Developmental Disorders Flashcards
What is a phenotype?
Phenotypes are the expression of diploid genotypes (dominant, recessive, additive)
How many protein-encoding genes do we have in our genome?
around 20,000
Describe the variation each individual has in terms of polymorphisms, non-synonymous variants, missense mutations and protein-truncating variants.
Our genomes are highly variant. Each of us differ from each other by ~2 million genetic variants. Each healthy person has:
• 22,000 genetic variants in exons
• 9,000 non-synonymous (Non-synonymous variant = change that leads to an amino acid change, while synonymous variant = variant that doesn’t lead to an amino acid change)
• 100 rare missense variants (i.e., MAF <1%).
• 5 rare protein-truncating variants (they stop the protein at some point on the code).
Describe the ‘Epigenetic Landscape’ model of Waddington (1957)
The idea here is that we inherit a genetic landscape from which is our base. The “epigenetic landscape” model of Waddington is an interesting way of visualising epigenetic mechanisms.
• In this model, a person’s neuro-development is represented by a ball and as development proceeds, the ball moves through an landscape of hills and valleys, with the valleys representing potential phenotypic pathways and endpoints.
• The shape of this landscape is determined by the genotype, with the gradient of the slopes acting to increase or decrease the likelihood of passage into any particular valley.
• The genotypic effects determine the probability of various phenotypes, but in any individual, the final phenotype is also influenced by the direction taken at developmental “choice points”, that can push an brain into a particular phenotypic valley.
What are the three main stages in cortical development?
The development of human cerebral cortex consists of 3 key stages:
- Neuronal proliferation in the germinal zone situated around the ventricular cavity
- Migration of the neurons outwards along radially arranged glial cells to take up superficial positions. At the surface, neurons migrate in an inside-out gradient, superficial migrating past the deeper layers.
- Organisation of neurons to form the six layered cortex.
Describe the conditions that result from abnormalities of cortical neuronal proliferation?
- Abnormalities of glial and neuronal proliferation and differentiation may produce Taylor-type focal cortical dysplasia (FCD). Such lesions are sometimes seen in neurofibromatosis type 1 or tuberous sclerosis, but more commonly occur as isolated lesions due to somatic mutations in mTOR pathways genes.
- Abnormally high levels of proliferation can cause hemimegalencephaly, where half of the brain is much larger, leading to drug-resistant seizures.
- A germline mutation affecting proliferation causes microcephaly.
Describe the conditions that result from abnormalities of cortical neuronal migration?
- Mutations of the X-linked gene Filamin 1, an actin binding protein, causes periventricular nodular heterotopia. BPNH (Bilateral periventricular nodular heterotopia) is a not uncommon cause of epilepsy and mild learning difficulty. Usually but not always fatal in males.
- Double cortin, or DCX is an X-linked microtubule associated protein that is widely expressed in migrating neurons. Mutations in DCX cause subcortical band heteropia (SBH) in heterozygous females, and anterior predominant LISSENCEPHALY in hemizygous males (much more serious).
- Lissencephaly 1 gene, LIS1 codes for platelet activating factor acetylhydrolase (PAF-AH), which stabilises microtubules, and mutations in this gene produce posterior predominant lissencephaly.
Describe the conditions that result from abnormalities of cortical neuronal organisation?
Abnormalities of cortical organisation may result in polymicrogyria, meaning an excessive number of abnormally small gyri. Several specific subsyndromes of polymicrogyria have been described, the best characterised of which is bilateral perisylvian polymicrogyria. These patients have some combination of pseudobulbar palsy, spastic quadreparesis, learning disability and epilepsy. A genetic basis is indicated on the basis of a number of reported families, with various modes of inheritance, but so far no genes.
What is the distribution of the aetiology of epilepsy?
The relative contributions to epilepsy are shown to the right. It is roughly thought that around 3⁄4 of epilepsy has no cause other than an inherited genetic predisposition – These epilepsies used be termed idiopathic, and indeed some of us still use this term, but they should perhaps more accurately be termed genetic epilepsies. The other 1/4 are caused by trauma, infection, metabolic disorders etc. Of the genetic epilepsies, a small proportion run in families – the so-called Mendelian epilepsies.
What theories try to explain the non-mendelian genetic contribution of epilepsy?
For the more common genetic epilepsies where there is little/no apparent family history, there are two non-mutually exclusive genetic hypotheses; the rare variant common disease hypothesis and the common variant common disease hypothesis
• The rare variant hypothesis proposes that individually rare (<1%) but highly penetrant variants which can occur in many different genes so that there is an overall high mutation rate to the disease class. This rare variant may be a de-novo mutation.
• Whereas the common variant common disease hypothesis proposes that risk variants confer a low relative risk for the disease, but because these variants are so common (>5% of population) in the population they account for a larger number of cases.
At the moment we don’t know whether common epilepsies are a rare or common variant disease, or a mixture of the two.
What mutations produce mendelian inherited epilepsies?
Mendelian epilepsy opened the door to gene discovery in idiopathic epilepsy, and therefore the pathophysiology in epilepsy. There are lots of patients with mutations in:
• Ion channel subunit genes:
○ Voltage-gated (Sodium, Potassium)
○ Ligand-gated (Nicotinic, GABA)
• Non-ion channel genes
○ LGI1 (lateral temporal lobe epilepsy)
○ GLUT1 (absence and other generalized epilepsies)
○ PCDH19 (female limited epilepsy)
○ DEPDC5 (‘variable foci’ - frontal, temporal)
What SNPs give an increased risk of idiopathic epilepsy?
The result of the ILAE meta-analysis of genome-wide association studies: Here, the significance of association between a SNP (Single Nucleotide Polymorphism) and epilepsy is plotted on the Y axis against its chromosomal position on the x-axis. After correction for multiple testing, three loci achieved genome-wide significance, SCN1A and PCDH7 for the phenotype all epilepsy, and the VRK2/FANCL locus for genetic generalized epilepsy. Interestingly, this locus is also a genome-wide significant susceptibility locus for schizophrenia, and we know that patients with epilepsy are at 11 fold higher risk of developing schizophrenia.
What was a significant finding of the 2014 GWAS study in Schizophrenia?
In 2014, this GWAS of schizophrenia was published, consisting of 37,000 cases and 113,000 controls
• Here, 128 associations achieved genome-wide significance, which although numerous explained only a very small proportion of heritability
• Importantly however, when the authors dug into the nature of the associations they discovered that only 8% of genome-wide significant SNPs were associated with a non-synonymous exonic polymorphism and concluded that most associated variants exert their effect by altering gene expression rather than protein structure. “My guess is that this will be true for epilepsy as well.”
How do we look for a de novo mutation in children with early and severe epilepsy?
The general approach to discovering genes for a disease by looking for de novo mutation is shown here - this is called the trio design.
So here, the exomes of an affected offspring and
their two unaffected parents are sequenced, with the idea that de novo mutations in the offspring may be causally related to the disease.
However, the interpretation of denovo mutations in trio exome sequencing studies is not as straightforward as you might think, due to the high rate of denovo mutation in healthy subjects. So, the exome point mutation rate of approximately 1.6 X10-8 per base per generation, translates to between 1 and 4 de novo protein disrupting mutations per child. This high rate of spontaneous mutation, most of which does not cause disease, means that there is no overall excess of de novo mutations in populations of epilepsy cases compared to healthy controls, and so just finding a gene with a protein disrupting de novo mutation in it does not necessarily implicate that gene in disease.
So the approach used to assigning causality to a gene using trio exome sequencing is the Rare variant collapsing framework illustrated. Qualifying de-novo mutations in a gene are summed in populations of cases and controls that then the significance of the difference in the frequency of de novo variants is calculated using a Fisher’s exact test. This approach can also be used to connect sets of genes to a disease.
