View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Am. J. Hum. Genet. 64:31–39, 1999 MOLECULAR EVOLUTION ’99 The Genomic Record of Humankind’s Evolutionary Roots Morris Goodman Department of Anatomy and Cell Biology, Wayne State University School of Medicine, Detroit In response to the call for a human genome–evolution zees are emotionally complex and intelligent. They use project (McConkey and Goodman 1997), the view has tools and have material cultures (McGrew 1992), are been expressed that what makes us human resides in the ecological generalists, are highly social (De Waal 1995; 1.5% difference in genomic DNA that separates us from McGrew et al. 1997), and apparently can learn and use chimpanzees (Gibbons 1998). This view is far too nar- rudimentary forms of language (Savage-Rumbaugh and row. Features that we associate with being human did Lewin 1994; Fouts 1997). In agreement with the newer not just arise de novo in the past 6 million years since information on the social lives and intelligence of chim- the lineage to humans separated from that to chimpan- panzees and other apes (McGrew et al. 1997), the results zees. Rather, some of the most striking human features, of molecular studies of primate phylogeny (Goodman et such as greatly enlarged brains and prolonged child- al. 1998, and in press) challenge the traditional anthro- hoods in social nurturing societies, have deep roots in pocentric view that humans are very different from all our evolutionary history. Forty to 30 million years ago other animals. Rather, the molecular results reveal that (Ma) neocortical portions of the brain increased in the genetically we humans are only slightly remodeled apes. two emerging branches of anthropoid primates—the We share with our most distant living ape relatives (the platyrrhines (or New World monkeys) and the catar- gibbons and siamangs) 195% identity in genomic DNA, rhines. Within the catarrhine branch, additional marked and with our closest relatives (the chimpanzees and bo- enlargements occurred by 18–6 Ma in the lineage to the nobos, or pygmy chimpanzees) 198.3% identity in typ- ancestors of modern hominids, and the largest neocor- ical noncoding DNA and probably ∼99.5% identity in tical increases occurred in the past 3 million years in the the active coding sequences of functional nuclear genes. lineage to modern humans. Traditional primate classifications, still favored by A parallel evolutionary trend prolonged fetal life and many anthropologists, use the nebulous concept of the periods of postnatal life needed to reach full maturity. grades of evolutionary advancement to place both ex- We may surmise that the genetic program for our en- tinct and living small-brained primates—those of the Pa- larged neocortex has both ancient conserved features leocene and Eocene epochs of 65–35 Ma—along with and more-recently derived features—in particular, the the living lemurs, lorises, bush babies, pottos, and tar- anthropoid-specific features shared with New and Old siers, in the suborder Prosimii, the primitive grade. In World monkeys and apes, the hominid-specific features turn, these traditional primate classifications place the shared with apes, and some human-specific features. Al- larger-brained primates in the suborder Anthropoidea, though many mutations in the past 40 million years have the advanced grade. Moreover, within Anthropoidea, a shaped the neurogenetic program for an enlarged neo- gradistic grouping places the African great apes (chim- cortex, it is possible that just a small number of regu- latory mutations in the past 6 million years have brought panzees, bonobos, and gorillas) with the Asian great apes about the final enlargement of our neocortex compared (orangutans) in subfamily Ponginae of family Pongidae, with that of chimpanzees. whereas humans, viewed as the most advanced primates, Behaviorally, the separation between chimpanzees and are the sole living members of family Hominidae. This humans is much smaller than once thought. Chimpan- traditional anthropocentric view of our place in the or- der Primates ignores (1) the overwhelming evidence that the African great apes share their more recent common Received October 19, 1998; accepted for publication November 13, ancestry with humans rather than with orangutans and 1998; electronically published January 13, 1999. (2) the mounting evidence that the clade of chimpanzees Address for correspondence and reprints: Dr. Morris Goodman, De- partment of Anatomy and Cell Biology, Wayne Sate University, 540 and bonobos is the sister group of humans—that is, E. Canfield Ave., Detroit, MI 48201. E-mail: [email protected] shares a more recent common ancestry with humans This article represents the opinion of the author and has not been peer than with gorillas (Goodman et al. 1998). In contrast, reviewed. ᭧ 1999 by The American Society of Human Genetics. All rights reserved. the cladistic view of how to classify organisms calls for 0002-9297/99/6401-0007$02.00 classifying our species, Homo sapiens, in a radically new 31 32 Am. J. Hum. Genet. 64:31–39, 1999 but strictly objective way, one without arbitrary anthro- provide evidence on primate phylogeny represent 61 pri- pocentric biases. In this new way, the cladistic evidence mate species belonging to 41 of the 60 recognized genera from molecules and morphology, as well as the fossil and almost all primate clades with ages older than that and molecular evidence on branch times in primate phy- of genera (Goodman et al., in press). More than 85% logeny, favors a phylogenetic classification (Goodman et of the b-globin gene cluster consists of flanking and in- al. 1998, and in press) in which humans are very close tergenic noncoding sequences that separate the cluster’s to apes, especially to chimpanzees and bonobos. b-type globin genes (e, g, wh, d, and b) from one another, In the derivation of this classification, three principles and these b-type globin genes have twice as much se- proposed by Hennig (1966) were followed. The first quence in their two introns (all noncoding) as in their principle is that each taxon should represent a mono- three exons. Because noncoding sequences evolve at a phyletic group or clade—that is, it should represent all relatively rapid rate and because a majority of the b- species descended from a common ancestor. The second globin gene–cluster sequence data are noncoding, max- principle is that the hierarchical groupings of lower- imum-parsimony trees constructed for each series of ranked taxa into higher-ranked taxa should describe the aligned orthologous sequences in these data have pro- phylogenetic relationships of the clades. The third prin- vided a well-resolved picture of phylogenetic relation- ciple is that, ideally, taxa at the same hierarchical level ships among primate clades. Maximum-likelihood trees, or rank should represent clades that are equally constructed for these aligned sequences, and so-called old—that is, at an equivalent evolutionary age. When neighbor-joining trees, constructed from matrices of these principles are followed, not only do all the apes pairwise distances among the aligned sequences, have group with humans within the family Hominidae, but depicted the same phylogenetic relationships among pri- also chimpanzees and bonobos as one subgenus and hu- mate clades as are depicted in the maximum-parsimony mans as the other subgenus group together into the same trees. genus, Homo. The percentages of sequence change on the branches Below I briefly discuss this classification of primates, of the globin phylogenetic trees were used in conjunction in which humans share their genus with chimpanzees with fossil evidence (reviewed in Goodman et al. 1998) and bonobos. Phylogenetic relationships depicted in the to estimate lineage divergence dates, by means of the classification are supported by DNA hybridization data model of local molecular clocks (Bailey et al. 1992). On and DNA sequence data from the mitochondrial genome the basis of fossil evidence, the lineage divergence date and from a growing number of unlinked nuclear ge- or last common ancestor (LCA) of Old World monkeys nomic loci (reviewed in Goodman et al. 1998, and in (family Cercopithecidae) and humans and apes (family press). Some of the best data come from noncoding se- Hominidae) was placed at 25 Ma, the LCA of platyr- quences of the nuclear genomic region called the “b- rhines and catarrhines at 40 Ma, and the LCA of strep- globin gene cluster.” Phylogenetic analyses of these non- sirhines and haplorhines (i.e., of all living primates) at coding sequences have provided evidence not only on 63 Ma. The paleontologically based age of 25 Ma for primate phylogeny (Goodman et al. 1998) but also on the LCA of cercopithecids and hominids served as the functionally important cis-regulatory elements that con- starting reference date for estimating, from relevant trol the developmental expression of the cluster’s func- branch lengths of the globin phylogenetic trees, the di- tional b-type globin genes, including e, g, and b (Gu- vergence dates for lineages within the hominid clade and mucio et al. 1996). In particular, mutations in cis- separately within the cercopithecid clade. The age of 40 regulatory elements during the emergence and evolution Ma for the LCA of platyrrhines and catarrhines served of the anthropoid primates changed the pattern of g- as the starting date for estimating the divergence dates globin gene expression, from being strictly embryonic to for lineages within the platyrrhine clade. The age of 63 being primarily fetal. These results, too, will be reviewed, Ma for the LCA of strepsirhines and haplorhines served since they illustrate how molecular evolution has shaped as the starting date for estimating the divergence dates functionally important components in the human ge- for lineages within the strepsirhine clade and also for nome.