Guest blog within the series ‘Ion channel drug discovery insights from the Channelogist’ (@Albion Drug Discovery Services) for Metrion Biosciences
Introduction
Cure CLCN4 is a charity founded by a patient family to provide support, raise awareness and fund medical research for effective treatments for CLCN4, a rare genetic condition which causes intellectual disability, behavioural and movement disorders and microcephaly. Each year, this important charity holds a conference to highlight the condition, bring together thought leaders in the field (including international academics and clinicians), provide input from patients and their families, and share research insights with the aim of discovering a cure for this rare disease and generating a strong patient-orientated research community and support network.
A link to this years’ conference can be found below:
This years’ event, held earlier this Summer, was organised by Paul McGuiness and Rebeca Ridings Figuero, alongside Dr. Peter Trill, a UK bio-entrepreneur and his wife Dr. Gina Tan, a clinician who founded Cure CLCN4 as their youngest daughter Daphne has a mutation in the CLCN4 gene and suffers from neuro-developmental symptoms. There were around 40 in-person attendees in addition to a varied online delegation of ~30 attendees, plus a closed session focused solely on patient families.
Selected talks are available in video here:
https://www.youtube.com/playlist?list=PLMK5QcuaGueUK3Yea-qiHdc-Ch7wRgrH1
Scientific background
CLCN4-related neurodevelopmental disorder (CLCN4-NDD) is a rare X-linked genetic condition associated with intellectual disability, psychiatric disorders, gastrointestinal issues and epilepsy. The extent and severity of the condition is extremely variable and dependent upon the type of gene change and the gender of the individual, amongst other factors.
CLCN4-NDD is caused by mutations of the CLCN4 gene encoding the CLC-4 channel which is expressed across various tissue types, but most prominently in the brain and skeletal muscle. Significantly, CLC4 is expressed on endosomes and the ER, and likely interacts with related CLC-3 channels which facilitate sorting of CLC-4 proteins into endosomal compartments.
Key points:
The importance of diverse intracellular organellar ion channels in a wide variety of diseases is now gaining greater attention (e.g. TMEM175 in Parkinson’s Disease), making them a new class of ion channel drug discovery target. Cure CLCN4 is also a great example of the growing trend for patient families to get involved in finding treatments for such rare disease channelopathies.
A summary of the talks presented across the two days is provided below:
Vera Kalscheuer - Molecular Genetics of CLCN4
Vera is one of the original and leading academics studying CLCN4 genetics (through her focus on X-linked genetic disorders) and is a major part of the CLCN4 foundation. Her group has identified many examples of patients and families with CLCN4 mutations, expanding the picture to include de novo and inherited missense mutations in females and males, respectively. She co-authored a comprehensive review in 2023 (Palmer et al., 2023) that brought together many international researchers and highlighted the emerging genetic diversity in CLCN4 patients. Her previous publications have attracted the attention of other clinicians trying to diagnose and treat rare CLCN4 patients scattered across the globe; Elizabeth Emma Palmer is an Australian clinical researcher who is now also a big part of the CLCN4 collaborative network and another major force in the CLCN4 foundation.
Vera and collaborators now have a growing database of CLCN4 gene mutations from the ClinVar database and publications, which currently shows 90 variants from 49 families which comprise 41 unique (typically de novo) and 18 recurrent (typically missense) mutations which have been predicted to be benign (1), pathogenic (36), or a variant of unknown significance (VUS, 53) which are the most perplexing and require further work to determine if/how they affect CLCN4 protein function, channel activity and patient symptoms. Michael Pusch is working through many of these mutations using Xenopus oocyte TEVC electrophysiology (next talk), and Vera highlighted that many of the in silico predictions for new mutations do not match with functional studies: for example, 30 of the 90 known variants have predicted ‘low’ or ‘benign’ effect, but 13 of these were shown to be either loss- or gain-of-function on CLCN4 currents.
The new data is also significant for showing g-o-f mutations, as CLCN4 was previously thought to be l-o-f (missense and truncating mutations, especially in females);
In males they see truncating l-o-f mutations (de novo, inherited), as well as new g-o-f variants (e.g. Val317x residue);
In females, de novo & inherited mutations are ALL missense, and there are new g-o-f variants (Ala555 residue)
Interestingly, although CLCN4 is an-X-linked genetic disease expected to be limited to females and inherited from the maternal line, males can both pass on inherited mutations and be affected by de novo mutations as X chromosome inactivation can be incomplete, and genome imprinting can occur from the fertilised (i.e. male XY) egg. However, de novo mutations are much less common in males (27%) than females (78%). 90% of variants are missense, and all affected females carry such mutations as truncating or intragenic chromosomal deletions appear to be compensated for by the un-affected X chromosome allele.
Michael Pusch – Functional studies of CLCN4 channel mutants
His interest initially stemmed from the fact he trained in Prof. Thomas Jensch’s lab, who cloned the first member of the CLC family. His specialist contribution is Xenopus oocyte electrophysiology of CLCN4 mutants.
There is a major distinction between members of the CLCN family; CLC-1, CLC-2 and the two kidney and inner ear CLC-K variants are ion channels expressed on the plasma membrane where they regulate anion transport, membrane potential and counter-ion movements. The other members (CLC-3, 4, 5, 6 and 7) are expressed in intracellular organelles (endosomes and lysosomes) where they act predominantly as co-transporters (exchangers) of Cl- and H+ ions to regulate pH, membrane potential and counter-ion permeability. There are rare human channelopathy diseases associated with 7/8 members of the CLCN family.
Each of the organellar CLC proteins are specialised/localised to different endocytotic, sorting and recycling endosomes and degradative lysosomes, where they all help to control pH and function.
Interestingly, CLC proteins seem to form channel dimers, with each subunit contributing an independent pathway for the exchanger transport of Cl- and protons (similar to dual ‘pore’ Fluc fluoride anion channels).
CLC4 may be unique in the CLCN family as its function, but certainly its trafficking and localisation to endosomes from the ER, requires heteromerization with CLC3 proteins (other CLCs homodimerise). Previous electrophysiology work showed all/most CLCN4 mutations were loss of function, but newer data shows that some (pathological) mutations are like wildtype, while other predicted benign variants are actually loss-of-function, and some rare new mutations are in fact gain-of-function.
New gain-of-function variants, revealed by studying electrophysiology with altered (acidic, i.e. physiological) pH would be lethal in males, so are only seen in females. Further biophysical work shows that some loss-of-function mutations actually shift voltage-dependence, but could have a dominant negative effect in heteromers.
Emma Palmer – CLCN4 clinical genetics and rare patient families
Emma is a key player in the CLCN4 community, and has lead the development of (new) clinical guidelines for treating patients with CLCN4 mutations with Vera and others. She remains a major initial point of contact for many GPs and clinicians who come across patients with CLCN4 mutations after genetic screening for undiagnosed diseases. Therefore, gaining knowledge about new variants, and testing their actual vs predicted significance for function and diagnosis and patient stratification, is key.
In this talk Emma outlined the typical rare disease family experience and clinical journey, patient and symptom progression, and the challenges remaining.
CLCN4 is a complex syndrome, with some sex-specific differences. Common symptoms involve:
Yann Herault – Animal models of CLCN4
Yann (Inserm) is a key part of the CLCN4 foundation, successfully creating several mouse and rat models of CLCN4 mutations for translational disease modelling and pharmacological screening; this works occurs alongside development of iPSC cell-based models of CLCN4 condition (e.g. by South Korean researchers featured on day 2).
The interesting issue is that the chromosome location of CLCN4 and neighbouring genes is more homologous in rat than mouse, so he decided to make global (exon 4 STOP) and Cre/Lox conditional CLCN4-/- knockout rats for more specific disease modelling, as well as a g-o-f A55V (A449V in rat) knock-in rat (using CRISPR). They have stablished breeding colonies for all 3 rat models and these are now available for collaborators, for free. Also significant is the fact that previous CLCN3 and CLCN4 knockout mice models show no gross symptoms, although other talks at this conference do report some subtle CNS deficits in CLCN4 knockout mice (e.g. Raul Guzman’s work on pyramidal dendrite morphology and function, South Korean autism researcher Yeni Kim).
Anselm Zdebik - Zebrafish CLCN4 model
Anselm works at UCL, but was previously in Thomas Jentsch’s lab working on CLCN channels. He is currently trying to develop zebrafish loss-of-function (V536M) and gain-of-function (A555V) CLCN4 genetic models, as many symptoms can be measured in this species (psycholocomotor, seizure, cerebral atrophy, anxiety, depression, autism). He is using a CRISPresso technique for zebrafish genetic editing; work is still in progress. He previously made a Kv (KCNJ10) channel knockout for electrophysiological and behavioural studies in zebrafish.
Joseph Mindell (NINDS) – CLC7 gain-of-function mutations in lysosomal storage disease
Joseph is another leading academic researcher in the CLC community and well known to senior participants. He became interested in this channel family as a postdoc in Chris Miller’s lab that studied CLC channels. His speciality is lysosomal disease, and includes study of CLC7 mutants and patient families; CLC7/Ostim1 is implicated in bone disease, but he didn’t talk about this today. Jo favours a pH-sensitive fluorescent dye assay (Oregon Green dextran) to study endosome function, as well as whole-cell and single channel patch clamp recordings from CLC channels expressed in heterologous cells. His group has discovered several new patient families with a g-o-f mutation in CLC7 (Y715C) which affect lysosome acidification, resulting in hypopigmentation (dysfunction in melanocytes) and systemic lysosome storage disease. Three new patients have similar g-o-f mutations but milder disease phenotypes. His recent work suggests that these g-o-f mutations affect the interaction of lysosome-specific phosphoinositide PI(3,5)P2 with CLC7 proteins, which normally inhibits transport activity.
Day 2:
CNS electrophysiology and pharmacology
Jinju Han (KAIST, Korea) – Excitatory neuron survival
The key data in this talk is the development and study of human ESC CLCN4 mutant and knockout stem cell lines, neural progenitors, and (immature) neurons to create translational disease-in-a-dish reagents and assays. Stem cell models using immature human CNS neurons may be more relevant for CLCN4 syndrome than many other CNS diseases (e.g. schizophrenia, depression, neurodegeneration) as CLCN4 is a neuro-developmental syndrome that affects infants and young children, rather than adults. However, the use of ES rather than iPS cells may affect drug discovery efforts outside of Korea due to ethical issues and regulatory requirements in the US, EU and UK.
Introducing the A555V g-o-f function into stem cell neurons actually decreased CLCN4 mRNA expression by 50%, and lead to rapid cell death (TUJ1+ neurons), indicating severe effects on neurogenesis, especially for excitatory neurons (identified by scRNA transcriptome profiling). Similar effects were seen in 2D hES and 3D organoid mini-brains, with altered dendrite morphology. It may be possible that this loss of excitatory neurons could impact neuronal development and cognition and behaviours, but it remains unclear how a similar mechanism could underlie the seizures that are prominent in CLCN4 patients.
Yeni Kim (Dongguk University, Korea) – Synaptic dendrite function and developmental deficit
Yeni works with Jinju Kim. The key finding here is that the anti-psychotic Risperidone, a mixed DA and 5-HT GPCR modulator, can reverse or ameliorate the CNS neuron morphological and behavioural effects of CLCN4 knockout mice. Previous studies found no gross neurological symptoms in knockout mice, but her group has identified autism-related effects on social interaction, anxiety and repetitive behaviour in 7-8 week old knockout mice. As CLCN4 patients have neuro-developmental deficits and some key features of autism, she postulated that an agent effective in her models of autism may also work for the CLCN4 condition. Knockout of CLCN4 differentially affected cortical expression of genes associated with neuronal plasticity and synapses (PSD95, CDK5), and dendritic morphology (Golgi staining). Risperidone could ‘correct’ these genetic deficits, as well as rescue (excitatory) neuron differentiation, dendritic and synaptic morphology, and behavioural symptoms.
The relevance of a complete and global CLCN4 knockout in mice to patients with gene-modifying mutations that may reduce or increase channel function is not clear, as-is the mechanism of Risperidone action or the signalling cascades affected, so further work is clearly required and is underway with human ES neuron translational models. Also, there was no link shown between these CNS and behavioural effects and the known function of CLCN4 to regulate organellar patho-physiology, although in their talk at the 2020 CLCN4 Symposium they said they saw increased LysoTracker dye labelling, suggesting increased organellar cycling, perhaps due to process and degrade dying neurons?
CLC channel biophysics and function:
Thomas Jentsch (FMP Berlin) - Physical and functional interactions of CLC4 and CLC3
Thomas is famous in the field for cloning the first CLC channel gene in 1990 (CLC1 in Torpedo electric organ), and several leading researchers in the current field are linked to his laboratories in Germany. He has looked at a range of CLC and VRAC anion channels, and their channelopathies and functions in various and varied human diseases.
The fact that human patients with single CLC gene deletions and mutations manifest significant (neurological) symptoms suggest that the CLC gene family members cannot fully compensate for each other, likely as each of them are expressed and function in different organellar compartments and pH states of the endo-lysosomal pathway (see Figure below). CLC3, 4 and 5 seem able to form heterodimers, so knockout of CLC3 probably causes more serious neurological symptoms than CLCN4 knockouts as the former also leads to reductions in CLCN4 protein, but the reverse is not the case as CLC3 is a dominant negative partner, and CLCN4 stability, trafficking and lysosomal function depends on CLC3 but not the other way round. Significantly, uncoupling or deletion of CLC3 H+ transport in knockout mice only leads to neurological dysfunction in the absence of CLCN4, suggesting the latter can contribute a functional transport function if it forms a dimer with non-functional CLC3 proteins, which can still help with protein complex stability and trafficking.
Taken from Jentsch & Pusch (2018): 10.1152/physrev.00047.2017
Thomas reiterated the theme that in many mouse CLC gene knockouts there are few discernible neurological deficits, and he (but not all others) employs an extensive battery of tests and assays (e.g. elevated maze, Barnes maze and open field for anxiety, home cage activity and sleep: wake cycle) to detect behavioural deficits. The only difference between wild type and CLC4 mutant mice was in a social interaction assay (in a test tube), which may correlate with the autism symptoms detected by South Korean researchers.
Merritt Maduke (Stanford) – CLC proton exchange transporter structure-function
Merritt runs a molecular and cellular physiology lab, where she uses a prokaryotic CLC gene and protein model. Her lab has obtained several new and detailed cryo-EM views, as well as carried extensive mutational analysis, to refine molecular dynamic models to try and determine each of the 4 steps of the Cl-/H+ counter-transport cycle, disproving several previous models and proposed protein conformations. These 4 states or configurations require the Glu residues in the permeation pathway to be in the ‘up’, ‘middle’ or ‘down’ orientation, as well as a new ‘out’ position.
During questions, it was asked whether CLCN researchers had used Google’s AlphaFold software to predict and model CLCN4 protein structure-function and the potential impact and effect of patient mutations. Several speakers replied that they had tried this, but it turns out that the crucial dimer stoichiometry of functional CLC proteins is not well modelled by this software (which was surprising and interesting to hear). Thus, it would appear as though there is still a real bonus to having empirical x-ray and cryo-EM structures and to use these in detailed computational MD studies of ion channel structure-function, rather than rely on purely in silico techniques.
Cecilia George (Nanion) – Investigating CLCs using APC methods
Cecilia showed the attendees how they could augment their current electrophysiology and imaging methods to include automated patch clamp (APC) and Solid State Membrane (SSM) technology and platforms for their CLC and lysosomal disease research, and how these could help them span the gap between basic research and drug discovery. Her talk revealed recent high quality APC data from CLC-1 channels (on the SP384 platform) and CLC-7 channel/transporter recordings (on the Surfe2R SSM platform), and included some data from native, purified lysosome membranes rather than plasma membranes from (mutated) over-expression systems. This is an important feature, as there are marked differences in the membrane lipid composition and protein-interacting partners between organellar and plasma membranes which can affect biophysical behaviour and pharmacology, with important implications for disease modelling and drug discovery efforts.
Rare disease patient organisations:
It is becoming clear in the field of channelopathies and rare disease that there is an increasing role for patient recognition, and family and charity lobbying, fundraising and involvement in the drug discovery process. Academics and clinicians already tap into these resources through diagnosis and treatment, especially given the rapid rise and reduced cost of genome sequencing, and parents in turn are more active in searching for specialists and treatments from the literature and online after their child is diagnosed (sometimes after several frustrating years without much useful information). It is also crucial to acknowledge the potential for drug discovery companies to also get involved with rare disease individuals, families and organisations, as many in this space are now doing. Several talks on Day 1 and Day 2 addressed some of these issues, but more still needs to be done.
Also worth mentioning is that fact that Gina Tan (mother of Daphne and wife and co-founder with Peter Trill of the CLCN4 Foundation) started on her rare disease journey after Daphne’s diagnosis with a single Facebook post that initially connected her to 5 other CLCN4 families. This group was found by the clinician Emma Palmer in Sydney who was looking for new patients and mutations, and has now grown to a support group with over 120 families and an extensive network of academics and clinicians, and an established rare disease charity based in the UK but with connections to groups around the world.
Paul Wang – Simon’s Searchlight Foundation
Paul provided an overview of Simon’s Searchlight Foundation, an organisation based in the US which deals mostly with rare disease patient groups. Paul is a neurodevelopmental disease clinician who worked in industry at Pfizer, before getting involved in clinical trial advocacy and rare disease patient groups. Simon’s Searchlight is a charity set up to help bring together the various stakeholders in rare disease, but especially to help patient families and foundations put together the necessary toolbox of experts, databases and reagents to facilitate the search and clinical testing of new treatments for rare disease patients. Their services include patient and family contact lists and co-ordinators, family and medical history surveys, blood and cell sampling to allow creation of iPSC disease models, development of cell line and animal genetic models and reagents (e.g. in academic labs, or by commercial CRO partners paid for by the charities and founders), through to clinical trial design and execution. A current focus with CLCN4 patients is to acquire clinical EEG recordings, as epilepsy is a common symptom but this data can also help assess other neurological changes, and may provide a biomarker for drug trials. Currently the Foundation is working on over 170 rare disease genes and thousands of patients and families involved with neurodevelopment disease.
Jessica Duis (Colorado Children’s Hospital) – Clinical trial readiness for rare disease
Jessica is involved with Angelman, Prader-Willi, and Pitt Hopkins syndrome patients and families, all of which are rare neurological conditions. She is a clinician who has led and advised investigator-driven clinical trials and now also works with industry to design better rare disease clinical trials and regulatory filings. The main issues are the small patient populations and difficulty in profiling, classifying and stratifying patients to enable a suitably robust and statistically powered trial (more likely a small Ph II than a larger Ph III), and the need for new outcome measures and biomarkers to be developed, validated, and approved by the FDA.
For Angelman patients she helped develop a ‘disease concept’ model using data from care givers, patients and their families, and clinicians to identify the most disruptive symptoms, so that these can be specifically addressed in treatment regimes and drug discovery research and clinical testing. These highlighted speech and motor problems, as well as seizures. Detailed case histories also identified changes in gait, which could then be correlated with genetic changes and aid in genotyped patient classification.
Dr Pangkong Fox - CACNA1 Foundation
Dr Fox’s son was diagnosed with a mutation in the Cav2.1 ion channel gene in 2021, and after looking for information online she found the CACNA1 Foundation and by 2022 she was their Scientific Engagement Director. She has a PhD in Cell and Molecular Biology and uses her scientific background to create education support and resources. Cav2.1 mutations are linked to various forms of ataxia, migraine and epilepsies, as well as neurodevelopmental conditions such as intellectual disability and autism. There is a complex geno-phenotype with over 300 pathogenic mutations that lead to loss- or gain-of-function. The CACNA1A Foundation is doing many of the same things as the CLCN4 Foundation and Simon’s Searchlight (via their Research network) to develop a ‘preclinical toolbox’ of patient histories, biobanking, iPS and heterologous cell line generation, animal models, etc. These are and will be used for drug re-purposing screens and the search for new treatments (small molecules, ASOs, gene therapy). They are utilising collaborations with many other charities and disease and research institutions to achieve this cost-effectively and quickly (e.g. NORD, RareX, Citizen Invitae platform), as well as working with industry partners (e.g. Metabolom in Ireland for biomarkers), and receiving funding from US agencies and organisations (e.g. Chan Zuckerberg Initiative).
Many other ion channel channelopathy parents have followed this pathway, but without an established rare disease organisation they created their own charity, advocacy group or foundation – examples include KCNT1, KCNH10, CLCN4, SCN2A, etc. These groups help to identify and bring together patients, academics, clinicians and industry to facilitate collaboration, raise awareness, increase understanding of the disease and patient and family needs, discover new treatments, and raise funds.
A Family Story
Whilst the main patient and family engagement session on Day 2 was limited to the key clinicians, attendees were able to gain a personal insight into what it was like to have a child diagnosed with CLCN4 syndrome in a short online session with a mum in the Bay Area. Her child was first taken to the GP at 6 months with feeding problems and at 12 months with developmental delays, but was not diagnosed until after genotyping and neurological clinical profiling at 2 years of age. It was a very touching presentation – the child was asleep I think, but we were shown family videos and also some slides that the Mum had created about disease progression, symptoms, and family challenges and highlights. This was a very moving, personal and emotionally affecting presentation.
There was a similar video montage and slideshow of patient family pictures and information on Day 1 (also shown at the 2022 Symposium) illustrating patients and families who were diagnosed at varied ages from months-years-decades old, spread across the globe. This type of patient family insight is a very powerful way to bring R&D into personal focus.
Drug Discovery case studies:
Michael Schwake (University of Bielefeld) - Parkinson’s Disease affects lysosomal function
Michael is another alumni from the Jentsch lab, where he studied CLC5 in Dent kidney disease. He is now working on Parkinson’s Disease, which can be classified as an Inherited Disease of membrane Trafficking (IDT) as there is genetic linkage to TMEM175 and TRPML1 lysosomal channels, as well as to GCase enzyme involved in autophagy and endo-lysosomal cycling and function. Several academic and industry groups are working on GCase treatments for PD, but he noted that while GCase activity is reduced in the brains of PD patients and animal models, there are also PD patient mutations which show no effect on enzyme activity. I wasn’t aware of this but Ambroxol (a non-selective Nav channel inhibitor, cough medicine) is being tested clinically for PD as it has been shown to act as a chaperone for GCase, to boost protein trafficking and function.
Michael is working on human patient-derived iPS neuron models of PD, differentiating TH+ dopa neurons; he stressed the importance of using isogenic controls to confirm disease and gene differences, which he creates using CRISPR gene editing (baculovirus vectors being more effective than other expression methods). His models replicate the loss of GCase expression and activity and formation of a-synuclein fibrils, and also found that another lysosomal protein LIMP2 is reduced in PD. Significantly, he found that LIMP2 activates GCase and it’s associated protein GBA1, so loss of LIMP2 may lead to reductions in GCase in PD. His group is developing cell-penetrating peptides of the LIMP2 helix to promote GCase and GBA1 activity and function as a novel therapeutic approach for PD. The obvious corollary for potential CLCN4 disease treatments is to use small molecules or genetic treatments to modulate upstream regulators as well as the CLCN4 gene and protein itself, such as CLC3 and other proteins which may augment CLCN4 trafficking and activity.
Laksh Aithani - Charm Therapeutics
Charm Therapeutics is a London start-up with US connections; Laksh previously worked at leading AI drug discovery company Exscientia (UK), and the other co-founder is a US academic and entrepreneur focused on in silico deep learning protein folding applications. Other team members have come from Google DeepMind and big pharma.
Their main argument is that genetic sequences do not reliably predict 2D or 3D protein structure (i.e. the AlphaFold approach), so high resolution protein crystal structures should be used; additionally, ligand binding can alter protein structures (i.e. apo state may not be useful for drug discovery).
They have developed a fast deep learning algorithm called DragonFold that is trained on the 25% of PDB crystal structures with ligands bound (GPCRs, ion channels) and which can be used in a virtual screening workflow to create virtual crystals of the target protein and small molecule libraries, using supercomputer hardware (100 NVIDIA GPUs) and optimised AI/ML software.
Their therapeutic interest is mainly oncology and Laksh presented 2 case studies of small virtual screens against a kinase and a nuclear polymerase. Only 100-500 compounds could be virtually bound owing to computing and cost limitations, with each screen identifying a handful of hits with improved selectivity or mM potency respectively, which were validated in functional binding and biological assays.
Marlen Lauffer (Dutch Centre for RNA Therapeutics) - ASO genetic treatments for rare disease
I really enjoyed Marlen’s talk and her enthusiasm to find effective treatments for patients suffering from ultra-rare diseases, for which academic drug discovery is too slow and commercial interest is extremely limited, and any drug to reach market would be extremely expensive.
Her group is a non-profit consortium supported by the Dutch government and EU funding to rapidly find treatments for rare neurodegenerative disease, as effective drugs need to be delivered to young children before life-limiting symptoms take hold. To this end, they are using patient-derived fibroblasts to directly-induce human stem cell iNeurons, as traditional iPS engineering techniques are much slower and more expensive. Their therapeutic focus is on RNA antisense oligonucleotides (ASOs) to modulate faulty gene expression through exon skipping, RNAi or allele-specific mechanisms. She acknowledged that these approaches work best for g-o-f mutants, which have now been identified in a small subset of CLCN4 patients, but others have shown it can work to up-regulate gene and protein expression as well which may be suitable for treating l-o-f CLCN4 mutations as well. Rather than trying to group related patients with similar mutant genotypes for traditional small clinical trials, she suggested that defining each patient as their own population enabled N=1 Ph 0 clinical trials with fewer regulatory requirements, essentially breaking down each rare disease into multiple ‘nano’ rare diseases (also called 1 Mutation 1 Medicine, 1M1M, or N=1 diseases). ASOs can be delivered locally by i.c.v, i.t or injection into the eye, but it is unclear if a single application (injection or transgenic expression) can ‘cure’ a disease (one-and-done), or whether repeated applications may be needed during a child’s development.
This is a high risk and revolutionary approach; there have been recent patient deaths with such ultra-personalised genetic approaches (e.g. a DMD patient family-funded effort in the US), but Marlen is using the example of Milasen which was an ASO developed and tested to treat a single young female patient with Batten’s disease (CLC lysosomal storage disorder). She presented a case study on a patient with the retinal disease Stargardt Syndrome, with ASOs designed to correct intron/exon splicing that are being tested in iPS retinal organoids. I was concerned by the fact that limited safety testing is carried out with these ASOs (as many assume they are incredibly specific and selective, and thereby safe). For example, limited toxicology testing is done in non-human species, which presumably have different gene sequences to the human protein, so the action and efficacy of human-specific ASOs on the target protein and it’s ‘ome partners and signalling cascades may be reduced and under-estimated. For example, several pharma groups and CRO companies (e.g. Amgen, Metrion Biosciences) have shown that siRNA and ASOs and other genetic modulators can have unanticipated safety pharmacology effects, such as long-term (rather than acute) effects on hERG channel activity. The FDA and EMEA are aware of thee issues and may be increasing the threshold for safety testing of such RNA treatments, but as with oncology the seriousness of the disease and small patient numbers (in genetically stratified patients) may allow a more relaxed approval process.
Conclusion
In conclusion, this key conference promoted collaboration and rare disease patient and family support and participation, clearly showing that a varied and muti-disciplinary approach to channelopathy drug discovery can lead to tangible results. Above all there was a feeling of satisfaction that contributions, however small, are not going un-noticed and there is an overwhelming number of people and companies who are dedicating their drug discovery endeavours to accelerate treatments for rare conditions such as CLCN4.
Guest blog within the series ‘Ion channel drug discovery insights from the Channelogist’ (@Albion Drug Discovery Services) for Metrion Biosciences
Introduction
Cure CLCN4 is a charity founded by a patient family to provide support, raise awareness and fund medical research for effective treatments for CLCN4, a rare genetic condition which causes intellectual disability, behavioural and movement disorders and microcephaly. Each year, this important charity holds a conference to highlight the condition, bring together thought leaders in the field (including international academics and clinicians), provide input from patients and their families, and share research insights with the aim of discovering a cure for this rare disease and generating a strong patient-orientated research community and support network.
A link to this years’ conference can be found below:
This years’ event, held earlier this Summer, was organised by Paul McGuiness and Rebeca Ridings Figuero, alongside Dr. Peter Trill, a UK bio-entrepreneur and his wife Dr. Gina Tan, a clinician who founded Cure CLCN4 as their youngest daughter Daphne has a mutation in the CLCN4 gene and suffers from neuro-developmental symptoms. There were around 40 in-person attendees in addition to a varied online delegation of ~30 attendees, plus a closed session focused solely on patient families.
Selected talks are available in video here:
https://www.youtube.com/playlist?list=PLMK5QcuaGueUK3Yea-qiHdc-Ch7wRgrH1
Scientific background
CLCN4-related neurodevelopmental disorder (CLCN4-NDD) is a rare X-linked genetic condition associated with intellectual disability, psychiatric disorders, gastrointestinal issues and epilepsy. The extent and severity of the condition is extremely variable and dependent upon the type of gene change and the gender of the individual, amongst other factors.
CLCN4-NDD is caused by mutations of the CLCN4 gene encoding the CLC-4 channel which is expressed across various tissue types, but most prominently in the brain and skeletal muscle. Significantly, CLC4 is expressed on endosomes and the ER, and likely interacts with related CLC-3 channels which facilitate sorting of CLC-4 proteins into endosomal compartments.
Key points:
The importance of diverse intracellular organellar ion channels in a wide variety of diseases is now gaining greater attention (e.g. TMEM175 in Parkinson’s Disease), making them a new class of ion channel drug discovery target. Cure CLCN4 is also a great example of the growing trend for patient families to get involved in finding treatments for such rare disease channelopathies.
A summary of the talks presented across the two days is provided below:
Vera Kalscheuer - Molecular Genetics of CLCN4
Vera is one of the original and leading academics studying CLCN4 genetics (through her focus on X-linked genetic disorders) and is a major part of the CLCN4 foundation. Her group has identified many examples of patients and families with CLCN4 mutations, expanding the picture to include de novo and inherited missense mutations in females and males, respectively. She co-authored a comprehensive review in 2023 (Palmer et al., 2023) that brought together many international researchers and highlighted the emerging genetic diversity in CLCN4 patients. Her previous publications have attracted the attention of other clinicians trying to diagnose and treat rare CLCN4 patients scattered across the globe; Elizabeth Emma Palmer is an Australian clinical researcher who is now also a big part of the CLCN4 collaborative network and another major force in the CLCN4 foundation.
Vera and collaborators now have a growing database of CLCN4 gene mutations from the ClinVar database and publications, which currently shows 90 variants from 49 families which comprise 41 unique (typically de novo) and 18 recurrent (typically missense) mutations which have been predicted to be benign (1), pathogenic (36), or a variant of unknown significance (VUS, 53) which are the most perplexing and require further work to determine if/how they affect CLCN4 protein function, channel activity and patient symptoms. Michael Pusch is working through many of these mutations using Xenopus oocyte TEVC electrophysiology (next talk), and Vera highlighted that many of the in silico predictions for new mutations do not match with functional studies: for example, 30 of the 90 known variants have predicted ‘low’ or ‘benign’ effect, but 13 of these were shown to be either loss- or gain-of-function on CLCN4 currents.
The new data is also significant for showing g-o-f mutations, as CLCN4 was previously thought to be l-o-f (missense and truncating mutations, especially in females);
In males they see truncating l-o-f mutations (de novo, inherited), as well as new g-o-f variants (e.g. Val317x residue);
In females, de novo & inherited mutations are ALL missense, and there are new g-o-f variants (Ala555 residue)
Interestingly, although CLCN4 is an-X-linked genetic disease expected to be limited to females and inherited from the maternal line, males can both pass on inherited mutations and be affected by de novo mutations as X chromosome inactivation can be incomplete, and genome imprinting can occur from the fertilised (i.e. male XY) egg. However, de novo mutations are much less common in males (27%) than females (78%). 90% of variants are missense, and all affected females carry such mutations as truncating or intragenic chromosomal deletions appear to be compensated for by the un-affected X chromosome allele.
Michael Pusch – Functional studies of CLCN4 channel mutants
His interest initially stemmed from the fact he trained in Prof. Thomas Jensch’s lab, who cloned the first member of the CLC family. His specialist contribution is Xenopus oocyte electrophysiology of CLCN4 mutants.
There is a major distinction between members of the CLCN family; CLC-1, CLC-2 and the two kidney and inner ear CLC-K variants are ion channels expressed on the plasma membrane where they regulate anion transport, membrane potential and counter-ion movements. The other members (CLC-3, 4, 5, 6 and 7) are expressed in intracellular organelles (endosomes and lysosomes) where they act predominantly as co-transporters (exchangers) of Cl- and H+ ions to regulate pH, membrane potential and counter-ion permeability. There are rare human channelopathy diseases associated with 7/8 members of the CLCN family.
Each of the organellar CLC proteins are specialised/localised to different endocytotic, sorting and recycling endosomes and degradative lysosomes, where they all help to control pH and function.
Interestingly, CLC proteins seem to form channel dimers, with each subunit contributing an independent pathway for the exchanger transport of Cl- and protons (similar to dual ‘pore’ Fluc fluoride anion channels).
CLC4 may be unique in the CLCN family as its function, but certainly its trafficking and localisation to endosomes from the ER, requires heteromerization with CLC3 proteins (other CLCs homodimerise). Previous electrophysiology work showed all/most CLCN4 mutations were loss of function, but newer data shows that some (pathological) mutations are like wildtype, while other predicted benign variants are actually loss-of-function, and some rare new mutations are in fact gain-of-function.
New gain-of-function variants, revealed by studying electrophysiology with altered (acidic, i.e. physiological) pH would be lethal in males, so are only seen in females. Further biophysical work shows that some loss-of-function mutations actually shift voltage-dependence, but could have a dominant negative effect in heteromers.
Emma Palmer – CLCN4 clinical genetics and rare patient families
Emma is a key player in the CLCN4 community, and has lead the development of (new) clinical guidelines for treating patients with CLCN4 mutations with Vera and others. She remains a major initial point of contact for many GPs and clinicians who come across patients with CLCN4 mutations after genetic screening for undiagnosed diseases. Therefore, gaining knowledge about new variants, and testing their actual vs predicted significance for function and diagnosis and patient stratification, is key.
In this talk Emma outlined the typical rare disease family experience and clinical journey, patient and symptom progression, and the challenges remaining.
CLCN4 is a complex syndrome, with some sex-specific differences. Common symptoms involve:
Yann Herault – Animal models of CLCN4
Yann (Inserm) is a key part of the CLCN4 foundation, successfully creating several mouse and rat models of CLCN4 mutations for translational disease modelling and pharmacological screening; this works occurs alongside development of iPSC cell-based models of CLCN4 condition (e.g. by South Korean researchers featured on day 2).
The interesting issue is that the chromosome location of CLCN4 and neighbouring genes is more homologous in rat than mouse, so he decided to make global (exon 4 STOP) and Cre/Lox conditional CLCN4-/- knockout rats for more specific disease modelling, as well as a g-o-f A55V (A449V in rat) knock-in rat (using CRISPR). They have stablished breeding colonies for all 3 rat models and these are now available for collaborators, for free. Also significant is the fact that previous CLCN3 and CLCN4 knockout mice models show no gross symptoms, although other talks at this conference do report some subtle CNS deficits in CLCN4 knockout mice (e.g. Raul Guzman’s work on pyramidal dendrite morphology and function, South Korean autism researcher Yeni Kim).
Anselm Zdebik - Zebrafish CLCN4 model
Anselm works at UCL, but was previously in Thomas Jentsch’s lab working on CLCN channels. He is currently trying to develop zebrafish loss-of-function (V536M) and gain-of-function (A555V) CLCN4 genetic models, as many symptoms can be measured in this species (psycholocomotor, seizure, cerebral atrophy, anxiety, depression, autism). He is using a CRISPresso technique for zebrafish genetic editing; work is still in progress. He previously made a Kv (KCNJ10) channel knockout for electrophysiological and behavioural studies in zebrafish.
Joseph Mindell (NINDS) – CLC7 gain-of-function mutations in lysosomal storage disease
Joseph is another leading academic researcher in the CLC community and well known to senior participants. He became interested in this channel family as a postdoc in Chris Miller’s lab that studied CLC channels. His speciality is lysosomal disease, and includes study of CLC7 mutants and patient families; CLC7/Ostim1 is implicated in bone disease, but he didn’t talk about this today. Jo favours a pH-sensitive fluorescent dye assay (Oregon Green dextran) to study endosome function, as well as whole-cell and single channel patch clamp recordings from CLC channels expressed in heterologous cells. His group has discovered several new patient families with a g-o-f mutation in CLC7 (Y715C) which affect lysosome acidification, resulting in hypopigmentation (dysfunction in melanocytes) and systemic lysosome storage disease. Three new patients have similar g-o-f mutations but milder disease phenotypes. His recent work suggests that these g-o-f mutations affect the interaction of lysosome-specific phosphoinositide PI(3,5)P2 with CLC7 proteins, which normally inhibits transport activity.
Day 2:
CNS electrophysiology and pharmacology
Jinju Han (KAIST, Korea) – Excitatory neuron survival
The key data in this talk is the development and study of human ESC CLCN4 mutant and knockout stem cell lines, neural progenitors, and (immature) neurons to create translational disease-in-a-dish reagents and assays. Stem cell models using immature human CNS neurons may be more relevant for CLCN4 syndrome than many other CNS diseases (e.g. schizophrenia, depression, neurodegeneration) as CLCN4 is a neuro-developmental syndrome that affects infants and young children, rather than adults. However, the use of ES rather than iPS cells may affect drug discovery efforts outside of Korea due to ethical issues and regulatory requirements in the US, EU and UK.
Introducing the A555V g-o-f function into stem cell neurons actually decreased CLCN4 mRNA expression by 50%, and lead to rapid cell death (TUJ1+ neurons), indicating severe effects on neurogenesis, especially for excitatory neurons (identified by scRNA transcriptome profiling). Similar effects were seen in 2D hES and 3D organoid mini-brains, with altered dendrite morphology. It may be possible that this loss of excitatory neurons could impact neuronal development and cognition and behaviours, but it remains unclear how a similar mechanism could underlie the seizures that are prominent in CLCN4 patients.
Yeni Kim (Dongguk University, Korea) – Synaptic dendrite function and developmental deficit
Yeni works with Jinju Kim. The key finding here is that the anti-psychotic Risperidone, a mixed DA and 5-HT GPCR modulator, can reverse or ameliorate the CNS neuron morphological and behavioural effects of CLCN4 knockout mice. Previous studies found no gross neurological symptoms in knockout mice, but her group has identified autism-related effects on social interaction, anxiety and repetitive behaviour in 7-8 week old knockout mice. As CLCN4 patients have neuro-developmental deficits and some key features of autism, she postulated that an agent effective in her models of autism may also work for the CLCN4 condition. Knockout of CLCN4 differentially affected cortical expression of genes associated with neuronal plasticity and synapses (PSD95, CDK5), and dendritic morphology (Golgi staining). Risperidone could ‘correct’ these genetic deficits, as well as rescue (excitatory) neuron differentiation, dendritic and synaptic morphology, and behavioural symptoms.
The relevance of a complete and global CLCN4 knockout in mice to patients with gene-modifying mutations that may reduce or increase channel function is not clear, as-is the mechanism of Risperidone action or the signalling cascades affected, so further work is clearly required and is underway with human ES neuron translational models. Also, there was no link shown between these CNS and behavioural effects and the known function of CLCN4 to regulate organellar patho-physiology, although in their talk at the 2020 CLCN4 Symposium they said they saw increased LysoTracker dye labelling, suggesting increased organellar cycling, perhaps due to process and degrade dying neurons?
CLC channel biophysics and function:
Thomas Jentsch (FMP Berlin) - Physical and functional interactions of CLC4 and CLC3
Thomas is famous in the field for cloning the first CLC channel gene in 1990 (CLC1 in Torpedo electric organ), and several leading researchers in the current field are linked to his laboratories in Germany. He has looked at a range of CLC and VRAC anion channels, and their channelopathies and functions in various and varied human diseases.
The fact that human patients with single CLC gene deletions and mutations manifest significant (neurological) symptoms suggest that the CLC gene family members cannot fully compensate for each other, likely as each of them are expressed and function in different organellar compartments and pH states of the endo-lysosomal pathway (see Figure below). CLC3, 4 and 5 seem able to form heterodimers, so knockout of CLC3 probably causes more serious neurological symptoms than CLCN4 knockouts as the former also leads to reductions in CLCN4 protein, but the reverse is not the case as CLC3 is a dominant negative partner, and CLCN4 stability, trafficking and lysosomal function depends on CLC3 but not the other way round. Significantly, uncoupling or deletion of CLC3 H+ transport in knockout mice only leads to neurological dysfunction in the absence of CLCN4, suggesting the latter can contribute a functional transport function if it forms a dimer with non-functional CLC3 proteins, which can still help with protein complex stability and trafficking.
Taken from Jentsch & Pusch (2018): 10.1152/physrev.00047.2017
Thomas reiterated the theme that in many mouse CLC gene knockouts there are few discernible neurological deficits, and he (but not all others) employs an extensive battery of tests and assays (e.g. elevated maze, Barnes maze and open field for anxiety, home cage activity and sleep: wake cycle) to detect behavioural deficits. The only difference between wild type and CLC4 mutant mice was in a social interaction assay (in a test tube), which may correlate with the autism symptoms detected by South Korean researchers.
Merritt Maduke (Stanford) – CLC proton exchange transporter structure-function
Merritt runs a molecular and cellular physiology lab, where she uses a prokaryotic CLC gene and protein model. Her lab has obtained several new and detailed cryo-EM views, as well as carried extensive mutational analysis, to refine molecular dynamic models to try and determine each of the 4 steps of the Cl-/H+ counter-transport cycle, disproving several previous models and proposed protein conformations. These 4 states or configurations require the Glu residues in the permeation pathway to be in the ‘up’, ‘middle’ or ‘down’ orientation, as well as a new ‘out’ position.
During questions, it was asked whether CLCN researchers had used Google’s AlphaFold software to predict and model CLCN4 protein structure-function and the potential impact and effect of patient mutations. Several speakers replied that they had tried this, but it turns out that the crucial dimer stoichiometry of functional CLC proteins is not well modelled by this software (which was surprising and interesting to hear). Thus, it would appear as though there is still a real bonus to having empirical x-ray and cryo-EM structures and to use these in detailed computational MD studies of ion channel structure-function, rather than rely on purely in silico techniques.
Cecilia George (Nanion) – Investigating CLCs using APC methods
Cecilia showed the attendees how they could augment their current electrophysiology and imaging methods to include automated patch clamp (APC) and Solid State Membrane (SSM) technology and platforms for their CLC and lysosomal disease research, and how these could help them span the gap between basic research and drug discovery. Her talk revealed recent high quality APC data from CLC-1 channels (on the SP384 platform) and CLC-7 channel/transporter recordings (on the Surfe2R SSM platform), and included some data from native, purified lysosome membranes rather than plasma membranes from (mutated) over-expression systems. This is an important feature, as there are marked differences in the membrane lipid composition and protein-interacting partners between organellar and plasma membranes which can affect biophysical behaviour and pharmacology, with important implications for disease modelling and drug discovery efforts.
Rare disease patient organisations:
It is becoming clear in the field of channelopathies and rare disease that there is an increasing role for patient recognition, and family and charity lobbying, fundraising and involvement in the drug discovery process. Academics and clinicians already tap into these resources through diagnosis and treatment, especially given the rapid rise and reduced cost of genome sequencing, and parents in turn are more active in searching for specialists and treatments from the literature and online after their child is diagnosed (sometimes after several frustrating years without much useful information). It is also crucial to acknowledge the potential for drug discovery companies to also get involved with rare disease individuals, families and organisations, as many in this space are now doing. Several talks on Day 1 and Day 2 addressed some of these issues, but more still needs to be done.
Also worth mentioning is that fact that Gina Tan (mother of Daphne and wife and co-founder with Peter Trill of the CLCN4 Foundation) started on her rare disease journey after Daphne’s diagnosis with a single Facebook post that initially connected her to 5 other CLCN4 families. This group was found by the clinician Emma Palmer in Sydney who was looking for new patients and mutations, and has now grown to a support group with over 120 families and an extensive network of academics and clinicians, and an established rare disease charity based in the UK but with connections to groups around the world.
Paul Wang – Simon’s Searchlight Foundation
Paul provided an overview of Simon’s Searchlight Foundation, an organisation based in the US which deals mostly with rare disease patient groups. Paul is a neurodevelopmental disease clinician who worked in industry at Pfizer, before getting involved in clinical trial advocacy and rare disease patient groups. Simon’s Searchlight is a charity set up to help bring together the various stakeholders in rare disease, but especially to help patient families and foundations put together the necessary toolbox of experts, databases and reagents to facilitate the search and clinical testing of new treatments for rare disease patients. Their services include patient and family contact lists and co-ordinators, family and medical history surveys, blood and cell sampling to allow creation of iPSC disease models, development of cell line and animal genetic models and reagents (e.g. in academic labs, or by commercial CRO partners paid for by the charities and founders), through to clinical trial design and execution. A current focus with CLCN4 patients is to acquire clinical EEG recordings, as epilepsy is a common symptom but this data can also help assess other neurological changes, and may provide a biomarker for drug trials. Currently the Foundation is working on over 170 rare disease genes and thousands of patients and families involved with neurodevelopment disease.
Jessica Duis (Colorado Children’s Hospital) – Clinical trial readiness for rare disease
Jessica is involved with Angelman, Prader-Willi, and Pitt Hopkins syndrome patients and families, all of which are rare neurological conditions. She is a clinician who has led and advised investigator-driven clinical trials and now also works with industry to design better rare disease clinical trials and regulatory filings. The main issues are the small patient populations and difficulty in profiling, classifying and stratifying patients to enable a suitably robust and statistically powered trial (more likely a small Ph II than a larger Ph III), and the need for new outcome measures and biomarkers to be developed, validated, and approved by the FDA.
For Angelman patients she helped develop a ‘disease concept’ model using data from care givers, patients and their families, and clinicians to identify the most disruptive symptoms, so that these can be specifically addressed in treatment regimes and drug discovery research and clinical testing. These highlighted speech and motor problems, as well as seizures. Detailed case histories also identified changes in gait, which could then be correlated with genetic changes and aid in genotyped patient classification.
Dr Pangkong Fox - CACNA1 Foundation
Dr Fox’s son was diagnosed with a mutation in the Cav2.1 ion channel gene in 2021, and after looking for information online she found the CACNA1 Foundation and by 2022 she was their Scientific Engagement Director. She has a PhD in Cell and Molecular Biology and uses her scientific background to create education support and resources. Cav2.1 mutations are linked to various forms of ataxia, migraine and epilepsies, as well as neurodevelopmental conditions such as intellectual disability and autism. There is a complex geno-phenotype with over 300 pathogenic mutations that lead to loss- or gain-of-function. The CACNA1A Foundation is doing many of the same things as the CLCN4 Foundation and Simon’s Searchlight (via their Research network) to develop a ‘preclinical toolbox’ of patient histories, biobanking, iPS and heterologous cell line generation, animal models, etc. These are and will be used for drug re-purposing screens and the search for new treatments (small molecules, ASOs, gene therapy). They are utilising collaborations with many other charities and disease and research institutions to achieve this cost-effectively and quickly (e.g. NORD, RareX, Citizen Invitae platform), as well as working with industry partners (e.g. Metabolom in Ireland for biomarkers), and receiving funding from US agencies and organisations (e.g. Chan Zuckerberg Initiative).
Many other ion channel channelopathy parents have followed this pathway, but without an established rare disease organisation they created their own charity, advocacy group or foundation – examples include KCNT1, KCNH10, CLCN4, SCN2A, etc. These groups help to identify and bring together patients, academics, clinicians and industry to facilitate collaboration, raise awareness, increase understanding of the disease and patient and family needs, discover new treatments, and raise funds.
A Family Story
Whilst the main patient and family engagement session on Day 2 was limited to the key clinicians, attendees were able to gain a personal insight into what it was like to have a child diagnosed with CLCN4 syndrome in a short online session with a mum in the Bay Area. Her child was first taken to the GP at 6 months with feeding problems and at 12 months with developmental delays, but was not diagnosed until after genotyping and neurological clinical profiling at 2 years of age. It was a very touching presentation – the child was asleep I think, but we were shown family videos and also some slides that the Mum had created about disease progression, symptoms, and family challenges and highlights. This was a very moving, personal and emotionally affecting presentation.
There was a similar video montage and slideshow of patient family pictures and information on Day 1 (also shown at the 2022 Symposium) illustrating patients and families who were diagnosed at varied ages from months-years-decades old, spread across the globe. This type of patient family insight is a very powerful way to bring R&D into personal focus.
Drug Discovery case studies:
Michael Schwake (University of Bielefeld) - Parkinson’s Disease affects lysosomal function
Michael is another alumni from the Jentsch lab, where he studied CLC5 in Dent kidney disease. He is now working on Parkinson’s Disease, which can be classified as an Inherited Disease of membrane Trafficking (IDT) as there is genetic linkage to TMEM175 and TRPML1 lysosomal channels, as well as to GCase enzyme involved in autophagy and endo-lysosomal cycling and function. Several academic and industry groups are working on GCase treatments for PD, but he noted that while GCase activity is reduced in the brains of PD patients and animal models, there are also PD patient mutations which show no effect on enzyme activity. I wasn’t aware of this but Ambroxol (a non-selective Nav channel inhibitor, cough medicine) is being tested clinically for PD as it has been shown to act as a chaperone for GCase, to boost protein trafficking and function.
Michael is working on human patient-derived iPS neuron models of PD, differentiating TH+ dopa neurons; he stressed the importance of using isogenic controls to confirm disease and gene differences, which he creates using CRISPR gene editing (baculovirus vectors being more effective than other expression methods). His models replicate the loss of GCase expression and activity and formation of a-synuclein fibrils, and also found that another lysosomal protein LIMP2 is reduced in PD. Significantly, he found that LIMP2 activates GCase and it’s associated protein GBA1, so loss of LIMP2 may lead to reductions in GCase in PD. His group is developing cell-penetrating peptides of the LIMP2 helix to promote GCase and GBA1 activity and function as a novel therapeutic approach for PD. The obvious corollary for potential CLCN4 disease treatments is to use small molecules or genetic treatments to modulate upstream regulators as well as the CLCN4 gene and protein itself, such as CLC3 and other proteins which may augment CLCN4 trafficking and activity.
Laksh Aithani - Charm Therapeutics
Charm Therapeutics is a London start-up with US connections; Laksh previously worked at leading AI drug discovery company Exscientia (UK), and the other co-founder is a US academic and entrepreneur focused on in silico deep learning protein folding applications. Other team members have come from Google DeepMind and big pharma.
Their main argument is that genetic sequences do not reliably predict 2D or 3D protein structure (i.e. the AlphaFold approach), so high resolution protein crystal structures should be used; additionally, ligand binding can alter protein structures (i.e. apo state may not be useful for drug discovery).
They have developed a fast deep learning algorithm called DragonFold that is trained on the 25% of PDB crystal structures with ligands bound (GPCRs, ion channels) and which can be used in a virtual screening workflow to create virtual crystals of the target protein and small molecule libraries, using supercomputer hardware (100 NVIDIA GPUs) and optimised AI/ML software.
Their therapeutic interest is mainly oncology and Laksh presented 2 case studies of small virtual screens against a kinase and a nuclear polymerase. Only 100-500 compounds could be virtually bound owing to computing and cost limitations, with each screen identifying a handful of hits with improved selectivity or mM potency respectively, which were validated in functional binding and biological assays.
Marlen Lauffer (Dutch Centre for RNA Therapeutics) - ASO genetic treatments for rare disease
I really enjoyed Marlen’s talk and her enthusiasm to find effective treatments for patients suffering from ultra-rare diseases, for which academic drug discovery is too slow and commercial interest is extremely limited, and any drug to reach market would be extremely expensive.
Her group is a non-profit consortium supported by the Dutch government and EU funding to rapidly find treatments for rare neurodegenerative disease, as effective drugs need to be delivered to young children before life-limiting symptoms take hold. To this end, they are using patient-derived fibroblasts to directly-induce human stem cell iNeurons, as traditional iPS engineering techniques are much slower and more expensive. Their therapeutic focus is on RNA antisense oligonucleotides (ASOs) to modulate faulty gene expression through exon skipping, RNAi or allele-specific mechanisms. She acknowledged that these approaches work best for g-o-f mutants, which have now been identified in a small subset of CLCN4 patients, but others have shown it can work to up-regulate gene and protein expression as well which may be suitable for treating l-o-f CLCN4 mutations as well. Rather than trying to group related patients with similar mutant genotypes for traditional small clinical trials, she suggested that defining each patient as their own population enabled N=1 Ph 0 clinical trials with fewer regulatory requirements, essentially breaking down each rare disease into multiple ‘nano’ rare diseases (also called 1 Mutation 1 Medicine, 1M1M, or N=1 diseases). ASOs can be delivered locally by i.c.v, i.t or injection into the eye, but it is unclear if a single application (injection or transgenic expression) can ‘cure’ a disease (one-and-done), or whether repeated applications may be needed during a child’s development.
This is a high risk and revolutionary approach; there have been recent patient deaths with such ultra-personalised genetic approaches (e.g. a DMD patient family-funded effort in the US), but Marlen is using the example of Milasen which was an ASO developed and tested to treat a single young female patient with Batten’s disease (CLC lysosomal storage disorder). She presented a case study on a patient with the retinal disease Stargardt Syndrome, with ASOs designed to correct intron/exon splicing that are being tested in iPS retinal organoids. I was concerned by the fact that limited safety testing is carried out with these ASOs (as many assume they are incredibly specific and selective, and thereby safe). For example, limited toxicology testing is done in non-human species, which presumably have different gene sequences to the human protein, so the action and efficacy of human-specific ASOs on the target protein and it’s ‘ome partners and signalling cascades may be reduced and under-estimated. For example, several pharma groups and CRO companies (e.g. Amgen, Metrion Biosciences) have shown that siRNA and ASOs and other genetic modulators can have unanticipated safety pharmacology effects, such as long-term (rather than acute) effects on hERG channel activity. The FDA and EMEA are aware of thee issues and may be increasing the threshold for safety testing of such RNA treatments, but as with oncology the seriousness of the disease and small patient numbers (in genetically stratified patients) may allow a more relaxed approval process.
Conclusion
In conclusion, this key conference promoted collaboration and rare disease patient and family support and participation, clearly showing that a varied and muti-disciplinary approach to channelopathy drug discovery can lead to tangible results. Above all there was a feeling of satisfaction that contributions, however small, are not going un-noticed and there is an overwhelming number of people and companies who are dedicating their drug discovery endeavours to accelerate treatments for rare conditions such as CLCN4.
Guest blog within the series ‘Ion channel drug discovery insights from the Channelogist’ (@Albion Drug Discovery Services) for Metrion Biosciences
Introduction
Cure CLCN4 is a charity founded by a patient family to provide support, raise awareness and fund medical research for effective treatments for CLCN4, a rare genetic condition which causes intellectual disability, behavioural and movement disorders and microcephaly. Each year, this important charity holds a conference to highlight the condition, bring together thought leaders in the field (including international academics and clinicians), provide input from patients and their families, and share research insights with the aim of discovering a cure for this rare disease and generating a strong patient-orientated research community and support network.
A link to this years’ conference can be found below:
This years’ event, held earlier this Summer, was organised by Paul McGuiness and Rebeca Ridings Figuero, alongside Dr. Peter Trill, a UK bio-entrepreneur and his wife Dr. Gina Tan, a clinician who founded Cure CLCN4 as their youngest daughter Daphne has a mutation in the CLCN4 gene and suffers from neuro-developmental symptoms. There were around 40 in-person attendees in addition to a varied online delegation of ~30 attendees, plus a closed session focused solely on patient families.
Selected talks are available in video here:
https://www.youtube.com/playlist?list=PLMK5QcuaGueUK3Yea-qiHdc-Ch7wRgrH1
Scientific background
CLCN4-related neurodevelopmental disorder (CLCN4-NDD) is a rare X-linked genetic condition associated with intellectual disability, psychiatric disorders, gastrointestinal issues and epilepsy. The extent and severity of the condition is extremely variable and dependent upon the type of gene change and the gender of the individual, amongst other factors.
CLCN4-NDD is caused by mutations of the CLCN4 gene encoding the CLC-4 channel which is expressed across various tissue types, but most prominently in the brain and skeletal muscle. Significantly, CLC4 is expressed on endosomes and the ER, and likely interacts with related CLC-3 channels which facilitate sorting of CLC-4 proteins into endosomal compartments.
Key points:
The importance of diverse intracellular organellar ion channels in a wide variety of diseases is now gaining greater attention (e.g. TMEM175 in Parkinson’s Disease), making them a new class of ion channel drug discovery target. Cure CLCN4 is also a great example of the growing trend for patient families to get involved in finding treatments for such rare disease channelopathies.
A summary of the talks presented across the two days is provided below:
Vera Kalscheuer - Molecular Genetics of CLCN4
Vera is one of the original and leading academics studying CLCN4 genetics (through her focus on X-linked genetic disorders) and is a major part of the CLCN4 foundation. Her group has identified many examples of patients and families with CLCN4 mutations, expanding the picture to include de novo and inherited missense mutations in females and males, respectively. She co-authored a comprehensive review in 2023 (Palmer et al., 2023) that brought together many international researchers and highlighted the emerging genetic diversity in CLCN4 patients. Her previous publications have attracted the attention of other clinicians trying to diagnose and treat rare CLCN4 patients scattered across the globe; Elizabeth Emma Palmer is an Australian clinical researcher who is now also a big part of the CLCN4 collaborative network and another major force in the CLCN4 foundation.
Vera and collaborators now have a growing database of CLCN4 gene mutations from the ClinVar database and publications, which currently shows 90 variants from 49 families which comprise 41 unique (typically de novo) and 18 recurrent (typically missense) mutations which have been predicted to be benign (1), pathogenic (36), or a variant of unknown significance (VUS, 53) which are the most perplexing and require further work to determine if/how they affect CLCN4 protein function, channel activity and patient symptoms. Michael Pusch is working through many of these mutations using Xenopus oocyte TEVC electrophysiology (next talk), and Vera highlighted that many of the in silico predictions for new mutations do not match with functional studies: for example, 30 of the 90 known variants have predicted ‘low’ or ‘benign’ effect, but 13 of these were shown to be either loss- or gain-of-function on CLCN4 currents.
The new data is also significant for showing g-o-f mutations, as CLCN4 was previously thought to be l-o-f (missense and truncating mutations, especially in females);
In males they see truncating l-o-f mutations (de novo, inherited), as well as new g-o-f variants (e.g. Val317x residue);
In females, de novo & inherited mutations are ALL missense, and there are new g-o-f variants (Ala555 residue)
Interestingly, although CLCN4 is an-X-linked genetic disease expected to be limited to females and inherited from the maternal line, males can both pass on inherited mutations and be affected by de novo mutations as X chromosome inactivation can be incomplete, and genome imprinting can occur from the fertilised (i.e. male XY) egg. However, de novo mutations are much less common in males (27%) than females (78%). 90% of variants are missense, and all affected females carry such mutations as truncating or intragenic chromosomal deletions appear to be compensated for by the un-affected X chromosome allele.
Michael Pusch – Functional studies of CLCN4 channel mutants
His interest initially stemmed from the fact he trained in Prof. Thomas Jensch’s lab, who cloned the first member of the CLC family. His specialist contribution is Xenopus oocyte electrophysiology of CLCN4 mutants.
There is a major distinction between members of the CLCN family; CLC-1, CLC-2 and the two kidney and inner ear CLC-K variants are ion channels expressed on the plasma membrane where they regulate anion transport, membrane potential and counter-ion movements. The other members (CLC-3, 4, 5, 6 and 7) are expressed in intracellular organelles (endosomes and lysosomes) where they act predominantly as co-transporters (exchangers) of Cl- and H+ ions to regulate pH, membrane potential and counter-ion permeability. There are rare human channelopathy diseases associated with 7/8 members of the CLCN family.
Each of the organellar CLC proteins are specialised/localised to different endocytotic, sorting and recycling endosomes and degradative lysosomes, where they all help to control pH and function.
Interestingly, CLC proteins seem to form channel dimers, with each subunit contributing an independent pathway for the exchanger transport of Cl- and protons (similar to dual ‘pore’ Fluc fluoride anion channels).
CLC4 may be unique in the CLCN family as its function, but certainly its trafficking and localisation to endosomes from the ER, requires heteromerization with CLC3 proteins (other CLCs homodimerise). Previous electrophysiology work showed all/most CLCN4 mutations were loss of function, but newer data shows that some (pathological) mutations are like wildtype, while other predicted benign variants are actually loss-of-function, and some rare new mutations are in fact gain-of-function.
New gain-of-function variants, revealed by studying electrophysiology with altered (acidic, i.e. physiological) pH would be lethal in males, so are only seen in females. Further biophysical work shows that some loss-of-function mutations actually shift voltage-dependence, but could have a dominant negative effect in heteromers.
Emma Palmer – CLCN4 clinical genetics and rare patient families
Emma is a key player in the CLCN4 community, and has lead the development of (new) clinical guidelines for treating patients with CLCN4 mutations with Vera and others. She remains a major initial point of contact for many GPs and clinicians who come across patients with CLCN4 mutations after genetic screening for undiagnosed diseases. Therefore, gaining knowledge about new variants, and testing their actual vs predicted significance for function and diagnosis and patient stratification, is key.
In this talk Emma outlined the typical rare disease family experience and clinical journey, patient and symptom progression, and the challenges remaining.
CLCN4 is a complex syndrome, with some sex-specific differences. Common symptoms involve:
Yann Herault – Animal models of CLCN4
Yann (Inserm) is a key part of the CLCN4 foundation, successfully creating several mouse and rat models of CLCN4 mutations for translational disease modelling and pharmacological screening; this works occurs alongside development of iPSC cell-based models of CLCN4 condition (e.g. by South Korean researchers featured on day 2).
The interesting issue is that the chromosome location of CLCN4 and neighbouring genes is more homologous in rat than mouse, so he decided to make global (exon 4 STOP) and Cre/Lox conditional CLCN4-/- knockout rats for more specific disease modelling, as well as a g-o-f A55V (A449V in rat) knock-in rat (using CRISPR). They have stablished breeding colonies for all 3 rat models and these are now available for collaborators, for free. Also significant is the fact that previous CLCN3 and CLCN4 knockout mice models show no gross symptoms, although other talks at this conference do report some subtle CNS deficits in CLCN4 knockout mice (e.g. Raul Guzman’s work on pyramidal dendrite morphology and function, South Korean autism researcher Yeni Kim).
Anselm Zdebik - Zebrafish CLCN4 model
Anselm works at UCL, but was previously in Thomas Jentsch’s lab working on CLCN channels. He is currently trying to develop zebrafish loss-of-function (V536M) and gain-of-function (A555V) CLCN4 genetic models, as many symptoms can be measured in this species (psycholocomotor, seizure, cerebral atrophy, anxiety, depression, autism). He is using a CRISPresso technique for zebrafish genetic editing; work is still in progress. He previously made a Kv (KCNJ10) channel knockout for electrophysiological and behavioural studies in zebrafish.
Joseph Mindell (NINDS) – CLC7 gain-of-function mutations in lysosomal storage disease
Joseph is another leading academic researcher in the CLC community and well known to senior participants. He became interested in this channel family as a postdoc in Chris Miller’s lab that studied CLC channels. His speciality is lysosomal disease, and includes study of CLC7 mutants and patient families; CLC7/Ostim1 is implicated in bone disease, but he didn’t talk about this today. Jo favours a pH-sensitive fluorescent dye assay (Oregon Green dextran) to study endosome function, as well as whole-cell and single channel patch clamp recordings from CLC channels expressed in heterologous cells. His group has discovered several new patient families with a g-o-f mutation in CLC7 (Y715C) which affect lysosome acidification, resulting in hypopigmentation (dysfunction in melanocytes) and systemic lysosome storage disease. Three new patients have similar g-o-f mutations but milder disease phenotypes. His recent work suggests that these g-o-f mutations affect the interaction of lysosome-specific phosphoinositide PI(3,5)P2 with CLC7 proteins, which normally inhibits transport activity.
Day 2:
CNS electrophysiology and pharmacology
Jinju Han (KAIST, Korea) – Excitatory neuron survival
The key data in this talk is the development and study of human ESC CLCN4 mutant and knockout stem cell lines, neural progenitors, and (immature) neurons to create translational disease-in-a-dish reagents and assays. Stem cell models using immature human CNS neurons may be more relevant for CLCN4 syndrome than many other CNS diseases (e.g. schizophrenia, depression, neurodegeneration) as CLCN4 is a neuro-developmental syndrome that affects infants and young children, rather than adults. However, the use of ES rather than iPS cells may affect drug discovery efforts outside of Korea due to ethical issues and regulatory requirements in the US, EU and UK.
Introducing the A555V g-o-f function into stem cell neurons actually decreased CLCN4 mRNA expression by 50%, and lead to rapid cell death (TUJ1+ neurons), indicating severe effects on neurogenesis, especially for excitatory neurons (identified by scRNA transcriptome profiling). Similar effects were seen in 2D hES and 3D organoid mini-brains, with altered dendrite morphology. It may be possible that this loss of excitatory neurons could impact neuronal development and cognition and behaviours, but it remains unclear how a similar mechanism could underlie the seizures that are prominent in CLCN4 patients.
Yeni Kim (Dongguk University, Korea) – Synaptic dendrite function and developmental deficit
Yeni works with Jinju Kim. The key finding here is that the anti-psychotic Risperidone, a mixed DA and 5-HT GPCR modulator, can reverse or ameliorate the CNS neuron morphological and behavioural effects of CLCN4 knockout mice. Previous studies found no gross neurological symptoms in knockout mice, but her group has identified autism-related effects on social interaction, anxiety and repetitive behaviour in 7-8 week old knockout mice. As CLCN4 patients have neuro-developmental deficits and some key features of autism, she postulated that an agent effective in her models of autism may also work for the CLCN4 condition. Knockout of CLCN4 differentially affected cortical expression of genes associated with neuronal plasticity and synapses (PSD95, CDK5), and dendritic morphology (Golgi staining). Risperidone could ‘correct’ these genetic deficits, as well as rescue (excitatory) neuron differentiation, dendritic and synaptic morphology, and behavioural symptoms.
The relevance of a complete and global CLCN4 knockout in mice to patients with gene-modifying mutations that may reduce or increase channel function is not clear, as-is the mechanism of Risperidone action or the signalling cascades affected, so further work is clearly required and is underway with human ES neuron translational models. Also, there was no link shown between these CNS and behavioural effects and the known function of CLCN4 to regulate organellar patho-physiology, although in their talk at the 2020 CLCN4 Symposium they said they saw increased LysoTracker dye labelling, suggesting increased organellar cycling, perhaps due to process and degrade dying neurons?
CLC channel biophysics and function:
Thomas Jentsch (FMP Berlin) - Physical and functional interactions of CLC4 and CLC3
Thomas is famous in the field for cloning the first CLC channel gene in 1990 (CLC1 in Torpedo electric organ), and several leading researchers in the current field are linked to his laboratories in Germany. He has looked at a range of CLC and VRAC anion channels, and their channelopathies and functions in various and varied human diseases.
The fact that human patients with single CLC gene deletions and mutations manifest significant (neurological) symptoms suggest that the CLC gene family members cannot fully compensate for each other, likely as each of them are expressed and function in different organellar compartments and pH states of the endo-lysosomal pathway (see Figure below). CLC3, 4 and 5 seem able to form heterodimers, so knockout of CLC3 probably causes more serious neurological symptoms than CLCN4 knockouts as the former also leads to reductions in CLCN4 protein, but the reverse is not the case as CLC3 is a dominant negative partner, and CLCN4 stability, trafficking and lysosomal function depends on CLC3 but not the other way round. Significantly, uncoupling or deletion of CLC3 H+ transport in knockout mice only leads to neurological dysfunction in the absence of CLCN4, suggesting the latter can contribute a functional transport function if it forms a dimer with non-functional CLC3 proteins, which can still help with protein complex stability and trafficking.
Taken from Jentsch & Pusch (2018): 10.1152/physrev.00047.2017
Thomas reiterated the theme that in many mouse CLC gene knockouts there are few discernible neurological deficits, and he (but not all others) employs an extensive battery of tests and assays (e.g. elevated maze, Barnes maze and open field for anxiety, home cage activity and sleep: wake cycle) to detect behavioural deficits. The only difference between wild type and CLC4 mutant mice was in a social interaction assay (in a test tube), which may correlate with the autism symptoms detected by South Korean researchers.
Merritt Maduke (Stanford) – CLC proton exchange transporter structure-function
Merritt runs a molecular and cellular physiology lab, where she uses a prokaryotic CLC gene and protein model. Her lab has obtained several new and detailed cryo-EM views, as well as carried extensive mutational analysis, to refine molecular dynamic models to try and determine each of the 4 steps of the Cl-/H+ counter-transport cycle, disproving several previous models and proposed protein conformations. These 4 states or configurations require the Glu residues in the permeation pathway to be in the ‘up’, ‘middle’ or ‘down’ orientation, as well as a new ‘out’ position.
During questions, it was asked whether CLCN researchers had used Google’s AlphaFold software to predict and model CLCN4 protein structure-function and the potential impact and effect of patient mutations. Several speakers replied that they had tried this, but it turns out that the crucial dimer stoichiometry of functional CLC proteins is not well modelled by this software (which was surprising and interesting to hear). Thus, it would appear as though there is still a real bonus to having empirical x-ray and cryo-EM structures and to use these in detailed computational MD studies of ion channel structure-function, rather than rely on purely in silico techniques.
Cecilia George (Nanion) – Investigating CLCs using APC methods
Cecilia showed the attendees how they could augment their current electrophysiology and imaging methods to include automated patch clamp (APC) and Solid State Membrane (SSM) technology and platforms for their CLC and lysosomal disease research, and how these could help them span the gap between basic research and drug discovery. Her talk revealed recent high quality APC data from CLC-1 channels (on the SP384 platform) and CLC-7 channel/transporter recordings (on the Surfe2R SSM platform), and included some data from native, purified lysosome membranes rather than plasma membranes from (mutated) over-expression systems. This is an important feature, as there are marked differences in the membrane lipid composition and protein-interacting partners between organellar and plasma membranes which can affect biophysical behaviour and pharmacology, with important implications for disease modelling and drug discovery efforts.
Rare disease patient organisations:
It is becoming clear in the field of channelopathies and rare disease that there is an increasing role for patient recognition, and family and charity lobbying, fundraising and involvement in the drug discovery process. Academics and clinicians already tap into these resources through diagnosis and treatment, especially given the rapid rise and reduced cost of genome sequencing, and parents in turn are more active in searching for specialists and treatments from the literature and online after their child is diagnosed (sometimes after several frustrating years without much useful information). It is also crucial to acknowledge the potential for drug discovery companies to also get involved with rare disease individuals, families and organisations, as many in this space are now doing. Several talks on Day 1 and Day 2 addressed some of these issues, but more still needs to be done.
Also worth mentioning is that fact that Gina Tan (mother of Daphne and wife and co-founder with Peter Trill of the CLCN4 Foundation) started on her rare disease journey after Daphne’s diagnosis with a single Facebook post that initially connected her to 5 other CLCN4 families. This group was found by the clinician Emma Palmer in Sydney who was looking for new patients and mutations, and has now grown to a support group with over 120 families and an extensive network of academics and clinicians, and an established rare disease charity based in the UK but with connections to groups around the world.
Paul Wang – Simon’s Searchlight Foundation
Paul provided an overview of Simon’s Searchlight Foundation, an organisation based in the US which deals mostly with rare disease patient groups. Paul is a neurodevelopmental disease clinician who worked in industry at Pfizer, before getting involved in clinical trial advocacy and rare disease patient groups. Simon’s Searchlight is a charity set up to help bring together the various stakeholders in rare disease, but especially to help patient families and foundations put together the necessary toolbox of experts, databases and reagents to facilitate the search and clinical testing of new treatments for rare disease patients. Their services include patient and family contact lists and co-ordinators, family and medical history surveys, blood and cell sampling to allow creation of iPSC disease models, development of cell line and animal genetic models and reagents (e.g. in academic labs, or by commercial CRO partners paid for by the charities and founders), through to clinical trial design and execution. A current focus with CLCN4 patients is to acquire clinical EEG recordings, as epilepsy is a common symptom but this data can also help assess other neurological changes, and may provide a biomarker for drug trials. Currently the Foundation is working on over 170 rare disease genes and thousands of patients and families involved with neurodevelopment disease.
Jessica Duis (Colorado Children’s Hospital) – Clinical trial readiness for rare disease
Jessica is involved with Angelman, Prader-Willi, and Pitt Hopkins syndrome patients and families, all of which are rare neurological conditions. She is a clinician who has led and advised investigator-driven clinical trials and now also works with industry to design better rare disease clinical trials and regulatory filings. The main issues are the small patient populations and difficulty in profiling, classifying and stratifying patients to enable a suitably robust and statistically powered trial (more likely a small Ph II than a larger Ph III), and the need for new outcome measures and biomarkers to be developed, validated, and approved by the FDA.
For Angelman patients she helped develop a ‘disease concept’ model using data from care givers, patients and their families, and clinicians to identify the most disruptive symptoms, so that these can be specifically addressed in treatment regimes and drug discovery research and clinical testing. These highlighted speech and motor problems, as well as seizures. Detailed case histories also identified changes in gait, which could then be correlated with genetic changes and aid in genotyped patient classification.
Dr Pangkong Fox - CACNA1 Foundation
Dr Fox’s son was diagnosed with a mutation in the Cav2.1 ion channel gene in 2021, and after looking for information online she found the CACNA1 Foundation and by 2022 she was their Scientific Engagement Director. She has a PhD in Cell and Molecular Biology and uses her scientific background to create education support and resources. Cav2.1 mutations are linked to various forms of ataxia, migraine and epilepsies, as well as neurodevelopmental conditions such as intellectual disability and autism. There is a complex geno-phenotype with over 300 pathogenic mutations that lead to loss- or gain-of-function. The CACNA1A Foundation is doing many of the same things as the CLCN4 Foundation and Simon’s Searchlight (via their Research network) to develop a ‘preclinical toolbox’ of patient histories, biobanking, iPS and heterologous cell line generation, animal models, etc. These are and will be used for drug re-purposing screens and the search for new treatments (small molecules, ASOs, gene therapy). They are utilising collaborations with many other charities and disease and research institutions to achieve this cost-effectively and quickly (e.g. NORD, RareX, Citizen Invitae platform), as well as working with industry partners (e.g. Metabolom in Ireland for biomarkers), and receiving funding from US agencies and organisations (e.g. Chan Zuckerberg Initiative).
Many other ion channel channelopathy parents have followed this pathway, but without an established rare disease organisation they created their own charity, advocacy group or foundation – examples include KCNT1, KCNH10, CLCN4, SCN2A, etc. These groups help to identify and bring together patients, academics, clinicians and industry to facilitate collaboration, raise awareness, increase understanding of the disease and patient and family needs, discover new treatments, and raise funds.
A Family Story
Whilst the main patient and family engagement session on Day 2 was limited to the key clinicians, attendees were able to gain a personal insight into what it was like to have a child diagnosed with CLCN4 syndrome in a short online session with a mum in the Bay Area. Her child was first taken to the GP at 6 months with feeding problems and at 12 months with developmental delays, but was not diagnosed until after genotyping and neurological clinical profiling at 2 years of age. It was a very touching presentation – the child was asleep I think, but we were shown family videos and also some slides that the Mum had created about disease progression, symptoms, and family challenges and highlights. This was a very moving, personal and emotionally affecting presentation.
There was a similar video montage and slideshow of patient family pictures and information on Day 1 (also shown at the 2022 Symposium) illustrating patients and families who were diagnosed at varied ages from months-years-decades old, spread across the globe. This type of patient family insight is a very powerful way to bring R&D into personal focus.
Drug Discovery case studies:
Michael Schwake (University of Bielefeld) - Parkinson’s Disease affects lysosomal function
Michael is another alumni from the Jentsch lab, where he studied CLC5 in Dent kidney disease. He is now working on Parkinson’s Disease, which can be classified as an Inherited Disease of membrane Trafficking (IDT) as there is genetic linkage to TMEM175 and TRPML1 lysosomal channels, as well as to GCase enzyme involved in autophagy and endo-lysosomal cycling and function. Several academic and industry groups are working on GCase treatments for PD, but he noted that while GCase activity is reduced in the brains of PD patients and animal models, there are also PD patient mutations which show no effect on enzyme activity. I wasn’t aware of this but Ambroxol (a non-selective Nav channel inhibitor, cough medicine) is being tested clinically for PD as it has been shown to act as a chaperone for GCase, to boost protein trafficking and function.
Michael is working on human patient-derived iPS neuron models of PD, differentiating TH+ dopa neurons; he stressed the importance of using isogenic controls to confirm disease and gene differences, which he creates using CRISPR gene editing (baculovirus vectors being more effective than other expression methods). His models replicate the loss of GCase expression and activity and formation of a-synuclein fibrils, and also found that another lysosomal protein LIMP2 is reduced in PD. Significantly, he found that LIMP2 activates GCase and it’s associated protein GBA1, so loss of LIMP2 may lead to reductions in GCase in PD. His group is developing cell-penetrating peptides of the LIMP2 helix to promote GCase and GBA1 activity and function as a novel therapeutic approach for PD. The obvious corollary for potential CLCN4 disease treatments is to use small molecules or genetic treatments to modulate upstream regulators as well as the CLCN4 gene and protein itself, such as CLC3 and other proteins which may augment CLCN4 trafficking and activity.
Laksh Aithani - Charm Therapeutics
Charm Therapeutics is a London start-up with US connections; Laksh previously worked at leading AI drug discovery company Exscientia (UK), and the other co-founder is a US academic and entrepreneur focused on in silico deep learning protein folding applications. Other team members have come from Google DeepMind and big pharma.
Their main argument is that genetic sequences do not reliably predict 2D or 3D protein structure (i.e. the AlphaFold approach), so high resolution protein crystal structures should be used; additionally, ligand binding can alter protein structures (i.e. apo state may not be useful for drug discovery).
They have developed a fast deep learning algorithm called DragonFold that is trained on the 25% of PDB crystal structures with ligands bound (GPCRs, ion channels) and which can be used in a virtual screening workflow to create virtual crystals of the target protein and small molecule libraries, using supercomputer hardware (100 NVIDIA GPUs) and optimised AI/ML software.
Their therapeutic interest is mainly oncology and Laksh presented 2 case studies of small virtual screens against a kinase and a nuclear polymerase. Only 100-500 compounds could be virtually bound owing to computing and cost limitations, with each screen identifying a handful of hits with improved selectivity or mM potency respectively, which were validated in functional binding and biological assays.
Marlen Lauffer (Dutch Centre for RNA Therapeutics) - ASO genetic treatments for rare disease
I really enjoyed Marlen’s talk and her enthusiasm to find effective treatments for patients suffering from ultra-rare diseases, for which academic drug discovery is too slow and commercial interest is extremely limited, and any drug to reach market would be extremely expensive.
Her group is a non-profit consortium supported by the Dutch government and EU funding to rapidly find treatments for rare neurodegenerative disease, as effective drugs need to be delivered to young children before life-limiting symptoms take hold. To this end, they are using patient-derived fibroblasts to directly-induce human stem cell iNeurons, as traditional iPS engineering techniques are much slower and more expensive. Their therapeutic focus is on RNA antisense oligonucleotides (ASOs) to modulate faulty gene expression through exon skipping, RNAi or allele-specific mechanisms. She acknowledged that these approaches work best for g-o-f mutants, which have now been identified in a small subset of CLCN4 patients, but others have shown it can work to up-regulate gene and protein expression as well which may be suitable for treating l-o-f CLCN4 mutations as well. Rather than trying to group related patients with similar mutant genotypes for traditional small clinical trials, she suggested that defining each patient as their own population enabled N=1 Ph 0 clinical trials with fewer regulatory requirements, essentially breaking down each rare disease into multiple ‘nano’ rare diseases (also called 1 Mutation 1 Medicine, 1M1M, or N=1 diseases). ASOs can be delivered locally by i.c.v, i.t or injection into the eye, but it is unclear if a single application (injection or transgenic expression) can ‘cure’ a disease (one-and-done), or whether repeated applications may be needed during a child’s development.
This is a high risk and revolutionary approach; there have been recent patient deaths with such ultra-personalised genetic approaches (e.g. a DMD patient family-funded effort in the US), but Marlen is using the example of Milasen which was an ASO developed and tested to treat a single young female patient with Batten’s disease (CLC lysosomal storage disorder). She presented a case study on a patient with the retinal disease Stargardt Syndrome, with ASOs designed to correct intron/exon splicing that are being tested in iPS retinal organoids. I was concerned by the fact that limited safety testing is carried out with these ASOs (as many assume they are incredibly specific and selective, and thereby safe). For example, limited toxicology testing is done in non-human species, which presumably have different gene sequences to the human protein, so the action and efficacy of human-specific ASOs on the target protein and it’s ‘ome partners and signalling cascades may be reduced and under-estimated. For example, several pharma groups and CRO companies (e.g. Amgen, Metrion Biosciences) have shown that siRNA and ASOs and other genetic modulators can have unanticipated safety pharmacology effects, such as long-term (rather than acute) effects on hERG channel activity. The FDA and EMEA are aware of thee issues and may be increasing the threshold for safety testing of such RNA treatments, but as with oncology the seriousness of the disease and small patient numbers (in genetically stratified patients) may allow a more relaxed approval process.
Conclusion
In conclusion, this key conference promoted collaboration and rare disease patient and family support and participation, clearly showing that a varied and muti-disciplinary approach to channelopathy drug discovery can lead to tangible results. Above all there was a feeling of satisfaction that contributions, however small, are not going un-noticed and there is an overwhelming number of people and companies who are dedicating their drug discovery endeavours to accelerate treatments for rare conditions such as CLCN4.