MCPH1 has experienced strong Darwinian positive selection during primate evolution, but there are no data showing functional divergence. Here we demonstrated evidence of functional divergence of MCPH1 between humans and nonhuman primates. Most of the human-specific amino acid changes could alter the regulatory effects of MCPH1 on the transcription of the down-stream genes, and a similar effect was observed for one of the four great-ape-specific amino acid changes. Accordingly, our data support the hypothesis that selection on MCPH1 has resulted in functional divergence at the protein level, which potentially contributes to changes in the development and evolution of brain size.
Absolute brain size has increased in parallel across primate evolution. Along the two branches we focused on in this analysis, absolute brain volume increased from 70 to 152 ml to 230 to 565 ml during the transition between lesser apes and great apes [30, 31], and from 230 to 565 ml to 1,129 to 1,685 ml during the transition between great apes and humans [30, 31]. Previous studies indicated that there were accelerated amino acid substitutions during both the origin of Hominidae’s ancestor and of our own species, paralleling the two brain enlargements , suggesting that the amino acid substitutions of MCPH1 were probably adaptive and may have contributed to the brain expansion during primate evolution. In addition, the gradient change of MCPH1’s transcription regulation from macaque to gibbon, and to humans (Figure 2D) seems to imply a continuum of functional divergence rather than a number of discrete shifts, which calls for further functional tests in extensive primate lineages.
Interestingly, all the human- and great-ape-specific mutations are located in the non-BRCT domains (Figure 1). Since the three BRCT domains of MCPH1 are critical for protein-protein interaction, the amino acid changes during primate evolution seems not to have caused drastic functional alteration, but rather a modification of the existing function.
Inferring the exact functional alterations of the human-specific and great-ape-specific mutations is difficult. Previous studies have shown that the middle domain (residues 367 to 485) of MCPH1, where the four human-specific mutations are located, is the binding domain by Condensin II for homologous recombination repair [32, 33], an important mechanism for cell cycle checkpoints and genome integrity. Concurrently, all four human-specific sites located in this middle domain showed altered regulatory effects when mutated into ancestral amino acids, suggesting that the human-specific mutations may have changed the binding property with Condensin II. Additionally, all four human-specific mutations caused changes in physicochemical properties of amino acids (see Additional file 3: Table S1).
We also found that for the regulatory changes of the down-stream genes, almost half (three out of eight) of the tested genes (p73, CyclinE1 and p14
) had significant differences between humans and rhesus macaques, either in the enhancing assay with MCPH1-E2F1 or in the repressing assay with MCPH1 alone (Figure 3), indicating a functional divergence between humans and non-human primates. The protein p73 is involved in both cell cycle regulation and induction of apoptosis . E2F1 is an important regulator of p73, especially during brain development . CyclinE1, meanwhile, is involved in cell cycle and is a key target gene of E2F1 and has been shown to take part in the determination of the number of neurons during mouse corticogenesis by regulating the G1 mode of cell division . p14
is an alternate reading frame product of CDKN2A involved in cell cycle regulation that is also involved in self-renewal of neural stem cells and neural development [38–41]. Additionally, human population studies have reported that the MCPH1 sequence variants were associated with brain volume in a sex-specific manner [24, 25]. Recently, it was also reported that MCPH1 might have contributed to the evolution of sexual dimorphism in brain mass across anthropoid primates . In fact, two of the down-stream genes regulated by MCPH1, p73 and cyclinE1, were reported to be associated with sex dimorphism during germ line development [43, 44], suggesting that the regulation of MCPH1 on brain development may differ between males and females. Taken together, the strengthened transactivation effect of human MCPH1 on these down-stream genes may contribute to the greatly enlarged neuro-progenitor pool in the human brain during neurogenesis, which is in line with recent studies that suggest MCPH1’s functional role in neuro-progenitor cells through the Chk1-Cdc25-Cdk1 pathway [45, 46].
Conversely, when MCPH1 acts alone as a transcription repressor, there were also differences between humans and rhesus macaques on the repressing effect of the down-stream genes (p73, CyclinE1 and p14
), implying that a homeostasis of gene expression regulation by MCPH1 is required during neurogenesis.
Although we observed functional divergence of MCPH1 due to its protein sequence changes during primate evolution and human origin, it should be stressed that we did not establish a direct link between the adaptive changes of MCPH1 and the ever-increasing brain size in primates. As shown in the MCPH1 knock-out mice analysis, the truncated MCPH1 not only caused a reduction in brain size, but also resulted in a reduction of testis size , suggesting that MCPH1 may also play a role during testis development. Accordingly, we cannot rule out the possibility that the adaptive evolution of MCPH1 in primates may be caused by selection on other phenotypes, though the current evidences mostly favor enlargement of the brain.
Initially proposed by King and Wilson , the importance of cis-regulatory changes in human evolution has recently been tested and confirmed . However, our functional data of MCPH1 suggests that protein sequence changes may also have significant phenotypic effects. Hence, the evolution of an important trait like brain function may require genetic alterations at multiple regulatory levels.