and MeCP2 mutated cells. Following this, the  iPS cells with  CDKL5 (the female) and MeCP2 mutations were then differentiated through Neuronal Precursor Cells (NPC’s) into mature neurons. When the authors then analysed the expression of the GRID1 gene in both the NPC's and the mature neurons for both mutations, they found (apparently to their surprise) that the GRID1 gene was now upregulated or overexpressed in all cells. Their data therefore suggests that the GRID1 gene is dysregulated in some way by both CDKL5 and MeCP2. They go on to suggest a mechanism through which MeCP2 might cause this change in regulation but do not appear to offer a mechanism for CDKL5.

Note - So .... "what does this all mean??" I hear you cry.....! Well, this study provides further evidence that at the molecular level CDKL5 and MeCP2 share some common pathways, and in particular, that they both regulate expression of the GRID1 gene - although we know from the mouse model study below that CDKL5 and MeCP2 have distinct pathways as well. A nice study but a little disappointing that the authors still tended to describe CDKL5 as a variant of Rett syndrome and they didn’t really suggest a mechanism by which CDKL5 also regulates the GRID1 gene. I think this work was presented in Bologna and had I understood it then I might have asked the odd question or two  ... missed opportunities ..  oh well, at least another small piece of the jigsaw. Also, for your information - the GRID1 gene codes for something called a Glutamate delta-1 (GluD1) receptor. GluD1 receptors are expressed throughout the forebrain during development with high levels in the hippocampus in adulthood. A recent article from the US proposed that GluD1 receptors were crucial for normal functioning of synapses and that the absence of GluD1 might lead to specific abnormalities in learning and memory - so there are some correlates here with what we see in the CDKL5 disorder. Oh, and if you don't quite understand what iPS cells are, then have a look at the "Future Research" section at the bottom of The CDKL5 Protein .

28. Mapping pathological phenotypes in a mouse model of CDKL5 Disorder.       
PLOS 2014.   Link to text.

This is a paper from Italy on a Cdkl5 knockout mouse model. The authors present a number of studies on the cellular and behavioural aspects of Cdkl5 deficient mice. Absence of the Cdkl5 protein in the mice was confirmed using 2 standard techniques - western blotting and immunofluorescence.

Cell biology 

Localisation of the Cdkl5 protein - further immunofluorescent studies also showed a number of features. In normal (wild-type) mice, the Cdkl5 protein is found both in the nucleus of neurons and in their cytoplasm. Furthermore, there is a specific type of neuron - called a pyramidal neuron - and although the Cdkl5 protein is present in the cytoplasm of pyramidal neurons of the cortex, it seems to be more prominent in the cytoplasm of pyramidal neurons in a specific area of the brain called the hippocampus -which is located under the cerebral cortex.

Dendrites - which are the branched projections of neurons, are particularly important as Cdkl5 activity seems to focus within these structures. Dendritic arborisation (the complexity of their appearance and therefore function) was found to be significantly reduced in knockout mice, and this was associated with a reduction in the thickness of the cortex and of the hippocampus.

Signalling - cell signalling (also known as signalling pathways) are the many complex processes that control and co-ordinate the different activities of cells. Some signalling factors known to be affected in Rett syndrome were examined in Cdkl5 knockout mice. One such factor, called BDNF, was found to be normal, as were levels of the Mecp2 protein itself. However, another signalling factor, known as “Akt” was found to be reduced along with levels of phosphorylated “ribosomal protein S6”  - also called  “rpS6” - a protein that is a component of  ribosomes, the protein factories that convert RNA into proteins - see The Genetics of CDKL5.

Cdkl5 and Mecp2 - Previous studies have shown a link between Cdkl5 and Mecp2 (Rett) function in the brain. The authors of this study were able to show that they probably also have distinct functions in the brain as well.


Behaviour - a number of behavioural tests were performed on the knockout mice and their responses compared to normal (wild-type) mice. Abnormal clasping was seen in a significant number of the knockout mice whereas no or very low levels were seen in the wild-type mice. Knockout mice showed significantly decreased locomotion in a familiar home-cage environment but normal locomotion in a new unfamiliar environment, suggesting that the capacity for locomotion itself may not be the source of their deficit.  Knockout mice also showed a reduction in head tracking (their responses to visual stimuli) which appeared to be due to deficient visual processing in their visual cortex that is, their eyes are working but their brains can’t fully organise the visual signals - presumably representing cortical visual blindness.

Seizures - None of the knockout mice showed evidence of spontaneous seizures - as with other knockout mice. However, when epileptic producing drugs were given they did show abnormal EEG responses compared to wild-type mice.

Mapping phenotypes - we know that different parts of the brain are involved in different functions (speech, hearing, vision, movement, etc). Different parts of the brain are also characterised by different types of neurons. In the final part of their paper, the authors describe the development of so called “conditional” knockout mice, in which the Cdkl5 gene has been selectively removed from specific types of neurons found in different areas of the brain. The idea here was to see whether some of the characteristics we see in the CDKL5 disorder could individually be mapped to different types of neurons. The researchers found that deficits in behaviour were associated with absent Cdkl5 in one type of neuron in the forebrain, while limb control and eye tracking changes occurred with Cdkl5 deficiency in a different type of neuron of the motor and visual cortical pathways. The authors suggest that therapies aimed at re-expressing CDKL5 in cortical pyramidal neurons may be successful in reversing the most debilitating behavioural phenotypes of the disorder - possibly a plug for PRT !

Overall, the authors concluded that their knockout mouse models exhibits several core features of the CDKL5 disorder and therefore serve as useful animal models of the disorder. The authors go on to discuss the signalling pathways that may be involved and suggest that the down regulation of the “rpS6” pathway may be a common signalling deficit in both the CDKL5 disorder and Rett syndrome, indicating that defective translational regulation, in other words - a problem with the control of protein production -  is a potential core mechanisms for the common pathological features of both disorders. However, their studies also suggest that Cdkl5 and Mecp2 may, in addition, act through different signaling mechanisms to affect common targets.

Note - this is great stuff and covers various aspects which is a little difficult to do justice to.  Large parts of the paper are fairly technical as the authors are obliged to describe the techniques involved in developing their mouse model and their subsequent cellular studies. Some of their results corroborate the results of previous studies and there are also some new findings presented. The work they present on signalling is particularly technical but extremely important. What they are saying here is - right, we know where the Cdkl5 protein is in the brain, now let's see what other pathways are missing as a result of its deficit. I imagine that studying related signalling pathways will be the way a lot of future research into CDKL5 will go. We may also see more mouse models of specific mutations being produced soon - let's hope!! Finally, the study represents the combined efforts of various individuals from a number of centres in Italy and Switzerland. I think it is to their credit that this work has been published as an open-access article.

27. Clinical features and gene mutational spectrum of CDKL5-related diseases in a cohort of Chinese patients.
BMC Medical Genetics 2014.   Link to text.

This study from Peeking University is the first detailed report from China on children with a CDKL5 disorder. Some of the authors were involved in a previous report that included a male with a CDKL5 mutation but no specific details were included then. In this study, 102 individuals who had a diagnosis of either early-onset epileptic encephalopathy or of Rett syndrome were screened for a CDKL5 mutation. There were 71 female and 31 males whose ages ranged from 1 month to 12 years (average 16 months). There were 64 who presented with early onset epileptic encephalopathy (36 with infantile spasms, 8 with Ohtahara syndrome and 20 with an unknown epileptic syndrome), while 38 had a diagnosis of Rett syndrome (16 classical, 10 congenital, 3 preserved speech variant and 9 Hanefeld variety). All of those with early onset epileptic encephalopathy had developed seizures within 6 months of age while none of those with a diagnosis of Rett syndrome had a MeCP2 mutation. 

Mutations were identified in 10 (9.8%) of the 102 who were screened. There were 9 females and 1 male. Of the 9 females, 7 were from the 9 with the Hanefeld variant while the other 2 had previously been diagnosed with an unknown epileptic syndrome. The 1 male identified had infantile spasms. As no CDKL5 mutations were found in parents these were all considered de novo mutations. A variety of mutation types were seen with 5 occurring in that part of the gene coding for the kinase domain while the other 5 affected the C-terminal. Patterns of X-inactivation were fairly random with a range from 50:50 to 63:37.

All affected individuals developed seizures by 4 months of age, and various seizure types presented during the course of their development, including epileptic spasms, partial seizures, myoclonic seizures and tonic seizures. All types of seizure were said to be intractable and resistant to antiepileptic drugs - although 1 patient with a splice mutation affecting the kinase domain was said to be seizure free from 3 years of age. 
EEG studies revealed a hypsarrhythmia in 5 patients. Some Rett-like features had developed in the 9 female patients, such as hand-mouthing, hand washing or clapping. All individuals showed severe psychomotor developmental delay. In addition, hypotonia, poor to absent acquisition of language, limited hand skills, poor eye contact, features of autonomic dysfunction and autistic symptoms were present. There was 1 child who had taken a few steps by the age of 6 months and another who had apparently spoken one word. The male patient in the study is relatively young and his development is still being monitored. MRI, blood and urine amino acid and organic acid investigations were all normal.   

Note - a nice paper and reasonably comprehensive review of 10 patients that endorses the features we all now recognise.  It is interesting that none of those patients who had initially been diagnosed with the classical, congenital or preserved speech variant form of Rett were found to carry a CDKL5 mutation - and none of them had a MeCP2 mutation either. It is not apparent whether mutations affecting other genes - eg FOXG1 - were sought. Conversely, 7 of the 9 with the Hanefeld variant had a CDKL5 mutation. In 1985, Hanefeld described cases of infantile spasms in patients who later developed many characteristics of Rett syndrome - and so he ultimately may have been the first to have reported individuals with a CDKL5 disorder.

26. Gastrointestinal problems in children with a CDKL5 Disorder: a parent-led survey.     
CDKL5 UK  Poster Presentation at the 3rd European Rett Syndrome Conference, Maastricht October 2013.      Link to poster

25. Identification of amphiphysin 1 as an endogenous substrate for CDKL5, a protein kinase associated with X-linked neurodevelopmental disorder.
Archives of Biochemistry and Biophysics 2013.   Link to abstract.                                                     

The molecules on which any enzyme acts are known as its substrates and you will remember that CDKL5 is a kinase that energises other molecules through a process called phosphorylation. Identification of substrates that are phosphorylated by CDKL5 will help in the understanding of its role and function. Previous studies have so far identified 2 substrates, MeCP2 and DNA methyltransferase 1 (Dnmt1). In this paper from Japan, the authors describe a new technique with which they were able to identify a further substrate - amphiphysin 1 (amph 1) - a protein coded by a gene on chromosome 7 which is associated with the cytoplasmic surface of synaptic vesicles and is thought to have important roles in neuronal transmission and development.

The bulk of the paper describes specific details of the various stages of the technique which are fairly technical (for me anyway!). However, the bottom line is that amph 1 appears to be a much stronger substrate than either MeCP2 or Dnmt1, and by this, the authors mean and show that amphi 1 is strongly phosphorylated by CDKL5. Previous studies have shown that CDKL5 is localized not only in the nucleus but also in the cytoplasm, especially in the early developmental stages. Amph1 is also known to be present in the cytoplasm in significant amounts and has an important role in neuronal function.  It is therefore suggested that phosphorylation of Amph1 in the cytoplasm by CDKL5 plays a crucial role during the early stage of neuronal development. 

In order to further show that amphi 1 may be a substrate of CDKL5, the authors created 3 “mutation” proteins. Each protein had a point (missense) mutation that has been described in the literature in children with a CDKL5 disorder. The mutations were within the kinase domain of the protein and each of these 3 “mutated” proteins failed to show any phosphorylating activity towards amphi 1. The authors therefore conclude that Amph1 is a potential endogenous substrate for CDKL5 and that it is potentially a critical molecular component of the pathogenesis underlying CDKL5-related neurodevelopmental disorders.

Note - more technical stuff!!  The term “endogenous” is important here as enzymes (proteins) can show far wider activity in the laboratory (in vitro) than they actually have in the body (in vivo). We are clearly more interested in endogenous substrates and it may therefore be necessary for this study (and other in vitro studies) to be corroborated in vivo by perhaps, using knockout or knockin mouse technology.

24. Mutations in the C-terminus of CDKL5: proceed with caution.                                                     
European Journal of Human Genetics 2013.    Link to abstract.

This study from France looks at mutations affecting the last 3 exons (19 - 21) of the CDKL5 gene.   The authors investigated 30 female patients with a clinically heterogeneous phenotype, ranging from nonspecific mental retardation to a severe neonatal encephalopathy. They were all screened for CDKL5 mutations which 2 were found to have. The first was a 2 year old girl who had severe encephalopathy with severe developmental delay, and seizures with apneas that started at the age of 3 months. She had a previously unreported mutation affecting exon 20. The second was an  11 year old female who presented with moderate developmental delay,  intellectual disability, speech delay and a number of other features. She was found to have a mutation in exon 21 that had previously been reported.

In each case parental screening identified that the asymptomatic father of each female also carried the same mutation.  The 11 year old was subsequently found to have a deletion involving the SOX5 gene on chromosome 12. The authors also review the literature where mutations affecting exons 19 - 21 have been reported. Their interpretation of these cases lead them to discuss the possibility that mutations in this region of the CDKL5 gene may not be pathogenic. The authors therefore suggest that screening for mutations in exons 19 - 21, and specifically after codon 938, may not be useful in establishing a diagnosis of atypical Rett syndrome.

Note - a nice little study validating the previous paper from Williamson (below) suggesting that the main CDKL5 protein in the human brain is translated from exons 2 - 18 and part of intron 18. I just wish they had said a CDKL5 disorder instead of atypical Rett syndrome!

23. Palmitoylation-dependent CDKL5-PSD-95 interaction regulates synaptic targeting of CDKL5 and dendritic spine development.       PNAS 2013.     Link to text.

This is a nice laboratory-based study from China looking at how the CDKL5 protein localises to its target area within developing neurons. Post-synaptic density protein-95 (PSD-95) has an essential role in the development and function of synapses where it is seen to aggregate in little clusters. Palmitoylation is a process whereby a chain of fatty acids is reversibly linked to a protein, in this case the PSD-95 protein. This then facilitates the localisation of PSD-95 to the cell membrane - which is again essential to synapse development. 

This study has brought all these factors together to show that the CDKL5 protein links with PSD-95 through an interaction that is dependent on palmitoylation. The distribution of CDKL5 is shown to correlate very closely with that of PSD-95 - suggesting a strong link between the two. It appears that the C-terminal of the CDKL5 protein interacts with the PSD-95-palmitoylated complex and this then allows the CDKL5 protein to get to the right place in the neuron where its kinase activity is required for dendritic and spine development.  The authors also go on to show that interference with the CDKL5-PSD-95 interaction inhibits dendritic formation and growth - thereby establishing the importance of this interaction in delivering the CDKL5 protein to the right part of the neuron.

Note - This study is telling us that the PSD-95-palmitoylation complex is responsible for targeting the CDKL5 protein to the right place in the neuron - a very detailed study as you might expect so I have just tried to present the key points - ie the bits I understand!! These findings also suggests a way that mutations of that part of the CDKL5 gene coding for the C-terminal might consequently affect the function of the CDKL5 protein. Mutations affecting the C-terminal can interfere with the CDKL5-PSD-95-palmitoylated interaction, and therefore, even if the CDKL5 protein has normal kinase activity, its kinase action will be ineffective because it is just in the wrong place.  Good study.

22. CDKL5 and ARX mutations in males with early-onset epilepsy.        
Pediatric Neurology 2013.      Link to abstract.    

This is a retrospective review from the University of Chicago of males who were referred for mutation analysis for CDKL5 and ARX. The clinical, molecular and neuroimaging findings were discussed for the two groups. For the CDKL5 group, there were 266 boys referred for analysis, of which 8 were found to have mutations. 

Clinical - The boys with CDKL5 were aged 2 months to 14 years at the time of evaluation. All 8 experienced seizures with at least 6 displaying the characteristic 3 stages of epilepsy. All had profound developmental delay with minimal or no language and motor skills, severe tone abnormalities and cortical visual impairment.  Many also had GI problems with gastroesophageal reflux and at least 7 required g-tube placement. The authors also noted that one boy had a maternal half-sister with the same missense mutation although her epilepsy was milder. 

Molecular – There were 4 nonsense and 2 missense mutations identified. The remaining 2 were deletions, of exon 3 and a larger deletion of exons 10 – 15. Maternal testing revealed that 6 mutations were de novo. The mother of the boy with a maternal half-sister with the same missense mutation did not herself have the mutation, suggesting the possibility of germline mosaicism.

Neuroimaging – Early brain imaging was normal, but later imaging showed progressive changes including cerebral and cerebellar atrophy in several boys. Some changes were evident as early as 6 weeks of age in one boy.

In their discussion, the authors compare the results of this review with those other accounts of CDKL5 in males. They state that males with CDKL5 mutations do appear to be more severely affected than females. This review includes the second case where germline mosaicism may have played a role in the transmission of a CDKL5 mutation and the authors therefore emphasize the need for genetic counselling.

Note – The ARX gene is also located on the X-chromosome and is positioned relatively close to the CDKL5 gene. Mutations in this gene are also a cause of early-onset epilepsy and severe neurodevelopmental delay. This study presents considerable corroborative information about CDKL5 in boys. The role of germline mosaicism in transmitting a mutation arises again. One interesting comment the authors make is in relation to the term “epileptic encephalopathy” which perhaps most of us have heard in relation to our own child. The authors suggest that “developmental epilepsy” is more appropriate as the neurodevelopmental delay we see is more likely to be related to the underlying cause of the epilepsy than to the epilepsy itself - an issue that arises in the knockout mouse study below. The results of neuroimaging might be something that makes us all think uh-oh..! and there is another published report in relation to this aspect for which I have not yet produced a review. One of the tables in this paper also clearly reports that one of the subjects in the study is deceased. I may have missed it but I can't find a reference  to this in the text -  I have made an enquiry to the author about this.

21. CDKL5 Regulates Flagellar Length and Localises to the Base of the Flagella in Chlamydomonas.                                               
Molecular Biology of the Cell 2013.     
Link to text.

This is a study from the Department of Plant Biology at the University of Minnesota, on a type of green algae called Chlamydomonas  - stick with me as it does get interesting.. honestly!!  Chlamydomonas is a unicellular organism that is characterised by having flagella (hair-like structures) which have both a sensory role and a role in locomotion. The length of the flagella is controlled by a set of genes, and mutations in 4 particular genes (LF1-4) cause flagella to grow up to three times their normal length. In this study the authors describe a fifth gene (LF5) that encodes a protein kinase with a sequence that is very similar to CDKL5. The associated protein appears to localise towards the base of a growing flagella, and its localisation is affected by mutations in genes LF1-3. The authors suggest that the LF5 protein may regulate the entry of proteins into the growing flagella including proteins that control length. Based on studies of the distribution of flagella lengths that occur with mutations, they also suggest that the LF5 gene may be involved in setting the length of the flagella rather than enforcing length control. 

Having set the scene with their basic research, the authors then go on to review the role of CDKL5 in humans. In particular, they discuss the interaction of CDKL5 with Rac1 (see The CDKL5 Protein) and point out that inhibition of Rac1 has been shown experimentally to alter the localisation of basal bodies, which are the foundations of structures called cilia, which in turn are structurally similar to flagella....are you still with me?   So....given this, and the results of their study, the authors suggest the possibility of a link between cilia and CDKL5There are indeed a group of conditions that are known as ciliopathies which are genetic disorders of cilia. It has been suggested that a particular form of epilepsy, juvenile myoclonic epilepsy (Janz syndrome) might also be a ciliopathy, although this remains unproven. The authors conclude, however, by suggesting that CDKL5 might be a ciliary protein and that what we now call a CDKL5 disorder might eventually also turn out to be a ciliopathy.

Note - when I started to read this, I thought I was going to be reading about plants.... botany...or perhaps even gardening! However, it actually turned out to be a fascinating read, with suggestions that, while a little speculative, nevertheless open up a whole new potential avenue of research which may ultimately influence our thought processes into the role of CDKL5. The one thing to bear in mind here is that the coding sequence of the CDKL5 gene is said to be highly preserved, appearing in many other species. The point about this, is that it implies that the functional part of the CDKL5 protein, the kinase domain at least, has survived through evolution and may therefore have a fundamentally important role necessary for basic normal function in life.  Therefore, although the Chlamydomonas may seem a long way away from us humans in the evolutionary tree, the role of a "CDKL5-like" protein in this relatively primitive organism may actually have more relevance to us that we might otherwise give credit for.

20. Novel Mutations in Cyclin-Dependent Kinase-Like 5 (CDKL5) Gene in Indian Cases of Rett Syndrome.                   
Neuromolecular Medicine 2013.   
Link to abstract.

This is the first report from India on the identification of mutations in the CDKL5 gene in children who had presented with atypical Rett syndrome. There were 44 cases presenting with features of atypical Rett syndrome. On screening, 5 children were found to have CDKL5 mutations with one child having 2 mutations. All the mutations were substitutions, 3 were novel while 3 were known. Details of these children are presented. Various clinical features are described including delayed milestones, hand stereotypies, hand wringing, clapping and mouthing. Only 4 of the children had seizures and in one child seizures did not start until 8 months of age. One child is described as having a severe degree of mental retardation while another showed a milder phenotype but with global developmental delay. This latter child was also the one without seizures and both children had mutations in exon 21. There was a third child with a mutation in exon 21, who also had a mutation in exon 12, which is described as being silent. The last 2 children had mutations affecting exons 15 and 16. The child with a mutation of exon 15 developed severe gastro-oesophageal reflux and also had changes on an MRI scan of her brain. The authors use an interesting on-line tool, PolyPhen-2, that predicts the effects of a substitution on the structure and function of a human protein. They also assessed the degree of X-inactivation but found no relation between this and the severity of the phenotype.

Note -this is a small study and some of the clinical detail described is a little vague and not presented in a way that allows for easy or adequate comparison with known data from other studies. It is also worth noting that mutations beyond intron 18 may not be clinically significant if you consider the paper by Sarah Williamson et al (review 8) below. However, having said all that, this paper can only help improve the recognition of the CDKL5 disorder around the world, and that can only be good for those children affected.

19. Neonatal Exposure to Antiepileptic Drugs Disrupts Striatal Synaptic Development
Annals of Neurology 2012.     Link to text.                                                                                                             

This is a very interesting piece of research from Georgetown University in Washington, DC, on the effects of exposure to anti-epileptic drugs (AED) during critical periods of brain development. This is a laboratory-based study on rats. The authors initially discuss how the developing brain is highly vulnerable to modifications of the molecular environment of neurons and that even transient interventions during sensitive developmental periods can have long-lasting functional consequences. In relation to AED’s, they note that phenobarbital (PB) exposure during gestation or early infancy has been associated with reduced intelligence quotient and decreased regional brain volumes in humans.  There is also some evidence that exposure to PB might contribute to the increased risk for neuropsychiatric disorders associated with early life seizures. Unfortunately, the interpretation of results from clinical studies is limited because of the difficulty in distinguishing the long-term effects of drug treatment from the effects associated with the underlying condition. However, despite this and other evidence, PB still remains the drug of choice in the first-line treatment of early-onset epilepsy.

In this study, the researchers look at striatal medium spiny neurons (MSN) which are a specific type of neuron found in a part of the brain called the Corpus Striatum. They play a key role in initiating and controlling movements of the body, limbs, and eyes. The authors studied the effects of PB and 2 other AED’s (phenytoin and lamotrigine) on the development of MSN’s in rats that were exposed to the drugs very early in life corresponding to their neonatal period. They looked at the electrical and histological effects, as well as some clinical effects.

Electrical studies – their results provide the first evidence that neonatal exposure to AED’s not only stunts the development of synapse function between neurons but may also produce a neurotoxic effect on surviving neurons which may in turn affect their function.
Histological studies – these showed that the appearance of dendritic spines were subtly altered in the brains of rates treated with PB. Spine width was slightly reduced and there were also relatively more dendritic filodopia – immature spines - which may represent failure to establish synapse formation.
Clinical studies – rats were exposed to what is called a “reversal learning task” a validated test of cognitive function and behaviour. Those rats treated with PB showed impaired reversal learning compared to control rats.

The authors go on to propose a potential mechanism of action for the detrimental effects observed with these AED’s which is distinct from their anti-epileptic activity. The 3 AED’s studied all have this property whereas other AED’s, such as Levetiracetam (Keppra) do not. The authors therefore conclude that AED’s used for the first line treatment of seizures in the newborn should be selected more carefully.

Note – a very interesting study for us personally as Ellie was put on relatively high doses of phenobarbitone at the age of 5 weeks on the basis that it was a “safe” drug. When we arrived in Australia her blood levels were checked and the local lab in Adelaide contacted us in a bit of a panic as her levels were so high. The idea that some AED’s used as first-line treatment in the first few months of life may actually be causing harm is something that clearly needs exploring further. The authors here make the point that although their study focused on neurons in one specific part of the brain, the AED’s studied have effects across many other parts of the brain and so there is potential for more widespread changes. For a genetic condition like CDKL5, we would obviously need further studies to differentiate the effects of the drugs from the effects of the underlying mutation. Further research beckons…..

18. Treatment of Neurodevelopmental Disorders in Adulthood.
The Journal of Neuroscience 2012.                    Link to article

This is an excellent review of a concept that appears to be growing in popularity. Neuronal plasticity is defined as “the capacity of neurons and neural circuits in the brain to change structurally and functionally in response to experience”. During our development, it was thought that there was a “critical period of plasticity” after which, the brain became less responsive (eg. it is easier to learn a foreign language as a child than as an adult). However, studies looking at the visual cortex in mouse models have shown that the state of plasticity observed in the “critical period” can be reactivated in adulthood through drug treatment and specific environmental manipulation - also called environmental enrichment (EE). Combine this with appropriate rehabilitation and it is then possible to demonstrate recovery of function to a certain extent. 

The authors review some of the evidence that suggests that the phenotypes associated with neurodevelopmental disorders might be improved, even in adulthood, with these approaches. They also focus on 2 particular conditions, Fragile X Syndrome and Neurofibromatosis type 1, both of which are genetically inherited conditions. They review experimental data on adult mouse models that have demonstrated the reversal of some of the features of the underlying condition. They conclude by saying that this line of research may eventually lead to the development of treatments that might alleviate or cure symptoms and disabilities that occur secondary to neurodevelopmental disorders.

Note - Although CDKL5 is NOT specifically mentioned in this review, it is still pretty exciting stuff!  Clearly there is a long way to go with this line of research, and as always, the big question will be can results from animal studies be translated to humans? If it does become possible to induce a state of plasticity such that a ”learning programme” can be instituted to improve cognitive function, then, this would go a long way to helping many adults and children  with conditions like CDKL5.

17. Proximal large bowel volvulus in children: 6 new cases and review of the literature. 
Journal of Pediatric Surgery  2012.                    Link to abstract

This is a review from 3 centres in the UK, Ireland and Hungary of an extremely rare condition that affects the bowel. A volvulus is a condition where the bowel twists on its mesentry (a fold of tissue that attaches it to the wall of the abdomen). This can then cause obstruction of the bowel and can even result in irreversible damage to the blood supply of the bowel. If diagnosed early enough it is possible to treat the condition without surgery although most patients do require some sort of surgical procedure. The authors review 36 cases that have previously been reported and an additional 6 cases of their own. Males and females were equally affected but 29 (69%) were described as having neurodevelopmental delay. The mean age of the group was 10 years (range 0 to 18 years). The presenting symptoms were colicky abdominal distension (100%), abdominal pain (98%), vomiting (83%) and constipation (69%). All but one of the patients with constipation had neurodevelopmental delay. All 42 patients required surgery and there were 6 (14%) deaths. The authors also noted that neurodevelopmental delay and severe chronic constipation were the commonest associated disorders. They suggest that in mentally impaired children, insufficiently treated chronic constipation results in a heavy and extremely dilated bowel that stretches the bowel mesentry and allows a volvulus to occur. They also point out that although children with a volvulus present with pain, distension and vomiting, these symptoms also frequently occur in children with neurodevelopmental delay and constipation without volvulus, and should respond to enemas and laxatives. The rarity of volvulus and the inability of a child with neurodevelopmental delay to adequately communicate their symptoms, may explain why there may be a delay in establishing the diagnosis.

Note - The first thing to say is that this is an extremely rare condition in children. The authors found only 36 reported cases in the English literature from 1965 to 2010. Although any child can be affected by this condition, it would appear that children with neurodevelopmental delay are more prone, presumably due at least in part to problems with communication - an issue which I am sure we all recognise. They do not, however, specify the nature of the neurodevelopmental delay present in the children of this study. Many of us will be only too familiar with the problems around bowel function and constipation - see Living with CDKL5,  so as I stated there, it is very important that parents and carers keep a close eye on this issue, and in my view, do not delay in involving your family doctor / specialist physician if you have a concern. Also - see article below regarding GI disorders....


16. Validation of high-resolution DNA melting analysis for mutation scanning of the CDKL5 gene: 
Identification of novel mutations. 
Gene  2012.                    Link to abstract

This study from France describes the validation of a technique called high-resolution melting analysis (HRMA) used to detect genetic mutations. HRMA was developed in the US by the University of Utah in conjunction with Idaho Technology, and introduced in 2003. The authors point out that hitherto, identification of mutations in CDKL5 has been time consuming, laborious and expensive. Their study reports the validation of HRMA and the results of a comparison with another technique called denaturing high performance liquid chromatography (dHPLC). Their results showed that point mutations and small insertions and deletions can be reliably detected by HRMA, and compared to dHPLC, HRMA profiles are more discriminating. They conclude for mutation screening, HRMA appears cost-effective, easy to set up, highly sensitive, non-toxic and rapid. The technique is ideally suited for large genes with mutations located along the whole coding sequence, such as occurs with the CDKL5 gene.

Note - I read this with interest because Ellie, like many other children who were eventually diagnosed with a CDKL5 disorder, initially had a false-negative result. As technology advances we would hope that testing for all mutations becomes more sensitive and reliable, quicker and cheaper.

15. Partial rescue of Rett syndrome by ω -3 polyunsaturated fatty acids (PUFAs) oil.                                                                                              
Genes and Nutrition  2012.    
Link to text. 

This is a small pilot clinical study from Italy looking at the role of anti-oxidants in the treatment of Rett syndrome (RTT). In using oxygen during energy production, the human body produces by-products known as free-radicals. These can damage cellular components, and have been implicated in many disease processes and conditions, including the ageing process, the result of which has ignited an anti-ageing anti-oxidant explosion - yes eat plenty of tomatoes and you too could look like me! ...  Ok ... a build up of free-radicals can occur because of an imbalance between their production and the detoxifying mechanisms normally present to deal with them. This imbalance produces what is known as oxidative stress (OS) and the authors of this paper state that an enhanced state of OS has been identified in Rett syndrome. They therefore investigated the use of ω-3 polyunsaturated fatty acids (PUFAs), which are anti-oxidants, in children with Rett syndrome.

They recruited 20 children into the study who had all recently been diagnosed with stage 1 RTT. Half the children were treated for 6 months with ω-3 PUFAs in the form of a fish oil, whilst the other half acted as a control group. The authors used the Rett clinical severity score (CSS) to look at clinical changes, and blood-markers of OS as a secondary outcome.

Clinical - Improvements were noted in various aspects of the CSS in the treated group compared to the control group. Significant improvements were observed for motor/independent sitting, ambulation, hands use, nonverbal communication, and respiratory dysfunction, while trends (improvements that were not statistically significant) were seen for language. There was no difference in seizure activity, autonomic symptoms or in the onset of hand stereotypies

Secondary - The authors also noted significant improvements in 5 of the 6 blood markers of OS in the treated group compared to the control group.   Furthermore, the levels of certain markers returned to the values found in gender and age matched healthy controls. 

The authors discuss the idea that RTT as a pure neuronal disease has been recently challenged. Studies have implicated the involvement of glial cells in the pathogenesis of MeCP2 deficiency, and this appears to be corroborated by the present study. The increase in OS markers previously identified in RTT, are the oxidation products of a specific component of myelin which is also present in several organs and tissues. Some neurological signs in stage 1 RTT, overlap with those of an X-linked condition called adreno-leukodystrophy (X-ALD), a rare inherited disorder mainly affecting the brain’s white matter in males, and leading to progressive brain damage. There is some evidence that OS related damage has been reported in X-ALD. The authors also discuss that the one blood marker that did not improve is considered an index of generalized systemic lipid peroxidation (a process that can lead to cell damage due to the free radicals) and it is possible that subjects in this study were not treated long enough to establish a significant change for this particular marker.

Note - The way I read and interpreted this  study was that although Rett syndrome is primarily a neurological condition, there may also be secondary effects contributing to the phenotype. This is on the basis that the associated direct biological impairment of a MeCP2 mutation is associated with OS and related free-radical damage. Treatment with ω-3 PUFAs may therefore go some way to reducing these secondary effects but presumably would not affect the primary underlying neurological condition. What may not be clear, however, is how much of a child's phenotype is related to the direct consequences of the underlying genetic disorder as opposed to the secondary effects, including the effects of OS. This is obviously still a very exciting piece of research for families with children who have Rett syndrome.  Whether this can translate to more definitive treatment for them or act as a starting point for the development of possible therapies for children with CDKL5 disorders clearly remains to be seen. One apparent point would seem to be that better outcomes will depend on instigating treatment as early as possible, and I suspect that a much larger trial is now on the cards.

14. Loss of CDKL5 disrupts kinome profile and event-related potentials leading to autistic-type phenotypes in mice.  
PNAS 2012.    Link to text.

This is the first CDKL5 study to be published based on research using knockout mice and was supported by charitable funds from the IFCR. The authors describe the generation of mice with a deletion of exon 6, which causes the production of a shortened protein truncated within the kinase domain. In order to remove the potential effects of X-inactivation, male mice were studied and compared to those with a normal Cdkl5 gene, termed wild-type (WT) mice. The results of the study were broadly presented in 3 domains, clinical, EEG analysis, and molecular biology.

Clinical - using standard tests of locomotion, sociability, and fear conditioning, Cdkl5 mice demonstrated hyperactivity, impaired motor control and decreased anxiety compared to WT mice. They also demonstrated impaired social behaviour, such as social interaction and nesting, and impairment of learning and memory. Many of these features are comparable to those observed in individuals with autism spectrum disorders and in those with atypical Rett syndrome.

EEG analysis - despite the prevalence of seizures in children with a CDKL5 disorder, no spontaneous epileptic activity was recorded in Cdkl5 mice compared to the WT controls. There were 2 theories put forward for this. Firstly, although there are huge similarities between the CDKL5 protein in humans and that in mice, CDKL5 in humans may have an additional distinct function that is absent in mice and therefore renders humans more susceptible to epilepsy. Secondly, the particular strain of mice used in this study, are known to be more resistant to seizure activity…oh !  However, there was a positive side to this... yes read on. When studying the function of neurons and neural circuits, if a deficit or malfunction is found, it can be difficult to determine whether this is primarily due to the genetic defect or secondary to the effects of any associated seizure activity. The authors of this study did observe disruption in neural circuit communication and in a particular activity of neurons, known as low-frequency event-related oscillations. The fact that seizure activity was absent in these mice suggests that these abnormalities are directly due to the Cdkl5 mutation.

Molecular biology - (the complicated bit) the molecules on which an enzyme acts are known as its substrates. The substrates of CDKL5 are not known. CDKL5 is a kinase and the set of all kinases in the body is known as the kinome. The authors were able to look at a range of potential CDKL5 substrates by studying changes in the kinome profile of the mice. A number of potential substrates were identified, including the kinases AMPK, PKA and AKT. The eventual conclusion of the authors is that loss of CDKL5 function causes disruption of multiple signal transduction pathways - which are effectively the controlling networks of cell function.

Note - this is clearly an important study in the evolution of our knowledge of CDKL5 disorders. Although the absence of seizures is a little disconcerting, we see such variation in the phenotypes of our own children that we shouldn't be too surprised if mice themselves turn out to be different in the way they are affected. I think the most exciting part of this study is in relation to the studies on the molecular biology, where our current knowledge is so sparse. Now, direct effects of mutations can be studied in more detail. Ultimately, this may be of  far more value, in terms of relating to humans that the clinical studies, and the results of this study should establish starting points for further research to give us a better understanding of the exact role of CDKL5 in brain physiology.

13. Respiratory and sleep disorders in female children with atypical Rett syndrome caused by mutations in the CDKL5 gene. Developmental Medicine & Child Neurology 2012.   Link to text.                                                     

This is a small study from the Netherlands looking at sleep and respiratory disturbances in 4 females with CDKL5 mutations. Their ages ranged from 2 to 15 years. The 3 younger children (ages 2, 4 and 5) had mutations affecting exons 5 and 9 and had severe phenotypes. The 15 year old had a mutation affecting exon 18 and a milder phenotype, being able to walk and having less severe seizure activity. The study involved the use of a questionnaire (Sleep Disturbance Scale for Children) and Polysomnography. REM sleep is thought to be the most important phase of our sleep cycle. In young children about 50% of sleep is spent in the REM phase while this drops to about 20% in adults. The 3 youngest in this study all spent less than 20% in REM with one recorded as having no REM sleep. Sleep efficiency - the percentage of time in bed spent asleep - was reduced in all 4 females - mainly due to frequent and long-lasting awakenings. Apnoea is the term for cessation of breathing, and it has a number of causes. Respiratory centres in the brain are responsible for controlling and driving the effort to breath, and central apnoea occurs when these control centres don’t function properly. Two of the females, the youngest and oldest, had central apnoea, which occurred during the awake state. The authors conclude by suggesting that Polysomnography  should  more-or-less be mandatory for the assessment of sleep and respiratory disturbances in children with a CDKL5 disorder.

Note - Whilst we recognise the pattern of sleep disturbances described in this study (yawn...don't we just...), we have not seen respiratory disturbances in Ellie, although I think others have in their children. As the authors themselves admit, this is a very small study which also has incomplete data. It is therefore difficult to draw any wider conclusions.  It does, however, establish some baseline data for children with CDKL5 mutations that may be of use in later studies.  Just one other thing - "atypical Rett syndrome" caused by CDKL5 mutations is now called the CDKL5 disorder.

12. CDKL5, a novel MYCN-repressed gene, blocks cell cycle and promotes differentiation of neuronal cells.  
Biochimica et Biophysica Acta 2012.   Link to text.

This paper from Italy reports on cell culture studies looking at the role of CDKL5 in neuron development, and in relation to another gene called MYCN. The MYCN gene is known to have an important role in the embryonic development of various tissues, including the brain. The MYCN protein regulates the activity of other genes through a “transcriptional” action, where it binds to specific regions of DNA. This study suggests that the expression of the CDKL5 gene may also be regulated by MYCN.  

Cell proliferation is defined as the increase in numbers of cells as a result of cell growth and division, while cell differentiation is the process whereby a cell develops into a more specialised type of cell such as a nerve cell. Cells will typically go through a period of proliferation and then differentiation. The authors observed that cells in which the CDKL5 protein was expressed, developed longer neuritic processes compared to control cells without CDKL5 - that is, showed better differentiation.  Furthermore,  the proportion of CDKL5 cells present in the resting phase of the cell cycle, when there is no proliferation, was significantly greater compared to control cells without CDKL5These results therefore, not only support previous studies that have suggested an important role of CDKL5 in the differentiation of neurons, but also suggest a role for CDKL5 in the control of neuron proliferation - possibly through blocking cells in the "rest phase" of the cell cycle.

The authors then investigated the relation between the expression of CDKL5 and MYCN, and found an inverse relationship between the two. That is, an over expression of MYCN reduced the expression of CDKL5, while a decrease in MYCN expression led to an increase in CDKL5. It was noted that MYCN has a role in cell expansion through proliferation, a function that would fit with its inverse relation to CDKL5 levels. Furthermore, the authors suggest that the MYCN-CDKL5 axis might be particularly important in the development of the cerebellum in the early post-natal period through regulating a shift from proliferation to differentiation of cerebellar granule cell precursors - these are the small cells destined to make up large parts of the cerebellum. The cerebellum is that part of the brain involved in motor control, balance and co-ordination.

Note - This again is quite a technical paper, and took me a while before I felt I'd done it any sort of justice. One theme that seems to be emerging is the possible role of CDKL5 in the development of tubular structures. The neurites described in this study are long tubular structures as are axons, dendrites and even the spines of dendrites. I would refer you to the article above on the study of flagella to see how the CDKL5 protein might ultimately have a fundamental role in the development of tubular structures across species and hence throughout evolution.

11. The CDKL5 disorder is an independent clinical entity associated with early-onset encephalopathy.                                             
European Journal of Human Genetics  2012.   
Link to text.

This is a study from Australia, which reports on the phenotypes of 86 individuals with a pathogenic or potentially pathogenic CDKL5 mutation. These individuals are located in Australia, Europe and the USA. The clinical aspects studied included seizures and motor function, hand function and speech, Rett-like features and co-morbidities, and dysmorphism. New criteria for the diagnosis of atypical Rett syndrome, including the early-onset seizure variant (ESV RTT) have recently been published and this study also looks the fulfilment of these in those studied.

There were 86 individuals studied with a variety of mutations. There were 77 females with an average age of 6.1 years (range 6 months to 22.4 years) and 9 males, average age 5.2 years (range 1.1 to 14.9 years). The main findings were that early-onset epilepsy, severe developmental delay and severely impaired gross motor function are key features of CDKL5. In this study, 90% of children with a CDKL5 disorder developed their seizures by 3 months of age. At the time of the study, 52 females were having daily seizures, 5 were having weekly seizures and 10 were having monthly seizures. Severe gross motor development was reported in all but 3 females with just over half able to sit. There were 8 females who had achieved independent walking and 9 who could walk with assistance. Males tended to be more severely affected, all had seizures and none could walk. There was a shared physical resemblance between affected females, and an overall spectrum of similar features in males and females. The majority of patients in this study did not meet the new criteria for ESV RTT. This was mainly because of the absence of regression in all males and 67.5% of females. A comparison between females with a CDKL5 disorder and Rett syndrome, identified that females with a CDKL5 disorder were more likely to have seizures, sleep disturbances, and less likely to have breathing disturbances, a spinal curvature, gastrointestinal problems, hand stereotypies, or to lose hand and speech skills. In summary, the authors point out that although in some instances, mutations in the CDKL5 gene have been found in individuals with Rett-like features , the majority of cases of CDKL5 are different. They therefore advocate that the CDKL5 disorder should be considered as a distinct clinical entity.

Note - This is the first study to undertake a detailed clinical analysis of a large group of individuals affected by a CDKL5 mutation. As such it provides the first meaningful information about the spectrum of the CDKL5 disorder.  One interesting point is that data for some of the study came from the InterRett database. It is not a great leap of imagination to think that a similar register for CDKL5 could and should be established - this study might act as a basis for such a register. One might hope that the various international genetic societies could co-operate in developing an international database or registry to capture all known cases of CDKL5 - there are said to be less than 300 known cases so far - and future cases as they present. This would then provide a wealth of data that could be used for future research projects. There is always going to be an issue with funding, but I would refer you to the European Commission website on rare diseases, as one such potential source. This study, at last, establishes the CDKL5 disorder as an distinct clinical entity which is an excellent starting point. The hard task is now to move this forward. Thank you Stephanie!

10. Recurrent mutations in the CDKL5 gene: Genotype-phenotype relationship.
American Journal of Medical Genetics  2012.   Link to abstract.

This study from France looks at whether there is a relationship between mutation and severity of phenotype. A total of 26 CDKL5 mutations were identified in 358 unrelated females who had been referred with encephalopathy and early seizures. Most of these mutations were unique but 8 were recurrent mutations that had previously been described in other studies (including studies published by this group). By studying females with recurrent mutations in both this study and those from previous studies, the authors have drawn some conclusions about the relationship between site of mutation and severity of phenotype. They looked at mutations occurring in 3 particular areas of the CDKL5 protein, the ATP binding site and the ST kinase active site, both of which lie in the catalytic or kinase domain, and the large C-terminal domain of the protein. They found that mutations affecting the ST kinase site and the C-terminal domain produced more severe phenotypes (refractory epilepsy, limited hand skills, non-walkers) than mutations affecting the ATP binding site. The mutations identified were missense and truncating mutations in the ST kinase segment and frameshift mutations in the C-terminal domain. In the ATP binding site, the recurrent mutations were all missense mutations at amino acid position 40, and when compared to girls with other CDKL5 mutations (presumably in this same area), those with this particular mutation (which is termed p.Ala40Val) tended to present with better hand use and better ability to walk. They also looked at whether X-inactivation (see The story of CDKL5)  might also have an effect on phenotype but were unable to draw any useful conclusion.

Note - Overall I found this quite compelling reading really and it certainly builds on previous studies that have suggested links between position of mutation and phenotype. I was a little unsure about the  exact number of girls in this study who actually had CDKL5, but overall I think it is a good and useful study. The fact that certain mutations at the C-terminal of the protein (which is well away from the kinase domain) are also associated with more severe phenotypes presumably suggests something about the 3-D structure of the CDKL5 protein in terms of how the C-terminal interacts with the kinase domain. There is a suggestion that part of the C-terminal might stabilise the protein against degradation and mutations here might therefore accelerate degradation -a view perhaps supported by the study below regarding a novel transcript. Who knows!

9. Variant of Rett Syndrome and CDKL5 gene: Clinical and autonomic description of 10 cases.                                                     
Neuropediatrics 2012.  
 Link to abstract.                                                                                                                                     

This study from Italy describes features in 10 girls, 8 from Italy and 2 from Sweden, aged between 2 and 13 years. All had established CDKL5 mutations (4 missense, 3 frameshift, 1 nonsense and 2 deletions). Early-onset epilepsy was the predominant feature with 9 developing seizures by the 4th month of life and one at 6 months. Speech was absent in all 10 girls, while 8 had some sort of disturbance of their vision. All 10 had experienced disturbances in their sleep patterns during the first 4 months of life and 7 continued to do so at the time of the review. The autonomic aspect of the study focussed on breathing patterns. There are 4 recognised disorders of breathing in Rett Syndrome. Forceful breathing patterns were noted in 8 of the girls with CDKL5 while the remaining 2 girls exhibited the apneustic type.

Note - A study which again reinforces many of the features we are coming to recognise as part of the CDKL5 spectrum. The description of the breathing patterns is interesting as I have not encountered this in any other studies I have read (I may have missed something!). They also discuss other autonomic dysfunction such as cold hands and feet, all of which we recognise in Ellie.

8. A novel transcript of cyclin-dependent kinase-like 5 (CDKL5) has an alternative C-terminus and is the predominant transcript in brain.      Human Genetics 2012.    Link to abstract.

The molecular characterisation of the CDKL5 protein is still largely unknown. However, previous research has identified 3 different variants of the CDKL5 protein. This paper describes the theoretical existence of a 4th variant which the authors then go on to show not only exists, but is likely to be the predominant form expressed in the human brain.  Although the CDKL5 gene contains 21 coding exons, this novel variant is shorter, only incorporating amino acids coded in exons up to 18. They conclude that this novel variant is likely to be of primary pathogenic importance in CDKL5-related conditions.

Note - This is quite a technical-based paper and took me several malts to get through!!  It is estimated that the human body is made up of over 100,000 different proteins yet only contains some 25 - 30,000 genes. This means that on average, one gene may code for perhaps 4 - 5 proteins, and this seems to be the case with CDKL5. The authors of this paper are suggesting that they may have identified one particular form of the CDKL5 protein which is slightly shorter in length, amino acids coded in exons 19 - 21 are not incorporated,  and may be the relevant form of the protein in CDKL5-related conditions. This therefore needs to be considered when screening children who present with features of CDKL5.

7. CDKL5 in different atypical Rett syndrome variants:  Description of the first eight cases from Spain.                                          
Journal of Pediatric Epilepsy 2012.   Link to text.

This study is from the Saint Joan de Deu Children's Hospital in Barcelona. Over a 26 year period, 408 patients were diagnosed with classical or variant Retts. There were 258 children identified who had mutations of the MECP2 gene. Of the remaining 150 without a genetic diagnosis, testing for CDKL5 mutations was offered to 53 girls and 9 boys who had either early epilepsy, polymorphic seizures or drug-resistant seizures.  They found 8 (15%)  of the girls, but none of the boys, had novel pathogenic mutations of the CDKL5 gene. There were variations in the phenotypes of the 8 children, compared to those described in other studies and which have been taken to be characteristic of CDKL5. For instance, the 3 stages of epilepsy as described by Bahi-Buisson were not observed, with only  4 children developing seizures within the first few months of life. There were 3 who developed their seizures at 11 months, 2 and 3 years of age and 1 who did not appear to have developed seizures at all. There were 2 children who regained some speech having previously lost it.  The authors report some association between the severity of the phenotype and the location of the mutation with the more severely affected having a mutation at the proximal region of the gene. The authors conclude that testing for CDKL5 must be mandatory in patients with Rett syndrome and early epilepsy,  in those with epileptic encephalopathy with onset in the first year of life as well as in those with classical or variant Retts without an identifiable MECP2 gene mutation.

Note - This article clearly throws a slight spanner in the works in that some variation is introduced into what, up to now, have been considered fairy constant characteristics of children with CDKL5, particularly in relation to the presence of early seizures.  It therefore at least raises the possibility that the spectrum of CDKL5 phenotypes may be broader than first thought. This issue of the Journal also has an editorial by Jeffrey Neul from the Texas Children's Hospital, which includes a short review of this article. He concludes that perhaps CDKL5 should now be recognised as a distinct clinical disorder with specific unique characteristics rather than as a variant of Rett syndrome, and that not to do so might be a disservice to both clinical medicine and to those with the condition.  Bravo!

Further Note - A subsequent Letter to the Editor regarding this article has pointed out that one of the described mutations, which affects almost the very end of the C-terminal domain, may only have "minimal" clinical effect. This is because a mutation at this site (which is in exon 20) affects that part of the CDKL5 gene which is not used to make the particular form of the protein that may be relevant in CDKL5 disorders (see last paper above). This may explain some of the variation in presentation described in the paper, in that the individual with this mutation may not in fact have a CDKL5 related disorder. The plot thickens!!!


6. Historic, clinical and prognostic features of epileptic encephalopathies caused by CDKL5 mutations.                                               
Paediatric Neurology 2012.   Link to abstract.

This is a retrospective study from the Mayo Clinic, looking at the clinical features and treatment of seizures in 6 children (4 girls, 2 boys) with CDKL5. The onset of seizures ranged from 1 to 3 months of age. Particular features are discussed including dysphagia (difficulty swallowing) and cortical visual impairment. It is also noted that the 2 males in the group appeared less affected by certain features than the females. No particular treatment eliminated seizures, but topiramate, vigabatrin and the ketogenic diet were the most helpful at reducing their frequency.

Note -  This is a study of a small number of children with a relatively short follow-up period, less than 3 years in 2 children and no more than a year in the other 4. The children studied are all young (eldest less than 3 years of age at review). The prognostic features referred to in the title seem to be the presence of the CDKL5 mutation itself and the resultant early-onset of seizures. Vagal Nerve Stimulation (VNS) does not appear to have been used as a treatment option, possibly because of the age of the children involved, although a recent study has suggested that VNS is a safe and effective treatment for resistant epilepsy in young children.  The authors of the above study rightly conclude that more research is needed.


5. Adult Phenotypes in Angelman- and Rett- Like Syndromes.                                                                                                                                               
Molecular Syndromology.  Published online: January 13 2012.   
Link to text.

This paper looks at the evolution of the clinical characteristics (phenotype) in adults with various syndromes, and includes a female who was diagnosed with a  splice site  mutation  in intron 7  of  CDKL5 at the age of 47.  She had developed seizures after 6 months and a subsequent delay in psychomotor development. Her motor development had only been mildly delayed, but she hadn't learned to speak and she had profound intellectual disability. Her seizures had persisted despite multiple drug treatments. She was small, being at the lower end of  centile charts  and displayed a number of facial characteristics including prominent cheekbones, mild prognathism, deep-set eyes and blepharochalasis.  A number of other characteristics are also described. The authors state that this is the oldest reported patient with a CDKL5 mutation. They also then review adult phenotypes reported in other published studies, which, with the adult  in this study comprises a group of 10 adults with an age range from 18 to 47  years.  From this review they surmise that apart from difficulty with seizure control, adults with CDKL5 may also have medical problems such as secondary scoliosis, contractures, feeding difficulties and airway infections.

Note - It is probably with some caution that we should apply the findings in this relatively small group of adults to every child now growing up with CDKL5, because, as many of us are aware, although our children share many characteristics, they also display variable levels of function and development. Furthermore, the fact that CDKL5 is now recognised as a separate condition may improve their longer-term outcome through earlier diagnosis and subsequent improvement in treatment and care.  Although factors that may influence their longer term  development are still to be determined, we are sure to learn more as longitudinal  studies continue.


4. Genes of Early-Onset Epileptic Encephalopathies: From Genotype to Phenotype.                                                                                 
Paediatric Neurology 2012.  
Link to abstract.

This is a review of genetic disorders associated with early-onset epileptic encephalopathies and the associated phenotypes, including CDKL5. The authors report that to-date 53 pathologic mutations of CDKL5 have been reported consisting of 12 missense, 4 nonsense, 8 splice-site, 22 deletions and 7 frameshift mutations. They state that no correlation has been established between the site of mutation and clinical severity and then go on to discuss a more severe epileptic encephalopathy as described by Bahi-Buisson et al They review the 3 stages of epilepsy in CDKL5 as described by Bahi-Buisson et al and the EEG findings as described in studies by Pintaudi et al and  Melani et al.

Note - A good review article although a lot of information to digest. There also seems to be some ambiguity here about what is known so far about the site of mutation and the severity of the phenotype. This paper states that there is no correlation between severity and site of mutation (giving references), but then refers to the study by Bahi-Buisson which suggests that there may be. This is the ambiguity I can't quite sort out - the CDKL5 gene has been  designated Online Mendelian Inheritance in Man OMIM® no. *300203   (the * indicates that this designation applies to the gene and its mutations) and it is to this designation that there is said to be no relation between site and severity. However, there is also another designation, OMIM® #300672  which is a descriptive designation for a phenotype (indicated by the #)  and refers to a more severe form of epileptic encephalopathy also associated with CDKL5 -and which according to the study by Bahi-Busson has a phenotype that may well be related to the site of mutation. If the genetic basis for both designations lies within the CDKL5 gene, why have 2 designations or are they being considered as distinct entities?  There  must be a straight-forward answer here but for now this remains slightly confusing (to me at least) and although I contacted  OMIM®  to clarify this,  I am not sure that I am any the wiser for the response (I would be grateful for clarification from anybody!!)  Clearly, the bottom line is that more children with CDKL5 mutations need to be studied to understand the relationship between site of mutation and severity of phenotype, and ultimately, severity will probably depend on several factors, and not just the site of mutation.  

3. A survey of parents of children with cortical or cerebral visual impairment.     
Journal of Visual Impairment and Blindness 2010.                     Link to text

This study summarises the results of a survey of 80 parents who have children with Cortical or Cerebral Visual Impairment (CVI). The various aetiologies (causes) are reviewed along with rehabilitation and intervention strategies. The most effective rehabilitation is said to be that which involves parents and other family members. Motion, particularly of a coloured object is the most important factor in improving the visual response. The survey also looks at 3 issues - how do parent's receive the diagnosis of CVI, what educational support do children with CVI receive, and what do parents of children with CVI feel about the support they and their children receive.

Note - This article has been written by parents of children with Cortical Visual Impairment (CVI) . This is not something that  we read so much about in CDKL5 articles, as the focus tends to be on seizures, developmental delay and mobility. However, children with CDKL5 certainly do suffer from CVI and although this article covers a broad spectrum of issues, many are relevant to us which I think is very helpful.


2. Gastrointestinal Disorders in Children with Neurodevelopmental Disabilities.                                                                           
Developmental Disabilities Research Reviews 2008.                    Link to abstract

As the title suggests, this article reviews the various gastrointestinal problems that children with neurodevelopmental disabilities can suffer. It systematically describes problems from the mouth and swallowing, through to the lower bowel and constipation. Underlying causes are discussed and the various treatment options reviewed.   The author points out how difficulties with eating may be confused with behavioural food avoidance or aversion, and that abdominal discomfort can present as chronic irritability, crying or dystonic (involuntary twisting or repetitive) movements of the face and neck. More than 80% of children with gastro-oesophagel reflux will have recurrent vomiting. Constipation is discussed and defined as opening of bowels less frequently than 3 times per week or the regular need for laxatives. Constipation is a common problem in children with disabilities and can go unrecognised for years. Treatment requires a consistent approach by willing parents and carers.

Note - Although this review tends to refer frequently to children with cerebral palsy,  many of the problems discussed will apply to children with CDKL5. I have highlighted 2 particular issues that are discussed, gastro-oesophageal reflux  and constipation, as these will be familiar to most of us. There is one particular telling line in the discussion about gastro-oesophagel reflux, which says  "It is ironic that the learning deficit that usually accompanies neurological impairment also impairs the ability of the individual to communicate the main symptom... which is pain."    A good and very useful review article - I have a full copy if anyone would like to read it


1. Melatonin therapy for circadian rhythm sleep disorders in children with multiple disabilities: what have we learned in the last decade?        Developmental Medicine & Child Neurology 2004.                    Link to text

Sleep is thought to be a learned neurological process that develops in response to cues from the environment. This is a process that we all go through, but which is impaired in children with neurodevelopmental problems. As a result they can retain their fragmented "baby" pattern of sleeping, potentially throughout their life. Circadian rhythm sleep disorders (CRSD) occur when there is a dissociation between sleep-wake  behaviour and the environment. Parents should promote the sleep learning process through "highly regulated strict sleep hygiene".  Melatonin is produced by the pineal gland in the brain, and although previous research has established that it can induce and maintain sleep,  it's exact mechanism of action remains unclear. It does not promote sleep in the same way as a hypnotic (sleeping tablet) does,  but may act by inhibiting neuronal excitability in the central nervous system. Exposure to light tends to decrease melatonin secretion while darkness promotes it. CRSD's tend to be associated with disturbance in melatonin secretion. There are also non-circadian rhythm sleep disorders due to pain from such things as reflux or orthopaedic problems, or from nocturnal seizures or related drug treatment. In the treatment of CRSD it is important to choose the correct melatonin formulation depending of the particular sleep disorder. Discontinuation for a few days is also recommended as children can acquire normal sleep patterns over time. Some studies suggest that children with neurodevelopmental disabilities, who have CRSD  can show a 70-90% quick response. However, not all children respond equally and there may be some with certain conditions who may not benefit. No major side-effects of melatonin have been reported in paediatric studies. 

Note -   This review contains a lot of information about sleep disorders and the role of melatonin. One of the authors published the original study on the role of melatonin in sleep disorders, and so the enthusiasm for melatonin is understandable. They do state that much of the skepticism regarding melatonin was based on lack of quality control, unsubstantiated benefit claims, misconceptions about the action of melatonin and oversimplification about the relationship between melatonin and CRSD's.  Perhaps the most relevant points for parents and carers are made in the first few paragraphs where they state that the majority of children with neurodevelopmental disabilities are said to have long-standing sleep disorders, and that the management is often neglected due to the inadequate training of doctors in these problems.