The development of the brain involves many thousands of different gene products acting in hundreds of distinct molecular and cellular processes, all tightly coordinated in space and time – from patterning and proliferation to cell migration, axon guidance, synapse formation and many others. Large numbers of proteins are involved in the biochemical pathways and networks underlying each cell biological process. Each of these systems has evolved not just to do a particular job, but to do it robustly – to make sure this process happens even in the face of diverse challenges.

Robustness is an emergent and highly adaptive property of complex systems that can be selected for in response to particular pressures. These include extrinsic factors, such as variability in temperature, supply of nutrients, etc., but also intrinsic factors. A major source of intrinsic variation is noise in gene expression – random fluctuations in the levels of all proteins in all cells. These fluctuations arise due to the probabilistic nature of gene transcription – whether a messenger RNA is actively being made from a gene at any particular moment. The system must be able to deal with these fluctuations and it can be argued that the noise in the system actually acts as a buffer. If the system only worked within a narrow operating range for each component then it would be very vulnerable to failure of any single part.

Natural selection will therefore favour system architectures that are more robust to environmental and intrinsic variation. In the process, such systems also indirectly become robust to the other major source of variation – mutations.

Many individual components can be deleted entirely with no discernible effect on the system (which is why looking exhaustively for a phenotype in mouse mutants can be so frustrating – many gene knockouts are irritatingly normal). You could say that if the knockout of a gene does not affect a particular process, that that means the gene product is not actually involved in that process, but that is not always the case. One can often show that a protein is involved biochemically and even that the system is sensitive to changes in the level of that protein – increased expression can often cause a phenotype even when loss-of-function manipulations do not.

Direct evidence for robustness of neurodevelopmental systems comes from examples of genetic background effects on phenotypes caused by specific mutations. While many components of the system can be deleted without effect, others do cause a clear phenotype when mutated. However, such phenotypes are often modified by the genetic background. This is commonly seen in mouse experiments, for example, where the effect of a mutation may vary widely when it is crossed into various inbred strains. The implication is that there are some genetic differences between strains that by themselves have no effect on the phenotype, but that are clearly involved in the system or process, as they strongly modify the effect of another mutation.

How is this relevant to understanding so-called complex disorders? There are two schools of thought on the genetic architecture of these conditions. One considers the symptoms of, say, autism or schizophrenia or epilepsy as the consequence of mutation in any one of a very large number of distinct genes. This is the scenario for intellectual disability, for example, and also for many other conditions like inherited blindness or deafness. There are hundreds of distinct mutations that can result in these symptoms. The mutations in these cases are almost always ones that have a dramatic effect on the level or function of the encoded protein.

The other model is that complex disorders arise, in many cases, due to the combined effects of a very large number of common polymorphisms – these are bases in the genome where the sequence is variable in the population (e.g., there might be an “A” in some people but a “G” in others). The human genome contains millions of such sites and many consider the specific combination of variants that each person inherits at these sites to be the most important determinant of their phenotype. (I disagree, especially when it comes to disease). The idea for disorders such as schizophrenia is that at many of these sites (perhaps thousands of them), one of the variants may predispose slightly to the illness. Each one has an almost negligible effect alone, but if you are unlucky enough to inherit a lot of them, then the system might be pushed over the level of burden that it can tolerate, into a pathogenic state.

These are the two most extreme positions – there are also many models that incorporate effects of both rare mutations and common polymorphisms. Models incorporating common variants as modifiers of the effects of rare mutations make a lot of biological sense. What I want to consider here is the model that the disease is caused in some individuals purely by the combined effects of hundreds or thousands of common variants (without what I call a “proper mutation”).

Ironically, robustness has been invoked by both proponents and opponents of this idea. I have argued that neurodevelopmental systems should be robust to the combined effects of many variants that have only very tiny effects on protein expression or function (which is the case for most common variants). This is precisely because the system has evolved to buffer fluctuations in many components all the time. In addition to being an intrinsic, passive property of the architecture of developmental networks, robustness is also actively promoted through homeostatic feedback loops, which can maintain optimal performance in the face of variations, by regulating the levels of other components to compensate. The effects of such variants should therefore NOT be cumulative – they should be absorbed by the system. (In fact, you could argue that a certain level of noise in the system is a “design feature” because it enables this buffering).

Others have argued precisely the opposite – that robustness permits cryptic genetic variation to accumulate in populations. Cryptic genetic variation has no effect in the context in which it arises (allowing it to escape selection) but, in another context – say in a different environment, or a different genetic background – can have a large effect. This is exactly what robustness allows to happen – indeed, the fact that cryptic genetic variation exists provides some of the best evidence that we have that the systems are robust as it shows directly that mutations in some components are tolerated in most contexts. But is there any evidence that such cryptic variation comprises hundreds or thousands of common variants?

To be fair, proving that is the case would be very difficult. You could argue from animal breeding experiments that the continuing response to selection of many traits means that there must be a vast pool of genetic variation that can affect them, which can be cumulatively enriched by selective breeding, almost ad infinitum. However, new mutations are known to make at least some contribution to this continued response to selection. In addition, in most cases where the genetics of such continuously distributed traits have been unpicked (by identifying the specific factors contributing to strain differences for example) they come down to perhaps tens of loci showing very strong and complex epistatic interactions (1, 2, 3). Thus, just because variation in a trait is multigenic, does not mean it is affected by mutations of small individual effect – an effectively continuous distribution can emerge due to very complex epistatic interactions between a fairly small number of mutations which have surprisingly large effects in isolation.

(I would be keen to hear of any examples showing real polygenicity on the level of hundreds or thousands of variants).

In the case of genetic modifiers of specific mutations – say, where a mutation causes a very different phenotype in different mouse strains – most of the effects that have been identified have been mapped to one or a small number of mutations which have no effect by themselves, but which strongly modify the phenotype caused by another mutation.

These and other findings suggest that (i) cryptic genetic variation relevant to disease is certainly likely to exist and to have important effects on phenotype, but that (ii) such genetic background effects can most likely be ascribed to one, several, or perhaps tens of mutations, as opposed to hundreds or thousands of common polymorphisms.

This is already too long, but it begs the question: if neurodevelopmental systems are so robust, then why do we ever get neurodevelopmental disease? The paradox of systems that are generally robust is that they may be quite vulnerable to large variation in a specific subset of components. Why specific types of genes are in this set, while others can be completely deleted without effect, is the big question. More on that in a subsequent post…

These studies provide an explanation for what seems like a paradox: on the one hand, twin studies show that autism is very strongly genetic (identical twins are much more likely to share a diagnosis than fraternal twins) – on the other, many cases are sporadic, with no one else in the family affected. How can the condition be “genetic” but not always run in the family? The explanation is that many cases are caused by new mutations – ones that arise in the germline of the parents. (This is similar to conditions like Down syndrome). The studies reported in Nature are trying to find those mutations and see which genes are affected.

They are only possible because of the tremendous advances in our ability to sequence DNA. The first genome cost three billion dollars to sequence and took ten years – we can do one now for a couple thousand dollars in a few days. That means you can scan through the entire genome in any affected individual for mutated genes. The problem is we each carry hundreds of such mutations, making it difficult to recognise the ones that are really causing disease.

The solution is to sequence the DNA of large numbers of people with the same condition and see if the same genes pop up multiple times. That is what these studies aimed to do, with samples of a couple hundred patients each. They also concentrated on families where autism was present in only one child and looked specifically for mutations in that child that were not carried by either parent – so-called de novo mutations, that arise in the generation of sperm or eggs. These are the easiest to detect because they are likely to be the most severe. (Mutations with very severe effects are unlikely to be passed on because the people who carry them are far less likely to have children).

There is already strong evidence that de novo mutations play an important role in the etiology of autism – first, de novo copy number variants (deletions or duplications of chunks of chromosomes) appear at a significantly higher rate in autism patients compared to controls (in 8% of patients compared to 2% of controls). Second, it has been known for a while that the risk of autism increases with paternal age – that is, older fathers are more likely to have a child with autism. (Initial studies suggested the risk was up to five-fold greater in fathers over forty – these figures have been revised downwards with increasing sample sizes, but the effect remains very significant, with risk increasing monotonically with paternal age). This is also true of schizophrenia and, in fact, of dominant Mendelian disorders in general (those caused by single mutations). The reason is that the germ cells generating sperm in men continue to divide throughout their lifetime, leading to an increased chance of a mutation having happened as time goes on.

The three studies in Nature were looking for a different class of mutation – point mutations or changes in single DNA bases. They each provide a list of genes with de novo mutations found in specific patients. Several of these showed a mutation in more than one (unrelated) patient, providing strong evidence that these mutations are likely to be causing autism in those patients. The genes with multiple hits include CHD8, SCN2A, KATNAL2 and NTNG1. Mutations in the last of these, NTNG1, were only found in two patients but have been previously implicated as a rare cause of Rett syndrome. This gene encodes the protein Netrin-G1, which is involved in the guidance of growing nerves and the specification of neuronal connections. CHD8 is a chromatin-remodeling factor and is involved in Wnt signaling, a major neurodevelopmental pathway, as well as interacting with p53, which controls cell growth and division. SCN2A encodes a sodium channel subunit; mutations in this gene are involved in a variety of epilepsies. Not much is known about KATNAL2, except by homology – it is related to proteins katanin and spastin, which sever microtubules – mutations in spastin are associated with hereditary spastic paraplegia. How the specific mutations observed in these genes cause the symptoms of autism in these patients (or contribute to them) is not clear – these discoveries are just a starting point, but they will greatly aid the quest to understand the biological basis of this disorder.

The fact that these studies only got a few repeat hits also means that there are probably many hundreds or even thousands of genes that can cause autism when mutated (if there were only a small number, we would see more repeat hits). Some of these will be among the other genes on the lists provided by these studies and will no doubt be recognisable as more patients are sequenced. Interestingly, many of the genes on the lists are involved in aspects of nervous system development or function and encode proteins that interact closely with each other – this makes it more likely that they are really involved.

These studies reinforce the fact that autism is not one disorder – not clinically and not genetically either. Like intellectual disability or epilepsy or many other conditions, it can be caused by mutations in any of a very large number of genes. The ones we know about so far make up around 30% of cases – these new studies add to that list and also show how far we have to go to complete it.

We should recognise too that the picture will also get more complex – in many cases there may be more than one mutation involved in causing the disease. De novo mutations are likely to be the most severe class and thus most likely to cause disease with high penetrance themselves. But many inherited mutations may cause autism only in combination with one or a few other mutations.

These complexities will emerge over time, but for now we can aim to recognise the simpler cases where a mutation in a particular gene is clearly implicated. Each new gene discovered means that the fraction of cases we can assign to a specific cause increases. As we learn more about the biology of each case, those genetic diagnoses will have important implications for prognosis, treatment and reproductive decisions. We can aim to diagnose and treat the underlying cause in each patient and not just the symptoms.

Complex disorders are so named because, while it is clear that they are highly heritable (risk to an individual increases the more closely related they are to someone who has the disorder), their mode of inheritance is far more difficult to discern. Unlike classical Mendelian disorders (such as cystic fibrosis or Huntington’s disease), these disorders do not show simple patterns of segregation within families that would peg them as recessive or dominant, nor can they be linked to mutations in a single gene. This has led people to propose two very different explanations for how they are inherited.

One theory is that such disorders arise due to unfortunate combinations of large numbers of genetic variants that are common in the population. Individually, such variants would have little effect on the phenotype, but collectively, if they surpass some threshold of burden, they could tip the balance into a pathological state. This has been called the common disease/common variant (CD/CV) model.

The alternative model is that these “disorders” are not really single disorders at all – rather they are umbrella terms for collections of a large number of distinct genetic disorders, which happen to result in a similar set of symptoms. Within any individual or family, the disorder may indeed be caused by a particular mutation. Because many of the disorders in question are very severe, with high mortality and reduced numbers of offspring, these mutations will be rapidly selected against in the population. They will therefore remain very rare and many cases of the disorder may arise from new, or de novo, mutations. This has therefore been called the multiple rare variants (MRV) model.

Lately, a number of mixed models have been proposed by various researchers, including myself. Even classical Mendelian disorders rarely show strictly Mendelian inheritance – instead the effects of the major mutations are invariably affected by modifiers in the genetic background. (These are variants with little effect by themselves but which may have a strong effect in combination with some other mutation). If this sounds like a return to the CD/CV model, there are a couple important distinctions to keep in mind. One is the nature of the mutations involved – the mixed model would still invoke some rare mutation that has a large effect on protein function. It may not always cause the disorder by itself (i.e., not every one who carries it will be affected), but could still be called causative in the sense that if the affected individual did not carry it one would expect they would not suffer from the disorder. The other is the number of mutations or variants involved – under the CD/CV model this could number in the thousands (a polygenic architecture), while under the mixed model one could expect a handful to be meaningfully involved (an oligogenic architecture – see diagram from review in Current Opinion in Neurobiology).

The new study, from the lab of Jeff Noebels, aimed to test these models in the context of epilepsy. Epilepsy is caused by an imbalance in excitation and inhibition within brain circuits. This can arise due to a large number of different factors, including alterations in the structural organisation of the brain, which may be visible on magnetic resonance imaging. Many neurodevelopmental disorders are therefore associated with epilepsy as a symptom (usually one of many). But it can also arise due to more subtle changes, not in the gross structure of the brain or the physical wiring of different circuits, but in the way the electrical activity of individual neurons is controlled.

The electrical properties of any neuron – how excitable it is, how long it remains active, whether it fires a burst of action potentials or single ones, what frequency it fires at and many other important parameters – are determined in large part by the particular ion channel proteins it expresses. These proteins form a pore crossing the membrane of the cell, through which electrically charged ions can pass. Different channels are selective for sodium, potassium or calcium ions and can be activated by different types of stimuli – binding a particular neurotransmitter or a change in the cell’s voltage for example. Many channels are formed from multiple subunits, each of which may be encoded by a different gene. There are hundreds of these genes in several large families, so the resultant complexity is enormous.

Many familial cases of epilepsy have been found to be caused by mutations in ion channel genes. However, most epilepsy patients outside these families do not carry these particular mutations. Therefore, despite these findings and despite the demonstrated high heritability, the particular genetic cause of the vast majority of cases of epilepsy has remained unknown. Large genome-wide association studies have looked for common variants that are associated with risk of epilepsy but have turned up nothing of note. The interpretation has been that common variants do not play a major role in the etiology of idiopathic epilepsy (epilepsy without a known cause).

The rare variants model suggests that many of these cases are caused by single mutations in any of the very large number of ion channel genes. A straightforward experiment to test that would be to sequence all these candidate genes in a large number of epilepsy patients. The hope is that it would be possible to shake out the “low hanging fruit” – obviously pathogenic mutations in some proportion of cases. The difficulty lies in recognising such a mutation as pathogenic when one finds it. This generally relies on some statistical evidence – any individual mutation, or such mutations in general, should be more frequent in epilepsy patients than in unaffected controls. The experiment must therefore involve as large a sample as possible and a control comparison group as well as patients.

Klassen and colleagues sequenced 237 ion channel genes in 152 patients with idiopathic epilepsy and 139 healthy controls. What they found was surprising in several ways. They did find lots of mutations in these genes, but they found them at almost equal frequency in controls as in patients. Even the mutations predicted to have the most severe effects on protein function were not significantly enriched in patients. Indeed, mutations in genes already known to be linked to epilepsy were found in patients and controls alike (though 96% of patients had such a mutation, so did 67% of controls). Either these specific mutations are not pathogenic or their effects can be strongly modified by the genetic background.

More interesting results emerged from looking at the occurrence of multiple mutations in these genes in individuals. 78% of patients vs 30% of controls had two or more mutations in known familial epilepsy genes. A similar trend was observed when looking at specific ion channel gene families, such as GABA receptors or sodium channels.

These data would seem to fit with the idea that an increasing mutational load pushes the system over a threshold into a pathological state. The reality seems more complicated, however, and far more nuanced. Though the average load was lower, many controls had a very high load and yet were quite healthy. It seems that the specific pattern of mutations is far more important than the overall number. This fits very well with the known biology of ion channels and previous work on genetic interactions between mutations in these genes.

Though one might expect a simple relationship between number of mutations and severity of phenotype, that is unlikely to be the case for these genes. It is well known that the effects of a mutation in one ion channel gene can be suppressed by mutation in another gene – restoring the electrical balance in the cell, at least to a degree sufficient for performance under normal conditions. The system is so complex, with so many individual components, that these interactions are extremely difficult to predict. This is complicated further by the fact that there are active processes within the system that act to normalise its function. It has been very well documented, especially by Eve Marder and colleagues, that changes to one ion channel in a neuron can be compensated for by homeostatic mechanisms within the cell that aim to readjust the electrical set-points for optimal physiological function. In fact, these mechanisms do not just happen within one cell, but across the circuit.

The upshot of the study is that, though some of the mutations they discovered are indeed likely to be the pathogenic culprits, it is very difficult to discern which ones they are. It is very clear that there is at least an oligogenic architecture for so-called “channelopathies” – the phenotype is determined by several mutations in each individual. (Note that this is not evidence for a highly polygenic architecture involving hundreds or thousands of genetic variants with tiny individual effects). The important insight is that it is not the overall number or mutational load that matters but the pattern of specific mutations in any individual that is crucial. Unfortunately, given how complicated the system is, this means it is currently not possible to predict an individual’s risk, even with this wealth of data. This will likely require a lot more biological information on the interactions between these mutations from experimental approaches and computational modelling.

What are the implications for other complex disorders? Should we expect a similarly complicated picture for diseases like schizophrenia or autism? Perhaps, though I would argue against over-extrapolating these findings. For the reasons described above, mutations in ion channel genes will show especially complex genetic interactions – it is, for example, even possible for two mutations that are individually pathogenic to suppress each other’s effects in combination. This is far less likely to occur for classes of mutations affecting processes such as neurodevelopment, many of which have been implicated in psychiatric disorders. Though by no means unheard of, it is far less common for the effects of one neurodevelopmental mutation to be suppressed by another – it generally just makes things worse. So, while modifying effects of genetic background will no doubt be important for such mutations, there is some hope that the interactions will be more straightforward to elucidate (mostly enhancing, far fewer suppressing). Others may see it differently of course (and I would be pleased to hear from you if you do); similar sequencing efforts currently underway for these disorders may soon tell whether that prediction is correct.

Klassen T, Davis C, Goldman A, Burgess D, Chen T, Wheeler D, McPherson J, Bourquin T, Lewis L, Villasana D, Morgan M, Muzny D, Gibbs R, & Noebels J (2011). Exome sequencing of ion channel genes reveals complex profiles confounding personal risk assessment in epilepsy. Cell, 145 (7), 1036-48 PMID: 21703448

Kasperaviciute, D., Catarino, C., Heinzen, E., Depondt, C., Cavalleri, G., Caboclo, L., Tate, S., Jamnadas-Khoda, J., Chinthapalli, K., Clayton, L., Shianna, K., Radtke, R., Mikati, M., Gallentine, W., Husain, A., Alhusaini, S., Leppert, D., Middleton, L., Gibson, R., Johnson, M., Matthews, P., Hosford, D., Heuser, K., Amos, L., Ortega, M., Zumsteg, D., Wieser, H., Steinhoff, B., Kramer, G., Hansen, J., Dorn, T., Kantanen, A., Gjerstad, L., Peuralinna, T., Hernandez, D., Eriksson, K., Kalviainen, R., Doherty, C., Wood, N., Pandolfo, M., Duncan, J., Sander, J., Delanty, N., Goldstein, D., & Sisodiya, S. (2010). Common genetic variation and susceptibility to partial epilepsies: a genome-wide association study Brain, 133 (7), 2136-2147 DOI: 10.1093/brain/awq130

Mitchell KJ (2011). The genetics of neurodevelopmental disease. Current opinion in neurobiology, 21 (1), 197-203 PMID: 20832285

Mirrored from http://wiringthebrain.blogspot.com