Molecular Studies
The term "molecular anthropology" was coined in 1962 by Emile Zuckerkandl, who, with Linus Pauling, invented the notion of using molecular evidence to uncover evolutionary histories (see unit 8). At the time, Zuckerkandl had already discerned a hint of what was to unfold in the science when he compared enzymic digests of proteins from humans, gorillas, chimpanzees, and orangutans. As mentioned earlier
FIGURE 15.2 Anatomy of the feet: The human foot is a platform, built for bipedalism, while the gorilla foot is more of a grasping organ. A key difference, therefore, is in the relationship of the great toe to the other toes of the foot. In humans, the great toe is parallel with the other toes; in apes, it is opposable.
FICURE 15.3 Palate and tooth anatomy: In apes, the jaw is
U-shaped; in modern humans and later extinct hominins, it is parabolic. The jaws of early hominins such as Australopithecus afarensis are somewhat intermediate in shape. Ape incisors are large and spatulate; a gap, the diastema, separates the second incisors from the large canine; the premolars and molars have high cusps. In humans, the incisors are small; no diastema appears; the canines are small; the premolars and molars have low cusps. In Australopithecus species, the incisors are larger than in modern humans, as are the canines; a diastema is sometimes present in early species; the premolars and molars are large with low cusps. The very earliest hominin species are more chimplike in their dentition. (Courtesy of Luba Gudz.)
FICURE 15.3 Palate and tooth anatomy: In apes, the jaw is
U-shaped; in modern humans and later extinct hominins, it is parabolic. The jaws of early hominins such as Australopithecus afarensis are somewhat intermediate in shape. Ape incisors are large and spatulate; a gap, the diastema, separates the second incisors from the large canine; the premolars and molars have high cusps. In humans, the incisors are small; no diastema appears; the canines are small; the premolars and molars have low cusps. In Australopithecus species, the incisors are larger than in modern humans, as are the canines; a diastema is sometimes present in early species; the premolars and molars are large with low cusps. The very earliest hominin species are more chimplike in their dentition. (Courtesy of Luba Gudz.)
Cercopithe' cinae (99)
Cercopithe cidae
Hominini (42) Homininae (99) Hominidae (87)
Cercopithe cidae
Cercopithe' cinae (99)
Hominoidea (99)
Catarrhini (86)
FICURE 15.4 Cladogram of catarrhine relations: This analysis of 264 morphological characters leads to a chimpanzee/human association as the most parsimonious tree; a tree with a hominoid trichotomy is less parsimonious. This study is one of very few morphological analyses that identifies chimpanzees and humans as one other's closest relatives. (Adapted from Shoshani et al.)
Hominoidea (99)
Catarrhini (86)
FICURE 15.4 Cladogram of catarrhine relations: This analysis of 264 morphological characters leads to a chimpanzee/human association as the most parsimonious tree; a tree with a hominoid trichotomy is less parsimonious. This study is one of very few morphological analyses that identifies chimpanzees and humans as one other's closest relatives. (Adapted from Shoshani et al.)
(see unit 3), first Morris Goodman and then Allan Wilson and Vincent Sarich actually went on to establish the new field of research. They used immunological reactions of certain blood proteins to measure genetic distances among the living hominoids. In the early 1960s, Goodman established the human/African ape affinity, while in the late 1960s Wilson and Sarich used the genetic distances to identify times of divergence between the ape and human lineages.
As with all such calculations, Wilson and Sarich calibrated their molecular clock using known (or assumed) divergence times derived from the fossil record. They applied the then-accepted divergence time of Old World monkeys (super-family Cercopithecoidea) and Hominoidea of 30 million years ago. According to their research, the genetic distance between humans and African apes was one-sixth of that between living African hominoids and Old World monkeys. This finding implied that African apes and humans diverged 5 million years ago (one-sixth of the 30 million years that anthropologists believed to be the case, based on fossil evidence, namely Ramapithecus; see figure 15.5).
In the nearly four decades since this first calculation of human/ape divergence based on molecular data, many different techniques have been applied to the problem, including electrophoresis of proteins, amino acid sequencing of proteins, restriction enzyme mapping of various types of DNA, sequencing of mitochondrial and nuclear DNA, and DNA-DNA hybridization. Although their results are by no means unanimous, the great majority of these techniques
FICURE 15.5 Ramapithecus reconstructed: In the original reconstruction of the two fragments of upper jaw (maxilla) of Lewis's Ramapithecus specimen, the shape appeared to be humanlike. This partly explains why the Miocene ape was thought to be an early hominin. The reconstruction was inaccurate, in part because of missing portions of the specimen.
FICURE 15.5 Ramapithecus reconstructed: In the original reconstruction of the two fragments of upper jaw (maxilla) of Lewis's Ramapithecus specimen, the shape appeared to be humanlike. This partly explains why the Miocene ape was thought to be an early hominin. The reconstruction was inaccurate, in part because of missing portions of the specimen.
have supported the human/African ape linkage and have yielded a divergence time of between 5 and 7 million years ago. This finding is in good accord with the known fossil record (see unit 19).
Much controversy surrounded this work, and not all disagreements pitted molecular biologists against morpholo-gists. For instance, considerable debate surrounded the issue of the rate at which genetic change in the hominoid lineages accumulated. Supporters of the molecular clock (such as Wilson) argued that the rate was constant and universal. Others (such as Goodman) believed that accumulation rates could change over time and in different lineages. Indeed, Goodman initially attributed some of the surprisingly small genetic distance between humans and African apes to a slowdown in the clock. A slowdown could, of course, affect calculations of divergence times: a small genetic distance might disguise a long evolutionary separation. By now, fluctuations in the clock's rate in general have been accepted, and a slowdown among hominoids in particular. Nevertheless, as long as such fluctuations are taken into account, it remains possible to use genetic data for calculations of divergence times via local clocks (see unit 8). For instance, using extensive DNA sequences of certain globin genes, Goodman (previously a critic of the clock) and his colleagues recently calculated the human/chimpanzee divergence as 5.9 million years.
During the first two decades of molecular anthropology, the vast majority of work agreed on two things: the reality of a human/African ape affinity and an inability to break the trichotomy. The latter factor implied that either the trichotomy was real or the techniques were not sensitive enough to detect what might be rather short branches in a tree with two divergence points. In the mid-1980s, evidence began to build in favor of a tree with two divergence points: the separation of the gorilla, followed later by a human/ chimpanzee split. During the subsequent decade, most molecular data sets of various types supported the same pattern. Cladistic analysis requires specific characters (not genetic distance); in this context, it means gene sequences. Of 10
such independent data sets collected to date, eight support a human/chimpanzee link, two a chimpanzee/gorilla link, and none a human/gorilla link. (Humans are known to share 98.3 percent identity in nuclear, noncoding DNA sequence and more than 99.5 percent identity in nuclear coding sequences, or genes.)
Molecular phylogenetics involves several potentially confounding complications, in particular the gene tree/species tree problem (see unit 8). This can yield a phylogenetic pattern of the sort now heavily supported, even though the evolutionary reality is a simple trichotomy. A thought experiment will illuminate the point.
Imagine that an ancestral species possessed a gene A. Now imagine that a variant of the gene, A', arose 10 million years ago, making the gene polymorphic. Individuals in the population of the common ancestor may now possess two copies of variant A (that is, homozygous for A), two copies of variant A' (homozygous for A'), or one copy of each variant (heterozygous). Suppose that 5 million years ago the ancestral species split into three daughter species, X, Y, and Z. In the population that leads to X, the variant A' is lost, leaving just A. In the population that leads to Z, variant A is lost, leaving just A'. A comparison of the sequences of this gene in species X and Z would indicate that they diverged 10 million years ago, despite the fact the speciation event occurred only 5 million years ago. This erroneous dating, based on conflation of so-called gene trees and species trees, would follow from the gene polymorphism.
What about species Y? If its population lost variant A, a comparison of all three species would imply that Y is more closely related to species Z than to species X; similarly, if Y lost variant A , it would appear to be more closely related to species X than to species Z. In fact, all three species are equally related. (See figure 15.6.)
As this model indicates, for ancestral species possessing many highly polymorphic genes, no simple, single picture will emerge in a comparison of its descendants' genes. This complexity, suggests Jeffrey Rogers, of the Southwest Foun-
FICuRE 15.6 Gene trees versus species trees: Gene polymorphism in an ancestral species followed by differential sorting of variants can lead to erroneous conclusions, regarding both the timing of divergence and the relationship among descendant species. (a) Genetic analysis would make species Y look more closely related to species X than to species Z. (b) Y looks more closely related to Z than to X. The reality is a trichotomy. (See text for details.)
10 million years ago
5 million years ago rrii
rrii
Species Y appears to be more closely related to species X
rrii
Species Y appears to be more closely related to species X
- Species Y appears to be more closely related to species Z
Figure 15.7 Morphological versus molecular views: The cladograms show the current views that most paleoanthropologists take on the two approaches. Most morphological analyses favor either a chimpanzee/gorilla clade or a trichotomy. Most molecular analyses favor a human/chimpanzee clade.
Figure 15.7 Morphological versus molecular views: The cladograms show the current views that most paleoanthropologists take on the two approaches. Most morphological analyses favor either a chimpanzee/gorilla clade or a trichotomy. Most molecular analyses favor a human/chimpanzee clade.
Majority molecular view
Majority morphological view
Majority molecular view or dation for Biomedical Research, San Antonio, Texas, explains the mixed data for the hominoids, stating that a trichotomy is the most likely pattern.
It is true that the gene tree/species tree problem can lead to an erroneously old divergence date. It is also true that the problem can yield a pattern of two divergences apparently separated in time whereas the reality is a trichotomy. How is hominoid history to be assessed, given the data to hand?
The processes involved are stochastic, in terms of the timing of the origin of polymorphisms and the subsequent sorting of variants. As a result, many data sets are required to test hypotheses. The fact that so many data sets point to a similar divergence time for the inferred human/chimpanzee split provides some confidence in that date, unless all genes just happened to have produced polymorphisms at the same time in the ancestral species prior to speciation—an unlikely event.
The same principle can be applied to the putative two-divergence pattern, as Maryellen Ruvolo, of Harvard University, has argued. Given the stochastic nature of the sorting of variants, there is a one-third probability of genetic data implying a human/chimpanzee alliance and a two-thirds probability of seeing chimpanzee/gorilla or human/gorilla alliances. Statistically speaking, Ruvolo calculates, the probability of eight human/chimpanzee alliances emerging from 10 data sets as a matter of chance is close to 1 in 3000. In other words, the observed pattern is very likely to reflect history rather than being a statistical quirk. (See figure 15.7.)
Average user rating: 5 stars out of 1 votes
Post a comment