Sunday, June 6, 2010

schizophrenia, autism, genes

There are at least two models of the genetic architecture of schizophrenia characterized by both the frequency and penetrance of risk alleles. The common disease-common allele hypothesis emphasizes the importance of relatively common alleles, each with a small effect, acting together to increase disease risk. The common disease-rare alleles model emphasizes the impact of individually rare, yet highly penetrant alleles. Most likely is that both common and rare alleles contribute to the risk of schizophrenia, although the relative impact of each remains unknown.

Recent data have shown that rare structural mutations leading to an altered copy number of dosage-sensitive genes can lead to the development of neuropsychiatric disorders. For instance, large-scale genome scans have identified several schizophrenia-associated CNVs at 1q21.1, 2p16.3, 15q11.2 and 15q11.3, 16p11.2, 17p12 and 22q11.2.

The report of increased CNV burden in schizophrenia has led to the suggestion that perhaps the same phenomenon may hold for rare de novo highly penetrant single gene mutations.

A Canadian group recently published a paper in PNAS that explores this idea. We'll talk about it some more tomorrow.

Thursday, June 3, 2010

Schizophenia, autism, genes....(2)

I will briefly summarize a recent PNAS paper dealing with the possible involvement of a rare disease allele in the causation of a common disease, schizophrenia.

De novo mutations in the gene encoding the synaptic scaffolding protein SHANK3 in patients ascertained for schizophrenia.
Gauthier J, Champagne N, Lafrenière RG, Xiong L, Spiegelman D, Brustein E, et al
Proc Natl Acad Sci U S A. 2010 Apr 27;107(17):7863-8

Hypothesis: a “significant” fraction of schizophrenia cases are due to new mutations. This hypothesis derives from 1) a reported association between paternal age and schizophrenia 2) constant 1% schizophrenia incidence in various populations despite decreased reproductive fitness 3) report of increased de novo CNV rate in schizophrenia. Note that there are alternate explanations for the first two scenarios, for instance, since schizophrenia is a spectrum disorder the relevant genes may be propagated by individuals with schizotypal or schizoid personality disorders, rather than schizophrenics; secondarily, these individuals may marry or have children later due to their behavioral peculiarities.

This paper reports on screening of one of many rare genetic causes of autism, the SHANK3 gene, in a large (185 probands) cohort of individuals ascertained for schizophrenia. 6 nonsynonymous variants were found only in the schizophrenia cohort, and 4 (H494Q, S952T, G1011V, and P1134H) were transmitted from an unaffected parent and excluded from further analysis. This quoted comment "Therefore, these four transmitted nonsynonymous variants can be excluded from a direct role as dominant mutations in SCZ," is not necessarily accurate from a genetics standpoint, as dominant mutations may show incomplete penetrance or variable expressivity, etc. The authors appear to be assuming that penetrance of a dominant disease gene must be high. In any case, two cases had de novo mutations not found in 285 controls. One of these individuals had two affected brothers with the same mutation, likely due to paternal gonadal mosaicism. All individuals had premorbid MR. It seems that this gene may be responsible for mental retardation primarily, with associated behavioral phenotypes such as autistic behaviors and nonspecific psychosis. I think its unlikely to account for schizophrenia in the healthy premorbid population, i.e., it is unlikely to be a schizophrenia-specific gene. Note that classical schizophrenia is characterized by usually normal premorbid development, with a later psychotic break (think John Forbes Nash).

The authors do some interesting neurobiological assays that show that one of the mutations (nonsense mutation found in the three brothers) fails to promote somatic sprouting of neurites compared with control neurons, while the missense mutation was not shown to affect neurite outgrowth. Note that there have been other groups who've reported lesions in genes in the same area causing neurite outgrowth defects and schizophrenia (e.g, Budel et al, J Neurosci. 2008 Dec 3;28(49):13161-72; Genetic variants of Nogo-66 receptor with possible association to schizophrenia block myelin inhibition of axon growth....a paper which provides evidence that four human NgR1 variants from schizophrenic individuals are functionally inactive in myelin-induced growth cone collapse assay, may possess dominant negative function in vitro, and may severely disrupt NgR1 signaling in individuals bearing these variants.) Nogo-66 is in the 22q11 region. Shank 3 is in the 22q13 region and is not deleted in the 22q11 deletion syndrome.

Note that there is already a well-described 22q13 deletion syndrome (Phelan-McDermid syndrome), which is related to moderate to severe developmental delay and mental retardation. The Shank3 gene is thought to be responsible for the neurological deficits of the syndrome (Wilson et al., 2003).

MR is already though to be a lesion of dendritic and synaptic pathology in some cases, perhaps diverse schizophrenia genetic pathologies will be shown to coalesce around a final common pathway of dysfunctional neurite outgrowth......

Wednesday, September 23, 2009

Where's the missing heritability?

Epigenetic inheritance and the missing heritability
Slatkin, M. Genetics 182, 845–850 (2009)

Summary of my reading of the paper:

The author derives the standard population genetic model for locus contribution to average risk and recurrence risk......

Derives the same sets of equations for epigenetic marks with the difference being that there are loss and gain rates of epigenetic marks to be taken into account

For traditional genetic loci: Describe the relation between two variables 1. allele frequency and 2. locus contribution to risk leading to two outcomes 3. recurrence risk and 4. average risk

Do the same for ‘epigenetic loci’ with the difference that for epigenetic loci in their model:
The equilibrium frequency of a two-state Markov process incorporating the transition probabilities α and β, the rates of loss and gain of epigenetic marks per meiosis, acts as a proxy for population frequency of epigenetic marks and has great bearing on average risk

α + β, the turnover rate of the marks, affects recurrence risk in sibs
i. Analogous to the genetic situation, alpha has to be low and epialleles have to be relatively stable for recurrence risk to be elevated
ii. Stables epialleles are likely to act like mutations and be detected in linkage studies
iii. High frequency epialleles with low effects (and even if they change with every generation, i.e, are not stable) may affect average risk substantially
5. Stated Conclusion: Epialleles, in this model, may account for missing causality but unlikely to account for missing heritability because the higher rate of loss of epigenetic modifications means that identity by descent does not imply identity in state; consequently, it will be difficult for epigenetic changes to account for the missing heritability of complex diseases unless they are more common than mutations or have more pronounced effects on disease risk.

is transgenerational
does not allow for epistatic interactions between loci

Saturday, July 11, 2009

Lithium and inositol signaling

Anyone following the literature on inositol signaling and Lithium? The plot thickens. Here is a brief summary and later I'll talk about things in detail in future posts for those more interested in learning more.

In brief, there's been a longstanding notion that perturbation of phosphoinositide signaling may well be the underlying cause of bipolar disorder. The reason for this is that Lithium, the most well-known mood stabilizer drug is known to decrease intracellular inositol, and more recently two other well-known anticonvulsant mood stabilizers have been shown to have the same effect. However, functional studies since then have been contradictory. Some groups have shown that intracellular inositol depletion per se does not perturb phosphoinositide signaling, using knockout animal models; recently another group published data using the ameba Dictyostelium, and human white blood cells, consistent with the idea that Lithium perturbs PIP3 signaling. The question that emerges is whether there is some methodological reason for these contradictory findings. Stay posted!

Thursday, July 9, 2009

Cerebral folate deficiency (CFD) a phenocopy of Rett syndrome

CFD syndrome is a newly described condition characterized by low levels of 5-methyltetrahydrofolate (the biologically active form of folates) in the cerebrospinal fluid and normal folate levels in the plasma and red blood cells. In infancy the onset of symptoms is around 4 to 6 months of age, with initially normal development followed by delayed development, deceleration of head growth, hypotonia, ataxia, and in one-third of children, dyskinesias, spasticity, speech difficulties, and epilepsy. Autistic features may be present. Visual disturbances and sensorineural hearing loss may develop around the age of 6 years. In some patients fronto-temporal atrophy with signs of periventricular and subcortical demyelination can become apparent from the age of 18 months.

This bears more than a passing resemblance with the temporal course of classic Rett syndrome. The latter condition is characterized by apparently normal psychomotor development during the first six to 18 months of life, with head growth deceleration as early as age three months followed by a period of developmental stagnation and regression in language and motor skills. Other features include autistic symptoms, episodic apnea and/or hyperpnea, gait ataxia and apraxia and tremors. Neurologic manifestations become relatively stable, although patients will likely develop dystonia and foot and hand deformities later. Seizures are reported in up to 90% of Rett syndrome cases. The hallmark of classic Rett syndrome is the loss of purposeful hand use and its replacement with repetitive stereotyped hand movements.

Both conditions have in common normal early development, regression at 4-6 months, deceleration of head growth, neurologic and autistic phenomena, and seizures. There are differences in the natural history of the two conditions. Classic Rett hand stereotypies are not generally seen in CFD. There is no gender predilection for the latter. Still, in a very young new patient below the age of 3 or 4 years with no family history, the two conditions can appear indistinguishable. Note that Cerebral folate deficiency is associated with low levels of 5-methyltetrahydrofolate in the cerebrospinal fluid with normal folate levels in the plasma and red blood cells. The low level of 5-methyltetrahydrofolate in the CSF can result from decreased transport across the blood–brain barrier, which is most probably because of the blocking of folate transport into the CSF by the binding of folate receptor antibodies to the folate receptors in the choroid plexus. Note that this condition is different from folinic-acid responsive seizure disorder and pyridoxine-dependent epilepsy, two treatable causes of neonatal epileptic encephalopathy. The former is diagnosed by characteristic peaks on cerebrospinal fluid (CSF) monoamine metabolite analysis; its genetic basis has remained elusive. The latter is due to alpha-aminoadipic semialdehyde (alpha-AASA) dehydrogenase deficiency, associated with pathogenic mutations in the ALDH7A1 (antiquitin) gene. Treatment of CFD with folinic acid for prolonged periods can result in significant improvement of clinical symptoms and a return of 5-methyltetrahydrofolate levels in the CSF to normal. There is no treatment for classic Rett syndrome. The true incidence of CFD is unknown; in view of the response to treatment in CFD and allied conditions, however, it probably merits consideration in the differential diagnosis of a patient with regression, infantile seizures, microcephaly and autistic phenomena.

Sunday, June 28, 2009