NEWS & VIEWS doi:10.1038/nature11380

NEUROSCIENCE and human brain evolution Several genes were duplicated during . It seems that one such duplication gave rise to a that may have helped to make human brains bigger and more adaptable than those of our ancestors.

DANIEL H. GESCHWIND from the surface of neurons that have an & GENEVIEVE KONOPKA important role in the transmission of nerve a SRGAP2 gene impulses. Previous studies had documented he decoding of the human and chim- the fact that humans have greater numbers and panzee genomes was heralded as an Mammalian ancestor densities of dendritic spines than other pri- opportunity to truly understand how mates and rodents7,8, but the molecular mecha- Tchanges in DNA resulted in the evolution of nisms driving this feature were unknown. By our cognitive features. However, more than studying cells from genetically modified mice A B a decade and much detective work later, the and human brain tissue, the authors show functional consequences of such changes that SRGAP2A promotes the maturation of 1,2 have proved elusive, with a few exceptions . A B C spines and slows down the migration of neu- Now, writing in Cell, Dennis et al.3 and Charrier rons within the developing , et al.4 describe the evolutionary history and whereas the human-specific SRGAP2C has function of the human gene SRGAP2 and A BD C opposite effects and so favours the formation provide evidence for molecular and cellular of further spines (Fig. 1b). These results sug- mechanisms that may link the gene’s evolution Non-human Humans gest that the emergence of SRGAP2C could with that of our brain. have contributed to expansion of the cortex It was already known that SRGAP2 is and increase in spine numbers in human b involved in brain development5 and that ancestors, and therefore to the evolution humans have at least three similar copies of of brain function and plasticity (the ability the gene, whereas non-human carry to alter neural connections in response to only one6. However, the study of duplicated, new experiences). or very similar, segments of DNA is hampered Interestingly, Dennis et al. studied a group by the fact that most human cells carry two Dendritic of young patients who had developmental dis- sets of (one inherited from each spine orders, and identified a few individuals — out parent), which makes it difficult to distinguish Non-human Humans of several thousand — who had DNA dupli- duplicated copies from the different parental mammals cations or deletions that affected SRGAP2A forms of the gene. To circumvent this problem, Neuron or SRGAP2C. Although such numbers are Dennis et al.3 searched for copies of SRGAP2 too small to be statistically significant, they in the genome of a hydatidiform mole — an Figure 1 | Evolution and function of a human indicate that mutations in SRGAP2A and abnormal, non-viable human embryo that gene. a, Dennis et al.3 and Charrier et al.4 detail SRGAP2C could be linked to disease. This results from the fusion of a sperm with an egg how the SRGAP2 gene, which is found as a hypothesis is consistent with previous sug- that has lost its genetic material; it therefore has single copy in the genomes of most mammals, gestions that the evolution of a bigger and chromosomes derived from a single parent. was duplicated three times during the evolution more complex brain in the human lineage was The authors showed that humans carry of human ancestors to give rise to four similar accompanied by an increased susceptibility to four non-identical copies (named A–D) of versions of the gene, named A–D. b, Charrier neurological disorders9. SRGAP2 at different locations on chromo- et al. demonstrate that the ‘ancestral’ version, Taken together, the findings reported by SRGAP2A, stimulates the maturation of dendritic some 1. By comparing the genes’ sequences spines (protuberances) on the surfaces of neurons, Dennis et al. and Charrier et al. significantly with that of the SRGAP2 gene from the orang- whereas SRGAP2C promotes an increased number add to the current working version of the utan and chimpanzee, the authors estimated of immature spines in humans. This development , and provide an example of that SRGAP2 was duplicated in the human might have contributed to the evolution of human how human-specific gene duplications can lineage about 3.4 million years ago, resulting cognitive abilities. modulate brain function. They also give func- in SRGAP2A (the ancestral version that we tional context to the long-known prolonged share with other primates) and SRGAP2B. the Homo lineage 2 million to 3 million years immaturity of the developing human brain Further duplications of SRGAP2B gave rise to ago, when human brain specializations (such (a process known as neoteny)10,11. By slowing SRGAP2C about 2.4 million years ago and to as those leading to the development of lan- spine maturation and promoting neuronal SRGAP2D about 1 million years ago (Fig. 1a). guage, social cognition and problem solving) migration, SRGAP2C might allow the environ- Dennis et al. found that SRGAP2B and were probably evolving. ment to have a more protracted influence on SRGAP2D are expressed at much lower levels, In a complementary study, Charrier and col- brain development than is possible during the and are more prone to sequence variations leagues4 investigated the role of the SRGAP2 shorter brain maturation times seen in other among humans, than are the A and C copies. genes in the development of spines — not mammals, providing additional malleability. They therefore suggest that SRGAP2C might those that permit us to walk upright, but Furthermore, Charrier et al. demon- have played a major part in the emergence of dendritic spines. These are small protrusions strate how evolutionary hypotheses derived

| NATURE | 1 © 2012 Macmillan Publishers Limited. All rights reserved RESEARCH NEWS & VIEWS from the comparative genomics of humans David Geffen School of Medicine at UCLA, 4. Charrier, C. et al. Cell 149, 923–935 (2012). and other primates can be tested in cell cul- University of California, Los Angeles, 5. Guerrier, S. et al. Cell 138, 990–1004 (2009). 6. Sudmant, P. H. et al. Science 330, 641–646 (2010). ture and animal models. This elegant work Los Angeles, California 90095-1761, USA. 7. Elston, G. N., Benavides-Piccione, R. & DeFelipe, J. provides a launching point for unravelling Genevieve Konopka is in the Department J. Neurosci. 21, RC163 (2001). a more detailed mechanistic understand- of Neuroscience, University of Texas 8. Benavides-Piccione, R., Ballesteros-Yan˜ez, I., DeFelipe, J. & Yuste, R. J. Neurocytol. 31, 337–346 ing of the role of human-specific duplicated Southwestern Medical Center, Dallas, (2002). genes in brain development, and of their Texas 75390-9111, USA. 9. Preuss, T. M., Cáceres, M., Oldham, M. C. & potential contribution to brain evolution and e-mails: [email protected]; Geschwind, D. H. Nature Rev. Genet. 5, 850–860 neurodevelopmental disorders. ■ [email protected] (2004). 10. Petanjek, Z. et al. Proc. Natl Acad. Sci. USA 108, 1. Konopka, G. et al. Nature 462, 213–217 (2009). 13281–13286 (2011). Daniel H. Geschwind is in the Department 2. Enard, W. et al. Cell 137, 961–971 (2009). 11. Somel, M. et al. Proc. Natl Acad. Sci. USA 106, of Neurology and the Semel Institute, 3. Dennis, M. Y. et al. Cell 149, 912–922 (2012). 5743–5748 (2009).

2 | NATURE | © 2012 Macmillan Publishers Limited. All rights reserved