Early fossil primates older than 40 million years have small body sizes and small brain sizes, which reflect their recent divergence from other mammalian orders (Martin, 1990
). One hallmark of primate evolutionary history is the origin of new species and new evolutionary groups that exhibit both increased body size and increased brain size relative to body size. As a result, the primate species living today exhibit a broad range of brain sizes and brain:body size ratios (Martin, 1990
; Preuss, 2007b
). In parallel with increasing brain size, the cerebral cortex of more recently appearing primate groups such as apes and humans shows much more extensive folding or gyrification. Little is known about the specific genetic mechanisms that underlie this suite of parallel evolutionary changes in brain structure, but this pattern is fundamental to the history of primate brain evolution.
Our results indicate that genetic variation now segregating in both baboon and human populations does influence individual differences in gyrification. The specific genes or genetic mechanisms that produce the cortical folding characteristic of Old World monkeys, apes and humans have not yet been identified (but see ref. (Piao et al., 2004)
). The embryonic development of cortical convolutions (sulci and gyri) in the human and nonhuman primate brain has been attributed to tension-based developmental mechanisms involving axonal connections between functional regions within the developing cortex (Van Essen, 1997
). Under this model, sulci and gyri result from the differential strength of connections in the developing brain, with gyri forming as a result of strong connections between the two cortical regions that grow into close proximity, while sulci form between regions that lack strong connections. This mechanical tension-based model, which is supported by recent studies in nonhuman primates (Hilgetag and Barbas, 2006
), connects increases in folding to changes in cortical volume, but does not address the extent of genetic control over gyrification (heritability), or predict whether the same genetic variation that influences cerebral volume also influences degree of cortical folding.
Evidence from congenital malformations in humans does not provide support for either a positive or a negative genetic correlation between these phenotypes (Stevenson, 2006a
). Known genetic mutations can cause profound congenital malformations of cerebral gyrification such as lissencephaly and poly-microgyria (Stevenson, 2006a
), but these conditions are often observed along with apparently normal brain volume. Microcephaly and megalencephaly are caused by processes that interfere with normal brain growth (Stevenson, 2006b
) and alter brain volume, but the pattern of gyrification is often well-preserved. These clinical syndromes show that severe pathological mutations can affect one without affecting the other. However, this lack of correlation will not necessarily hold for normal (non-pathological) genetic variation. Overall, the clinical data do not argue for either a positive or negative intra-species genetic correlation between brain volume and gyrification.
We believe that our results have important implications concerning the evolutionary forces that have acted on primate brain structure. It is reasonable to presume that the differences in brain size between hominoids (e.g. living apes and humans) as compared to less encephalized primates (e.g. lemurs and lorises), resulted from natural selection favoring increased cerebral volume, neuron number and overall functional complexity (Schoenemann, 2006b
; Vallender et al., 2008
). The evolution of larger brain size has been attributed to its positive effects on cognitive and behavioral capability, especially in complex social groups (Dunbar, 2003
; Dunbar and Shultz, 2007
). Note that the increase in brain size was not a global increase affecting all structures equally, but involved specific additions at specific evolutionary stages (Preuss, 2007b
Our studies indicate that genetic variation segregating among humans and presumably independent but functionally similar genetic variation segregating among baboons appears to have opposite effects on brain volume and gyrification. The results suggest that in either humans or baboons, selection favoring increased brain volume would produce a correlated reduction in GI. The separate observation of this pattern of genetic architecture in two primate species strongly implies that this genetic architecture may be characteristic of many primate species, and that it was also true for human ancestors. Consequently, assuming that natural selection acted predominantly on brain size (Schoenemann, 2006b
), evolutionary genetic changes that resulted in increased brain size during primate evolution would have simultaneously produced less folded, more lissencephalic brains. However comparative primate neuroanatomy and the history of primate brain evolution shows the opposite pattern, with a strong positive correlation between volume and gyrification across species and clades. Therefore we infer that there was a second independent selective force that did not negate the progressive evolutionary increases in brain size, but did generate additional developmental changes resulting in brains with greater degrees of cortical folding. One possibility for this second selective force is based in parturition and reproductive constraints (Rosenberg and Trevathan, 2002
). The pelvic inlet and outlet of many anthropoid primates and especially humans are restricted in size by biomechanical constraints related to locomotion. Consequently, the size of the birth canal cannot increase freely to accommodate ever larger neonatal brains and skulls. Indeed, humans are not the only primate for which the size of a newborn's skull can create problems during birth (Leutenegger, 1974
As adult brain volume increased during various stages of primate evolution, there may have been an advantage within specific primate lineages to maintain neonatal brain volume below certain limits. This natural selection for retaining small neonatal brain size may have resulted in the second set of genetic changes that increased cortical folding in humans, great apes and to a lesser extent in Old World monkeys. This progressive increase in cortical folding could have been achieved through tension-based mechanisms (Van Essen, 1997
). Regardless of the specific selective forces or developmental mechanisms at work, our results constitute initial evidence that one set of selective pressures and genetic changes can explain progressively increasing brain size, and that an independent second set of pressures and genetic changes may be needed to explain greater cortical folding. One prediction of the tension-based model of cortical folding (Van Essen, 1997
) is that at least some of the genes that influence the degree of cortical folding act through effects on axon tension, either by altering the strength of cortical-cortical connections or affecting the ability of axons to resist elongation as cortical segments grow apart during embryogenesis. Our data predict that detailed analysis of the genetic bases of brain volume and gyrification will identify two sets of evolutionary genetic changes distinguishing humans from smaller brained primates, one set that produced increased cerebral volume and another that leads to greater gyrification.