Everyone has a birthday right. Its the day you (your infant self) popped into the world and started breathing, right?  But what about the day “you” were born – that is – “you” in the more philosophical, Jungian, spiritual, social, etc. kind of a way when you became aware of being in some ways apart from others and the world around you.  In her 1997 paper, “The Basal Ganglia and Cognitive Pattern Generators“, Professor Ann Graybiel writes,

The link between intent and action may also have a quite specific function during development. This set of circuits may provide part of the neural mechanism for building up cognitive patterns involving recognition of the self. It is well documented that, as voluntary motor behaviors develop and as feedback about the consequences of these behaviors occurs, the perceptuomotor world of the infant develops (Gibson 1969). These same correlations among intent, action, and consequence also offer a simple way for the young organism to acquire the distinction between actively initiated and passively received events. As a result, the infant can acquire the recognition of self as actor. The iterative nature of many basal ganglia connections and the apparent involvement of the basal ganglia in some forms of learning could provide a mechanism for this development of self-awareness.

As Professor Graybiel relates the “self” to function in the basal-ganglia and the so-called cortico-thalamic basal-ganglia loops – a set of parallel circuits that help to properly filter internal mental activity into specific actions and executable decisions – I got a kick out of a paper that describes how the development of the basal-ganglia can go awry for cells that are born at certain times.

Check out the paper, “Modular patterning of structure and function of the striatum by retinoid receptor signaling” by Liao et al.   It reveals that mice who lack a certain retinoic acid receptor gene (RARbeta) have a type of defective neurogenesis in late-born cells that make up a part of the basal ganglia (striatum) known as a striosome.  Normally, the authors say, retinoic acid helps to expand a population of late-born striosomal cells, but in the RARbeta mutant mice, the rostral striosomes remain under-developed.   When given dopaminergic stimulation, these mutant mice showed slightly less grooming and more sterotypic behaviors.

So when was “my self’s” birthday?  Was it when these late-born striosomal cells were, umm, born?  Who knows, but I’m glad my retinoic acid system was intact.

For a great many reasons, research on mental illness is focused on the frontal cortex.  Its just a small part of the brain, and certainly, many things can go wrong in other places during brain/cognitive development, but, it remains a robust finding, that when the frontal cortex is not working well, individuals have difficulties in regulating thoughts and emotions.  Life is difficult enough to manage, let alone without a well functioning frontal cortex.  So its no surprise that many laboratories look very closely at how this region develops prenatally and during childhood.

One of the more powerful genetic methods is the analysis of gene expression via microarrays (here is a link to a tutorial on this technology).  When this technology is coupled with extremely careful histological analysis and dissection of cortical circuits in the frontal cortex, it begins to become possible to begin to link changes in gene expression with the physiological properties of specific cells and local circuits in the frontal cortex. The reason this is an exciting pursuit is because the mammalian neocortex is organized in a type of layered fashion wherein 6 major layers have different types of connectivity and functionality.  The developmental origins of this functional specificity are thought to lie in a process known as radial migration (here is a video of a neuron as it migrates radially and finds its place in the cortical hierarchy).  As cells are queued out of the ventricular zone, and begin their migration to the cortical surface, they are exposed to all sorts of growth factors and morphogens that help them differentiate and form the proper connectivities.  Thus, the genes that regulate this process are of keen interest to understanding normal and abnormal cognitive development.

Here’s an amazing example of this – 2 papers entitled, “Infragranular gene expression disturbances in the prefrontal cortex in schizophrenia: Signature of altered neural development?” [doi:10.1016/j.nbd.2009.12.013] and “Molecular markers distinguishing supragranular and infragranular layers in the human prefrontal cortex [doi:10.1111/j.1460-9568.2007.05396.x] both by Dominique Arion and colleagues.  In both papers, the authors ask, “what genes are differentially expressed in different layers of the cortex?”.  This is a powerful line of inquiry since the different layers of cortex are functionally different in terms of their connectivity.  For example, layers II-III (the so-called supragranular layers) are known to connect mainly to other cortical neurons – which is different functionally than layers V-VI (the so-called infragranular layers) that connect mainly to the striatum (layer V) and thalamus (layer VI).  Thus, if there are genes whose expression is unique to a layer, then one has a clue as to how that gene might contribute to normal/abnormal information processing.

The authors hail from a laboratory that is well-known for work over many years on fine-scaled histological analysis of the frontal cortex at the University of Pittsburgh and used a method called, laser capture microdissection, where post-mortem sections of human frontal cortex (area 46) were cut to separate the infragraular layer from the supragranular layer.  The mRNA from these tissue sections was then used for DNA microarray hybridization.  Various controls, replicate startegies and in-situ tissue hybridizations were then employed to validate the initial microarray results.

In first paper, the where the authors compare infra vs. supragranular layers, they report that 40 genes were more highly expressed in the supragranular layers (HOP, CUTL2 and MPPE1 were among the most enriched) and 29 genes were highly expressed in the infragranular layers (ZNF312, CHN2, HS3ST2 were among the most enriched).  Other differentially expressed genes included several that have previously been implicated in cortical layer formation such as RLN, TLX-NR2E1, SEMA3E, PCP4, SERPINE2, NR2F2/ARP1, PCDH8, WIF1, JAG1, MBP.  Amazing!! A handful of genes that seem to label subpopulations of projection neurons in the frontal cortex.  Polymorphic markers for these genes would surely be powerful tools for imaging-genetic studies on cognitive development.

In the second paper, the authors compare infra vs. supragranular gene expression in post-mortem brains from patients with schizophrenia and healthy matched controls. Using the same methods, the team reports both supra- and infragranular gene expression changes in schizophrenia (400 & 1200 differences respectively) – more than 70% of the differences appearing to be reductions in gene expression in schizophrenia. Interestingly, the team reports that the genes that were differentially expressed in the infragranular layers provided sufficient information to discriminate between cases and controls, whilst the gene expression differences in the supragranular layers did not.  More to the point, the team finds that 51 genes that were differentially expressed in infra- vs. supragranular expression were also differentially expressed in cases vs. controls  (many of these are also found to be associated in population genetic association studies of schiz vs. control as well!).  Thus, the team has identified layer (function) -specific genes that are associated with schizophrenia.  These genes, the ones enriched in the infragranular layers especially, seem to be at the crux of a poorly functioning frontal cortex.

The authors point to 3 such genes (SEMA3E, SEMA6D, SEMA3C) who happen to members of the same gene family – the semaphorin gene family.  This gene family is very important for the neuronal guidance (during radial migration), morphology, pruning and other processes where cell shape and position are regulated.  The authors propose that the semaphorins might act as “integrators” of various forms of wiring during development and in adulthood.  More broadly, the authors provide a framework to understand how the development of connectivity on the frontal cortex is regulated by genetic factors – indeed, many suspected genetic risk factors play a role in the developmental pathways the authors have focused on.

The genome is just the A,G,T,C bases that encode proteins and other mRNA molecules.  The “epi”genome are various modification to the DNA – such as methylation (at C residues) – and acetylation of histone proteins.   These changes help the DNA form various secondary and tertiary structures that can facilitate or block the interaction of DNA with the transcriptional machinery.

When DNA is highly methylated, it generally is less accessible for transcription and hence gene expression is reduced.  When histone proteins (purple blobs that help DNA coil into a compact shape) are acetylated, the DNA is much more accessible and gene expression goes up.

We know that proper epigenetic regulation is critical for cognitive development because mutations in MeCP2 – a protein that binds to methylated C residues – leads to Rett syndrome.  MeCP2 is normally responsible for binding to methylated DNA and recruiting histone de-acetylases (HDACs) to help DNA coil and condense into a closed form that is inaccessible for gene expression (related post here).

When DNA is accessible for gene expression, then it appears that – during brain development – there are relatively more synaptic spines produced (related post here).  Is this a good thing? Rett syndrome would suggest that – NO – too many synaptic spines and too much excitatory activity during brain development may not be optimal.  Neither is too little excitatory (too much inhibitory) activity and too few synaptic spines.  It is likely that you need just the right balance (related post here). Some have argued (here) that autism & schizophrenia are consequences of too many & too few synapses during development.

The sketch above illustrates a theoretical conjecture – not a scenario that has been verified by extensive scientific study. It tries to explain why epigenetic effects can, in practice, be difficult to disentangle from true (changes in the A,G,T,C sequence) genetic effects.  This is because – for one reason – a mother’s experience (extreme stress, malnutrition, chemical toxins) can – based on some evidence – exert an effect on the methylation of her child’s genome.  Keep in mind, that methylation is normal and widespread throughout the genome during development.  However, in this scenario, if the daughter’s behavior or physiology were to be influenced by such methylation, then she could, in theory, when reaching reproductive age, expose her developing child to an environment that leads to altered methylation (shown here of the grandaughter’s genome).  Thus, an epigenetic change would look much like there is a genetic variant being passed from one generation to the next, but such a genetic variant need not exist (related post here, here) – as its an epigenetic phenomenon.  Genes such as BDNF have been the focus of many genetic/epigenetic studies (here, here) – however, much, much more work remains to determine and understand just how much stress/malnutrition/toxin exposure is enough to cause such multi-generational effects.  Disentangling the interaction of genetics with the environment (and its influence on the epigenome) is a complex task, and it is very difficult to prove the conjecture/model above, so be sure to read the literature and popular press on these topics carefully.

Other research covered (here, here) demonstrates the importance of the proper balance of excitatory and inhibitory signalling during cortical development.

pointer to symptommedia.org – fantastic video resource of specific symptoms of mental illness.

“The intention of these clips are to be used in the classroom setting as visual compliments to the written description of symptoms for psychological phenomena found in the DSM handbook.”

One of the difficult aspects of understanding mental illness, is separating the real causes of the illness from what might be secondary or tertiary consequences of having the illness.  If you think about a car whose engine is not running normally, there may be many observable things going wrong (pinging sound, stalling, smoke, vibration, overheating, loss of power, etc.) – but, what is the real cause of the problem?  What should be done to fix the car? – a faulty sparkplug or timing belt perhaps?  Such is often the problem in medicine, where a fundamental problem can lead to a complex, hard-to-disentangle, etiology of symptoms.  Ideally, you would fix the core problem and then expect the secondary and tertiary consequences to normalize.

This inherent difficulty, particularly in mental illness, is one of the reasons that genetic research is of such interest.  Presumably, the genetic risk factors are deeper and more fundamentally involved in the root causes of the illness – and hence – are preferable targets for treatment.  The recent paper, “Widespread Reductions of Cortical Thickness in Schizophrenia and Spectrum Disorders and Evidence of Heritability” [Arch Gen Psychiatry. 2009;66(5):467-477] seeks to ascertain whether one aspect of schizophrenia – a widespread and well-documented thinning of the neocortex – is due to genetic risk (hence something that is closer to a primary cause) or – rather – if cortical thinning is not due to genetics, and so more of a secondary consequence of things that go wrong earlier in the development of the illness.

To explore this idea, the team of Goldman et al., did something novel.  Rather than examine the differences in cortical thickness between patients and control subjects, the team evaluated the cortical thickness of 59 patients and 72 unaffected siblings as well as 196 unrelated, matched control participants.  If the cortical thickness of the siblings (who share 50% of their genetic variation) was more similar to the patients, then it would suggest that the cortical thinning of the patients was under genetic control and hence – perhaps – a biological trait that is more of a primary cause.  On the other hand, if the cortical thickness of the siblings (who share 0% of their genetic variation) was more similar to that of the healthy control participants, then it would suggest that cortical thinning was – perhaps more of a secondary consequence of some earlier deficit.

The high-resolution structural neuroimaging allowed the team to carefully assess cortical thickness - which is normally between a mere 2 and 4 millimeters – across different areas of the cortex.  The team reports that, for the most part, the cortical thickness measures of the siblings were more similar to the unrelated controls – thus suggesting that cortical thickness may not be a direct component of the genetic risk architecture for schizophrenia.  Still, the paper discusses several candidate mechanisms which could lead to cortical thinning in the illness – some of which might be assessed in the future using other imaging modalities in the context of their patient/sibling/control experimental design.

The neuregulin-1 (NRG1) gene is widely known as one of the most well-replicated genetic risk factors for schizophrenia.  Converging evidence shows that it is associated with schizophrenia at the gene expression and mouse model levels which are consistent with its molecular functions in neural development.   However, in several recent genome-wide association studies (GWAS), there appeared nary a blip of association at the 8p12 locus where NRG1 resides.  What gives?

While there are many possibilities for this phenomenon (some discussed here), the recent paper, “Support for NRG1 as a Susceptibility Factor for Schizophrenia in a Northern Swedish Isolated Population” by Maaike Alaerts and colleagues, suggest that the typical GWAS study may not adequately probe genetic variation at a fine enough scale – or, if you will, use a netting with sufficiently small holes.  By holes, I mean both the physical distance between genetic markers and the frequency with which they occur in populations.  While GWAS studies may use upwards of 500,000 markers – that’s a pretty fine scale net for a 3,000,000,000bp genome (about 6,000bp apart) – Alaerts and colleagues set forth with slightly finer-scale netting.  They focus on a 157kb region that is about 60kb upstream from the start of the NRG1 gene and construct a net consisting of 37 variants between the markers rs4268087 and rs17601950 (average spacing about 5kb).  They used the tagger program to select markers that account for all haplotypes whose frequency is higher than 1.5%.  Thus – even though there are still more than 500 possible snps in the region Alaerts and colleagues are exploring, they are using a slightly finer netting than a typical GWAS.

The results of their analysis (using GENEPOP) of 486 patients and 514 ethnically matched control participants from northern Sweden did reveal significant associations in an area slightly downstream (about 50kb closer to the start point of the NRG1 gene) than the location of the “previously often replicated variants”, suggesting that the region does confer some risk for schizophrenia, but, that diagnostic markers for such risk will be different for different populations.  More telling however are the very weak effects of the haplotypes that show significant association.  Those haplotypes with the most significance show meager differences in how often they are observed in patients vs. controls.  For example, one haplotype was observed in 5% of patients vs. 3% of controls. Others examples were, 11 vs. 9, 25 vs. 22 and 40% vs. 35% – revealing the very modest (krill sized) effects that single genetic variants can have in conferring risk toward mental illness.

However, there are potentially lots of krill in the genomic sea!

File this story under “the more you know, the more you don’t know” or simply under “WTF!“  The new paper, “Microduplications of 16p11.2 are associated with schizophrenia” [doi:10.1038/ng.474] reveals that a short stretch of DNA on chromosome 16p11.2 is – very rarely – duplicated and – more rarely – deleted.  In an analysis of 8,590 individuals with schizophrenia, 2,172 with developmental delay or autism, 4,822 with bipolar disorder and 30,492 controls, the the microduplication of 16p11.2 was strongly associated with schizophrenia, bipolar and autism while the reciprocal microdeletion was strongly associated with developmental delay or autism – but not associated with schizophrenia or bipolar disorder.

OK, so the title of my post is misleading (hey, its a blog) since there are clearly many additional factors that contribute to the developmental outcome of autism vs. schizophrenia, but this stretch of DNA seems to hold clues about early development of brain systems that go awry in both disorders.  Here is a list of the brain expressed genes in this 600 kbp region (in order from telomere-side to centromere-side): SPN, QPRT, C16orf54, MAZ, PRRT2, C16orf53, MVP, CDIPT, SEZ6L2, ASPHD1, KCTD13, TMEM219, TAOK2, HIRIP3, INO80E, DOC2A, FLJ25404, FAM57B, ALDOA, PPP4C, TBX6, YPEL3, GDPD3, MAPK3, CORO1A.

Any guess as to which one(s) are the culprits?  I’ll go with HIRIP3 given its role in chromatin structure regulation – and the consequent regulation of under- (schiz?)/over- (autism) growth of synapses. What an amazing mystery to pursue.

Daniel R. Weinberger, M.D., Chief of the Clinical Brain Disorders Branch and Director of the Genes, Cognition and Psychosis Program, National Institute of Mental Health  discusses the background, findings and general issues of genes and mental illness in this brief interview on his paper, “A primate-specific, brain isoform of KCNH2 affects cortical physiology, cognition, neuronal repolarization and risk of schizophrenia”.  Click  HERE for the podcast and HERE for the original post.

Thanks again to Dr. Weinberger for his generous participation!

pointer to: download Power Point presentation hosted on the HUGO website entitled, “From the human genome to human behaviour: how far have we travelled?” (both English and Russian text) – by Ian Craig and Nick Yankovsky, Education Council Human Genome Organisation.

Covers recent findings on MAOA and 5HTT several and others also covered here.

Congrats to Hsien on the new position!