Final Report Summary - GENCODYS (Genetic and Epigenetic Networks in Cognitive Dysfunction)
Executive Summary:
Scope: Cognitive disorders (CD) impose a major medical and socio-economical problem owing to their high incidence in our population. Intellectual disabilities alone account for 10% of the total health care expenditure in most European countries. Mutations in over 500 different genes have been associated with early onset (i.e. before 18 years of age) cognitive disorders, such as intellectual disability (ID), autism spectrum disorders (ASD), and schizophrenia. Yet, the molecular basis for approximately half of the patients with an early onset CD remains unknown, which creates an enormous burden to families confronted with such a disorder. Whereas early onset CD collectively impose a major medical and socio-economical problem, individual types of such CD are rare, which has precluded systematic studies that aim to provide insight into pathogenic mechanisms in CD. Our research is fundamentally different in that we aim to identify and investigate the key networks that are commonly disrupted in CD. We have used a systems biology approach to gain pathways-based insights into mechanisms leading to cognitive dysfunction in humans.
Conclusion: Research within Gencodys has been highly successful. The achievements are impressive and represent a major advancement in our understanding of the molecular underpinnings of cognitive disorders. As a consortium we have uncovered some 150 new genes that carry causative mutations in such neurodevelopmental disorders. We have also demonstrated beyond doubt that CD genes converge into common molecular networks. This insight is important for developing strategies for therapeutic interventions, which can be targeted to common molecular processes and networks, and not strictly at single genes. Gencodys has made major advances towards development of therapeutic strategies. An unparalleled collection of animal models, both Drosophila and mouse, have been generated and characterized for a wide variety of neurodevelopmental, behavioural and cognitive parameters. These models are invaluable for pre-clinical studies of pharmaceutical compounds. In fact, GENCODYS has already started in testing a number of compounds, both in newly developed in vitro and ex vivo systems as well as in selected mouse and fly models. We anticipate that the wealth of new knowledge and innovative resources developed within in GENCODYS will become a basis for a plethora of follow-up studies across Europe, both in fundamental research and in translational research for developing therapeutic interventions. Partners who have been collaborating in GENCODYS will be actively involved in such follow-up projects, reinforcing existing collaborations and establishing new ones. Renewed support from the EU will be instrumental to benefit optimally from the established resources and to continue to increase the impact of EU research teams in translational research of the large group of genetic cognitive disorders.
Project Context and Objectives:
Concept:
Cognitive disorders (CD) impose a major medical and socio‐economical problem owing to their high incidence in our population. Intellectual disabilities alone account for 10% of the total health care expenditure in most European countries. Mutations in over 400 different genes have been associated with early onset (i.e. before 18 years of age) cognitive disorders, such as mental retardation, autism, some neurodegenerative disorders, and even schizophrenia. Yet, the molecular basis for the majority of patients with an early onset CD remains unknown, which creates an enormous burden to families confronted with such a disorder. Whereas early onset CD collectively impose a major medical and socio-economical problem, individual types of such CD are rare, which has precluded systematic studies that aim to provide insight into pathogenic mechanisms in CD. So far, functional studies have been restricted to a few single gene defects, such as for the Fragile X syndrome. Our research will be fundamentally different in that we will identify the key networks that are commonly disrupted in CD. We follow a systems biology approach to gain pathways‐based insights into mechanisms leading to cognitive dysfunction in humans. The scale of our approach has been unprecedented world‐wide, and execution of this program and warrants the leading role of Europe in this important field of medical research. A schematic of the GENCODYS strategy is depicted in Figure 1. In short:
(1) Identification of genes involved in cognitive disorders. Despite considerable progress in the identification of genes underlying CD, the vast majority of the molecular defects underlying CD are still unknown. Therefore, our first objective is to identify new genes causative of CD by developing and implementing high‐throughput strategies applied to a unique large cohort of families with multiple CD patients.
(2) Elucidation of molecular networks that are commonly disrupted in CD. Emerging evidence from genetic and neurobiological research provides evidence for a limited number of molecular and cellular pathways that are shared by the various genetic CDs. These pathways are associated with abnormalities of dendritic structure, dendritic spine morphology and synaptic plasticity, suggesting a common pathophysiological mechanisms for the cognitive impairment associated with intellectual disability, autism and other CDs. Thus, the second stage of our strategy is to systematically explore this concept by elucidation of molecular networks that are commonly disrupted in CD. Besides elucidating new common pathways in CD, we will further expand the already recognizable common networks, pathways and physiological mechanisms that we and others have identified: Rho GTPases, epigenetic regulators of the activity of neuronal genes and postsynaptic glutamate receptor complexes using a variety of functional genomics strategies. We shall make optimal use of genetic models that are the cornerstones of neuroscience, such as mouse and fruit fly. Powerful tools and a comprehensive understanding of the molecular processes and mechanisms underlying cognitive (dys)function are key to the third, translational part of our research.
(3) Identify genetic modifiers and small compounds that modulate the disease phenotype. The third part of our research program is to resolve the molecular underpinnings of the large degree of clinical variability that is typical for all types of CD, even among patients carrying identical gene mutations. Genetic modifier screens in available Drosophila models for CD will be used to reveal phenotypically relevant genetic interactions and molecular networks. Moreover, drug screens shall be conducted in fly and cellular models for CD, and subsequent selected drug testing in mouse models will eventually potentiate pharmacological intervention in CD patients.
Project objectives:
The GENCODYS project aims at gaining comprehensive and conceptual advance in understanding the molecular basis of human brain development in health and disease. GENCODYS proposes a set of complementary ambitious objectives:
1. Identification of a plethora of novel genes involved in cognitive disorders by application of powerful cutting‐edge genetic research to unique patient material collected within the GENCODYS consortium.
2. Systematic generation of Drosophila melanogaster models to study molecular pathways and neurobiological mechanisms involving genes mutated in human CD.
3. Large scale generation of mouse models of CDs and systematic characterization of multiple key electrophysiological and behavioural properties.
4. A systems biology approach applied to functional neurogenomics analyses to gain insight into genetic and epigenetic networks involving CD genes.
5. Bioinformatics data integration involving collected data from the GENCODYS program with public genomics data.
6. Unbiased identification of small chemical compounds that have the propensity to improve a CD phenotype, initially in primary neuronal cultures and Drosophila, and subsequently in available mouse models.
7. Consolidation of a leading role of Europe in translational research on cognitive disorders by scientific excellence in complementary research areas and their integration into a systematic research pipeline.
Progress beyond the state of the art:
The main innovation of the GENCODYS program lies in the systematic and large‐scale integration of research fields that thus far have been largely separated. Medical genetics research, one the one hand, has been of great importance for diagnostics, prognosis and disease management. Yet, the elucidation of these genes in itself provides only limited insights into the associated disease mechanisms and molecular pathways that are disrupted. On the other hand, spectacular conceptual breakthroughs have been obtained in modern neuroscience research, through exploiting the powerful tools provided by genetic manipulation of model organisms. However, there are few examples where such research is directly applied to models for human CD, and only sporadic, if any, studies where comparative research is conducted for a larger number of CD genes. GENCODYS has the ambition to bridge this gap through the convergence of experimental efforts and the combination of an unprecedented set of tools directly relating to normal cognitive dysfunction. Clearly, such integration of experimental strategies is expected to yield information that is superior in value to the sum of its individual components.
GENCODYS will thus address neurobiological concepts fundamental to cognitive dysfunction, and will do so in an integrated manner using a number of novel and/or strongly innovative approaches, which in itself are beyond state‐of‐the art (Table 1). This partnership of groups at leading research institutions and selected biotech SMEs will also deliver new enabling technologies and proof of principle for modulation of CD disease phenotypes. This will be realized by the design and implementation of screening protocols whereby small molecule screening will be performed by following genome‐wide approaches and the design of assays that are amenable to small molecule screening in primary neuronal cultures and in hundreds of Drosophila models for CD. Furthermore, mouse models for CD will follow these studies, and are amenable to medium scale genetic manipulation and provide advantages for investigating the mechanisms in both physiological and neuropathological settings in a mammalian context.
Overall strategy and general description:
To achieve an optimal workflow for our translational research strategy, we have defined six work packages, which focus on the individual stages that cover the path from the clinic to model organisms and functional research and back again to the clinic (Figure 1).
1. HUMAN GENETICS & GENOMICS
The foundation to our strategy is an ever expanding collections of patient and family material: currently we have DNA samples from >600 families with X‐linked MR (www.euromrx.com) a unique collection of ~500 multiplex families of recessive forms of CD from populations with a high level (40‐70%) of consanguineous marriages, and >10,000 sporadic patients with various types of CD (MR, autism). Clinical data and massive amounts of genetics and genomics data of these patients will be registered in a patient database. Application of state of the art mapping and sequencing technologies should reveal a large number of novel CD genes.
2. FLY MODELS
Gene mutations known to be associated with CD or newly identified in subproject HUMAN GENETICS & GENOMICS will be modelled in powerful models for neuroscience research, Drosophila melanogaster and the mouse. Drosophila will be the model of choice for genome‐wide genetic screens, high throughput phenotyping, including assessment of learning and memory, high‐resolution analysis of gene expression. Genetic modifier screens and chemical compound screens will be carried out in selected Drosophila models to reveal phenotypically relevant genetic interactions and molecular networks that can be manipulated by endogenous compounds.
3. MOUSE MODELS
The mouse will be used for studying higher cognitive processes for subsets of CD genes. Over 50 mouse models are already available for CD‐associated genes involved in Rho GTPase signaling, epigenetic pathways and NMDA receptor complex and postsynaptic density and others can be readily obtained. The effects of gene mutations towards the general behaviour repertoire (e.g. motor function, novelty seeking, anxiety) and specific cognitive functions (e.g. attention, short and long term memory, strategy utilization) will be explored. In addition, effects of mutations on general architecture of the nervous system, neuronal networks, branching, synaptogenesis and synaptic plasticity and electrophysiological properties will be determined. Whenever appropriate, these models will also be used for in vivo validation of therapeutic approaches.
4. FUNCTIONAL NEUROGENOMICS
The studies constitute a variety of complementary High Throughput methodologies for systematic analysis of the molecular and cellular networks that are disrupted in CD. They will encompass the determination of expression patterns in the nervous system and the subcellular localization of CD- associated mRNAs and proteins, using expression profiling, proteomics, in situ hybridization and live imaging. Innovative strategies for systematic analysis of gene expression in mouse brain, and cluster analysis results will reveal genes with similar patterns of expression indicative of common pathways. Furthermore, neuron‐based assays will be exploited to conduct screens for small molecules that can rescue disrupted neurophysiological processes brought about by engineered CD gene mutations.
5. BIO‐INFORMATICS
A key component of our project will be the integration of the massive amounts of data that will be generated at all stages of the program. To optimally exploit the data generated by our systems biology approach we will harness the predictive power of a variety of bioinformatics techniques. Functional characteristics of CD genes and corresponding proteins, mutation patterns, homozygous loci, mutations, CNVs, expression levels, epigenetic data and phenotype information will be cross‐referenced to identify significant enrichments that are indicative of causation. This will be complemented by a systems biology approach to delineate the biological coherence of the 400+ known CD genes by analyzing their involvement in common pathways using public genomics data.
6. POTENTIATION
This subproject encompasses activities that encourage the formation, development, and exploitation of new ideas. Envisaged are training and networking activities, organization of workshops, dissemination of results through websites, meetings, and scientific publications. We will actively exploit the possibilities for patenting, using fair measures for protection of intellectual property rights.
A consortium of scientific complementarities and balance:
The partners have been carefully chosen for the GENCODYS consortium based on the expertise that they contribute to the project in the fields of genetics, neurobiology, Drosophila and mouse biology, bioinformatics and molecular and cellular biology. Partners were also chosen based on their outstanding knowledge and state‐of‐the‐art equipment. The project has a truly European dimension with 13 scientific partners from five European countries (Germany (2), France (3), Austria (1), UK (2), Netherlands (3)), two SMEs from the UK and a private entity representing Patient organization across Europe (seated in Belgium). Moreover, two major Institutes from Asia are full participants of GENCODYS, one from Iran and one from Pakistan. Finally, the inclusion of partners is based on several previous and ongoing collaborations among partners on projects related to the current Integrating project.
In order to achieve high synergy and conduct research as competitive and efficient as possible, the members of the consortium have been chosen dependent on their excellence in specific fields to reach the aim of successful molecular genetic neuroscience research. Therefore, the expertise which is present in GENCODYS is highly complementary and is balanced between the fields of clinical and molecular genetics, research with model organisms, bio‐informatics and molecular and cellular neurobiology. While each group has specific expertise in subfields of the general field of expertise, most fields/subfields of expertise are covered by more than one group so that the team with the lead expertise will serve as reference centre for the respective work.
Project Results:
GENCODYS research is conducted in five scientific work packages. Workpackage (WP) 1 aims to identify the molecular defects underlying early onset CD, which is the key to diagnosis, understanding the pathogenesis and eventually, treatment of this heterogeneous disorder. The genes identified in WP1 are the basis for investigation of the respective protein and associated functional networks. In WP2 we use the fruitfly and in WP3 the mouse as model organisms to gain insight into genetic networks and their role in normal and disrupted neurodevelopment, learning and memory and other behaviours. In WP4 we investigate the molecular and cellular mechanisms that are associated with disruptions of genetic networks that are key to CD: synaptic protein complexes, epigenetic pathways, cytoskeleton structure. Finally, bioinformatics in WP5 has an important role in reconstructing genetic networks underlying CD by large scale systematic integration of datasets generated in the other workpackages. In addition to these five scientific workpackages, two work packages dedicated to dissemination and management:
From the start, Gencodys has met the goals we have set, and in fact for many parts of the project scientific progress is well ahead of schedule. Collective output in terms of new data as well as scientific publications has been highly impressive for the entire GENCODYS program. Below, we discuss the highlights of GENCODYS for each of the workpackages. The dissemination of GENCODYS research to various stakeholders, i.e. the medical and scientific community, scholars and students, patients and patient organizations and society as a whole, has been the focus of WP6 (Potentiation).
WP1: Human Genetics and Genomics: Identification of both convergent and distinct genetic pathways genes in ID and in autism spectrum disorders
Highlights of achievements of workpackage 1:
- The GENCODYS biobank and clinical database of patients with a cognitive disorder contains >13,000 individuals with a CD, which significantly surpasses the number which we aimed for at the start of the GENCODYS project.
- We have developed protocols for targeted sequencing and whole-exome sequencing that enable the identification of causative mutations in 30-70% of the patients, depending on the familial occurrence of the disorder.
- We have implemented protocols for next generation sequencing in a diagnostic setting.
- The thus obtained genetic diagnosis provides better opportunities for (prenatal) diagnostic testing, counseling and disease management in the respective families.
- Besides mutations in genes already linked to CD previously, we have identified at least 150 new CD genes that already have been or will be described in a variety of high-impact journals.
- We identified many new genes underlying neural migration disorders. These studies have highlighted common molecular and cellular networks underlying such disorders. For example, the implication of microtubule disruptions is associated with malformations of cortical development (MCD; tubulinopathies)) and defective glycosylation of dystroglycan in disorders that are now collectively known as muscular dystrophy-dystroglycanopathies (MDGGs).
- We have uncovered a contribution of mosaic parental germ-line mutations in recurrent forms of neurodevelopmental disorders (MCD).
- We have revealed that de novo mutations and genomic disruptions underlie a large number of cognitive disorders, such as ID, ASD and rare disorders like Möbius syndrome.
WP1: Summary of the main results
This workpackage had 5 major tasks:
Task 1.1. Recruitment of families and patients.
The foundation of the GENCODYS program is an ever expanding collection of patient and family material. The recruitment of patients and their families has been focused on several groups of patients distinguished in five groups based on the pattern of inheritance of the cognitive dysfunction: 1. Autosomal Recessive Intellectual Disability (ARID); 2. ARID or X-linked ID (XLID); 3. XLID; 4. Isolated CD; 5. Isolated Autism Spectrum Disorder (ASD). In this classification, ID and CD are used interchangeably. Patients classified as isolated ASD do not have other cognitive deficits besides their ASD. Sporadic patients with ID/ASD have typically de novo dominant mutations or a multifactorial etiology (several genetic variants). The numbers of patients are indicated in the Table 1. The given numbers are an underestimate as families (categories 1-3) are listed by only one affected per family, whereas there are 2-20 patients for each of these families.
For all patients informed consent has been obtained using protocols that have been approved by the local and/or national competent authorities for research with human individuals. The consent comprised the collection of blood or other patient materials, genome-wide genetic analyses aimed at the elucidation of the molecular basis of the genetic disorder in each of the patients/families, and the approval to use the results of these studies in medical and scientific reports and presentations.
The gencodys program has exceeded its huge ambitions set at the start of the program. A valuable cohort of CD patient has been established, which is unique both in its size and in the extent of clinical information that has been collected on individuals within this cohort. This cohort has been the basis of successful genetic and fundamental research within the GENCODYS project and will remain a valuable resource for further studies of this kind as well as for translational research and future clinical trials.
Task 1.2. Clinical characterization, documentation and establishment of cell lines.
All families and patients ascertained until now have undergone extensive clinical examinations by the Clinicians connected to the various GENCODYS groups. These data have been recorded by a standardized form that was established in the first year of the project. The standardization allows the systematic entry of clinical records in the GENCODYS database. The virtue of the database is that queries can be made for patients that present a specific constellation of clinical features (i.e. the main cognitive defect as well as co-morbidities). Such queries allows the selection of patients carrying a mutation in a candidate gene-of-interest, which is often used for the identification of matching cases (mutation in the same gene) that support the qualification as a new “CD gene”.
The Clinical database is connected to the molecular database. The molecular database contains the results of molecular genetics analyses carried out in each of the patients/families, such as the presence of copy number variants (CNVs), homozygous regions, and the presence of rare DNA variants (rare variants and mutations) segregating with the phenotype. Also the molecular database has a high utility for identifying matching mutations.
Besides the collection of blood samples for genetics and genomics studies, we have also collected other biomaterials for other molecular and cell-based investigations. For almost 1200 patients we have established Epstein-Barr virus transformed lymphoblastoïd cells lines. These cell lines are highly useful for transcript analyses, transcriptome studies and epigenetic analyses. In addition, we have established over 150 fibroblast lines from skin biopsies of patients. Fibroblast lines from several patients with confirmed causative mutations, such as Kleefstra syndrome and Koolen-de Vries syndrome have been used to generate induced pluripotent stem cells (IPS). In addition, we have successfully implemented protocols for dedifferentiation of IPS cells into neural lineages for functional studies (link to other parts of the GENCODYS program). Such patient-derived excitatory cortical neurons are used for electrophysiological studies to measure synaptic activity and neuronal network activity that can shed light into the nature of the cognitive deficit.
Task 1.3. Exclusion of known genes for ID and and/or autism spectrum disorder (ASD).
The exclusion of genetic causes that have already been implicated in CDs previously has been conducted by the application of diagnostic protocols as well as the use of targeted innovative analyses that have been set up for this purpose within the GENCODYS program. Examples of diagnostic pre-screens include the interrogation of the genome for CNVs by Comparative Genomic Hybridization using SNP-microarrays (arrayCGH). Other standard tests include investigation of repeat expansions at the FMR1 locus (Fragile X syndrome) and phenotype-based sequence analysis of CD genes. CNV has provided a conclusive diagnosis (genomic deletion or duplication) in 10-15% of all patients and FMR1 repeat expansions in 2-5% of patients.
Specific analyses that have been set up within the GENCODYS program include a next generation sequencing pipeline for targeted exome sequencing of all exons encoded by the X chromosome. This pre-analysis has identified a large number of causative mutations in known ID genes and has uncovered some 10 new genes for X-linked ID.
Task 1.4. Mapping of novel genes for ID and/or ASD.
Mapping and CNV analysis has been conducted in 5929 and 7849 individuals, respectively, of the GENCODYS cohort (Table 1). Systematic whole-genome SNP genotyping has been conducted on High-Density SNP-array platforms (Affymetrix or Illumina), containing 250,000 to > 1 million SNPs. The results and some examples of SNP microarray mapping are discussed for families with an apparent autosomal recessive or X-linked genetic defect and for isolated cases.
CNV analysis in isolated cases: CNV analysis is a first step in the diagnostic analysis of patients with ID and associated phenotypic features and has replaced traditional (microscopic) karyotyping in many if not most diagnostic laboratories in Europe. CNV analysis has been conducted in 7849 individuals during the GENCODYS program. .In isolated patients, a causative CNV leading to a conclusive diagnosis can be identified in approximately 10-15% of cases.
Homozygosity mapping by SNP microarrays in CD families: With the exception of some X-linked families, the inheritance pattern of the large consanguineous families will be almost exclusively autosomal recessive. Since consanguinity results in a high degree of homozygosity, we expect that the majority of mutations in these families will be homozygous. Homozygosity mapping studies have been performed in hundreds of consanguineous families from Iran, Pakistan, Turkey and Arab countries. Homozygosity mapping has been followed by next generation sequencing, initially by targeted exome sequencing of homozygous intervals and later by whole-exome sequencing (WES). This approach has resulted in the identification of numerous new genes for recessive forms of CD
Task 1.5. Identification of novel molecular defects underlying ID and/or ASD.
Next Generation sequencing has been conducted in almost 2500 individuals of the GENCODYS cohort (Table 1). These high-throughput sequencing activities have been performed on various platforms (Roche, Illumina, Complete Genomics) following a variety of enrichment protocols. GENCODYS partners have pioneered the development of protocols for the selection of variants and interpretation of next generation sequencing data in CD patients with various inheritance patterns of their disorder (recessive, X-linked, dominant, de novo, oligogenic) and in combination with various sequencing procedures (targeted exome, whole exome, whole genome).
In the European population, the majority of individuals with a CD are seen as isolated cases. This is a consequence of the reduced fecundity of CD individuals. As a consequence, the associated mutation (copy number variant (CNV) or single nucleotide variant (SNV)) has usually occurred as a de novo event in the germline of one of the unaffected parents (Vissers et al., 2010). Therefore, for identification of such events we usually have performed trio sequencing, i.e. NGS of the patient and both of the healthy parents. Candidate SNVs are selected based on their de novo occurrence. Within GENCODYS we have conducted whole exome sequencing (WES) and whole genome sequencing (WGS) in trios.
Whole exome sequencing
In a groundbreaking study (De Ligt et al., 2012), we sequenced the coding regions of more than 21,000 genes obtained from 100 patients with an IQ below 50 and their unaffected parents. A data-analysis procedure was developed to identify and classify de novo, autosomal recessive, and X-linked mutations. In addition, we used high-throughput resequencing to confirm new candidate genes in 765 persons with intellectual disability (a confirmation series). All mutations were evaluated by molecular geneticists and clinicians in the context of the patients' clinical presentation.
We identified 79 de novo mutations in 53 of 100 patients. A total of 10 de novo mutations and 3 X-linked (maternally inherited) mutations that had been previously predicted to compromise the function of known intellectual-disability genes were found in 13 patients. Potentially causative de novo mutations in novel candidate genes were detected in 22 patients. Additional de novo mutations in 3 of these candidate genes were identified in patients with similar phenotypes in the confirmation series, providing support for mutations in these genes as the cause of intellectual disability. We detected no causative autosomal recessive inherited mutations in the discovery series. Thus, the total diagnostic yield of conclusive mutations was 16%, mostly involving de novo mutations. An even higher percentage of de novo mutations was seen in another study (Rauch et al., 2012). These findings show that de novo mutations represent an important cause of intellectual disability; exome sequencing was used as an effective diagnostic strategy for their detection.
De novo mutations have been identified in a number of patients presenting a recognizable syndrome, for which the etiology had been hitherto unknown. Examples of these include disorders with aberrant cortical migration (MCD; P3a), Moebius syndrome (P1a) and Kleefstra syndrome (P1a).
Whole genome sequencing
Despite the high percentage of de novo mutations identified by WES, the majority of cases remain undiagnosed. We therefore applied whole-genome sequencing to 50 patients with severe ID and their unaffected parents. All patients included had not received a molecular diagnosis after extensive genetic prescreening, including microarray-based CNV studies and exome sequencing (see above). Notwithstanding this prescreening, 84 de novo SNVs affecting the coding region were identified, which showed a statistically significant enrichment of loss-of-function mutations as well as an enrichment for genes previously implicated in ID-related disorders. In addition, we identified eight de novo CNVs, including single-exon and intra-exonic deletions, as well as interchromosomal duplications. These CNVs affected known ID genes more frequently than expected. On the basis of diagnostic interpretation of all de novo variants, a conclusive genetic diagnosis was reached in 20 patients. Together with one compound heterozygous CNV causing disease in a recessive mode, this results in a diagnostic yield of 42% in this extensively studied cohort, and 62% as a cumulative estimate in an unselected cohort. These results suggest that de novo SNVs and CNVs affecting the coding region are a major cause of severe ID. Genome sequencing can be applied as a single genetic test to reliably identify and characterize the comprehensive spectrum of genetic variation, providing a genetic diagnosis in the majority of patients with severe ID
WES in families with recessive CD
With the exception of some X-linked families, the inheritance pattern of the large consanguineous families will be almost exclusively autosomal recessive. Since consanguinity results in a high degree of homozygosity, we expected that the majority of mutations in these families will be homozygous. Homozygosity mapping studies have been performed in hundreds of consanguineous families from Iran, Pakistan, Turkey and Arab countries. Homozygosity mapping has been followed by next generation sequencing, initially by targeted exome sequencing of homozygous intervals (Najmabadi et al., 2011). In this landmark study Partners 2 and 4 performed enrichment of exons from homozygous regions followed by next-generation sequencing in 136 consanguineous families with autosomal-recessive ID. This study, has revealed additional mutations in 23 genes previously implicated in intellectual disability or related neurological disorders, as well as single, probably disease-causing variants in 50 novel candidate genes. Proteins encoded by several of these genes interact directly with products of known intellectual disability genes, and many are involved in fundamental cellular processes such as transcription and translation, cell-cycle control, energy metabolism and fatty-acid synthesis, which seem to be pivotal for normal brain development and function.
Soon after this work, several GENCODYS partners have performed WES in families in which prior homozygosity mapping was performed (1 individual per family) and in consanguineous families without prior homozygosity mapping (WES for 2-3 individuals per family). These studies have uncovered a plethora of new CD genes, which have resulted in an impressive number of high-impact publications. been published in a range of high-impact papers (e.g. Roscioli et al., 2012; Schuurs-Hoeijmakers et al., 2012; Wortmann et al., 2012). WES has also been applied to non-consanguineous families of European origin with multiple affected individuals in a sibship. This has also revealed several novel genes, such as SLC6A17 (Iqbal et al. 2015). In total, we have identified well over 100 new ARID genes. Causality of these genes is supported by the occurrence of recessive mutations in at least 2 families with overlapping CD phenotypes. Many of these genes have already been published, but many other papers are still in preparation, including two that will report the results of large-scale WES in hundreds of ARID families from Iran and Pakistan.
X-exome sequencing
X-linked intellectual disability (XLID) is a clinically and genetically heterogeneous disorder. During the past two decades in excess of 100 X-chromosome ID genes have been identified. Yet, a large number of families mapping to the X-chromosome remained unresolved suggesting that more XLID genes or loci are yet to be identified. Here, we have investigated 405 unresolved families with XLID. We employed massively parallel sequencing of all X-chromosome exons in the index males. The majority of these males were previously tested negative for copy number variations and for mutations in a subset of known XLID genes by Sanger sequencing. In total, 745 X-chromosomal genes were screened. After stringent filtering, a total of 1297 non-recurrent exonic variants remained for prioritization. Co-segregation analysis of potential clinically relevant changes revealed that 80 families (20%) carried pathogenic variants in established XLID genes. In 19 families, we detected likely causative protein truncating and missense variants in 7 novel and validated XLID genes (CLCN4, CNKSR2, FRMPD4, KLHL15, LAS1L, RLIM and USP27X) and potentially deleterious variants in 2 novel candidate XLID genes (CDK16 and TAF1). We show that the CLCN4 and CNKSR2 variants impair protein functions as indicated by electrophysiological studies and altered differentiation of cultured primary neurons from Clcn4-/- mice or after mRNA knock-down. The newly identified and candidate XLID proteins belong to pathways and networks with established roles in cognitive function and intellectual disability in particular. We suggest that systematic sequencing of all X-chromosomal genes in a cohort of patients with genetic evidence for X-chromosome locus involvement may resolve up to 58% of Fragile X-negative cases. (Hu H et al. 2015).
Autism spectrum disorders (ASD)
The group of P3b (Laumonnier) analyzed a cohort of 100 French families including at least one individual with ASD by arrayCGH (Agilent) followed by targeted exon capture and NGS of 216 genes encoding mostly proteins belonging to the NMDA receptor interacting proteome. The targeted exon capture and NGS allowed the identification of a number of candidate genes and genomic regions, novel and replicating previous loci involved in neurodevelopmental disorders. Of interest, two de novo CNVs involved the GRM5 gene in 2 unrelated patients. The GRM5 gene encodes the mGluR5 postsynaptic receptor that is thought to be involved in the pathophysiology of Fragile X syndrome.
Conclusion
Next generation sequencing methodologies such as WES and WGS have proven to be highly effective in the identification of mutations underlying CD of all inheritance patterns (de novo, recessive, X-linked and oligogenic as in ASD). These genome-wide approaches allow the analysis of all genes, or selected groups of genes (gene packages) in a single experiment. The costs of these NGS methods have dropped considerably over the last year and have reached a price of less than 1000 € per sample. This means that NGS is also a cost-effective approach in comparison to microarray analyses and Sanger sequencing protocols. As a consequence, WES and soon WGS will become the methodologies of choice for a first-step genetic analysis of patient with an unknown (cognitive) disorder. In fact, at the institute of P1a WES is already offered as a diagnostic test and soon this will be extended by WGS protocols. The latter have the potential to also pick up genetic and genomic variants in non-coding regions, but the assignment of causality of such variants needs further insight in biological mechanisms as well as the development of tools for bioinformatics analyses.
Workpackage 2: Fly Models
- An unbiased screen for learning and memory conducted for almost 2500 Drosophila genes
- Characterisation of 435 CD genes and candidate CD genes.
- Identification of an epigenetic module underlying CD by genetic interacting studies.
- Development and application of innovative imaging pipelines to visualise CD protein activity in the nervous system without interfering background fluorescence.
- Publication of the first systematic ID gene screen identifying conserved functional modules of interconnected ID genes.
- Functional screens of CD genes completed: phototaxis/eye morphology, behavior, synapse and lethality/flight performance, and memory.
- First drug approaches in Drosophila initiated, including several chemical compounds that have been tested in EHMT and other CD models such as ADHD.
- First successful drug rescue project in Drosophila models completed.
In WP2, the fruit fly Drosophila melanogaster and its powerful genetics have been exploited to systematically uncover CD gene function, generating and integrating cellular and subcellular expression profiling during development, neuroanatomical phenotyping, functional and behavioural screening approaches in a high throughput manner. Beyond these unprecedented functional studies in an intact nervous system, genetic and pharmacologic screens have been applied in Drosophila to identify novel cognitive genes providing an entry point for the development of novel diagnostic and therapeutic strategies.
WP2 Summary of the main results
This workpackage had 4 major tasks:
2.1. Systematic analysis of expression patterns of CD genes and subcellular localisation of their protein products.
We have generated 212 transgenic lines carrying tagged GENCODYS constructs marked with the fluorescent marker that cover altogether 131 GENCODYS genes. In addition to lines with fluorescent marker we prepared 216 retrofitted constructs marked with mini-white and obtained lines for another 14 genes that are not covered by lines with fluorescent marker. Altogether, we have generated 145 transgenic strains carrying tagged version of GENCODYS genes.
Throughout the GENCODYS grant we have developed technology for live imaging of fluorescent fosmid-based gene expression reporters in Drosophila embryos using multi-view Selective Plane Illumination Microscopy (SPIM). The complete pipeline involving the construction up of custom OpenSPIM microscope, registration of the SPIM views, fusion by multi-view deconvolution and visualisation using the BigDataViewer published and available as open source plugin. We found that only about 10% of GENCODYS transgenic lines show high enough endogenous expression to allow SPIM imaging in embryos.
We screened the GENCODYS transgenic lines for expression in adult brain. To date we screened 52 lines that is about one half of the obtained transgenics. Based on the image data obtained, we categorized GENCODYS genes expressed in adult brain into four patterns (Figure 3). The genes with expression localized to projections (1), to cell body membrane or cytoplasm (2) or to nuclei (3) of all neurons and genes with expression in nuclei of only several cells in a spatially restricted brain region (4).
The goal is to make the image data available to scientific community. To that end we developed two visualization tools CATMAID and BigDataViewer. We have recently showcased the capabilities of the CATMAID viewer by providing unprecedented level of access to raw data of endogenously tagged library of rab proteins imaged in various developmental contexts throughout Drosophila lifecycle. This dissemination strategy serves as model for distribution and analysis of the images of GENCODYS transgenic lines. The BigDataViewer (BDV) enables browsing through arbitrarily large image volumes and dynamically re-slicing them across arbitrary dimensions. Additionally, we have recently developed a BigDataServer that makes it possible to access remote data sources in the same way as if they were on a local computer. Both tools are open source and thus can be adapted and extended for the purposes of various image-based projects, even beyond the GENCODYS project.
2.2. Systematic generation and characterisation of fly CD models.
The function of most genes underlying CDs in the nervous system is poorly understood. Gencodys WP2 was devoted to determine defects in a minimum of 300 Drosophila CD models. For this purpose, unique resources have been exploited, most importantly two transgenic RNAi libraries. CD models have been characterized in analysis pipelines designed to determine defects on three different levels relevant to human mental disorders: neuronal and synaptic morphology, neuronal function and cognitive behavior/ learning and memory.
Synaptic morphology was known to be compromised in several cognitive disorders, as revealed by post‐mortem studies and mouse models, but the extent to which compromised synapse development occurs in these disorders was illusive. We have used >500 RNAi lines to pan‐neuronally knock‐down CD genes, and has analyzed multiple parameters of synapse development at the larval Neuromuscular junction in a high‐throughput imaged‐based screen. This screen comprised >10.000 tracked and quantitatively assessed synapses.
To test whether CD genes are required for neurodevelopmental and basal neurotransmission, 500 RNAi lines were analyzed for performance in two fundamental fly paradigms: adult flight/ escape response and phototaxis.
Habituation is a form of non‐associative learning, in which a response to a repeatedly applied stimulus gradually wanes. It is defective in classical learning and memory mutants and as reported in several Gencodys papers. Aktogen, an SME within GENCODYS, has developed a 32‐chamber semi‐automated version of the light‐off jump habituation paradigm. In this system more than 300 mutant Drosophila lines were screened and a series of novel learning mutants were isolated. A learning defect was observed in 30% of genotypes tested. The high‐throughput adapted habituation paradigm is therefore a powerful tool to screen disease genes, disease gene candidates or other genes for their role in learning.
Previously, one of the GENCODYS partners (Keleman, IMP Vienna) has established the courtship conditioning paradigm to study long‐term memory in Drosophila. In this paradigm, naïve males learn to selectively suppress futile courtship of unreceptive mated females. Learning, short‐term or long‐term memory can be evaluated in these males. Almost 2500 RNAi lines targeting ID genes, candidate ID genes and synaptic genes were analyzed for long term‐memory defects with use of a tracking system.
Gencodys has integrated these rich datasets to identify further molecular networks that form novel molecular networks underlying cognitive disorders.
2.3. Identification and characterization of novel cognitive genes (i.e. new CD candidate genes).
This task has been accomplished in close collaboration with GENCODYS bio-informaticians as described at WP5. The basis of candidate gene identification from Drosophila experiments included:
- the unbiased screen for learning and memory conducted for almost 2500 Drosophila genes
- Characterisation of 435 CD genes and candidate CD genes and associated phenotypes as characterized in our phenotyping pipeline.
- The identification of target genes for CD genes encoding transcription factors and epigenetic regulators.
- Genetic interaction studies using crossed of different fly CD gene mutants..
2.4. Pharmacological treatment of CD fly models to identify novel drug targets and compounds for the treatment of CD.
The development of strategies for therapeutic intervention in CDs has been a major ambition of the GENCODYS program. We have been highly committed to develop in vitro, in vivo and ex vivo methodologies for testing pharmaceutical or other sorts of intervention. Drosophila is highly suitable for use in medium-size screens for compounds.
We have developed a working prototype of an insect drug delivery system: a vacuum injection drug delivery system. It consists of a vacuum pump, a number of two- and three-way solenoid valves, an electronic atomizer nozzle, a syringe pump, a spraying chamber, National Instrument control hardware and control software implemented in the National Instrument LabVIEW package.
Various drug projects have been completed or are ongoing and have already attracted novel funding that secures the first follow up study on Gencodys achievements: project funded by the Jerome Lejeune foundation for testing Spermidine on CD fly models. Other examples of intervention studies in Drosophila CD models include a variety of compounds (transcriptional and translational inhibitors, antioxidants, high glucose diet and a library of epigenetic drugs) applied to EHMT-mutant flies, lamotrigine in the NF1 model, and a candidate drug for treatment of mitochondrial (complex I) disorders.
Workpackage 3: Mouse Models
- In total 54 mouse lines have been generated and made available to the GENCODYS consortium and the scientific community as a whole.
- Approximately forty mutant CD mouse lines have been characterised throughout a specific phenotyping Pipeline.
- In-depth molecular characterization of several mouse models has been completed (in collaboration with WP4).
- Complete electrophysiological evaluation of five novel lines of mice with mutations relevant to genetic changes in human patients with various forms of autism and intellectual disability.
- Pharmaceutical interventions have been tested in five CD mouse models.
The goal of WP3 was to develop, characterize and make available mouse models to elucidate the function of genes mutated in CDs, and the pathomechanisms linked to their dysfunction. While existing mouse models were used as reference, we have focused on genes recently identified or that will be identified by the GENCODYS consortium as part of WP1, or on key genes involved in the major CD pathways. This work package was highly complementary to WP1 (human genetics) and WP2 and WP4 (fly models and functional genomics) and systematic integration of data (WP5) from WP3 with others WPs will result in a powerful view of the effects of the alteration of these genes and their pathways at a cellular and whole organism level.
WP3 Summary of the main results
This workpackage had 5 major tasks:
3.1 Catalogue existing mouse models for monogenic, early onset CDs and generate novel ones. Establish a pipeline for the production of novel mutant mice to model CDs using prioritisation. Maintain and distribute the mouse models in and outside of the consortium.
Mouse models with the relevant human disease information, based on literature and database searches and on unpublished information from GENCODYS partners were identified and catalogued by the consortium. This collection was extended by models some of which were generated within GENCODYS, others. Identified outside of the consortium, a list models to be considered was agreed by the consortium in compliance with the following criteria:
- Type of mutations associated to the human phenotype.
- Frequency at which mutations have been identified.
- Involvement of the gene in known molecular or cellular pathways.
- Other models already available or under construction.
- Other reasons for selection (data from e.g. Drosophila).
- Possibilities to further investigate the model once established and phenotyped at ICS.
- Is or will the same gene being investigated at other levels
A total 54 mouse mutant lines have been either newly generated (34 lines) or recovered from outside (20 lines) and made available for the consortium by P10 (ICS, mouse Clinics Institute,, Strasbourg). Six other lines have been generated by P14 (Welcome Trust Sanger Institute/University of Edinburg).
All mouse lines generated during the Gencodys project at ICS (P10) are available for partners during and after the project. During the last 3 years, different type of materials (live mice, embryos, organs) from more than 20 lines were shared with GENCODYS partners for further studies. Lines issued from IKMC (International Knock out Mice Consortium) ES cells are available for wider scientific community via EMMA/INFRAFRONTIER database (https://www.infrafrontier.eu/). Partner 14 has archived and distributed CD models to the broader community beyond the GENCODYS timeframe. Provisions will be made for storing and distributing specific mouse tissues, especially selected brain regions to the other partners (notably for transcriptome studies in WP4).
3.2 Integrated neurological and behavioural phenotypic analysis focusing on evaluation of cognitive dysfunction
One of the objectives of WP3 consists of neurological and behavioural phenotypic analysis of mouse models for ID, focusing particularly on evaluation of cognitive dysfunction (ICS). The behavioural screen contains tests yielding an overview of a wide range of functions (Figure 2). This step is pivotal for observation of potential non cognitive phenotypes, and is also necessary to verify the specificity of the potential cognitive alterations observed. In addition, mice are subjected to specific evaluation of cognitive abilities, including different aspects of learning and memory processes, attention processes and sensorimotor gating.
Mouse phenotypes have been aligned to the corresponding human phenotypes and that of Drosophila mutants. Gencodys has integrated these rich datasets to identify further molecular networks that form novel molecular networks underlying cognitive disorders.
At the end of behavioural pipeline, half of the animals from each group (genotype, treatment) were perfused and brains dissected out, post-fixed and maintained for histological analysis. For the other half of mice, specific brain regions (Hippocampus, Cortex, cerebellum, olfactory bulb, and the rest of the brain) are sampled, snap frozen, then maintained at -80°C. In some cases, all animal brains are processed for structures dissections. The samples are shipped to other partners for subsequent biological analysis in WP4.
3.3 In depth physiological and neuromorphological analysis of these models, including synapse physiology and correlation with the behavioral assays
This task of the Gencodys project has been aimed to determine the physiological consequences of mutations in CD genes using the generated mouse models. These physiologicalcharacteristics are critical to highlight pre- and postsynaptic alterations and other neurobiological features. Comparison of such features between the various models may identify or reinforce common molecular and cellular mechanisms of disease, which may be useful in defining targets for intervention strategies.
Conclusion: Electrophysiological measurements of synaptic transmission parameters were performed in acute brain slices from genetically altered mice bearing mutations in genes homologous to those mutated in humans with brain disorders. Observed changes in the amplitude of synaptic transmission and its short- and long-term plasticity in several lines of mutant mice suggest that these alterations may underlie intellectual disability in humans with functionally similar mutations. Therefore, such changes may be used as a sensitive gauge for testing various pharmacological substances aimed at correction of cognitive impairments. Using mutation effect size values, we worked out a standardised comparative approach which allows an unambiguous ranking of the mutation severity with respect to its impact on synaptic transmission.
3.4 Evaluation of the potential effects of pharmacological or genetic treatments on cognitive functions in selected mouse models
The development of strategies for therapeutic intervention in CDs has been a major ambition of the GENCODYS program. We have been highly committed to develop in vitro, in vivo and ex vivo methodologies for testing pharmaceutical or other sorts of intervention. Mouse is highly suitable for testing specific compounds for their capacity to alleviate neurological and cognitive deficits resulting from CD gene mutations.
In the work towards this deliverable GENCODYS partners analysed behavioural and cognitive abnormalities in mouse genetic models of cognitive disability (CD) and attempted to reverse observed phenotypes by pharmacological means. The following experiments have been performed:
a) reversal of behavioural phenotypes of Ophn1-KO mice by fasudil (HA1077).
b) rescue of Mecp2-KO phenotypes by Epothilone D.
c) rescue of IlRapl1-KO behavioural phenotypes by alpha5 GABA inverse agonist.
d) rescue of mutant Spred1 and Syngap1 touchscreen pretraining phenotypes by the MEK inhibitor PD-0325901.
Chronic treatment with fasudil counteracted the increased vertical and horizontal activity, as well as memory deficits in the object recognition task in Ophn1-KO mice. In a follow up study using new sets of mice, several phenotypes were confirmed in Ophn1-KO mice, including altered motor coordination performance in the rotarod, altered learning and memory as observed in the Y-maze spontaneous alternation task, and delayed spatial learning in the water maze. Moreover, motor deficits were prevented with Fasudil.
On the other hand, Epothilone D, a Taxol analogue, had no effect on behavioural changes observed in Mecp2-KO mice.
We also reported preliminary data on the effects of alpha5 GABA inverse agonist on learning deficits observed in IL1RapL1-KO mice. Despite the treatment, Il1Rapl1-KO males had increased activity and decreased exploration of the centre of the open field, suggesting that treatment had no effect of general activity. On the other hand, alpha5 GABA inverse agonist reverses spatial learning deficits in the water maze observed in Il1Rapl1-KO males.
We attempted to rescue the behavioural phenotype of Spred1 and Syngap mutant mice by treatment with an inhibitor of MEK, the enzyme that phosphorylates ERK. Animals in the drug group received single injections of the MEK inhibitor PD-0325901 for 5 days, and the vehicle group received DMSO only. However, as revealed in the following Figure, this treatment failed to improve cognitive performance of Spred1 mutants. Experiments with Syngap mutants are still ongoing.
3.5 Creation of an accessible web‐based database with bioinformatic comparison of profiles and systems biology approaches to identify pathways, common mechanisms of action and potential therapeutic interventions (with WP2, 4 and 5).
The database has been setup to manage the generation of mutant mice status as well as the phenotyping data. This web-based module covers all status related to mouse models tasks of partner 10. Indeed, it displays the mouse lines generation update including 4 steps: Clones, microinjection, chimeras and GLT (germ line transmission). The status of every step is highlighted with a color code. All GENCODYS lines can be found via the search menu and the listing is available for export (CSV format). The Phenotyping data sub-module is linked to display all results from the Phenotyping pipeline. This will display the details of the tests performed (age of mice, number, etc), measured parameters and data analysis.
Workpackage 4: Functional Neurogenomics
- Dataset of more than 14 quantitative gene expression profiles, 300 sequenced transcriptomes and epigenomes of cognitive disorders (CDs) murine models will be available to the broader scientific and medical community.
- Novel platform for image capture and analyses of synaptic proteins at the level of the neuron, brain region and whole brain scale will have applications in academic and medical research, and technology development.
- Neuromorphological characterization of 43 CD genes in mammalian neurons.
- A novel in vitro system has been developed to screen pharmacological rescue of CD mutant phenotypes. Several candidate treatments were tested. Analysis of Transcriptome profiles may reveal further target pathways suitable for intervention.
- Pharmacological rescue of CD mutant phenotypes in mammalian neurons.
WP4 has used systems approach of mammalian neurons in vitro, integrating large-scale functional genomics and epigenomics, transcriptomics, cell biological, morphological and electrophysiological data. In addition, we used mouse brain tissues as sources of material for specific functional genomic assays. WP4 forms an important bridge between the high‐throughput and genome‐wide approaches of GENCODYS Human Genetics and Genomics research and investigations with Fly and mouse models. The mammalian cell based assays used in WP4 form the basis of assays for drug rescue and genetic modifier experiments and thus identification of therapeutic targets.
WP4 Summary of the main results
This workpackage had 6 major tasks:
4.1 Generation of in vitro neuronal models of CDs using primary neurons and stem cell derived neurons.
Availability of in vitro neuronal models is a vital resource to studying the intact animal and human brain. We have established conditions for studying CD gene mutations using mature neurons and in vivo preparations of the hippocampus slices. Here, we summarize our findings in developing primary hippocampi cultures, and in vitro multi-electrode arrays (MEA) performed in primary neurons with disease relevant mutations.
In the course of developing an in vitro assay for the purposes of studying the impact of disease-relevant mutations on neuronal circuits we unexpectedly observed evidence for mutational buffering. The technology we used to monitor the activity of neural circuits was a tissue culture chamber where a 64 electrode array (MEA, Multi-electrode array) was overlaid with primary cultures of mouse hippocampal neurons. We previously reported that the firing patterns and synchronisation of neuronal networks can be longitudinally recorded from their initial silent phase to a later stable and synchronised phase. Here we evidence for a mutation buffering process and show its general property in overcoming mutations in different classes of genes. We also establish that this buffering is a novel form of plasticity and does not require the classical mechanisms of glutamatergic synaptic plasticity or homeostatic plasticity. Moreover, the buffering process overcomes defects in glutamatergic signalling including those induced by pharmacological perturbations. These studies also demonstrate the utility of MEA systems for studying mutations and disease mechanisms on neuronal networks.
4.2 Characterisation and quantitation of gene expression and chromatin configuration including DNA methylation in inducible mouse KO ES and NSC cells and their differentiation into neurons.
Fourteen different mutants were characterized by quantitative RNA-sequencing for the hippocampus, a brain region which features a primary role in the establishment, maintenance and regulation of long term memory as well as learning, behavior and spatial navigation: in order to minimize sources of noise and variation, hippocampus at P30 days was used for all the samples and for each of the 14 mutated CD genes a specific set of wild-types was used: the latter device resulted into the availability of a remarkably large wild-type transcriptome baseline, which was fruitfully used to extract important information about the degree of inter-individual variation of gene expression.
In addition, for each of the CD mouse models generated in GENCODYS, a comprehensive dataset including genome-wide RNA expression and epigenetic marks distribution (namely DNA methylation, H3k4me1, H3k4me3, H3k27me3, H3k27ac, H3k9me2, H3k9me3). These epigenetic profiles complete and enrich the information provided by the sole RNA-seq. overall, approximately 350 different samples were sequenced for 16 different mutated CD genes. This large dataset constitutes a remarkable, optimal quality, database that will allow us to disclose and dissect the epigenetic foundations of cognitive disorders (CDs). The data is currently being processed for a high-impact publication and will thereafter be made available to the scientific community. This unparalleled in-depth CD gene regulatory database will allow us to disclose and dissect the molecular foundations of CDs.
4.3 Studies of subcellular localisation of CD proteins; and
4.4 Morphological characterisation of CD genes in mammalian neurons.
Following the identification of mutations in genes, the characterization of the neuronal function of the CD gene and encoded protein is a major step for a better understanding of the physiological role and pathophysiological impact during the development of the central nervous system (CNS). The objective of this deliverable was to build a collection of expression vectors for a selected group of CD genes and to establish their subcellular localization pattern after their transfection in the in vitro model for CNS development and differentiation that is widely used in neurobiology, namely the embryonic hippocampal primary neuronal cultures. This deliverable thus provides functional data supporting the deleterious impact of the mutations in CD genes identified in patients with neurodevelopmental disorder.
The analysis was conducted for 43 CD genes (wildtype) and 22 mutant alleles of CD genes. Strikingly, our results highlighted that more than 40% of the CD proteins are expressed in dendritic spines, suggesting that synapses are particularly targeted in neurodevelopmental disorders. To our knowledge, such an extensive functional work in primary neurons, has not been performed for this number of genes. We also demonstrated that most of the mutants analyzed in our study caused either an abnormal localization (e.g. for the ASD gene PTCHD1) or altered synaptic density or shape, thus validating the deleterious effect of the mutations identified in the patients included in GENCODYS project.
4.5 Electrophysiological characterisation of neural network abnormalities in genetic and RNAi models.
Neural networks are thought to be the target of many CD genes and disruption in the connectivity or functional properties of synapses or neurons can all result in changes to the overall network. Using multi-electrode array (MEA) recordings of neuronal networks in vitro are a rapid and efficient measure of gene function. We have investigated β/γ oscillations of the extracellular field potentials evoked by kainate in the CA3 area of hippocampal slices from mice with CD gene mutations. The neural network assays allow ex vivo characterization of brain physiology and connectivity in brains of mouse models, which can be used as a new tool for assessment of the efficacy of pharmaceutical compounds.
4.6 Pharmacological and RNAi rescue of CD mutant phenotypes in mammalian neurons.
We have contributed to the investigation in pharmacological rescue of CD mutant phenotypes. These studies complement the studies described in task 3.4 and Deliverable D35 in which we have attempted to reverse observed phenotypes by pharmacological means for behavioral and cognitive abnormalities in mouse genetic models of CD. In this task we have mainly used in vitro and ex vivo approaches to achieve pharmacological reversal of CD-associated phenotypes. In summary, the following interventions have been established:
• MEK inhibitor PD-0325901 SynGap LTP and input-output physiology;
We investigated whether administration of the mitogen-activated kinase kinase (MEK) inhibitor PD-0325901 could reverse electrophysiological consequences of the heterozygous loss-of-function mutation in the Syngap1 gene. In particular, we were interested to determine if MEK inhibition would be able to diminish LTP deficit in Syngap1+/- mice. We chose to administer this antagonist at a final dose of 20 mg/kg which was well tolerated by all experimental animals. We revealed that LTP was still significantly decreased in Syngap1+/- mice even after 6 day treatment with PD-0325901 (single administration daily by oral gavage). We also tested the compound for affecting input-output relationships reflecting the intensity of the basal synaptic transmission, which were significantly steeper in hippocampal slices prepared from Syngap1+/- mice. Here, our pharmacological reversal experiments demonstrated that abnormal synaptic transmission phenotypes can be rescued in mature mice by inhibiting MEK for 6 days.
• Isoproterenol in SynGAP;
In another series of experiments, we attempted to pharmacologically reverse the LTP deficit in Syngap1+/- mice by activating protein kinase A. Our hypothesis was that if induction of LTP occurs in the presence of elevated cAMP, and as a result, higher activity of PKA, the LTP deficit in Syngap1+/- mice would be diminished. Isoproterenol, a beta-adrenergic agonist was chosen as a drug able to up-regulate PKA activity. Incubation with 1 uM isoproterenol for 10 min before LTP induction by theta-burst stimulation let to a transient up-regulation of fEPSP amplitudes, but the extent of LTP in the test pathway was still significantly less as compared to LTP in WT slices. Thus, a pre-treatment with isoproterenol failed to rescue LTP deficiency in Syngap1+/- mice
• Activity-dependent synaptic mapping mouse line Arc-Venus after pharmacological treatment with ketamine.
Partner14 has employed the activity-dependent synaptic mapping mouse line Arc-Venus. Arc directly binds to PSD95 and assembles sets of proteins that control cognitive functions. Using unique synaptome mapping tools developed within GENCODYS, we treated Arc-Venus mice with different doses of ketamine or sterile saline. Low and high doses of ketamine both increase Arc-Venus synapse density in the short term. However they have opposite long-term effects, with the low dose increasing and the high dose decreasing Arc-Venus synapse density. This suggests that these two doses, despite both being the same NMDA-receptor antagonist, are able to activate different responses at the level of the composition of the post-synaptic density. This is interesting given the recently-identified long-lasting antidepressant effects of a low ketamine dose in both humans and mice. This demonstrates that genes involved with cognitive disorders can be modulated by pharmacological treatments
• BKCa channel opener molecule (BMS-204352) in Fmr1 KO mice.
We investigated a modulator of BKCa channels in a mouse model for Fragile X syndrome (Fmr1 knockout). Fragile X Syndrome (FXS) is the most common form of inherited intellectual disability and is also associated with autism spectrum disorders. Previous studies implicated BKCa channels in the neuropathogenesis of FXS, but the main question was whether pharmacological BKCa stimulation would be able to rescue FXS neurobehavioral phenotypes.
In a collaborative work, Partner 3b used a selective BKCa channel opener molecule (BMS-204352) to address this issue in Fmr1 KO mice, modeling the FXS pathophysiology. In vitro, acute BMS-204352 treatment restored the abnormal dendritic spine phenotype. In vivo, a single injection of BMS-204352 rescued the hippocampal glutamate homeostasis and the behavioral phenotype. Indeed, disturbances in social recognition and interaction, non-social anxiety, and spatial memory were corrected by BMS-204352 in Fmr1 KO mice.
These results demonstrate that the BKCa channel is a new therapeutic target for FXS. The GENCODYS investigations have been published and have been independently confirmed recently by another group. This pharmacological molecule might open new ways for FXS therapy.
Workpackage 5: Bio-informatics
- We have developed a new method to predict genetic interactions between genes, using cancer genome data, that we will apply to (intellectual disability) disease genes.
- We have developed a webserver to predict new genes for a pathway, based on the weighted co-expression of > 4000 gene expression datasets of man and mouse. The tool has been used succesfully to predict one new ID gene.
- Our gene network approaches have identified and validated in vivo that there epistatic/synergistic interactions between variant genes are likely to be a significant contributor to cognitive dysfunction.
- We have shown that the presence of multiple functionally-similar genes is an abnormally frequent feature of de novo structural variants, and provides a significant indicator of pathogenicity.
- The apparent polygenic pathogenticity of these variants again support a role for genetic interactions in ID.
The massive amount of OMICS data generated in a systems biology project like GENCODYS requires state-of-the-art bioinformatics to arrive at concrete, biomedically relevant hypotheses and conclusions. Top-notch specialized bioinformatics groups within GENCODYS have taken full advantage of internal and external data sources in order to obtain both quantitative and qualitative understanding of the biological systems disrupted in cognitive dysfunction, and therewith to derive hypotheses about genes and pathways that are amenable to experimental testing by the consortium. We have harnessed the predictive power of a variety of bioinformatics and systems biology techniques, including comparative genomics and network approaches that combine information from disease networks with protein networks, and quantitative approaches that link publicly available QTLs to molecular data and Genome Wide Association Studies (GWAS).
WP5 Summary of the main results
This workpackage had 3 major tasks:
5.1 Characterizing and predicting CD genes;
5.2 Identification of gene regulatory networks in CD;
5.3 Identification of phenotype networks and the modular nature of cognitive disorders.
Phenotype-genotype based bioinformatics
We have compared human phenotypes with mouse and Drosophila phenotypes in order to decide which aspects of either model can be used to identify human ID genes, basically searching for phenologs: genes that in human and in a model species give rise to the similar phenotypes within each species.
For Drosophila we have examined various phenotypes of the eye, the wing and of behavior. With respect to the eye we have systematically mapped all the eye phenotypes of a set of 270 FLY orthologs of human ID genes, of which 180 did give rise to a phenotype. Given the issues in large scale orthology determination using automated methods, these orthologs have been manually curated and are stored in the sysID database (http://sysid.cmbi.umcn.nl/) as have their Drosophila phenotypes. There are many patterns in the data that link specific Drosophila phenotypes to specific classes of ID genes. The strongest one that we observed was that Drosophila gene orthologs with morphological eye phenotypes, in contrast to genes without phenotypes, are relatively highly expressed in the human nervous system and are enriched for neuronal functions, suggesting that eye phenotyping can distinguish different classes of ID gene. Overall we find that also underlying molecular processes correlate with eye phenotypes. The combined data indicate that ID disorders, despite their extreme genetic diversity, are caused by disruption of a limited number of highly connected functional modules.
For mouse, we first established the generalized concordances between human and mouse phenotypes. Exploiting the structured ontologies that are available for both and for other molecular annotations, we genes being labelled with the same mouse phenotype also had similar human disease phenotypes. Indeed, while far from perfect, the relationship between mouse and human disease phenotypes was demonstrably move informative that the relation between molecular annotations and human disease phenotypes. Many mouse models have been made to study genes of interest to human disease, and thus there is a clear bias of interest towards looking at phenotypes that may have presumed face validity. To discover the mouse models relevant to cognitive dysfunction, we looked at the mouse model phenotypes amongst genes that were variant in individuals with cognitive dysfunction phenotypes. Using patients held within the DECIPHER database that presented with a wide range of developmental and neurodevelopmental abnormalities, we non-exclusively grouped patients according to each specific phenotype and asked whether the orthologues of genes found to be variant in each specific-phenotype patient group were associated with any particular phenotypes when disrupted in the mouse. Forming CNV sets for each of 1,088 distinct human abnormalities, we were able to associate a total of 143 (13%) human abnormalities with mouse model phenotypes. We then considered each of autism and ADHD specifically. For autism, we identified a large number of mouse model phenotypic-associations amongst the orthologues of human genes that, apart from 6 of 104 genes, had not previously been implicated in autism. When these phenotypes were then compared to the phenotypes of mouse models of human orthologues that had been implicated in autism, an excellent correspondence was observed. For ADHD, the cohort was much less powerful than that used to identify autism human-mouse phenotypic correspondences. Nonetheless, an association with genes variant in ADHD patients whose orthologues’ disruption in the mouse gave abnormal learning/memory/conditioning phenotypes was found. Using a gene co-expression network built from human brain region- and development-specific gene expression, we did observe that the orthologues of genes whose disruption in the mouse gave hyperactivity phenotypes were more associated with genes variant in ADHD patients that would be expected by chance, supporting the use of hyperactivity in the mouse as a face valid phenotype for this disorder.
Genetics and Genomics
These bioinformatics studies fall into two parts. Predicting the effect of mutation in non-coding DNA and developing new bio-informatics protocols for predicting ID genes. Regarding the first part: complete genome sequencing, rather than “just” exome sequencing, is rapidly becoming standard and there is an urgent need to be able to interpret variations in non-coding DNA in terms of their potential contribution to disease. The second part has capitalized on the avalanche of genomics data and on the availability on curated lists of ID genes within the consortium to develop and test methods to predict new ID genes.
A model for brain-specific mutation deleteriousness in non-coding DNA,
Regarding the tissue-specific mutation deleteriousness within non-coding DNA we have specifically searched for transcription factor binding sites that are functionally relevant during Brain development. To develop our model we have used the Brainspan data (www.brainspan.org) that measures gene expression in sixteen cortical and subcortical structures across the full course of human brain development. We have then modeled the activity of transcription factor binding sites using the linear modeling approach Ismara. This approach has in the past shown to be able to explain ~10% of gene expression based on the presence/absence of transcription factor binding site motifs. Brainspan RNAseq data have been reanalyzed in order to be able to map transcripts and their transcription start sites to promoters and nearby transcription factor binding site motifs. This then gives a score per transcription factor binding site motif of its involvement in brain development. As we have activity scores per motif and we have gene expression levels per transcription factor we can correlate those with each other, expecting to see that highly expressed transcription factors in certain brain tissues will also have highly active motif, even though the analyses are completely independent of each other. Indeed that is what we observe for a number of motifs and their associated transcription factors. Finally, we have stored the motifs that were predicted to be active during brain development and that are associated with known ID genes in the sysID database.
During this project is has become clear that, despite the enormous heterogeneity of the ID there are a limited number of underlying molecular processes that allow us to actually make predictions for likely ID candidate genes. Nevertheless, it also have become clear that, when one wants to make successful predictions, it is best to take that heterogeneity into account as much as possible via our phenotypic classification that to some extent reflects the underlying biology. Thus various datamining and data integration activities have given us new candidates for ID genes that we have integrated in our SysID database. We have put them in a single integrated list, and parallel to that have put them in separate lists that are based on the SysID clinical classification, as the latter captures part of the underlying biology and therewith increases the “predictability” of ID genes. Similarly, other tasks revolve around establishing links between fly and mouse on the one hand and ID phenotypes on the other hand that was essential in being able to predict new ID genes. Both deliverables have been used in D40 to predict new ID genes.
Conclusions
Capitalizing on the various methods we developed in the GenCodys project, the wealth of genomics data and other data that are becoming available and on the availability of a well annotated list of ID genes we have predicted new candidate ID genes that have been stored in our SysID server and are available to the consortium.
WP6: Potentation
The main objective of the WP6 „Potentiation“ is the dissemination of new ideas and knowledge generated through Gencodys project to the scientific community and the general public. For this purpose, we established various training courses related to the different approaches (i.e. genomics, bioinformatics, animal models, etc.) used in Gencodys, internal and external workshops, as well as a program of scientist/student visiting and exchange. We also created a dedicated Gencodys website.
- Training activities, practical and theoretical courses have been organized at host institutes.
- Research visits for PhD students, postdocs and MDs have been facilitated by offering travel grants and fellowships.
- A total of eight workshops have been organized.
- Two successful international conferences were organized, which were open for non-Gencodys participants.
- So far the GENCODYS has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. Interaction with patients and families organizations to disseminate new diagnostic opportunities and to establish tight interactions for future clinical studies in tight association with families.
WP6 Summary of the main results
This workpackage had 3 major tasks:
Task 6.1. Training activities, practical and theoretical courses at host institutes.
GENCODYS has actively encouraged the education of young researchers and doctors via exchanges between partner Institutes and the teaching of courses by specialized partners. These interactions can efficiently promote the dissemination of knowledge within the GENCODYS consortium, i.e. to transfer new techniques and methodologies from one group to another or to establish and reinforce uniform protocols for research and clinical investigations. In addition, research visits were expected to stimulate progress of specific research activities within the various work packages. Regular exchanges took place involving all partner Institutes.
Specific exchanges:
- Training activity related to WP2, practical and theoretical Drosophila courses (Nijmegen, Berlin).
- Three Courses/ workshops on Next Generation Sequencing, first two at MPI-MG, Berlin, third at USWR-GRC in Teheran (instructed by MPI-MG).
- Course on "Comparative Genomics: from evolution to function", Amsterdam.
- Two rounds of Electrophysiological techniques workshop, Synome, Cambridge
Task 6.2 Research visits for PhD students, postdocs and for MDs
Our aim here was to support exchange of young researchers and doctors between partner Institutes. These interactions can efficiently promote the dissemination of knowledge within the GENCODYS consortium, i.e. to transfer new techniques and methodologies from one group to another or to establish and reinforce uniform protocols for research and clinical investigations. In addition, research visits were expected to stimulate progress of specific research activities within the various work packages. Frequent research visits were envisaged for PhD students, doctors and postdocs from the partners from Iran (P4) and Pakistan (P13) to the European Institutes, in particular those of P1a and P2a. Three main types of exchanges can be depicted :
- Organization of research visits and training networks inside the consortium for Ph.D. students. Possibilities of shared Ph.D. supervision between two groups.
- Research visits for post docs (need for specific technique used in a group from the consortium) - Research visits, training opportunities for MDs from other participating groups.
- Visits of clinical investigators for clinical investigations of CD patients.
To initiate the exchange program and elaborate the guidelines, the GENCODYS Board has established 2 types of exchange allowances that could be requested by each partner, and on the basis of the duration of the visit :
- Short visit (< 3 months) : up to 1500 euros
- Long visit (> 3 months) : up to 500 euros /month + travel costs (similar to Marie Curie allowances).
In both cases, the allowance could only be used for compensation of costs for travel from one Institution to the other and for accommodation costs during the stay at the host Institution.
In total 12 PhD students (including MDs) and 10 postdocs have been engaged in exchanges. Although the Gencodys project is terminated on 1st May 2015, it is highly probable and expected that the exchanges and research visits will be continued since close collaborative and interacting projects have been well established between various groups, and we can assume that this will certainly offer novel opportunities to submit new proposals for future EU calls
Task 6.3 Organization of workshops
This deliverable has been completed with the organization of 8 workshops and conferences. GENCODYS has organized two international conferences, which were open to non-GENCODYS scientists. These conferences have been attended by top-scientists in various expertise areas involving cognitive disorders in order to stimulate multi-disciplinary networking.
Gencodys workshops hosted for the consortium members:
1 Workshop: Cognitive Mouse Models Binding WP1-WP4, 7-8 February 2011, IGBMC/ ICS (partner 10), Illkirch, France (APPENDIX I)
2 WP1 Meeting, 19 April 2011, Ateneo Garden Palace Hotel, Rome, Italy (APPENDIX II)
3 WP3-WP4 Workshop, 5-6 February 2013, University of Edinburgh (partner 14), (APPENDIX III)
4 Workshop: Workshop Strabourg WP1-WP5, “data integration’, 17-18 February 2015, IGBMC/ ICS (partner 10), Illkirch, France (APPENDIX IV)
Gencodys workshops hosted for students, scientist and specialist from Pakistan, Morocco and neighbouring countries:
5 Workshop on Genetic techniques, “New Trends in Genetic Diagnosis of intellectual Disability”, 16-17 September 2014, ISESCO, Rabat, Morocco (APPENDIX V)
6 International Symposium on genetic Diseases, 29-31 March 2015 and Training workshop, “New Trends in Molecular Diagnosis of genetic Diseases”, 1-5 April 2015, Allama Iqbal Medical College (partner 13), in Islamabad, Pakistan (APPENDIX VI)
Gencodys international Conferences hosted:
7 1st International Gencodys Conference, “Integrative Networks in Intellectual Disabilities”, 14-17 April 2013, Paphos, Cyprus (APPENDIX VII)
8 2nd International Gencodys Conference, “Integrative Networks in Intellectual Disabilities”, 27-29 April 2015, Chania, Crete, Greece (APPENDIX VIII)
Task 6.4 Dissemination activities, web‐site forum, interaction with patients organizations
So far the GENCODYS program has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. The impact of many of the publications has been very high and several landmark publications. Top publications have appeared in Nature (n=4), Nature Genetics (6), Nature Neuroscience (6), Nature Communications (2), Nature Protocols (3), Nature Cell Biology (1), Nature Methods (1), Science (1), Science Reports (1), Cell (1), Developmental Cell (1), Neuron (3), Lancet Neurology (1), New England Journal of Medicine (1), Molecular Psychiatry (6) and 15 publications in the American Journal of Human Genetics. In addition, Gencodys PIs have produced authoritative reviews and opinion papers in Annual Reviews in Neuroscience, Annual Reviews in Genetics, Current opinion in Neurobiology, and Trends in Genetics. The broad multidisciplinary and translational character of the GENCODYS program is reflected in the broad subject categories of the journals in which GENCODYS papers have been published.
Intellectual property: Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014
Task 6.4.4 Website and website forum
The GENCODYS website (http://www.gencodys.eu) has been designed at the start and is working and being improved by the consortium members on a regular basis with among others online updates of publications and meetings. A specific section provides info for the public in general and patients and their families in particular. It includes general information on the project and its aims, and background information on genetics and inheritance patterns, including answers to some frequently asked questions on technical aspects. Also, the ways in which patients might benefit from the work done by the Gencodys consortium is addressed. The website will link to other websites where more information can be found such as the Eurogentest website, and the Rare disease communities website.
Task 6.4.5 Patients organizations
Patients with a cognitive disorder are the begin and endpoint of the GENCODYS program. Large efforts have been made to identify the genetic underpinnings of their CD and to get better insight into the molecular and cellular mechanisms of disease. Cumulative knowledge is being used for the benefit of the patient, initially for improved opportunities for making (prenatal) genetic diagnoses, genetic counselling, clinical management and improved standards of care. In the long run, GENCODYS has led the foundation for development of methodologies for intervention studies. GENCODYS acknowledges the importance of an early and close involvement of patients, patient organizations and patient advocate groups in order to achieve optimal benefits and impact for our target group. To that end, patient organizations have been represented within GENCODYS by inclusion of The European Genetic Alliances Network (EGAN/VSOP; p12) in the project. This has stimulated, facilitated and shaped the interaction between the scientific partners in GENCODYS and patient organisations. There have been ample meetings between GENCODYS PIs and a multitude of patient organizations during Annual GENCODYS meetings, specifically organized meetings, and in parallel to European Conferences such as the European Society of Human Genetics (ESHG), other EU project meetings, like FP7 PatientPartner (www.patientpartner‐europe.eu) and EURORDIS.
Potential Impact:
The GENCODYS main goal was to gain better insight into molecular and cellular processes that are key to neurodevelopmental processes and the higher cognitive functions of the human brain. We have studied a group of human disorders that constitute one of the largest unsolved problems in modern medicine. The basis of our research have been the genes and their cognate proteins that are disrupted in human genetic disorders that are characterized by impaired cognitive functioning (Cognitive Disorders; CD). Since disruption of the functions of these genes and their associated molecular pathways are associated with CD, it is obvious that their normal functions are crucial for development, functioning and maintenance of the healthy human brain. We have considerably impacted the field through the identification of a plethora of new causative genes and their associated molecular networks. Such new knowledge can and will be applied immediately to improve the possibilities for genetic testing and counseling for patients and families who are confronted with disorders of CD. Not only has the GENCODYS consortium heavily invested in the development and implementation of diagnostic testing, our research activities were and still are also targeted to device new strategies for therapeutic intervention in patients, e.g. through application of ingenious screens in primary neuronal cultures and Drosophila to identify small molecules that suppress the biochemical or phenotypic defects associated with CD gene mutations.
Specific scientific and technical impact
GENCODYS united a multidisciplinary team of European research groups, each of which represent leading international players in their discipline, as evidenced by publication records in top journals. The complementary disciplines of the participants reflects the translational nature of the GENCODYS program. We have implemented innovative high-throughput technology to significantly speed up the search of genes that underlie human disease. In addition, extensive functional studies in genetic models has provided new insights into gene function and into the way that genes interact to each other in health and disease. Model organisms have been generated and used for drug screening, which will aid the design of drugs and treatments. The scale of our approach reaches far beyond the capacities of local or national networks and is unparalleled by other international collaborative efforts in this important field of medical research.
Projected outcomes:
1) Identification of approximately at least 50 novel CD genes. These genes will likely be involved a variety of different CD conditions, such as intellectual disability (ID), autism spectrum disorders (ASD), ADHD, epilepsy, schizophrenia and others, since these conditions have a shared molecular etiology. Therefore, GENCODYS has also provided many new targets for further investigation in multifactorial forms of CDs.
2) Diagnostic tools to be able to establish a molecular diagnosis of patients with this extremely heterogeneous disorder.
3) Establishment of pathways that will a) give vital information about the development of the human brain and b) enable therapeutic intervention in CD patients.
The GENCODYS program has exceeded the expectations for these outcome measures, which illustrate the enormous scientific and medical impact of GENCODYS research:
Ad 1) GENCODYS research has identified causative mutations in some 150 genes that had not been implicated in a cognitive disorder before. The number of new genes might actually be larger, as we have also a list of candidate CD genes that await confirmation in unrelated CD patients (matching cases). Thus, the GENCODYS program has contributed approximately 150 new genes to the ~700 CD genes that are known to date (350 genes at the start of the GENCODYS). Mutations were found to cause forms of CDs of all kinds of inheritance pattern (de novo, recessive, X-linked and oligogenic as in ASD). Moreover, we have made an effort to carefully catalog the CD phenotypes that are associated with these mutations, which is not usually done in next generation studies in other patient cohorts. Thus, for individual patient and familial cases we have diagnostic profiles that comprise the cognitive features (ID, ASD, epilepsy, motor function, behavior) as well as co-morbidities.
Ad 2) Next generation sequencing methodologies such as WES and WGS have proven to be highly effective in the identification of mutations underlying CD of all inheritance patterns (de novo, recessive, X-linked and oligogenic as in ASD). These genome-wide approaches allow the analysis of all genes, or selected groups of genes (gene packages) in a single experiment. As a consequence, WES and soon WGS will become the methodologies of choice for a first-step genetic analysis of patient with an unknown (cognitive) disorder. In fact, at the institute of P1a WES is already offered as a diagnostic test and soon this will be extended by WGS protocols. The latter have the potential to also pick up genetic and genomic variants in non-coding regions, but the assignment of causality of such variants needs further insight in biological mechanisms as well as the development of tools for bioinformatics analyses. To that end, GENCODYS partners are committed in setting up internationally shared databases (in collaboration with the Sanger-based DECIPHER database/UK DDD project and other databases) for the registration of DNA variants that have been identified in CD patients. In any case, based on our results and extensive experience, we propagate the implementation of next generation sequencing technology as a first step in diagnostic testing of unexplained forms of intellectual disability. This will reveal a conclusive diagnosis in 30-70% of the patients (depending on the inheritance pattern) and thus a diagnosis early in life. This will obviate a plethora of needless invasive and expensive investigations that can carry on for years if not decades. Instead, better registration of the specific genetic aetiology and the constellation of cognitive deficits and other co-morbidities will offer a better standard of care for the respective patients.
Ad 3) Bio-informatics analyses and functional studies in cell and model organisms within GENCODYS of the 150 new genes and previously identified ones has demonstrated beyond doubt that CD genes converge into common molecular networks. This insight is important for developing strategies for therapeutic interventions, which can be targeted to common molecular processes and networks, and not strictly at single genes. Gencodys has made major advances towards development of therapeutic strategies. An unparalleled collection of animal models, both Drosophila and mouse, have been generated and characterized at a wide variety of neurodevelopmental, behavioural and cognitive parameters. These models are invaluable for pre-clinical studies of pharmaceutical compounds. In fact, GENCODYS has already started in testing a number of compounds, both in newly developed in vitro and ex vivo systems as well as in selected mouse and fly models. We anticipate that the wealth of new knowledge and innovative resources developed within in GENCODYS will become a basis for a plethora of follow-up studies across Europe, both in fundamental research and in translational research for developing therapeutic interventions. Partners who have successfully collaborated in GENCODYS will be actively involved in such follow-up projects, reinforcing existing collaborations and establishing new ones. Renewed support from the EU will be instrumental to benefit optimally from the established resources and to continue to increase the impact of EU research teams in translational research of the large group of genetic cognitive disorders. At present, GENCODYS groups are engaged in renewed EU-applications within the frame of Horizon 2020, E-RARE, ERA-NET-Neuron, EU COST, Marie Curie ITN, and the Human Brain Project.
Dissemination activities
Dissemination has been an integral part of the GENCODYS program and details have been reported in the summary for workpackage 6 (Potentiation). In short:
- Gencodys website: www.Gencodys.eu
- Training activities, practical and theoretical courses have been organized at host institutes.
- Research visits for PhD students, postdocs and MDs have been facilitated by offering travel grants and fellowships.
- A total of eight workshops have been organized.
- Two successful international conferences were organized, which were open for non-Gencodys participants.
- So far the GENCODYS has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. The impact of many of the publications has been very high and several landmark publications. Top publications have appeared in Nature (n=4), Nature Genetics (6), Nature Neuroscience (6), Nature Communications (2), Nature Protocols (3), Nature Cell Biology (1), Nature Methods (1), Science (1), Science Reports (1), Cell (1), Developmental Cell (1), Neuron (3), Lancet Neurology (1), New England Journal of Medicine (1), Molecular Psychiatry (6) and 15 publications in the American Journal of Human Genetics.
- GENCODYS PIs have produced authoritative reviews and opinion papers in Annual Reviews in Neuroscience, Annual Reviews in Genetics, Current opinion in Neurobiology, and Trends in Genetics. The broad multidisciplinary and translational character of the GENCODYS program is reflected in the broad subject categories of the journals in which GENCODYS papers have been published.
- GENCODYS research has made highlights in the media, such as national newspapers, scientific commentaries, radio and television, and social media.
- GENCODYS PIs have actively contributed to teaching (primary and secondary schools), lecturing (Masterclasses, Summerschools etc) and other activities that created awareness of our research (Science fairs, books).
- Intellectual property: Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014
Resources developed within GENCODYS
The development and application of high-throughput technologies for CD research have been characteristic in the GENCODYS strategy. By that, we have developed knowledge and a range of resources that will serve the scientific community and which should facilitate further (EU-funded) research. Examples of such valuable resources are:
- A biobank containing ~13.000 DNA samples and other biomaterials from CD patients of known and still unknown etiology. Patients have been systematically investigated and phenotypes are collected in a database by using standardized classifications. This unique biobank is a highly attractive resource for follow-up collaborative research.
- Protocols for whole-exome and whole-genome sequencing and subsequent filtering of DNA variants. Such protocols have already been implemented in diagnostic testing at the Radboud university medical center, Nijmegen, The Netherlands and are being used for genetic investigation of patients that are referred from all over the world.
- Throughout the GENCODYS grant we have developed technology for live imaging of fluorescent fosmid-based gene expression reporters in Drosophila using multi-view Selective Plane Illumination Microscopy (SPIM). Additionally, we have developed a BigDataServer that makes it possible to access remote data sources in the same way as if they were on a local computer. Both tools are open source and thus can be adapted and extended for the purposes of various image-based projects, even beyond the GENCODYS project.
- Aktogen Ltd, an SME partner within GENCODYS, has developed a series of semi-automated screens for behavioral assessment of Drosophila mutants. In addition, they have developed a vacuum injection drug delivery system for Drosophila and a proof-of-principle drug target and compound screen in EHMT1 mutant fly lines. This work has transpired in Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014.
- Large collections of Drosophila mutants have been systematically investigated in a neurobiological pipeline (almost 300 CD gene mutants) and in an unbiased screen for learning and memory. This collection of fly mutants will serve the scientific community by allowing further studies into mechanisms of disease as well as pharmacological intervention projects.
- A unique collection of CD mouse models has been generated, consisting of 40 newly generated mutant lines and 20 recovered lines. These mutants have been analyzed (and still are) following systematic phenotyping pipelines, which is providing valuable insights into the corresponding human CDs. These mouse mutants represent an extremely rich collection of animals that will be made available to the scientific community for further mechanistic studies as well as pre-clinical investigations of novel intervention strategies.
- Several pharmaceutical interventions have been tested in our CD mouse and fly models. Some of these revealed phenotypic rescue of behavioral or neurobiological abnormalities and thus hold promise for further investigation as potential leads for intervention in CD patients.
- Development of an ex vivo protocol for testing various pharmacological substances aimed at correction of cognitive impairments. SME partner Synome showed that changes in the amplitude of synaptic transmission and its short- and long-term plasticity in acute brain slices of several lines of mutant mice may underlie intellectual disability in humans with functionally similar mutations. Therefore, such changes may be used as a sensitive gauge for testing various pharmacological substances aimed at correction of cognitive impairments.
- Development of a novel in vitro system to screen pharmacological rescue of CD mutant phenotypes based on primary neuronal cultures from a PSD-95-eGFP mouse line developed by Synome. The synaptic signals are analyzed in a reproducible 96-well format amenable for high-throughput screening, potentially allowing testing of pharmacological treatments.
- Innovative tools have been developed for bio-informatics analyses that facilitate the identification and refinement of common molecular networks that are affected by groups of CD genes.
Socio-economical impact
Based on GENCODYS results we strongly recommend the implementation of next generation sequencing (NGS) technology as a first step in diagnostic testing of unexplained forms of intellectual disability. The costs of NGS methods have dropped considerably over the last year and have reached a price of less than 1000 € per sample. This means that NGS is cost-effective approach in comparison to microarray analyses (which can only identify CNVs) and Sanger sequencing protocols (which are gene-specific). This will reveal a conclusive diagnosis in 30-70% of the patients depending on the inheritance pattern and depending on the used strategy. Alternative NGS strategies include whole-exome sequencing (WES), WES of selected genes (panels), and whole-genome sequencing (WGS). WGS appears to be most robust and effective in terms of diagnostic yield, but the choice of the strategy may be affected by other considerations (costs, locally available technology and expertise, and ethical issues such as the risk of chance findings).
In any case, application NGS has the potential to provide a conclusive diagnosis early in life. This will obviate a plethora of needless invasive and expensive investigations that can carry on for years if not decades. The economical advantage is obvious. In addition, for the patient and the family the knowledge of the exact cause of the disorder has an enormous impact: 1. It may change the standards of the medical care that is given to the patient; 2. It will have prognostic value in terms of co-morbidities that are typical for mutations in a specific gene; 3. And it strongly affects the choice for reproductive options of the parents of an affected child
Medical and Societal impact
As mentioned above, the results of the GENCODYS program are of immediate benefit for patients, their families and society as a whole, because of the improvement of diagnostics for CDs. Given the rarity of the genetic defects, international collaboration remains essential, both with medical professionals and the patient community to resolve medical research questions. Based on gathered knowledge and experience we are committed to develop a modular Standard of Care for specific CDs, such as those that are due to mutation in a group of gene affecting common epigenetic mechanisms (the chromatinopathies). To realise an infrastructure for this and to disseminate and implement Standards of Care, we aim to further integrate European Centres of Expertise into a European Neurodevelopmental Disorders (NDD)-network that will pave the way for, or become part of, an intended European Reference Network (ERN) for “rare diseases of brain development and rare intellectual disabilities”. It is our mbition to transform this into one of the expected 23 ERNs that will be established in 2016 for clusters of rare diseases, as stipulated by the European Commission.
In addition, we have set the stage for applying the generated knowledge and technology to foster the identification of lead compounds to mitigate the disease phenotype of genetic conditions. The conceptual advance to delineate molecular networks that are commonly disrupted in cognitive disorders opens perspective to treat patients with mutations in different genes with common compounds. For example, based on advances in the understanding of the neurobiology of chromatinopathies, targeted therapeutic agents might be designed to correct the underlying mechanisms of neural dysfunction, and thus will have to be assessed in patients.
For obtaining maximum impact for our target group, the CD patients and their families, we have maintained close interaction with patient organizations throughout the GENCODYS program. This was facilitated via the embedding of a patient umbrella organization withan GENCODYS: the European Genetic Alliance Network, represented by VSOP (Vereniging Samenwerkende Ouder - en Patiëntenorganisaties – The Netherlands). We aim to continue and extend such close interactions with patient networks, in order to be prepared for future clinical trials.
List of Websites:
www.GENCODYS.eu
Scope: Cognitive disorders (CD) impose a major medical and socio-economical problem owing to their high incidence in our population. Intellectual disabilities alone account for 10% of the total health care expenditure in most European countries. Mutations in over 500 different genes have been associated with early onset (i.e. before 18 years of age) cognitive disorders, such as intellectual disability (ID), autism spectrum disorders (ASD), and schizophrenia. Yet, the molecular basis for approximately half of the patients with an early onset CD remains unknown, which creates an enormous burden to families confronted with such a disorder. Whereas early onset CD collectively impose a major medical and socio-economical problem, individual types of such CD are rare, which has precluded systematic studies that aim to provide insight into pathogenic mechanisms in CD. Our research is fundamentally different in that we aim to identify and investigate the key networks that are commonly disrupted in CD. We have used a systems biology approach to gain pathways-based insights into mechanisms leading to cognitive dysfunction in humans.
Conclusion: Research within Gencodys has been highly successful. The achievements are impressive and represent a major advancement in our understanding of the molecular underpinnings of cognitive disorders. As a consortium we have uncovered some 150 new genes that carry causative mutations in such neurodevelopmental disorders. We have also demonstrated beyond doubt that CD genes converge into common molecular networks. This insight is important for developing strategies for therapeutic interventions, which can be targeted to common molecular processes and networks, and not strictly at single genes. Gencodys has made major advances towards development of therapeutic strategies. An unparalleled collection of animal models, both Drosophila and mouse, have been generated and characterized for a wide variety of neurodevelopmental, behavioural and cognitive parameters. These models are invaluable for pre-clinical studies of pharmaceutical compounds. In fact, GENCODYS has already started in testing a number of compounds, both in newly developed in vitro and ex vivo systems as well as in selected mouse and fly models. We anticipate that the wealth of new knowledge and innovative resources developed within in GENCODYS will become a basis for a plethora of follow-up studies across Europe, both in fundamental research and in translational research for developing therapeutic interventions. Partners who have been collaborating in GENCODYS will be actively involved in such follow-up projects, reinforcing existing collaborations and establishing new ones. Renewed support from the EU will be instrumental to benefit optimally from the established resources and to continue to increase the impact of EU research teams in translational research of the large group of genetic cognitive disorders.
Project Context and Objectives:
Concept:
Cognitive disorders (CD) impose a major medical and socio‐economical problem owing to their high incidence in our population. Intellectual disabilities alone account for 10% of the total health care expenditure in most European countries. Mutations in over 400 different genes have been associated with early onset (i.e. before 18 years of age) cognitive disorders, such as mental retardation, autism, some neurodegenerative disorders, and even schizophrenia. Yet, the molecular basis for the majority of patients with an early onset CD remains unknown, which creates an enormous burden to families confronted with such a disorder. Whereas early onset CD collectively impose a major medical and socio-economical problem, individual types of such CD are rare, which has precluded systematic studies that aim to provide insight into pathogenic mechanisms in CD. So far, functional studies have been restricted to a few single gene defects, such as for the Fragile X syndrome. Our research will be fundamentally different in that we will identify the key networks that are commonly disrupted in CD. We follow a systems biology approach to gain pathways‐based insights into mechanisms leading to cognitive dysfunction in humans. The scale of our approach has been unprecedented world‐wide, and execution of this program and warrants the leading role of Europe in this important field of medical research. A schematic of the GENCODYS strategy is depicted in Figure 1. In short:
(1) Identification of genes involved in cognitive disorders. Despite considerable progress in the identification of genes underlying CD, the vast majority of the molecular defects underlying CD are still unknown. Therefore, our first objective is to identify new genes causative of CD by developing and implementing high‐throughput strategies applied to a unique large cohort of families with multiple CD patients.
(2) Elucidation of molecular networks that are commonly disrupted in CD. Emerging evidence from genetic and neurobiological research provides evidence for a limited number of molecular and cellular pathways that are shared by the various genetic CDs. These pathways are associated with abnormalities of dendritic structure, dendritic spine morphology and synaptic plasticity, suggesting a common pathophysiological mechanisms for the cognitive impairment associated with intellectual disability, autism and other CDs. Thus, the second stage of our strategy is to systematically explore this concept by elucidation of molecular networks that are commonly disrupted in CD. Besides elucidating new common pathways in CD, we will further expand the already recognizable common networks, pathways and physiological mechanisms that we and others have identified: Rho GTPases, epigenetic regulators of the activity of neuronal genes and postsynaptic glutamate receptor complexes using a variety of functional genomics strategies. We shall make optimal use of genetic models that are the cornerstones of neuroscience, such as mouse and fruit fly. Powerful tools and a comprehensive understanding of the molecular processes and mechanisms underlying cognitive (dys)function are key to the third, translational part of our research.
(3) Identify genetic modifiers and small compounds that modulate the disease phenotype. The third part of our research program is to resolve the molecular underpinnings of the large degree of clinical variability that is typical for all types of CD, even among patients carrying identical gene mutations. Genetic modifier screens in available Drosophila models for CD will be used to reveal phenotypically relevant genetic interactions and molecular networks. Moreover, drug screens shall be conducted in fly and cellular models for CD, and subsequent selected drug testing in mouse models will eventually potentiate pharmacological intervention in CD patients.
Project objectives:
The GENCODYS project aims at gaining comprehensive and conceptual advance in understanding the molecular basis of human brain development in health and disease. GENCODYS proposes a set of complementary ambitious objectives:
1. Identification of a plethora of novel genes involved in cognitive disorders by application of powerful cutting‐edge genetic research to unique patient material collected within the GENCODYS consortium.
2. Systematic generation of Drosophila melanogaster models to study molecular pathways and neurobiological mechanisms involving genes mutated in human CD.
3. Large scale generation of mouse models of CDs and systematic characterization of multiple key electrophysiological and behavioural properties.
4. A systems biology approach applied to functional neurogenomics analyses to gain insight into genetic and epigenetic networks involving CD genes.
5. Bioinformatics data integration involving collected data from the GENCODYS program with public genomics data.
6. Unbiased identification of small chemical compounds that have the propensity to improve a CD phenotype, initially in primary neuronal cultures and Drosophila, and subsequently in available mouse models.
7. Consolidation of a leading role of Europe in translational research on cognitive disorders by scientific excellence in complementary research areas and their integration into a systematic research pipeline.
Progress beyond the state of the art:
The main innovation of the GENCODYS program lies in the systematic and large‐scale integration of research fields that thus far have been largely separated. Medical genetics research, one the one hand, has been of great importance for diagnostics, prognosis and disease management. Yet, the elucidation of these genes in itself provides only limited insights into the associated disease mechanisms and molecular pathways that are disrupted. On the other hand, spectacular conceptual breakthroughs have been obtained in modern neuroscience research, through exploiting the powerful tools provided by genetic manipulation of model organisms. However, there are few examples where such research is directly applied to models for human CD, and only sporadic, if any, studies where comparative research is conducted for a larger number of CD genes. GENCODYS has the ambition to bridge this gap through the convergence of experimental efforts and the combination of an unprecedented set of tools directly relating to normal cognitive dysfunction. Clearly, such integration of experimental strategies is expected to yield information that is superior in value to the sum of its individual components.
GENCODYS will thus address neurobiological concepts fundamental to cognitive dysfunction, and will do so in an integrated manner using a number of novel and/or strongly innovative approaches, which in itself are beyond state‐of‐the art (Table 1). This partnership of groups at leading research institutions and selected biotech SMEs will also deliver new enabling technologies and proof of principle for modulation of CD disease phenotypes. This will be realized by the design and implementation of screening protocols whereby small molecule screening will be performed by following genome‐wide approaches and the design of assays that are amenable to small molecule screening in primary neuronal cultures and in hundreds of Drosophila models for CD. Furthermore, mouse models for CD will follow these studies, and are amenable to medium scale genetic manipulation and provide advantages for investigating the mechanisms in both physiological and neuropathological settings in a mammalian context.
Overall strategy and general description:
To achieve an optimal workflow for our translational research strategy, we have defined six work packages, which focus on the individual stages that cover the path from the clinic to model organisms and functional research and back again to the clinic (Figure 1).
1. HUMAN GENETICS & GENOMICS
The foundation to our strategy is an ever expanding collections of patient and family material: currently we have DNA samples from >600 families with X‐linked MR (www.euromrx.com) a unique collection of ~500 multiplex families of recessive forms of CD from populations with a high level (40‐70%) of consanguineous marriages, and >10,000 sporadic patients with various types of CD (MR, autism). Clinical data and massive amounts of genetics and genomics data of these patients will be registered in a patient database. Application of state of the art mapping and sequencing technologies should reveal a large number of novel CD genes.
2. FLY MODELS
Gene mutations known to be associated with CD or newly identified in subproject HUMAN GENETICS & GENOMICS will be modelled in powerful models for neuroscience research, Drosophila melanogaster and the mouse. Drosophila will be the model of choice for genome‐wide genetic screens, high throughput phenotyping, including assessment of learning and memory, high‐resolution analysis of gene expression. Genetic modifier screens and chemical compound screens will be carried out in selected Drosophila models to reveal phenotypically relevant genetic interactions and molecular networks that can be manipulated by endogenous compounds.
3. MOUSE MODELS
The mouse will be used for studying higher cognitive processes for subsets of CD genes. Over 50 mouse models are already available for CD‐associated genes involved in Rho GTPase signaling, epigenetic pathways and NMDA receptor complex and postsynaptic density and others can be readily obtained. The effects of gene mutations towards the general behaviour repertoire (e.g. motor function, novelty seeking, anxiety) and specific cognitive functions (e.g. attention, short and long term memory, strategy utilization) will be explored. In addition, effects of mutations on general architecture of the nervous system, neuronal networks, branching, synaptogenesis and synaptic plasticity and electrophysiological properties will be determined. Whenever appropriate, these models will also be used for in vivo validation of therapeutic approaches.
4. FUNCTIONAL NEUROGENOMICS
The studies constitute a variety of complementary High Throughput methodologies for systematic analysis of the molecular and cellular networks that are disrupted in CD. They will encompass the determination of expression patterns in the nervous system and the subcellular localization of CD- associated mRNAs and proteins, using expression profiling, proteomics, in situ hybridization and live imaging. Innovative strategies for systematic analysis of gene expression in mouse brain, and cluster analysis results will reveal genes with similar patterns of expression indicative of common pathways. Furthermore, neuron‐based assays will be exploited to conduct screens for small molecules that can rescue disrupted neurophysiological processes brought about by engineered CD gene mutations.
5. BIO‐INFORMATICS
A key component of our project will be the integration of the massive amounts of data that will be generated at all stages of the program. To optimally exploit the data generated by our systems biology approach we will harness the predictive power of a variety of bioinformatics techniques. Functional characteristics of CD genes and corresponding proteins, mutation patterns, homozygous loci, mutations, CNVs, expression levels, epigenetic data and phenotype information will be cross‐referenced to identify significant enrichments that are indicative of causation. This will be complemented by a systems biology approach to delineate the biological coherence of the 400+ known CD genes by analyzing their involvement in common pathways using public genomics data.
6. POTENTIATION
This subproject encompasses activities that encourage the formation, development, and exploitation of new ideas. Envisaged are training and networking activities, organization of workshops, dissemination of results through websites, meetings, and scientific publications. We will actively exploit the possibilities for patenting, using fair measures for protection of intellectual property rights.
A consortium of scientific complementarities and balance:
The partners have been carefully chosen for the GENCODYS consortium based on the expertise that they contribute to the project in the fields of genetics, neurobiology, Drosophila and mouse biology, bioinformatics and molecular and cellular biology. Partners were also chosen based on their outstanding knowledge and state‐of‐the‐art equipment. The project has a truly European dimension with 13 scientific partners from five European countries (Germany (2), France (3), Austria (1), UK (2), Netherlands (3)), two SMEs from the UK and a private entity representing Patient organization across Europe (seated in Belgium). Moreover, two major Institutes from Asia are full participants of GENCODYS, one from Iran and one from Pakistan. Finally, the inclusion of partners is based on several previous and ongoing collaborations among partners on projects related to the current Integrating project.
In order to achieve high synergy and conduct research as competitive and efficient as possible, the members of the consortium have been chosen dependent on their excellence in specific fields to reach the aim of successful molecular genetic neuroscience research. Therefore, the expertise which is present in GENCODYS is highly complementary and is balanced between the fields of clinical and molecular genetics, research with model organisms, bio‐informatics and molecular and cellular neurobiology. While each group has specific expertise in subfields of the general field of expertise, most fields/subfields of expertise are covered by more than one group so that the team with the lead expertise will serve as reference centre for the respective work.
Project Results:
GENCODYS research is conducted in five scientific work packages. Workpackage (WP) 1 aims to identify the molecular defects underlying early onset CD, which is the key to diagnosis, understanding the pathogenesis and eventually, treatment of this heterogeneous disorder. The genes identified in WP1 are the basis for investigation of the respective protein and associated functional networks. In WP2 we use the fruitfly and in WP3 the mouse as model organisms to gain insight into genetic networks and their role in normal and disrupted neurodevelopment, learning and memory and other behaviours. In WP4 we investigate the molecular and cellular mechanisms that are associated with disruptions of genetic networks that are key to CD: synaptic protein complexes, epigenetic pathways, cytoskeleton structure. Finally, bioinformatics in WP5 has an important role in reconstructing genetic networks underlying CD by large scale systematic integration of datasets generated in the other workpackages. In addition to these five scientific workpackages, two work packages dedicated to dissemination and management:
From the start, Gencodys has met the goals we have set, and in fact for many parts of the project scientific progress is well ahead of schedule. Collective output in terms of new data as well as scientific publications has been highly impressive for the entire GENCODYS program. Below, we discuss the highlights of GENCODYS for each of the workpackages. The dissemination of GENCODYS research to various stakeholders, i.e. the medical and scientific community, scholars and students, patients and patient organizations and society as a whole, has been the focus of WP6 (Potentiation).
WP1: Human Genetics and Genomics: Identification of both convergent and distinct genetic pathways genes in ID and in autism spectrum disorders
Highlights of achievements of workpackage 1:
- The GENCODYS biobank and clinical database of patients with a cognitive disorder contains >13,000 individuals with a CD, which significantly surpasses the number which we aimed for at the start of the GENCODYS project.
- We have developed protocols for targeted sequencing and whole-exome sequencing that enable the identification of causative mutations in 30-70% of the patients, depending on the familial occurrence of the disorder.
- We have implemented protocols for next generation sequencing in a diagnostic setting.
- The thus obtained genetic diagnosis provides better opportunities for (prenatal) diagnostic testing, counseling and disease management in the respective families.
- Besides mutations in genes already linked to CD previously, we have identified at least 150 new CD genes that already have been or will be described in a variety of high-impact journals.
- We identified many new genes underlying neural migration disorders. These studies have highlighted common molecular and cellular networks underlying such disorders. For example, the implication of microtubule disruptions is associated with malformations of cortical development (MCD; tubulinopathies)) and defective glycosylation of dystroglycan in disorders that are now collectively known as muscular dystrophy-dystroglycanopathies (MDGGs).
- We have uncovered a contribution of mosaic parental germ-line mutations in recurrent forms of neurodevelopmental disorders (MCD).
- We have revealed that de novo mutations and genomic disruptions underlie a large number of cognitive disorders, such as ID, ASD and rare disorders like Möbius syndrome.
WP1: Summary of the main results
This workpackage had 5 major tasks:
Task 1.1. Recruitment of families and patients.
The foundation of the GENCODYS program is an ever expanding collection of patient and family material. The recruitment of patients and their families has been focused on several groups of patients distinguished in five groups based on the pattern of inheritance of the cognitive dysfunction: 1. Autosomal Recessive Intellectual Disability (ARID); 2. ARID or X-linked ID (XLID); 3. XLID; 4. Isolated CD; 5. Isolated Autism Spectrum Disorder (ASD). In this classification, ID and CD are used interchangeably. Patients classified as isolated ASD do not have other cognitive deficits besides their ASD. Sporadic patients with ID/ASD have typically de novo dominant mutations or a multifactorial etiology (several genetic variants). The numbers of patients are indicated in the Table 1. The given numbers are an underestimate as families (categories 1-3) are listed by only one affected per family, whereas there are 2-20 patients for each of these families.
For all patients informed consent has been obtained using protocols that have been approved by the local and/or national competent authorities for research with human individuals. The consent comprised the collection of blood or other patient materials, genome-wide genetic analyses aimed at the elucidation of the molecular basis of the genetic disorder in each of the patients/families, and the approval to use the results of these studies in medical and scientific reports and presentations.
The gencodys program has exceeded its huge ambitions set at the start of the program. A valuable cohort of CD patient has been established, which is unique both in its size and in the extent of clinical information that has been collected on individuals within this cohort. This cohort has been the basis of successful genetic and fundamental research within the GENCODYS project and will remain a valuable resource for further studies of this kind as well as for translational research and future clinical trials.
Task 1.2. Clinical characterization, documentation and establishment of cell lines.
All families and patients ascertained until now have undergone extensive clinical examinations by the Clinicians connected to the various GENCODYS groups. These data have been recorded by a standardized form that was established in the first year of the project. The standardization allows the systematic entry of clinical records in the GENCODYS database. The virtue of the database is that queries can be made for patients that present a specific constellation of clinical features (i.e. the main cognitive defect as well as co-morbidities). Such queries allows the selection of patients carrying a mutation in a candidate gene-of-interest, which is often used for the identification of matching cases (mutation in the same gene) that support the qualification as a new “CD gene”.
The Clinical database is connected to the molecular database. The molecular database contains the results of molecular genetics analyses carried out in each of the patients/families, such as the presence of copy number variants (CNVs), homozygous regions, and the presence of rare DNA variants (rare variants and mutations) segregating with the phenotype. Also the molecular database has a high utility for identifying matching mutations.
Besides the collection of blood samples for genetics and genomics studies, we have also collected other biomaterials for other molecular and cell-based investigations. For almost 1200 patients we have established Epstein-Barr virus transformed lymphoblastoïd cells lines. These cell lines are highly useful for transcript analyses, transcriptome studies and epigenetic analyses. In addition, we have established over 150 fibroblast lines from skin biopsies of patients. Fibroblast lines from several patients with confirmed causative mutations, such as Kleefstra syndrome and Koolen-de Vries syndrome have been used to generate induced pluripotent stem cells (IPS). In addition, we have successfully implemented protocols for dedifferentiation of IPS cells into neural lineages for functional studies (link to other parts of the GENCODYS program). Such patient-derived excitatory cortical neurons are used for electrophysiological studies to measure synaptic activity and neuronal network activity that can shed light into the nature of the cognitive deficit.
Task 1.3. Exclusion of known genes for ID and and/or autism spectrum disorder (ASD).
The exclusion of genetic causes that have already been implicated in CDs previously has been conducted by the application of diagnostic protocols as well as the use of targeted innovative analyses that have been set up for this purpose within the GENCODYS program. Examples of diagnostic pre-screens include the interrogation of the genome for CNVs by Comparative Genomic Hybridization using SNP-microarrays (arrayCGH). Other standard tests include investigation of repeat expansions at the FMR1 locus (Fragile X syndrome) and phenotype-based sequence analysis of CD genes. CNV has provided a conclusive diagnosis (genomic deletion or duplication) in 10-15% of all patients and FMR1 repeat expansions in 2-5% of patients.
Specific analyses that have been set up within the GENCODYS program include a next generation sequencing pipeline for targeted exome sequencing of all exons encoded by the X chromosome. This pre-analysis has identified a large number of causative mutations in known ID genes and has uncovered some 10 new genes for X-linked ID.
Task 1.4. Mapping of novel genes for ID and/or ASD.
Mapping and CNV analysis has been conducted in 5929 and 7849 individuals, respectively, of the GENCODYS cohort (Table 1). Systematic whole-genome SNP genotyping has been conducted on High-Density SNP-array platforms (Affymetrix or Illumina), containing 250,000 to > 1 million SNPs. The results and some examples of SNP microarray mapping are discussed for families with an apparent autosomal recessive or X-linked genetic defect and for isolated cases.
CNV analysis in isolated cases: CNV analysis is a first step in the diagnostic analysis of patients with ID and associated phenotypic features and has replaced traditional (microscopic) karyotyping in many if not most diagnostic laboratories in Europe. CNV analysis has been conducted in 7849 individuals during the GENCODYS program. .In isolated patients, a causative CNV leading to a conclusive diagnosis can be identified in approximately 10-15% of cases.
Homozygosity mapping by SNP microarrays in CD families: With the exception of some X-linked families, the inheritance pattern of the large consanguineous families will be almost exclusively autosomal recessive. Since consanguinity results in a high degree of homozygosity, we expect that the majority of mutations in these families will be homozygous. Homozygosity mapping studies have been performed in hundreds of consanguineous families from Iran, Pakistan, Turkey and Arab countries. Homozygosity mapping has been followed by next generation sequencing, initially by targeted exome sequencing of homozygous intervals and later by whole-exome sequencing (WES). This approach has resulted in the identification of numerous new genes for recessive forms of CD
Task 1.5. Identification of novel molecular defects underlying ID and/or ASD.
Next Generation sequencing has been conducted in almost 2500 individuals of the GENCODYS cohort (Table 1). These high-throughput sequencing activities have been performed on various platforms (Roche, Illumina, Complete Genomics) following a variety of enrichment protocols. GENCODYS partners have pioneered the development of protocols for the selection of variants and interpretation of next generation sequencing data in CD patients with various inheritance patterns of their disorder (recessive, X-linked, dominant, de novo, oligogenic) and in combination with various sequencing procedures (targeted exome, whole exome, whole genome).
In the European population, the majority of individuals with a CD are seen as isolated cases. This is a consequence of the reduced fecundity of CD individuals. As a consequence, the associated mutation (copy number variant (CNV) or single nucleotide variant (SNV)) has usually occurred as a de novo event in the germline of one of the unaffected parents (Vissers et al., 2010). Therefore, for identification of such events we usually have performed trio sequencing, i.e. NGS of the patient and both of the healthy parents. Candidate SNVs are selected based on their de novo occurrence. Within GENCODYS we have conducted whole exome sequencing (WES) and whole genome sequencing (WGS) in trios.
Whole exome sequencing
In a groundbreaking study (De Ligt et al., 2012), we sequenced the coding regions of more than 21,000 genes obtained from 100 patients with an IQ below 50 and their unaffected parents. A data-analysis procedure was developed to identify and classify de novo, autosomal recessive, and X-linked mutations. In addition, we used high-throughput resequencing to confirm new candidate genes in 765 persons with intellectual disability (a confirmation series). All mutations were evaluated by molecular geneticists and clinicians in the context of the patients' clinical presentation.
We identified 79 de novo mutations in 53 of 100 patients. A total of 10 de novo mutations and 3 X-linked (maternally inherited) mutations that had been previously predicted to compromise the function of known intellectual-disability genes were found in 13 patients. Potentially causative de novo mutations in novel candidate genes were detected in 22 patients. Additional de novo mutations in 3 of these candidate genes were identified in patients with similar phenotypes in the confirmation series, providing support for mutations in these genes as the cause of intellectual disability. We detected no causative autosomal recessive inherited mutations in the discovery series. Thus, the total diagnostic yield of conclusive mutations was 16%, mostly involving de novo mutations. An even higher percentage of de novo mutations was seen in another study (Rauch et al., 2012). These findings show that de novo mutations represent an important cause of intellectual disability; exome sequencing was used as an effective diagnostic strategy for their detection.
De novo mutations have been identified in a number of patients presenting a recognizable syndrome, for which the etiology had been hitherto unknown. Examples of these include disorders with aberrant cortical migration (MCD; P3a), Moebius syndrome (P1a) and Kleefstra syndrome (P1a).
Whole genome sequencing
Despite the high percentage of de novo mutations identified by WES, the majority of cases remain undiagnosed. We therefore applied whole-genome sequencing to 50 patients with severe ID and their unaffected parents. All patients included had not received a molecular diagnosis after extensive genetic prescreening, including microarray-based CNV studies and exome sequencing (see above). Notwithstanding this prescreening, 84 de novo SNVs affecting the coding region were identified, which showed a statistically significant enrichment of loss-of-function mutations as well as an enrichment for genes previously implicated in ID-related disorders. In addition, we identified eight de novo CNVs, including single-exon and intra-exonic deletions, as well as interchromosomal duplications. These CNVs affected known ID genes more frequently than expected. On the basis of diagnostic interpretation of all de novo variants, a conclusive genetic diagnosis was reached in 20 patients. Together with one compound heterozygous CNV causing disease in a recessive mode, this results in a diagnostic yield of 42% in this extensively studied cohort, and 62% as a cumulative estimate in an unselected cohort. These results suggest that de novo SNVs and CNVs affecting the coding region are a major cause of severe ID. Genome sequencing can be applied as a single genetic test to reliably identify and characterize the comprehensive spectrum of genetic variation, providing a genetic diagnosis in the majority of patients with severe ID
WES in families with recessive CD
With the exception of some X-linked families, the inheritance pattern of the large consanguineous families will be almost exclusively autosomal recessive. Since consanguinity results in a high degree of homozygosity, we expected that the majority of mutations in these families will be homozygous. Homozygosity mapping studies have been performed in hundreds of consanguineous families from Iran, Pakistan, Turkey and Arab countries. Homozygosity mapping has been followed by next generation sequencing, initially by targeted exome sequencing of homozygous intervals (Najmabadi et al., 2011). In this landmark study Partners 2 and 4 performed enrichment of exons from homozygous regions followed by next-generation sequencing in 136 consanguineous families with autosomal-recessive ID. This study, has revealed additional mutations in 23 genes previously implicated in intellectual disability or related neurological disorders, as well as single, probably disease-causing variants in 50 novel candidate genes. Proteins encoded by several of these genes interact directly with products of known intellectual disability genes, and many are involved in fundamental cellular processes such as transcription and translation, cell-cycle control, energy metabolism and fatty-acid synthesis, which seem to be pivotal for normal brain development and function.
Soon after this work, several GENCODYS partners have performed WES in families in which prior homozygosity mapping was performed (1 individual per family) and in consanguineous families without prior homozygosity mapping (WES for 2-3 individuals per family). These studies have uncovered a plethora of new CD genes, which have resulted in an impressive number of high-impact publications. been published in a range of high-impact papers (e.g. Roscioli et al., 2012; Schuurs-Hoeijmakers et al., 2012; Wortmann et al., 2012). WES has also been applied to non-consanguineous families of European origin with multiple affected individuals in a sibship. This has also revealed several novel genes, such as SLC6A17 (Iqbal et al. 2015). In total, we have identified well over 100 new ARID genes. Causality of these genes is supported by the occurrence of recessive mutations in at least 2 families with overlapping CD phenotypes. Many of these genes have already been published, but many other papers are still in preparation, including two that will report the results of large-scale WES in hundreds of ARID families from Iran and Pakistan.
X-exome sequencing
X-linked intellectual disability (XLID) is a clinically and genetically heterogeneous disorder. During the past two decades in excess of 100 X-chromosome ID genes have been identified. Yet, a large number of families mapping to the X-chromosome remained unresolved suggesting that more XLID genes or loci are yet to be identified. Here, we have investigated 405 unresolved families with XLID. We employed massively parallel sequencing of all X-chromosome exons in the index males. The majority of these males were previously tested negative for copy number variations and for mutations in a subset of known XLID genes by Sanger sequencing. In total, 745 X-chromosomal genes were screened. After stringent filtering, a total of 1297 non-recurrent exonic variants remained for prioritization. Co-segregation analysis of potential clinically relevant changes revealed that 80 families (20%) carried pathogenic variants in established XLID genes. In 19 families, we detected likely causative protein truncating and missense variants in 7 novel and validated XLID genes (CLCN4, CNKSR2, FRMPD4, KLHL15, LAS1L, RLIM and USP27X) and potentially deleterious variants in 2 novel candidate XLID genes (CDK16 and TAF1). We show that the CLCN4 and CNKSR2 variants impair protein functions as indicated by electrophysiological studies and altered differentiation of cultured primary neurons from Clcn4-/- mice or after mRNA knock-down. The newly identified and candidate XLID proteins belong to pathways and networks with established roles in cognitive function and intellectual disability in particular. We suggest that systematic sequencing of all X-chromosomal genes in a cohort of patients with genetic evidence for X-chromosome locus involvement may resolve up to 58% of Fragile X-negative cases. (Hu H et al. 2015).
Autism spectrum disorders (ASD)
The group of P3b (Laumonnier) analyzed a cohort of 100 French families including at least one individual with ASD by arrayCGH (Agilent) followed by targeted exon capture and NGS of 216 genes encoding mostly proteins belonging to the NMDA receptor interacting proteome. The targeted exon capture and NGS allowed the identification of a number of candidate genes and genomic regions, novel and replicating previous loci involved in neurodevelopmental disorders. Of interest, two de novo CNVs involved the GRM5 gene in 2 unrelated patients. The GRM5 gene encodes the mGluR5 postsynaptic receptor that is thought to be involved in the pathophysiology of Fragile X syndrome.
Conclusion
Next generation sequencing methodologies such as WES and WGS have proven to be highly effective in the identification of mutations underlying CD of all inheritance patterns (de novo, recessive, X-linked and oligogenic as in ASD). These genome-wide approaches allow the analysis of all genes, or selected groups of genes (gene packages) in a single experiment. The costs of these NGS methods have dropped considerably over the last year and have reached a price of less than 1000 € per sample. This means that NGS is also a cost-effective approach in comparison to microarray analyses and Sanger sequencing protocols. As a consequence, WES and soon WGS will become the methodologies of choice for a first-step genetic analysis of patient with an unknown (cognitive) disorder. In fact, at the institute of P1a WES is already offered as a diagnostic test and soon this will be extended by WGS protocols. The latter have the potential to also pick up genetic and genomic variants in non-coding regions, but the assignment of causality of such variants needs further insight in biological mechanisms as well as the development of tools for bioinformatics analyses.
Workpackage 2: Fly Models
- An unbiased screen for learning and memory conducted for almost 2500 Drosophila genes
- Characterisation of 435 CD genes and candidate CD genes.
- Identification of an epigenetic module underlying CD by genetic interacting studies.
- Development and application of innovative imaging pipelines to visualise CD protein activity in the nervous system without interfering background fluorescence.
- Publication of the first systematic ID gene screen identifying conserved functional modules of interconnected ID genes.
- Functional screens of CD genes completed: phototaxis/eye morphology, behavior, synapse and lethality/flight performance, and memory.
- First drug approaches in Drosophila initiated, including several chemical compounds that have been tested in EHMT and other CD models such as ADHD.
- First successful drug rescue project in Drosophila models completed.
In WP2, the fruit fly Drosophila melanogaster and its powerful genetics have been exploited to systematically uncover CD gene function, generating and integrating cellular and subcellular expression profiling during development, neuroanatomical phenotyping, functional and behavioural screening approaches in a high throughput manner. Beyond these unprecedented functional studies in an intact nervous system, genetic and pharmacologic screens have been applied in Drosophila to identify novel cognitive genes providing an entry point for the development of novel diagnostic and therapeutic strategies.
WP2 Summary of the main results
This workpackage had 4 major tasks:
2.1. Systematic analysis of expression patterns of CD genes and subcellular localisation of their protein products.
We have generated 212 transgenic lines carrying tagged GENCODYS constructs marked with the fluorescent marker that cover altogether 131 GENCODYS genes. In addition to lines with fluorescent marker we prepared 216 retrofitted constructs marked with mini-white and obtained lines for another 14 genes that are not covered by lines with fluorescent marker. Altogether, we have generated 145 transgenic strains carrying tagged version of GENCODYS genes.
Throughout the GENCODYS grant we have developed technology for live imaging of fluorescent fosmid-based gene expression reporters in Drosophila embryos using multi-view Selective Plane Illumination Microscopy (SPIM). The complete pipeline involving the construction up of custom OpenSPIM microscope, registration of the SPIM views, fusion by multi-view deconvolution and visualisation using the BigDataViewer published and available as open source plugin. We found that only about 10% of GENCODYS transgenic lines show high enough endogenous expression to allow SPIM imaging in embryos.
We screened the GENCODYS transgenic lines for expression in adult brain. To date we screened 52 lines that is about one half of the obtained transgenics. Based on the image data obtained, we categorized GENCODYS genes expressed in adult brain into four patterns (Figure 3). The genes with expression localized to projections (1), to cell body membrane or cytoplasm (2) or to nuclei (3) of all neurons and genes with expression in nuclei of only several cells in a spatially restricted brain region (4).
The goal is to make the image data available to scientific community. To that end we developed two visualization tools CATMAID and BigDataViewer. We have recently showcased the capabilities of the CATMAID viewer by providing unprecedented level of access to raw data of endogenously tagged library of rab proteins imaged in various developmental contexts throughout Drosophila lifecycle. This dissemination strategy serves as model for distribution and analysis of the images of GENCODYS transgenic lines. The BigDataViewer (BDV) enables browsing through arbitrarily large image volumes and dynamically re-slicing them across arbitrary dimensions. Additionally, we have recently developed a BigDataServer that makes it possible to access remote data sources in the same way as if they were on a local computer. Both tools are open source and thus can be adapted and extended for the purposes of various image-based projects, even beyond the GENCODYS project.
2.2. Systematic generation and characterisation of fly CD models.
The function of most genes underlying CDs in the nervous system is poorly understood. Gencodys WP2 was devoted to determine defects in a minimum of 300 Drosophila CD models. For this purpose, unique resources have been exploited, most importantly two transgenic RNAi libraries. CD models have been characterized in analysis pipelines designed to determine defects on three different levels relevant to human mental disorders: neuronal and synaptic morphology, neuronal function and cognitive behavior/ learning and memory.
Synaptic morphology was known to be compromised in several cognitive disorders, as revealed by post‐mortem studies and mouse models, but the extent to which compromised synapse development occurs in these disorders was illusive. We have used >500 RNAi lines to pan‐neuronally knock‐down CD genes, and has analyzed multiple parameters of synapse development at the larval Neuromuscular junction in a high‐throughput imaged‐based screen. This screen comprised >10.000 tracked and quantitatively assessed synapses.
To test whether CD genes are required for neurodevelopmental and basal neurotransmission, 500 RNAi lines were analyzed for performance in two fundamental fly paradigms: adult flight/ escape response and phototaxis.
Habituation is a form of non‐associative learning, in which a response to a repeatedly applied stimulus gradually wanes. It is defective in classical learning and memory mutants and as reported in several Gencodys papers. Aktogen, an SME within GENCODYS, has developed a 32‐chamber semi‐automated version of the light‐off jump habituation paradigm. In this system more than 300 mutant Drosophila lines were screened and a series of novel learning mutants were isolated. A learning defect was observed in 30% of genotypes tested. The high‐throughput adapted habituation paradigm is therefore a powerful tool to screen disease genes, disease gene candidates or other genes for their role in learning.
Previously, one of the GENCODYS partners (Keleman, IMP Vienna) has established the courtship conditioning paradigm to study long‐term memory in Drosophila. In this paradigm, naïve males learn to selectively suppress futile courtship of unreceptive mated females. Learning, short‐term or long‐term memory can be evaluated in these males. Almost 2500 RNAi lines targeting ID genes, candidate ID genes and synaptic genes were analyzed for long term‐memory defects with use of a tracking system.
Gencodys has integrated these rich datasets to identify further molecular networks that form novel molecular networks underlying cognitive disorders.
2.3. Identification and characterization of novel cognitive genes (i.e. new CD candidate genes).
This task has been accomplished in close collaboration with GENCODYS bio-informaticians as described at WP5. The basis of candidate gene identification from Drosophila experiments included:
- the unbiased screen for learning and memory conducted for almost 2500 Drosophila genes
- Characterisation of 435 CD genes and candidate CD genes and associated phenotypes as characterized in our phenotyping pipeline.
- The identification of target genes for CD genes encoding transcription factors and epigenetic regulators.
- Genetic interaction studies using crossed of different fly CD gene mutants..
2.4. Pharmacological treatment of CD fly models to identify novel drug targets and compounds for the treatment of CD.
The development of strategies for therapeutic intervention in CDs has been a major ambition of the GENCODYS program. We have been highly committed to develop in vitro, in vivo and ex vivo methodologies for testing pharmaceutical or other sorts of intervention. Drosophila is highly suitable for use in medium-size screens for compounds.
We have developed a working prototype of an insect drug delivery system: a vacuum injection drug delivery system. It consists of a vacuum pump, a number of two- and three-way solenoid valves, an electronic atomizer nozzle, a syringe pump, a spraying chamber, National Instrument control hardware and control software implemented in the National Instrument LabVIEW package.
Various drug projects have been completed or are ongoing and have already attracted novel funding that secures the first follow up study on Gencodys achievements: project funded by the Jerome Lejeune foundation for testing Spermidine on CD fly models. Other examples of intervention studies in Drosophila CD models include a variety of compounds (transcriptional and translational inhibitors, antioxidants, high glucose diet and a library of epigenetic drugs) applied to EHMT-mutant flies, lamotrigine in the NF1 model, and a candidate drug for treatment of mitochondrial (complex I) disorders.
Workpackage 3: Mouse Models
- In total 54 mouse lines have been generated and made available to the GENCODYS consortium and the scientific community as a whole.
- Approximately forty mutant CD mouse lines have been characterised throughout a specific phenotyping Pipeline.
- In-depth molecular characterization of several mouse models has been completed (in collaboration with WP4).
- Complete electrophysiological evaluation of five novel lines of mice with mutations relevant to genetic changes in human patients with various forms of autism and intellectual disability.
- Pharmaceutical interventions have been tested in five CD mouse models.
The goal of WP3 was to develop, characterize and make available mouse models to elucidate the function of genes mutated in CDs, and the pathomechanisms linked to their dysfunction. While existing mouse models were used as reference, we have focused on genes recently identified or that will be identified by the GENCODYS consortium as part of WP1, or on key genes involved in the major CD pathways. This work package was highly complementary to WP1 (human genetics) and WP2 and WP4 (fly models and functional genomics) and systematic integration of data (WP5) from WP3 with others WPs will result in a powerful view of the effects of the alteration of these genes and their pathways at a cellular and whole organism level.
WP3 Summary of the main results
This workpackage had 5 major tasks:
3.1 Catalogue existing mouse models for monogenic, early onset CDs and generate novel ones. Establish a pipeline for the production of novel mutant mice to model CDs using prioritisation. Maintain and distribute the mouse models in and outside of the consortium.
Mouse models with the relevant human disease information, based on literature and database searches and on unpublished information from GENCODYS partners were identified and catalogued by the consortium. This collection was extended by models some of which were generated within GENCODYS, others. Identified outside of the consortium, a list models to be considered was agreed by the consortium in compliance with the following criteria:
- Type of mutations associated to the human phenotype.
- Frequency at which mutations have been identified.
- Involvement of the gene in known molecular or cellular pathways.
- Other models already available or under construction.
- Other reasons for selection (data from e.g. Drosophila).
- Possibilities to further investigate the model once established and phenotyped at ICS.
- Is or will the same gene being investigated at other levels
A total 54 mouse mutant lines have been either newly generated (34 lines) or recovered from outside (20 lines) and made available for the consortium by P10 (ICS, mouse Clinics Institute,, Strasbourg). Six other lines have been generated by P14 (Welcome Trust Sanger Institute/University of Edinburg).
All mouse lines generated during the Gencodys project at ICS (P10) are available for partners during and after the project. During the last 3 years, different type of materials (live mice, embryos, organs) from more than 20 lines were shared with GENCODYS partners for further studies. Lines issued from IKMC (International Knock out Mice Consortium) ES cells are available for wider scientific community via EMMA/INFRAFRONTIER database (https://www.infrafrontier.eu/). Partner 14 has archived and distributed CD models to the broader community beyond the GENCODYS timeframe. Provisions will be made for storing and distributing specific mouse tissues, especially selected brain regions to the other partners (notably for transcriptome studies in WP4).
3.2 Integrated neurological and behavioural phenotypic analysis focusing on evaluation of cognitive dysfunction
One of the objectives of WP3 consists of neurological and behavioural phenotypic analysis of mouse models for ID, focusing particularly on evaluation of cognitive dysfunction (ICS). The behavioural screen contains tests yielding an overview of a wide range of functions (Figure 2). This step is pivotal for observation of potential non cognitive phenotypes, and is also necessary to verify the specificity of the potential cognitive alterations observed. In addition, mice are subjected to specific evaluation of cognitive abilities, including different aspects of learning and memory processes, attention processes and sensorimotor gating.
Mouse phenotypes have been aligned to the corresponding human phenotypes and that of Drosophila mutants. Gencodys has integrated these rich datasets to identify further molecular networks that form novel molecular networks underlying cognitive disorders.
At the end of behavioural pipeline, half of the animals from each group (genotype, treatment) were perfused and brains dissected out, post-fixed and maintained for histological analysis. For the other half of mice, specific brain regions (Hippocampus, Cortex, cerebellum, olfactory bulb, and the rest of the brain) are sampled, snap frozen, then maintained at -80°C. In some cases, all animal brains are processed for structures dissections. The samples are shipped to other partners for subsequent biological analysis in WP4.
3.3 In depth physiological and neuromorphological analysis of these models, including synapse physiology and correlation with the behavioral assays
This task of the Gencodys project has been aimed to determine the physiological consequences of mutations in CD genes using the generated mouse models. These physiologicalcharacteristics are critical to highlight pre- and postsynaptic alterations and other neurobiological features. Comparison of such features between the various models may identify or reinforce common molecular and cellular mechanisms of disease, which may be useful in defining targets for intervention strategies.
Conclusion: Electrophysiological measurements of synaptic transmission parameters were performed in acute brain slices from genetically altered mice bearing mutations in genes homologous to those mutated in humans with brain disorders. Observed changes in the amplitude of synaptic transmission and its short- and long-term plasticity in several lines of mutant mice suggest that these alterations may underlie intellectual disability in humans with functionally similar mutations. Therefore, such changes may be used as a sensitive gauge for testing various pharmacological substances aimed at correction of cognitive impairments. Using mutation effect size values, we worked out a standardised comparative approach which allows an unambiguous ranking of the mutation severity with respect to its impact on synaptic transmission.
3.4 Evaluation of the potential effects of pharmacological or genetic treatments on cognitive functions in selected mouse models
The development of strategies for therapeutic intervention in CDs has been a major ambition of the GENCODYS program. We have been highly committed to develop in vitro, in vivo and ex vivo methodologies for testing pharmaceutical or other sorts of intervention. Mouse is highly suitable for testing specific compounds for their capacity to alleviate neurological and cognitive deficits resulting from CD gene mutations.
In the work towards this deliverable GENCODYS partners analysed behavioural and cognitive abnormalities in mouse genetic models of cognitive disability (CD) and attempted to reverse observed phenotypes by pharmacological means. The following experiments have been performed:
a) reversal of behavioural phenotypes of Ophn1-KO mice by fasudil (HA1077).
b) rescue of Mecp2-KO phenotypes by Epothilone D.
c) rescue of IlRapl1-KO behavioural phenotypes by alpha5 GABA inverse agonist.
d) rescue of mutant Spred1 and Syngap1 touchscreen pretraining phenotypes by the MEK inhibitor PD-0325901.
Chronic treatment with fasudil counteracted the increased vertical and horizontal activity, as well as memory deficits in the object recognition task in Ophn1-KO mice. In a follow up study using new sets of mice, several phenotypes were confirmed in Ophn1-KO mice, including altered motor coordination performance in the rotarod, altered learning and memory as observed in the Y-maze spontaneous alternation task, and delayed spatial learning in the water maze. Moreover, motor deficits were prevented with Fasudil.
On the other hand, Epothilone D, a Taxol analogue, had no effect on behavioural changes observed in Mecp2-KO mice.
We also reported preliminary data on the effects of alpha5 GABA inverse agonist on learning deficits observed in IL1RapL1-KO mice. Despite the treatment, Il1Rapl1-KO males had increased activity and decreased exploration of the centre of the open field, suggesting that treatment had no effect of general activity. On the other hand, alpha5 GABA inverse agonist reverses spatial learning deficits in the water maze observed in Il1Rapl1-KO males.
We attempted to rescue the behavioural phenotype of Spred1 and Syngap mutant mice by treatment with an inhibitor of MEK, the enzyme that phosphorylates ERK. Animals in the drug group received single injections of the MEK inhibitor PD-0325901 for 5 days, and the vehicle group received DMSO only. However, as revealed in the following Figure, this treatment failed to improve cognitive performance of Spred1 mutants. Experiments with Syngap mutants are still ongoing.
3.5 Creation of an accessible web‐based database with bioinformatic comparison of profiles and systems biology approaches to identify pathways, common mechanisms of action and potential therapeutic interventions (with WP2, 4 and 5).
The database has been setup to manage the generation of mutant mice status as well as the phenotyping data. This web-based module covers all status related to mouse models tasks of partner 10. Indeed, it displays the mouse lines generation update including 4 steps: Clones, microinjection, chimeras and GLT (germ line transmission). The status of every step is highlighted with a color code. All GENCODYS lines can be found via the search menu and the listing is available for export (CSV format). The Phenotyping data sub-module is linked to display all results from the Phenotyping pipeline. This will display the details of the tests performed (age of mice, number, etc), measured parameters and data analysis.
Workpackage 4: Functional Neurogenomics
- Dataset of more than 14 quantitative gene expression profiles, 300 sequenced transcriptomes and epigenomes of cognitive disorders (CDs) murine models will be available to the broader scientific and medical community.
- Novel platform for image capture and analyses of synaptic proteins at the level of the neuron, brain region and whole brain scale will have applications in academic and medical research, and technology development.
- Neuromorphological characterization of 43 CD genes in mammalian neurons.
- A novel in vitro system has been developed to screen pharmacological rescue of CD mutant phenotypes. Several candidate treatments were tested. Analysis of Transcriptome profiles may reveal further target pathways suitable for intervention.
- Pharmacological rescue of CD mutant phenotypes in mammalian neurons.
WP4 has used systems approach of mammalian neurons in vitro, integrating large-scale functional genomics and epigenomics, transcriptomics, cell biological, morphological and electrophysiological data. In addition, we used mouse brain tissues as sources of material for specific functional genomic assays. WP4 forms an important bridge between the high‐throughput and genome‐wide approaches of GENCODYS Human Genetics and Genomics research and investigations with Fly and mouse models. The mammalian cell based assays used in WP4 form the basis of assays for drug rescue and genetic modifier experiments and thus identification of therapeutic targets.
WP4 Summary of the main results
This workpackage had 6 major tasks:
4.1 Generation of in vitro neuronal models of CDs using primary neurons and stem cell derived neurons.
Availability of in vitro neuronal models is a vital resource to studying the intact animal and human brain. We have established conditions for studying CD gene mutations using mature neurons and in vivo preparations of the hippocampus slices. Here, we summarize our findings in developing primary hippocampi cultures, and in vitro multi-electrode arrays (MEA) performed in primary neurons with disease relevant mutations.
In the course of developing an in vitro assay for the purposes of studying the impact of disease-relevant mutations on neuronal circuits we unexpectedly observed evidence for mutational buffering. The technology we used to monitor the activity of neural circuits was a tissue culture chamber where a 64 electrode array (MEA, Multi-electrode array) was overlaid with primary cultures of mouse hippocampal neurons. We previously reported that the firing patterns and synchronisation of neuronal networks can be longitudinally recorded from their initial silent phase to a later stable and synchronised phase. Here we evidence for a mutation buffering process and show its general property in overcoming mutations in different classes of genes. We also establish that this buffering is a novel form of plasticity and does not require the classical mechanisms of glutamatergic synaptic plasticity or homeostatic plasticity. Moreover, the buffering process overcomes defects in glutamatergic signalling including those induced by pharmacological perturbations. These studies also demonstrate the utility of MEA systems for studying mutations and disease mechanisms on neuronal networks.
4.2 Characterisation and quantitation of gene expression and chromatin configuration including DNA methylation in inducible mouse KO ES and NSC cells and their differentiation into neurons.
Fourteen different mutants were characterized by quantitative RNA-sequencing for the hippocampus, a brain region which features a primary role in the establishment, maintenance and regulation of long term memory as well as learning, behavior and spatial navigation: in order to minimize sources of noise and variation, hippocampus at P30 days was used for all the samples and for each of the 14 mutated CD genes a specific set of wild-types was used: the latter device resulted into the availability of a remarkably large wild-type transcriptome baseline, which was fruitfully used to extract important information about the degree of inter-individual variation of gene expression.
In addition, for each of the CD mouse models generated in GENCODYS, a comprehensive dataset including genome-wide RNA expression and epigenetic marks distribution (namely DNA methylation, H3k4me1, H3k4me3, H3k27me3, H3k27ac, H3k9me2, H3k9me3). These epigenetic profiles complete and enrich the information provided by the sole RNA-seq. overall, approximately 350 different samples were sequenced for 16 different mutated CD genes. This large dataset constitutes a remarkable, optimal quality, database that will allow us to disclose and dissect the epigenetic foundations of cognitive disorders (CDs). The data is currently being processed for a high-impact publication and will thereafter be made available to the scientific community. This unparalleled in-depth CD gene regulatory database will allow us to disclose and dissect the molecular foundations of CDs.
4.3 Studies of subcellular localisation of CD proteins; and
4.4 Morphological characterisation of CD genes in mammalian neurons.
Following the identification of mutations in genes, the characterization of the neuronal function of the CD gene and encoded protein is a major step for a better understanding of the physiological role and pathophysiological impact during the development of the central nervous system (CNS). The objective of this deliverable was to build a collection of expression vectors for a selected group of CD genes and to establish their subcellular localization pattern after their transfection in the in vitro model for CNS development and differentiation that is widely used in neurobiology, namely the embryonic hippocampal primary neuronal cultures. This deliverable thus provides functional data supporting the deleterious impact of the mutations in CD genes identified in patients with neurodevelopmental disorder.
The analysis was conducted for 43 CD genes (wildtype) and 22 mutant alleles of CD genes. Strikingly, our results highlighted that more than 40% of the CD proteins are expressed in dendritic spines, suggesting that synapses are particularly targeted in neurodevelopmental disorders. To our knowledge, such an extensive functional work in primary neurons, has not been performed for this number of genes. We also demonstrated that most of the mutants analyzed in our study caused either an abnormal localization (e.g. for the ASD gene PTCHD1) or altered synaptic density or shape, thus validating the deleterious effect of the mutations identified in the patients included in GENCODYS project.
4.5 Electrophysiological characterisation of neural network abnormalities in genetic and RNAi models.
Neural networks are thought to be the target of many CD genes and disruption in the connectivity or functional properties of synapses or neurons can all result in changes to the overall network. Using multi-electrode array (MEA) recordings of neuronal networks in vitro are a rapid and efficient measure of gene function. We have investigated β/γ oscillations of the extracellular field potentials evoked by kainate in the CA3 area of hippocampal slices from mice with CD gene mutations. The neural network assays allow ex vivo characterization of brain physiology and connectivity in brains of mouse models, which can be used as a new tool for assessment of the efficacy of pharmaceutical compounds.
4.6 Pharmacological and RNAi rescue of CD mutant phenotypes in mammalian neurons.
We have contributed to the investigation in pharmacological rescue of CD mutant phenotypes. These studies complement the studies described in task 3.4 and Deliverable D35 in which we have attempted to reverse observed phenotypes by pharmacological means for behavioral and cognitive abnormalities in mouse genetic models of CD. In this task we have mainly used in vitro and ex vivo approaches to achieve pharmacological reversal of CD-associated phenotypes. In summary, the following interventions have been established:
• MEK inhibitor PD-0325901 SynGap LTP and input-output physiology;
We investigated whether administration of the mitogen-activated kinase kinase (MEK) inhibitor PD-0325901 could reverse electrophysiological consequences of the heterozygous loss-of-function mutation in the Syngap1 gene. In particular, we were interested to determine if MEK inhibition would be able to diminish LTP deficit in Syngap1+/- mice. We chose to administer this antagonist at a final dose of 20 mg/kg which was well tolerated by all experimental animals. We revealed that LTP was still significantly decreased in Syngap1+/- mice even after 6 day treatment with PD-0325901 (single administration daily by oral gavage). We also tested the compound for affecting input-output relationships reflecting the intensity of the basal synaptic transmission, which were significantly steeper in hippocampal slices prepared from Syngap1+/- mice. Here, our pharmacological reversal experiments demonstrated that abnormal synaptic transmission phenotypes can be rescued in mature mice by inhibiting MEK for 6 days.
• Isoproterenol in SynGAP;
In another series of experiments, we attempted to pharmacologically reverse the LTP deficit in Syngap1+/- mice by activating protein kinase A. Our hypothesis was that if induction of LTP occurs in the presence of elevated cAMP, and as a result, higher activity of PKA, the LTP deficit in Syngap1+/- mice would be diminished. Isoproterenol, a beta-adrenergic agonist was chosen as a drug able to up-regulate PKA activity. Incubation with 1 uM isoproterenol for 10 min before LTP induction by theta-burst stimulation let to a transient up-regulation of fEPSP amplitudes, but the extent of LTP in the test pathway was still significantly less as compared to LTP in WT slices. Thus, a pre-treatment with isoproterenol failed to rescue LTP deficiency in Syngap1+/- mice
• Activity-dependent synaptic mapping mouse line Arc-Venus after pharmacological treatment with ketamine.
Partner14 has employed the activity-dependent synaptic mapping mouse line Arc-Venus. Arc directly binds to PSD95 and assembles sets of proteins that control cognitive functions. Using unique synaptome mapping tools developed within GENCODYS, we treated Arc-Venus mice with different doses of ketamine or sterile saline. Low and high doses of ketamine both increase Arc-Venus synapse density in the short term. However they have opposite long-term effects, with the low dose increasing and the high dose decreasing Arc-Venus synapse density. This suggests that these two doses, despite both being the same NMDA-receptor antagonist, are able to activate different responses at the level of the composition of the post-synaptic density. This is interesting given the recently-identified long-lasting antidepressant effects of a low ketamine dose in both humans and mice. This demonstrates that genes involved with cognitive disorders can be modulated by pharmacological treatments
• BKCa channel opener molecule (BMS-204352) in Fmr1 KO mice.
We investigated a modulator of BKCa channels in a mouse model for Fragile X syndrome (Fmr1 knockout). Fragile X Syndrome (FXS) is the most common form of inherited intellectual disability and is also associated with autism spectrum disorders. Previous studies implicated BKCa channels in the neuropathogenesis of FXS, but the main question was whether pharmacological BKCa stimulation would be able to rescue FXS neurobehavioral phenotypes.
In a collaborative work, Partner 3b used a selective BKCa channel opener molecule (BMS-204352) to address this issue in Fmr1 KO mice, modeling the FXS pathophysiology. In vitro, acute BMS-204352 treatment restored the abnormal dendritic spine phenotype. In vivo, a single injection of BMS-204352 rescued the hippocampal glutamate homeostasis and the behavioral phenotype. Indeed, disturbances in social recognition and interaction, non-social anxiety, and spatial memory were corrected by BMS-204352 in Fmr1 KO mice.
These results demonstrate that the BKCa channel is a new therapeutic target for FXS. The GENCODYS investigations have been published and have been independently confirmed recently by another group. This pharmacological molecule might open new ways for FXS therapy.
Workpackage 5: Bio-informatics
- We have developed a new method to predict genetic interactions between genes, using cancer genome data, that we will apply to (intellectual disability) disease genes.
- We have developed a webserver to predict new genes for a pathway, based on the weighted co-expression of > 4000 gene expression datasets of man and mouse. The tool has been used succesfully to predict one new ID gene.
- Our gene network approaches have identified and validated in vivo that there epistatic/synergistic interactions between variant genes are likely to be a significant contributor to cognitive dysfunction.
- We have shown that the presence of multiple functionally-similar genes is an abnormally frequent feature of de novo structural variants, and provides a significant indicator of pathogenicity.
- The apparent polygenic pathogenticity of these variants again support a role for genetic interactions in ID.
The massive amount of OMICS data generated in a systems biology project like GENCODYS requires state-of-the-art bioinformatics to arrive at concrete, biomedically relevant hypotheses and conclusions. Top-notch specialized bioinformatics groups within GENCODYS have taken full advantage of internal and external data sources in order to obtain both quantitative and qualitative understanding of the biological systems disrupted in cognitive dysfunction, and therewith to derive hypotheses about genes and pathways that are amenable to experimental testing by the consortium. We have harnessed the predictive power of a variety of bioinformatics and systems biology techniques, including comparative genomics and network approaches that combine information from disease networks with protein networks, and quantitative approaches that link publicly available QTLs to molecular data and Genome Wide Association Studies (GWAS).
WP5 Summary of the main results
This workpackage had 3 major tasks:
5.1 Characterizing and predicting CD genes;
5.2 Identification of gene regulatory networks in CD;
5.3 Identification of phenotype networks and the modular nature of cognitive disorders.
Phenotype-genotype based bioinformatics
We have compared human phenotypes with mouse and Drosophila phenotypes in order to decide which aspects of either model can be used to identify human ID genes, basically searching for phenologs: genes that in human and in a model species give rise to the similar phenotypes within each species.
For Drosophila we have examined various phenotypes of the eye, the wing and of behavior. With respect to the eye we have systematically mapped all the eye phenotypes of a set of 270 FLY orthologs of human ID genes, of which 180 did give rise to a phenotype. Given the issues in large scale orthology determination using automated methods, these orthologs have been manually curated and are stored in the sysID database (http://sysid.cmbi.umcn.nl/) as have their Drosophila phenotypes. There are many patterns in the data that link specific Drosophila phenotypes to specific classes of ID genes. The strongest one that we observed was that Drosophila gene orthologs with morphological eye phenotypes, in contrast to genes without phenotypes, are relatively highly expressed in the human nervous system and are enriched for neuronal functions, suggesting that eye phenotyping can distinguish different classes of ID gene. Overall we find that also underlying molecular processes correlate with eye phenotypes. The combined data indicate that ID disorders, despite their extreme genetic diversity, are caused by disruption of a limited number of highly connected functional modules.
For mouse, we first established the generalized concordances between human and mouse phenotypes. Exploiting the structured ontologies that are available for both and for other molecular annotations, we genes being labelled with the same mouse phenotype also had similar human disease phenotypes. Indeed, while far from perfect, the relationship between mouse and human disease phenotypes was demonstrably move informative that the relation between molecular annotations and human disease phenotypes. Many mouse models have been made to study genes of interest to human disease, and thus there is a clear bias of interest towards looking at phenotypes that may have presumed face validity. To discover the mouse models relevant to cognitive dysfunction, we looked at the mouse model phenotypes amongst genes that were variant in individuals with cognitive dysfunction phenotypes. Using patients held within the DECIPHER database that presented with a wide range of developmental and neurodevelopmental abnormalities, we non-exclusively grouped patients according to each specific phenotype and asked whether the orthologues of genes found to be variant in each specific-phenotype patient group were associated with any particular phenotypes when disrupted in the mouse. Forming CNV sets for each of 1,088 distinct human abnormalities, we were able to associate a total of 143 (13%) human abnormalities with mouse model phenotypes. We then considered each of autism and ADHD specifically. For autism, we identified a large number of mouse model phenotypic-associations amongst the orthologues of human genes that, apart from 6 of 104 genes, had not previously been implicated in autism. When these phenotypes were then compared to the phenotypes of mouse models of human orthologues that had been implicated in autism, an excellent correspondence was observed. For ADHD, the cohort was much less powerful than that used to identify autism human-mouse phenotypic correspondences. Nonetheless, an association with genes variant in ADHD patients whose orthologues’ disruption in the mouse gave abnormal learning/memory/conditioning phenotypes was found. Using a gene co-expression network built from human brain region- and development-specific gene expression, we did observe that the orthologues of genes whose disruption in the mouse gave hyperactivity phenotypes were more associated with genes variant in ADHD patients that would be expected by chance, supporting the use of hyperactivity in the mouse as a face valid phenotype for this disorder.
Genetics and Genomics
These bioinformatics studies fall into two parts. Predicting the effect of mutation in non-coding DNA and developing new bio-informatics protocols for predicting ID genes. Regarding the first part: complete genome sequencing, rather than “just” exome sequencing, is rapidly becoming standard and there is an urgent need to be able to interpret variations in non-coding DNA in terms of their potential contribution to disease. The second part has capitalized on the avalanche of genomics data and on the availability on curated lists of ID genes within the consortium to develop and test methods to predict new ID genes.
A model for brain-specific mutation deleteriousness in non-coding DNA,
Regarding the tissue-specific mutation deleteriousness within non-coding DNA we have specifically searched for transcription factor binding sites that are functionally relevant during Brain development. To develop our model we have used the Brainspan data (www.brainspan.org) that measures gene expression in sixteen cortical and subcortical structures across the full course of human brain development. We have then modeled the activity of transcription factor binding sites using the linear modeling approach Ismara. This approach has in the past shown to be able to explain ~10% of gene expression based on the presence/absence of transcription factor binding site motifs. Brainspan RNAseq data have been reanalyzed in order to be able to map transcripts and their transcription start sites to promoters and nearby transcription factor binding site motifs. This then gives a score per transcription factor binding site motif of its involvement in brain development. As we have activity scores per motif and we have gene expression levels per transcription factor we can correlate those with each other, expecting to see that highly expressed transcription factors in certain brain tissues will also have highly active motif, even though the analyses are completely independent of each other. Indeed that is what we observe for a number of motifs and their associated transcription factors. Finally, we have stored the motifs that were predicted to be active during brain development and that are associated with known ID genes in the sysID database.
During this project is has become clear that, despite the enormous heterogeneity of the ID there are a limited number of underlying molecular processes that allow us to actually make predictions for likely ID candidate genes. Nevertheless, it also have become clear that, when one wants to make successful predictions, it is best to take that heterogeneity into account as much as possible via our phenotypic classification that to some extent reflects the underlying biology. Thus various datamining and data integration activities have given us new candidates for ID genes that we have integrated in our SysID database. We have put them in a single integrated list, and parallel to that have put them in separate lists that are based on the SysID clinical classification, as the latter captures part of the underlying biology and therewith increases the “predictability” of ID genes. Similarly, other tasks revolve around establishing links between fly and mouse on the one hand and ID phenotypes on the other hand that was essential in being able to predict new ID genes. Both deliverables have been used in D40 to predict new ID genes.
Conclusions
Capitalizing on the various methods we developed in the GenCodys project, the wealth of genomics data and other data that are becoming available and on the availability of a well annotated list of ID genes we have predicted new candidate ID genes that have been stored in our SysID server and are available to the consortium.
WP6: Potentation
The main objective of the WP6 „Potentiation“ is the dissemination of new ideas and knowledge generated through Gencodys project to the scientific community and the general public. For this purpose, we established various training courses related to the different approaches (i.e. genomics, bioinformatics, animal models, etc.) used in Gencodys, internal and external workshops, as well as a program of scientist/student visiting and exchange. We also created a dedicated Gencodys website.
- Training activities, practical and theoretical courses have been organized at host institutes.
- Research visits for PhD students, postdocs and MDs have been facilitated by offering travel grants and fellowships.
- A total of eight workshops have been organized.
- Two successful international conferences were organized, which were open for non-Gencodys participants.
- So far the GENCODYS has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. Interaction with patients and families organizations to disseminate new diagnostic opportunities and to establish tight interactions for future clinical studies in tight association with families.
WP6 Summary of the main results
This workpackage had 3 major tasks:
Task 6.1. Training activities, practical and theoretical courses at host institutes.
GENCODYS has actively encouraged the education of young researchers and doctors via exchanges between partner Institutes and the teaching of courses by specialized partners. These interactions can efficiently promote the dissemination of knowledge within the GENCODYS consortium, i.e. to transfer new techniques and methodologies from one group to another or to establish and reinforce uniform protocols for research and clinical investigations. In addition, research visits were expected to stimulate progress of specific research activities within the various work packages. Regular exchanges took place involving all partner Institutes.
Specific exchanges:
- Training activity related to WP2, practical and theoretical Drosophila courses (Nijmegen, Berlin).
- Three Courses/ workshops on Next Generation Sequencing, first two at MPI-MG, Berlin, third at USWR-GRC in Teheran (instructed by MPI-MG).
- Course on "Comparative Genomics: from evolution to function", Amsterdam.
- Two rounds of Electrophysiological techniques workshop, Synome, Cambridge
Task 6.2 Research visits for PhD students, postdocs and for MDs
Our aim here was to support exchange of young researchers and doctors between partner Institutes. These interactions can efficiently promote the dissemination of knowledge within the GENCODYS consortium, i.e. to transfer new techniques and methodologies from one group to another or to establish and reinforce uniform protocols for research and clinical investigations. In addition, research visits were expected to stimulate progress of specific research activities within the various work packages. Frequent research visits were envisaged for PhD students, doctors and postdocs from the partners from Iran (P4) and Pakistan (P13) to the European Institutes, in particular those of P1a and P2a. Three main types of exchanges can be depicted :
- Organization of research visits and training networks inside the consortium for Ph.D. students. Possibilities of shared Ph.D. supervision between two groups.
- Research visits for post docs (need for specific technique used in a group from the consortium) - Research visits, training opportunities for MDs from other participating groups.
- Visits of clinical investigators for clinical investigations of CD patients.
To initiate the exchange program and elaborate the guidelines, the GENCODYS Board has established 2 types of exchange allowances that could be requested by each partner, and on the basis of the duration of the visit :
- Short visit (< 3 months) : up to 1500 euros
- Long visit (> 3 months) : up to 500 euros /month + travel costs (similar to Marie Curie allowances).
In both cases, the allowance could only be used for compensation of costs for travel from one Institution to the other and for accommodation costs during the stay at the host Institution.
In total 12 PhD students (including MDs) and 10 postdocs have been engaged in exchanges. Although the Gencodys project is terminated on 1st May 2015, it is highly probable and expected that the exchanges and research visits will be continued since close collaborative and interacting projects have been well established between various groups, and we can assume that this will certainly offer novel opportunities to submit new proposals for future EU calls
Task 6.3 Organization of workshops
This deliverable has been completed with the organization of 8 workshops and conferences. GENCODYS has organized two international conferences, which were open to non-GENCODYS scientists. These conferences have been attended by top-scientists in various expertise areas involving cognitive disorders in order to stimulate multi-disciplinary networking.
Gencodys workshops hosted for the consortium members:
1 Workshop: Cognitive Mouse Models Binding WP1-WP4, 7-8 February 2011, IGBMC/ ICS (partner 10), Illkirch, France (APPENDIX I)
2 WP1 Meeting, 19 April 2011, Ateneo Garden Palace Hotel, Rome, Italy (APPENDIX II)
3 WP3-WP4 Workshop, 5-6 February 2013, University of Edinburgh (partner 14), (APPENDIX III)
4 Workshop: Workshop Strabourg WP1-WP5, “data integration’, 17-18 February 2015, IGBMC/ ICS (partner 10), Illkirch, France (APPENDIX IV)
Gencodys workshops hosted for students, scientist and specialist from Pakistan, Morocco and neighbouring countries:
5 Workshop on Genetic techniques, “New Trends in Genetic Diagnosis of intellectual Disability”, 16-17 September 2014, ISESCO, Rabat, Morocco (APPENDIX V)
6 International Symposium on genetic Diseases, 29-31 March 2015 and Training workshop, “New Trends in Molecular Diagnosis of genetic Diseases”, 1-5 April 2015, Allama Iqbal Medical College (partner 13), in Islamabad, Pakistan (APPENDIX VI)
Gencodys international Conferences hosted:
7 1st International Gencodys Conference, “Integrative Networks in Intellectual Disabilities”, 14-17 April 2013, Paphos, Cyprus (APPENDIX VII)
8 2nd International Gencodys Conference, “Integrative Networks in Intellectual Disabilities”, 27-29 April 2015, Chania, Crete, Greece (APPENDIX VIII)
Task 6.4 Dissemination activities, web‐site forum, interaction with patients organizations
So far the GENCODYS program has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. The impact of many of the publications has been very high and several landmark publications. Top publications have appeared in Nature (n=4), Nature Genetics (6), Nature Neuroscience (6), Nature Communications (2), Nature Protocols (3), Nature Cell Biology (1), Nature Methods (1), Science (1), Science Reports (1), Cell (1), Developmental Cell (1), Neuron (3), Lancet Neurology (1), New England Journal of Medicine (1), Molecular Psychiatry (6) and 15 publications in the American Journal of Human Genetics. In addition, Gencodys PIs have produced authoritative reviews and opinion papers in Annual Reviews in Neuroscience, Annual Reviews in Genetics, Current opinion in Neurobiology, and Trends in Genetics. The broad multidisciplinary and translational character of the GENCODYS program is reflected in the broad subject categories of the journals in which GENCODYS papers have been published.
Intellectual property: Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014
Task 6.4.4 Website and website forum
The GENCODYS website (http://www.gencodys.eu) has been designed at the start and is working and being improved by the consortium members on a regular basis with among others online updates of publications and meetings. A specific section provides info for the public in general and patients and their families in particular. It includes general information on the project and its aims, and background information on genetics and inheritance patterns, including answers to some frequently asked questions on technical aspects. Also, the ways in which patients might benefit from the work done by the Gencodys consortium is addressed. The website will link to other websites where more information can be found such as the Eurogentest website, and the Rare disease communities website.
Task 6.4.5 Patients organizations
Patients with a cognitive disorder are the begin and endpoint of the GENCODYS program. Large efforts have been made to identify the genetic underpinnings of their CD and to get better insight into the molecular and cellular mechanisms of disease. Cumulative knowledge is being used for the benefit of the patient, initially for improved opportunities for making (prenatal) genetic diagnoses, genetic counselling, clinical management and improved standards of care. In the long run, GENCODYS has led the foundation for development of methodologies for intervention studies. GENCODYS acknowledges the importance of an early and close involvement of patients, patient organizations and patient advocate groups in order to achieve optimal benefits and impact for our target group. To that end, patient organizations have been represented within GENCODYS by inclusion of The European Genetic Alliances Network (EGAN/VSOP; p12) in the project. This has stimulated, facilitated and shaped the interaction between the scientific partners in GENCODYS and patient organisations. There have been ample meetings between GENCODYS PIs and a multitude of patient organizations during Annual GENCODYS meetings, specifically organized meetings, and in parallel to European Conferences such as the European Society of Human Genetics (ESHG), other EU project meetings, like FP7 PatientPartner (www.patientpartner‐europe.eu) and EURORDIS.
Potential Impact:
The GENCODYS main goal was to gain better insight into molecular and cellular processes that are key to neurodevelopmental processes and the higher cognitive functions of the human brain. We have studied a group of human disorders that constitute one of the largest unsolved problems in modern medicine. The basis of our research have been the genes and their cognate proteins that are disrupted in human genetic disorders that are characterized by impaired cognitive functioning (Cognitive Disorders; CD). Since disruption of the functions of these genes and their associated molecular pathways are associated with CD, it is obvious that their normal functions are crucial for development, functioning and maintenance of the healthy human brain. We have considerably impacted the field through the identification of a plethora of new causative genes and their associated molecular networks. Such new knowledge can and will be applied immediately to improve the possibilities for genetic testing and counseling for patients and families who are confronted with disorders of CD. Not only has the GENCODYS consortium heavily invested in the development and implementation of diagnostic testing, our research activities were and still are also targeted to device new strategies for therapeutic intervention in patients, e.g. through application of ingenious screens in primary neuronal cultures and Drosophila to identify small molecules that suppress the biochemical or phenotypic defects associated with CD gene mutations.
Specific scientific and technical impact
GENCODYS united a multidisciplinary team of European research groups, each of which represent leading international players in their discipline, as evidenced by publication records in top journals. The complementary disciplines of the participants reflects the translational nature of the GENCODYS program. We have implemented innovative high-throughput technology to significantly speed up the search of genes that underlie human disease. In addition, extensive functional studies in genetic models has provided new insights into gene function and into the way that genes interact to each other in health and disease. Model organisms have been generated and used for drug screening, which will aid the design of drugs and treatments. The scale of our approach reaches far beyond the capacities of local or national networks and is unparalleled by other international collaborative efforts in this important field of medical research.
Projected outcomes:
1) Identification of approximately at least 50 novel CD genes. These genes will likely be involved a variety of different CD conditions, such as intellectual disability (ID), autism spectrum disorders (ASD), ADHD, epilepsy, schizophrenia and others, since these conditions have a shared molecular etiology. Therefore, GENCODYS has also provided many new targets for further investigation in multifactorial forms of CDs.
2) Diagnostic tools to be able to establish a molecular diagnosis of patients with this extremely heterogeneous disorder.
3) Establishment of pathways that will a) give vital information about the development of the human brain and b) enable therapeutic intervention in CD patients.
The GENCODYS program has exceeded the expectations for these outcome measures, which illustrate the enormous scientific and medical impact of GENCODYS research:
Ad 1) GENCODYS research has identified causative mutations in some 150 genes that had not been implicated in a cognitive disorder before. The number of new genes might actually be larger, as we have also a list of candidate CD genes that await confirmation in unrelated CD patients (matching cases). Thus, the GENCODYS program has contributed approximately 150 new genes to the ~700 CD genes that are known to date (350 genes at the start of the GENCODYS). Mutations were found to cause forms of CDs of all kinds of inheritance pattern (de novo, recessive, X-linked and oligogenic as in ASD). Moreover, we have made an effort to carefully catalog the CD phenotypes that are associated with these mutations, which is not usually done in next generation studies in other patient cohorts. Thus, for individual patient and familial cases we have diagnostic profiles that comprise the cognitive features (ID, ASD, epilepsy, motor function, behavior) as well as co-morbidities.
Ad 2) Next generation sequencing methodologies such as WES and WGS have proven to be highly effective in the identification of mutations underlying CD of all inheritance patterns (de novo, recessive, X-linked and oligogenic as in ASD). These genome-wide approaches allow the analysis of all genes, or selected groups of genes (gene packages) in a single experiment. As a consequence, WES and soon WGS will become the methodologies of choice for a first-step genetic analysis of patient with an unknown (cognitive) disorder. In fact, at the institute of P1a WES is already offered as a diagnostic test and soon this will be extended by WGS protocols. The latter have the potential to also pick up genetic and genomic variants in non-coding regions, but the assignment of causality of such variants needs further insight in biological mechanisms as well as the development of tools for bioinformatics analyses. To that end, GENCODYS partners are committed in setting up internationally shared databases (in collaboration with the Sanger-based DECIPHER database/UK DDD project and other databases) for the registration of DNA variants that have been identified in CD patients. In any case, based on our results and extensive experience, we propagate the implementation of next generation sequencing technology as a first step in diagnostic testing of unexplained forms of intellectual disability. This will reveal a conclusive diagnosis in 30-70% of the patients (depending on the inheritance pattern) and thus a diagnosis early in life. This will obviate a plethora of needless invasive and expensive investigations that can carry on for years if not decades. Instead, better registration of the specific genetic aetiology and the constellation of cognitive deficits and other co-morbidities will offer a better standard of care for the respective patients.
Ad 3) Bio-informatics analyses and functional studies in cell and model organisms within GENCODYS of the 150 new genes and previously identified ones has demonstrated beyond doubt that CD genes converge into common molecular networks. This insight is important for developing strategies for therapeutic interventions, which can be targeted to common molecular processes and networks, and not strictly at single genes. Gencodys has made major advances towards development of therapeutic strategies. An unparalleled collection of animal models, both Drosophila and mouse, have been generated and characterized at a wide variety of neurodevelopmental, behavioural and cognitive parameters. These models are invaluable for pre-clinical studies of pharmaceutical compounds. In fact, GENCODYS has already started in testing a number of compounds, both in newly developed in vitro and ex vivo systems as well as in selected mouse and fly models. We anticipate that the wealth of new knowledge and innovative resources developed within in GENCODYS will become a basis for a plethora of follow-up studies across Europe, both in fundamental research and in translational research for developing therapeutic interventions. Partners who have successfully collaborated in GENCODYS will be actively involved in such follow-up projects, reinforcing existing collaborations and establishing new ones. Renewed support from the EU will be instrumental to benefit optimally from the established resources and to continue to increase the impact of EU research teams in translational research of the large group of genetic cognitive disorders. At present, GENCODYS groups are engaged in renewed EU-applications within the frame of Horizon 2020, E-RARE, ERA-NET-Neuron, EU COST, Marie Curie ITN, and the Human Brain Project.
Dissemination activities
Dissemination has been an integral part of the GENCODYS program and details have been reported in the summary for workpackage 6 (Potentiation). In short:
- Gencodys website: www.Gencodys.eu
- Training activities, practical and theoretical courses have been organized at host institutes.
- Research visits for PhD students, postdocs and MDs have been facilitated by offering travel grants and fellowships.
- A total of eight workshops have been organized.
- Two successful international conferences were organized, which were open for non-Gencodys participants.
- So far the GENCODYS has produced a total of 172 scientific publications. This number is expected to grow >200 as many manuscripts are still in the stage of submission and preparation. The impact of many of the publications has been very high and several landmark publications. Top publications have appeared in Nature (n=4), Nature Genetics (6), Nature Neuroscience (6), Nature Communications (2), Nature Protocols (3), Nature Cell Biology (1), Nature Methods (1), Science (1), Science Reports (1), Cell (1), Developmental Cell (1), Neuron (3), Lancet Neurology (1), New England Journal of Medicine (1), Molecular Psychiatry (6) and 15 publications in the American Journal of Human Genetics.
- GENCODYS PIs have produced authoritative reviews and opinion papers in Annual Reviews in Neuroscience, Annual Reviews in Genetics, Current opinion in Neurobiology, and Trends in Genetics. The broad multidisciplinary and translational character of the GENCODYS program is reflected in the broad subject categories of the journals in which GENCODYS papers have been published.
- GENCODYS research has made highlights in the media, such as national newspapers, scientific commentaries, radio and television, and social media.
- GENCODYS PIs have actively contributed to teaching (primary and secondary schools), lecturing (Masterclasses, Summerschools etc) and other activities that created awareness of our research (Science fairs, books).
- Intellectual property: Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014
Resources developed within GENCODYS
The development and application of high-throughput technologies for CD research have been characteristic in the GENCODYS strategy. By that, we have developed knowledge and a range of resources that will serve the scientific community and which should facilitate further (EU-funded) research. Examples of such valuable resources are:
- A biobank containing ~13.000 DNA samples and other biomaterials from CD patients of known and still unknown etiology. Patients have been systematically investigated and phenotypes are collected in a database by using standardized classifications. This unique biobank is a highly attractive resource for follow-up collaborative research.
- Protocols for whole-exome and whole-genome sequencing and subsequent filtering of DNA variants. Such protocols have already been implemented in diagnostic testing at the Radboud university medical center, Nijmegen, The Netherlands and are being used for genetic investigation of patients that are referred from all over the world.
- Throughout the GENCODYS grant we have developed technology for live imaging of fluorescent fosmid-based gene expression reporters in Drosophila using multi-view Selective Plane Illumination Microscopy (SPIM). Additionally, we have developed a BigDataServer that makes it possible to access remote data sources in the same way as if they were on a local computer. Both tools are open source and thus can be adapted and extended for the purposes of various image-based projects, even beyond the GENCODYS project.
- Aktogen Ltd, an SME partner within GENCODYS, has developed a series of semi-automated screens for behavioral assessment of Drosophila mutants. In addition, they have developed a vacuum injection drug delivery system for Drosophila and a proof-of-principle drug target and compound screen in EHMT1 mutant fly lines. This work has transpired in Patent applications submitted: United Kingdom Patent Application No. 1407483.5 and No. 1507178.0 Monitoring Insect Behaviour, Aktogen Limited, 29th April 2014.
- Large collections of Drosophila mutants have been systematically investigated in a neurobiological pipeline (almost 300 CD gene mutants) and in an unbiased screen for learning and memory. This collection of fly mutants will serve the scientific community by allowing further studies into mechanisms of disease as well as pharmacological intervention projects.
- A unique collection of CD mouse models has been generated, consisting of 40 newly generated mutant lines and 20 recovered lines. These mutants have been analyzed (and still are) following systematic phenotyping pipelines, which is providing valuable insights into the corresponding human CDs. These mouse mutants represent an extremely rich collection of animals that will be made available to the scientific community for further mechanistic studies as well as pre-clinical investigations of novel intervention strategies.
- Several pharmaceutical interventions have been tested in our CD mouse and fly models. Some of these revealed phenotypic rescue of behavioral or neurobiological abnormalities and thus hold promise for further investigation as potential leads for intervention in CD patients.
- Development of an ex vivo protocol for testing various pharmacological substances aimed at correction of cognitive impairments. SME partner Synome showed that changes in the amplitude of synaptic transmission and its short- and long-term plasticity in acute brain slices of several lines of mutant mice may underlie intellectual disability in humans with functionally similar mutations. Therefore, such changes may be used as a sensitive gauge for testing various pharmacological substances aimed at correction of cognitive impairments.
- Development of a novel in vitro system to screen pharmacological rescue of CD mutant phenotypes based on primary neuronal cultures from a PSD-95-eGFP mouse line developed by Synome. The synaptic signals are analyzed in a reproducible 96-well format amenable for high-throughput screening, potentially allowing testing of pharmacological treatments.
- Innovative tools have been developed for bio-informatics analyses that facilitate the identification and refinement of common molecular networks that are affected by groups of CD genes.
Socio-economical impact
Based on GENCODYS results we strongly recommend the implementation of next generation sequencing (NGS) technology as a first step in diagnostic testing of unexplained forms of intellectual disability. The costs of NGS methods have dropped considerably over the last year and have reached a price of less than 1000 € per sample. This means that NGS is cost-effective approach in comparison to microarray analyses (which can only identify CNVs) and Sanger sequencing protocols (which are gene-specific). This will reveal a conclusive diagnosis in 30-70% of the patients depending on the inheritance pattern and depending on the used strategy. Alternative NGS strategies include whole-exome sequencing (WES), WES of selected genes (panels), and whole-genome sequencing (WGS). WGS appears to be most robust and effective in terms of diagnostic yield, but the choice of the strategy may be affected by other considerations (costs, locally available technology and expertise, and ethical issues such as the risk of chance findings).
In any case, application NGS has the potential to provide a conclusive diagnosis early in life. This will obviate a plethora of needless invasive and expensive investigations that can carry on for years if not decades. The economical advantage is obvious. In addition, for the patient and the family the knowledge of the exact cause of the disorder has an enormous impact: 1. It may change the standards of the medical care that is given to the patient; 2. It will have prognostic value in terms of co-morbidities that are typical for mutations in a specific gene; 3. And it strongly affects the choice for reproductive options of the parents of an affected child
Medical and Societal impact
As mentioned above, the results of the GENCODYS program are of immediate benefit for patients, their families and society as a whole, because of the improvement of diagnostics for CDs. Given the rarity of the genetic defects, international collaboration remains essential, both with medical professionals and the patient community to resolve medical research questions. Based on gathered knowledge and experience we are committed to develop a modular Standard of Care for specific CDs, such as those that are due to mutation in a group of gene affecting common epigenetic mechanisms (the chromatinopathies). To realise an infrastructure for this and to disseminate and implement Standards of Care, we aim to further integrate European Centres of Expertise into a European Neurodevelopmental Disorders (NDD)-network that will pave the way for, or become part of, an intended European Reference Network (ERN) for “rare diseases of brain development and rare intellectual disabilities”. It is our mbition to transform this into one of the expected 23 ERNs that will be established in 2016 for clusters of rare diseases, as stipulated by the European Commission.
In addition, we have set the stage for applying the generated knowledge and technology to foster the identification of lead compounds to mitigate the disease phenotype of genetic conditions. The conceptual advance to delineate molecular networks that are commonly disrupted in cognitive disorders opens perspective to treat patients with mutations in different genes with common compounds. For example, based on advances in the understanding of the neurobiology of chromatinopathies, targeted therapeutic agents might be designed to correct the underlying mechanisms of neural dysfunction, and thus will have to be assessed in patients.
For obtaining maximum impact for our target group, the CD patients and their families, we have maintained close interaction with patient organizations throughout the GENCODYS program. This was facilitated via the embedding of a patient umbrella organization withan GENCODYS: the European Genetic Alliance Network, represented by VSOP (Vereniging Samenwerkende Ouder - en Patiëntenorganisaties – The Netherlands). We aim to continue and extend such close interactions with patient networks, in order to be prepared for future clinical trials.
List of Websites:
www.GENCODYS.eu
