Genomic regulation of natural variation in cortical and noncortical brain volume

Abstract

Background: The relative growth of the neocortex parallels the emergence of complex cognitive functions across species. To determine the regions of the mammalian genome responsible for natural variations in cortical volume, we conducted a complex trait analysis using 34 strains of recombinant inbred (Rl) strains of mice (BXD), as well as their two parental strains (C57BL/6J and DBA/2J). We measured both neocortical volume and total brain volume in 155 coronally sectioned mouse brains that were Nissl stained and embedded in celloidin. After correction for shrinkage, the measured cortical and noncortical brain volumes were entered into a multiple regression analysis, which removed the effects of body size and age from the measurements. Marker regression and interval mapping were computed using WebQTL.

Results: An ANOVA revealed that more than half of the variance of these regressed phenotypes is genetically determined. We then identified the regions of the genome regulating this heritability. We located genomic regions in which a linkage disequilibrium was present using WebQTL as both a mapping engine and genomic database. For neocortex, we found a genome-wide significant quantitative trait locus (QTL) on chromosome 11 (marker D11Mit19), as well as a suggestive QTL on chromosome 16 (marker D16Mit100). In contrast, for noncortex the effect of chromosome 11 was markedly reduced, and a significant QTL appeared on chromosome 19 (D19Mit22).

Conclusion: This classic pattern of double dissociation argues strongly for different genetic factors regulating relative cortical size, as opposed to brain volume more generally. It is likely, however, that the effects of proximal chromosome 11 extend beyond the neocortex strictly defined. An analysis of single nucleotide polymorphisms in these regions indicated that ciliary neurotrophic factor (Cntf) is quite possibly the gene underlying the noncortical QTL. Evidence for a candidate gene modulating neocortical volume was much weaker, but Otx1 deserves further consideration.

Figure 1

QTL Plot for Residual Neocortal Volume. The red upper line indicates the LRS value for genome-wide significant QTL and the lower grey line the level of suggestive QTL according to the criteria of Lander and Kruglyak [33]. There is a genome-wide significant linkage disequilibrium located in the proximal portion of chromosome 11. There is also a suggestive association on proximal chromosome 16. Notice that there is no significant association between neocortical volume and any marker on chromosome 19.

Figure 2

QTL Plot for Residual Noncortal Volume. Here, the effect of the QTL on proximal chromosome 11 is attenuated and no longer significant. There is a suggestive association on the distal portion of chromosome 15 and a significant QTL on chromosome 19. Thus, there is a double dissociation between neocortex and noncortex for significant QTLs on chromosomes 11 and 19.