Schizophrenia as genetic pelmanism
If you take a brand new pack of cards and start shuffling, it is not hard to appreciate that the longer you continue, the less likely it will be that you will find a series of cards in the same order as in the beginning. The European and Asian genomes are like a pack of cards that effectively started shuffling as humans first walked “Out of Africa” some 100,000 years ago. Meiotic recombination is the shuffling process and the result is a decreasing ability to predict at the gross level what combinations of marker alleles will be found together on a chromosome. African populations, with a longer “shuffling” time and without population bottlenecks (which effectively reorder the cards) show the least predictability (“linkage disequilibrium,” LD) across their genomes.

There are two counteracting forces to halt or even reverse this entropic breakdown. Firstly, if a particular region becomes strongly selected for, then its frequency increase in the population [...continued] will, in the medium-term, outrun the shuffling effect such that the region flanking the selected genetic variant will maintain its order (Gibson et al., 2006; Li et al., 2006). This is known as a “selective sweep,” and numerous post-HapMap studies have successfully fished out regions of our genomes under this selective pressure (e.g., the lactose tolerance variant in populations where milk became a part of the staple prehistoric diet (Tishkoff et al., 2007). Secondly, and rather more obscurely, there can be physical restraints to recombination shuffling. These usually involve the physical reordering of sequence on our chromosomes, for example, in the case of paracentric inversions. The physical alignment of normal and inverted DNA sequences during meiosis is thus prevented and so recombination is suppressed, leading to greater LD.

Now imagine the situation where reasonably common stretches of less-shuffled chromosomes exist in the population. These are more likely to be found as matching pairs in any given individual compared to other parts of the genome. This appears to the researcher as a long stretch or tract of homozygous DNA. Such tracts have been studied elsewhere, particularly in the context of mapping and identifying recessive disease genes in remote, consanguineous (inbred) populations where the recessive mutations in genomic DNA of reduced allelic complexity are not only more likely to be exposed but occur within prominent tracts which co-segregate with the diagnosis. A newly published paper by Lencz et al. takes all of these ideas and combines them into a single strategy to hunt for schizophrenia-causing genes. They took raw data from their recently published genomewide association study of schizophrenia (178 cases of schizophrenia and 144 healthy controls: Lencz et al., 2007) and reassessed it for the presence of long “runs of homozygosity” (ROH) restricted to the case group. Their hypothesis was that if these regions existed, they would contain recessive mutations contributing to the disease.

Three hundred thirty-nine common ROHs were identified in the study, making up 12-13 percent of the total genome. The largest of these were predominantly found spanning the chromosome centromeres. This is perhaps not surprising since recombination rates have long been known to be reduced (through repression rather than selective sweep) at centromeres (see Kong et al., 2002). Nine of the commonest ROHs neatly overlap with previously described regions from selective sweep studies, as would be predicted. The key finding, however, was that when ROHs were compared between cases and controls, nine were found significantly more frequently in schizophrenia. Within these tracts, numerous genes were identified and, of these, there is pre-existing evidence in support of a few of them as potential candidates including NOS1AP, ATF2, NSF, MAPT, PIK3C3, and SNTG1.

One caveat to these findings is that a region of homozygosity, a loss of heterozygosity, copy number variation (CNV), and a deletion can, in some instances, all refer to the same genomic lesion and are not simple to distinguish by chip-based genotyping. The authors are careful to spell out technical and biological reasons for believing that their findings are a reflection of true homozygosity, but further independent verification would be reassuring, particularly in the context of how CNVs/genomic rearrangements might complicate recombination rates.

The significance of these findings is that we now have the potential to explore a brand new mutation class in a complex genetic disorder. Until now, the major research techniques such as linkage, association, and cytogenetics have only identified (and perhaps can only identify) dominantly behaving variants, albeit mostly with reduced penetrance. These are presumed to act through gain-of-function or, more likely, loss-of-function/haploinsufficiency mechanisms. The ROH regions described here are predicted to house reasonably common recessive risk variants: such properties meaning that they are not likely to be present in ascertained families with high densities of affected individuals but rather sporadic cases of illness where these alleles have, by chance, been inherited from both parents. It is not entirely clear why some of the more common ROHs didn’t feature in the original association study based on this data, particularly in genotype frequency rather than allele frequency analyses.

Nevertheless, the authors also make an additional, intriguing claim that these ROHs are not only overrepresented in the schizophrenia cohort because they are causative but because they have also been subject to positive selection. They cite the discovery of these ROHs in previous selective sweep scans, their more recent derivation from ancestral haplotypes, the presence of genes within which show selection pressure through alternative analyses, and their restriction to Caucasian populations as good evidence for such a claim. This effect may be due to some form of “heterozygote advantage” (also known as “overdominance”) which maintains or promotes the deleterious allele in the population. Examples where this phenomenon has been observed include recessive mutations giving rise to sickle-cell anemia, cystic fibrosis, and triose phosphate isomerase deficiency. Others have previously hypothesized that selection for the greater cognitive abilities in Homo sapiens compared to earlier hominins might have been at the cost of the emergence of schizophrenia, although the timescales of this kind of selection and the kind resulting in selective sweep are likely to be vastly different. An alternative explanation discussed in the paper is that rare recessive mutations could have “hitchhiked” their way to prominence within the selective sweep driven by a favorable variant in a closely linked gene. This latter idea seems more reasonable, given the difficulty in trying to imagine what cognitive or neurodevelopmental features would have been exclusively beneficial for the Caucasian population. It might also tally with some of the phenotypic epiphenomena that may coexist with schizophrenia (e.g., altered risk of rheumatoid arthritis, etc).

Finally, as an aside, this represents the third method of analysis, after the principal case-control studies and prediction of copy number variants, which can be applied to the large genomewide genotyping datasets being produced in numerous labs. Are there other aces waiting to be found in the hand?

References:

Gibson J, Morton NE, Collins A. Extended tracts of homozygosity in outbred human populations. Hum Mol Genet. 2006 Mar 1;15(5):789-95. Abstract

Kong A, Gudbjartsson DF, Sainz J, Jonsdottir GM, Gudjonsson SA, Richardsson B, Sigurdardottir S, Barnard J, Hallbeck B, Masson G, Shlien A, Palsson ST, Frigge ML, Thorgeirsson TE, Gulcher JR, Stefansson K. A high-resolution recombination map of the human genome. Nat Genet. 2002 Jul 1;31(3):241-7. Abstract

Lencz T, Morgan TV, Athanasiou M, Dain B, Reed CR, Kane JM, Kucherlapati R, Malhotra AK. Converging evidence for a pseudoautosomal cytokine receptor gene locus in schizophrenia. Mol Psychiatry. 2007 Jun 1;12(6):572-80. Abstract

Li LH, Ho SF, Chen CH, Wei CY, Wong WC, Li LY, Hung SI, Chung WH, Pan WH, Lee MT, Tsai FJ, Chang CF, Wu JY, Chen YT. Long contiguous stretches of homozygosity in the human genome. Hum Mutat. 2006 Nov 1;27(11):1115-21. Abstract

Tishkoff SA, Reed FA, Ranciaro A, Voight BF, Babbitt CC, Silverman JS, Powell K, Mortensen HM, Hirbo JB, Osman M, Ibrahim M, Omar SA, Lema G, Nyambo TB, Ghori J, Bumpstead S, Pritchard JK, Wray GA, Deloukas P. Convergent adaptation of human lactase persistence in Africa and Europe. Nat Genet. 2007 Jan 1;39(1):31-40. Abstract

Lencz T, Lambert C, DeRosse P, Burdick KE, Morgan TV, Kane JM, Kucherlapati R, Malhotra AK. (2007) Runs of homozygosity reveal highly penetrant recessive loci in schizophrenia. PNAS.