Johnson, Norman A. Darwinian detectives

Подождите немного. Документ загружается.

each other’s closest relatives, with C as the outgroup, despite the fact that

B and C are actually the closest relatives in the true species phylogeny. The

gene tree—the phylogeny established from that particular gene—is not

wrong. It accurately reflects the history of that gene. The problem is that the

history of the particular gene may not reflect the history of the genome as a

whole, and thus the gene tree can lead us astray from determining the true

phylogenetic relationships of the species (the species tree).

The problem that these polymorphisms existing in the ancestral species

(and hence, called ancestral polymorphisms) can cause usually presents itself

when the distance between nodes  and  is relatively small.

When the

difference between the two nodes is large, there is enough time for one or

the other genetic variant to become fixed in the ancestor of species B and C,

and thus both B and C (the two closest related species) will have the same

variant. This time is usually on the order of several times the effective popu-

lation size in generations. If the effective population size is ,, then after

a , generations or so, the problem is mitigated; , generations in

an ape, however, is on the order of a million years. (If the effective popula-

tion sizes of humans and the nonhuman apes were greater, then this problem

would be worse because more variants would have been maintained during

the time between nodes  and .)

So how do population geneticists cope with this pesky problem raised by

the assortment of ancestral polymorphisms? The usual approach is to collect

more data—that is, more DNA sequences from as many genes as possible—

and then determine whether a consensus emerges among the genes. With the

increasing speed and decreasing cost of sequencing, this brute force method

has become progressively easier. In well-studied groups such as the great

apes, such power became available by the early and middle s. As the

technology progresses, sequencing multiple genes will become par for the

course even in groups of species from organisms that are not as well studied.

Returning to the human-chimp-gorilla story, the preponderance of genes

examined does support the close relationship between humans and chimps.

During the middle s, Maryellen Ruvolo examined the data from a

number of genes.

She first lumped genes that were very closely genetically

linked into a single data set, as linked genes may not be independent because

they tend to share the same evolutionary history and hence genealogy. As an

extreme example, all of the mitochondrial genes count as one independent

gene because the lack of recombination ties their fates together. She also

threw out those genes that failed to conclusively support any phylogenetic

hypothesis. After this culling of the data, Ruvolo was left with  independent

genes that each provided conclusive phylogenies. Eleven of these genes support

a tree with humans and chimpanzees as the closest relatives, two support a tree

that places gorillas and chimps together, and just one gene supports a tree that

has humans and gorillas as closest relatives. How strong is the support for

humans and chimpanzees being closest relatives? Very strong; if humans

and chimpanzees were not each other’s closest relatives, the likelihood of

DARWINIAN DETECTIVES

obtaining so many genes supporting such a relationship and so few supporting

alternative patterns just by chance is less than two parts in a thousand.

Subsequent studies strongly reinforce Ruvolo’s conclusion; humans and

chimpanzees are each other’s closest relatives.

Chimp Population Genetics

As with other aspects of their biology, the population genetics of common

chimpanzees is better known than that of the bonobo. Although genetic

variation in common chimpanzees is higher than it is in humans, the extent

to which chimps have more variation depends upon the specific genes

and the specific chimps sampled. Unlike humans, in whom little genetic

differentiation occurs among populations, chimps are highly genetically dif-

ferentiated. So much differentiation exists among chimpanzee populations

that biologists now recognize three subspecies of chimps: the western African

chimpanzees (Pan troglodytes verus), the central African chimpanzees (Pan

troglodytes troglodytes), and the eastern African chimpanzees (Pan troglodytes

schwein-furthii).

In the late s, Svante Pääbo’s group undertook the most intensive study

of chimp population genetics as of that date. They examined over ,

nucleotides of an X-linked noncoding region.

The average difference

between any two chimps is about four times that of humans, similar to the

pattern that had been previously observed with mitochondrial DNA. These

data illustrate that the long-term average effect of genetic drift on this

X-linked gene has been greater in humans than chimpanzees; the estimated

effective population size of chimpanzees is ,, whereas that of humans is

only about ,. The age of the most recent common ancestor of this gene

in chimpanzees is over two million years ago. By contrast, the most recent

common ancestor for these X-linked sequences in humans is almost ,

years old. (Note that this figure is far greater than the estimated age of the

mitochondrial Eve—the most-recent common ancestor varies considerably

among genes!) In yet another example of how different genes can have dif-

ferent natural histories, the central chimps were the most variable subspecies

of chimp at this X-linked locus, whereas earlier results of the mtDNA studies

had shown that the western chimpanzees were the most variable.

Pääbo’s group was also able to obtain some information on the population

genetics of the bonobo. This species is far less genetically variable than the

common chimpanzee. Based on these same X-linked sequences, the effective

population size of the bonobo was under ,, even lower than that

estimated for humans. Of course, the census population size of humans is

several hundred thousand times higher than that of the bonobo.

Mutation, genetic drift, and natural selection will cause isolated popula-

tions to diverge from each other. Gene flow, however, if present, will make

the populations be less genetically differentiated. The prospect of gene flow

Are We the Third Chimpanzee? 

can make it difficult to decide between two scenarios: two populations that

recently diverged and are no longer exchanging genes versus populations

that diverged a long time ago but still maintain a trickle of gene flow between

them. Jody Hey at Rutgers University and his colleagues recently developed

models that allow one to determine simultaneously how long ago two popu-

lations or two species diverged from each other and how much gene flow is

occurring between them. With his graduate student Rong-Ling Wong, Hey

applied these methods to chimpanzees and bonobos.

Wong and Hey’s analysis of a total of  genes supports a model of evolution

in which bonobos and chimpanzees are not currently exchanging genes and

have not exchanged genes since their split, which most likely occurred ,

years ago. The estimate of the separation date is within the same ballpark as the

best estimates for the time when humans and Neanderthals diverged.

In contrast to the lack of gene flow between chimpanzees and bonobos,

Wong and Hey’s analysis does support a low level of gene flow between the

subspecies of chimpanzees. That gene flow, however, is only in one direction.

On average, there is an estimated one migrant every four generations from

Pan troglodytes verus (western chimpanzee) into the gene pool of the central

chimpanzee (Pan troglodytes troglodytes). In contrast, no migration occurs

from the central chimpanzee to the western chimpanzee. Wong and Hey

estimate that these subspecies diverged about , years ago. They

also estimate that the effective population size of the central chimpanzee is

around , breeding individuals, but that the other subspecies and bonobos

have effective population sizes on the order of ,.

The Chimp Genome Project

Only three years after writing a “white paper” that argued that the National

Human Genome Research Institute should financially support the sequencing

of the chimpanzee genome, an international consortium achieved their first

major goal: publication of the “draft” genome sequence. This “draft sequence,”

along with several supporting scientific articles and commentaries, was

published on September , , in a special issue of Nature dedicated to the

chimpanzee genome.

Although what has been published is only a “draft” sequence subject to

further refinements, the work of the consortium represents a large advance

in what we know about the chimpanzee genome and the differences in the

genetic makeup between humans and chimpanzees. The vast majority ()

of the genome from the draft sequence should have an error rate on the

order of one in ,, which is sufficiently accurate to use in sequence

comparisons.

Based on the sequences from whole chimpanzee and human genomes, the

single nucleotide divergence between the species is .; after adjustments

for polymorphisms within each species, this figure is .. This is consistent

DARWINIAN DETECTIVES

with what had been found with analysis of single genes. But what is somewhat

of a surprise is the extent of divergence due to insertions and deletions—

genetic material that gained or lost by one or the other species since humans

and chimps diverged from a common ancestor. Both the human and the

chimpanzee genomes contain between  to  million nucleotides of DNA

sequence found in just one but not both species. This total divergence (almost

 million nucleotides) represents  of the genome. So how much do

chimps and humans differ? On the basis of changes at just the nucleotide

sites, they differ by just over . If instead one also examines insertions and

deletions, that divergence level climbs to .

The rate of nucleotide divergence varies considerably in different regions in

the genome. Some, but not all, of this variation stems from differences among

chromosomes. A two-fold difference exists in the divergence rate between

the X chromosome sequences (changed the least) and the Y chromosome

sequences (changed the most). The divergence rates for autosomes do differ

among themselves, but they all fall in between those for the X and Y chro-

mosome. What could explain such a pattern? A logical reason, one that the

consortium authors suggest, is that the pattern emerges due to more mutation

taking place in the male germ-line than in that of females. This phenomenon,

known as male-driven evolution, occurs because males make more sperm

than females make eggs. More cell divisions are thus needed in the develop-

ment of sperm than in egg development. Due to these extra cell divisions and

more rounds of copying DNA, more mutations occur in the male germ-line.

Recall that Y-chromosome genes are present only in males and are transmitted

only from father to son. Due to their male-limited transmission, Y-linked

DNA sequences would be expected to mutate faster than would the auto-

somes. This should lead to a higher rate of evolution than in autosomes,

which are present equally among females and males. By a similar logic, the

X-chromosome genes should experience lower rates of mutation and diver-

gence because the average X chromosome is present twice as often in females

(which have lower mutation) than in males. Based on the differences between

the rates of evolution of the chromosomes, the authors estimate that three to

six times as many mutations occur in the male germline for every one that

occurs in the female germline.

The chimpanzee genome study also provides insight into the patterns of

evolution of mobile genetic elements, reinforcing the notion that the genomes

of mammals are not static. For instance, since the split between humans and

chimps, , new copies of one of these elements (Alu) were inserted into the

human genome. By comparison, only , new Alu elements were inserted

into the chimp genome. Given a six-million-year divergence time, these

figures imply that one new Alu has entered our genome on average every 

years for our genome, and once every , years for the chimp’s. We will

discuss more about Alu and other mobile genetic elements in chapter .

From an evolutionary genetic perspective, perhaps the most interesting

aspect of the chimp genome is the insight that it can provide us about how

Are We the Third Chimpanzee? 

various forms of selection operate in apes at the level of the whole genome.

Knowing the differences between the human genome and chimp genome

enables evolutionary biologists to calculate the overall levels of functional

constraint and hence the extent of negative selection that has acted on our

species. As we have seen several times, the ratio of divergence at the sites that

change amino acids (replacement) to that of sites that do not change amino

acids (silent) provides a rough estimate of the functional constraint of a gene.

What the genome sequences provide is a global estimate of the intensity of

negative selection operating on the genome. This ratio for the human-chimp

divergence is .; the rate of divergence at replacement sites is just under one

quarter of that for silent sites. If we assume that no positive selection is occur-

ring and that no negative selection operates on the silent sites, this result

suggests that more than three-quarters of amino-acid changing sites are

under fairly strong negative selection. In actuality, this figure probably under-

estimates the degree of constraint operating on the sites that change amino

acid, because a substantial fraction of the silent sites themselves are actually

under weak negative selection. Negative selection is a pervasive force on the

protein-coding regions of the human genome.

Although an overwhelming majority of the replacement sites between

humans and chimpanzees are under negative selection, they appear to be less

constrained than replacement sites in rodents. In other words, negative

selection seems even more potent in rodents than in primates. Consider the

just over , genes for which copies have been identified in all four species

humans, chimpanzees, mice, and rats. The ratio of replacement to silent site

divergence between chimps and humans in these genes is .. (This ratio is

slightly less than the . ratio found across the genome; the difference

probably reflects the fact that the genes with recognizable counterparts in all of

these species are more apt to be more constrained.) For the divergence between

mice and rats, however, the same ratio is only .; thus negative selection

appears more potent in rodents than in primates. One plausible explanation for

the increased constraint from negative selection operating in rodents is that

rodents generally have higher effective population sizes than do apes. When

effective population sizes are high, selection—both positive and negative—is

more efficient relative to genetic drift. Mutations that are slightly deleterious

have a greater probability of becoming fixed in species with low effective

population sizes than they are in species with higher effective population sizes.

Other data support rodents having a larger effective population than humans

do. One caveat: comparing humans and chimps with mice and rats is not

the most appropriate comparison because the absolute divergence between

mice and rats is considerably larger than that between humans and chimps. If

the intensity of negative selection fluctuates over time, the ratio of replacement

to silent sites may vary with the divergence time between species. This problem

is likely to soon be moot; it is likely that in the near future, genomes will be

sequenced from other primates that are about as diverged from humans as

mice are to rats, and thus better comparisons will be available.

DARWINIAN DETECTIVES

In the next chapter, we’ll return to the inferences about the action of

selection that the data on chimp genome sequences allow us to draw; there

we will focus on positive selection.

We’re About  Chimp—What Does That Mean?

“Sameness/otherness is a philosophical paradox that is resolved by argument,

not by data. Genetic data can tell us precisely what we already knew, that

humans are both very similar to and different from the great apes.”

speaks Jonathan Marks in his provocative book What It Means to Be 

Chimpanzee. Although I disagree with several statements in Marks’ book—

for instance, Marks thinks Neanderthals should be classified as subspecies of

Homo sapiens—I agree with his preceding statement.

Long before the sequencing of the human and chimpanzee genomes, long

before the sequencing of any genes, long even before the discovery of

Mendelian genetics, biologists knew that humans and chimpanzees were

closely related to one another. This relationship is based upon similarities at

a huge suite of phenotypic (observable morphological, physiological, embry-

ological, and other) characters. We didn’t need DNA to tell us that we are

closely related to chimpanzees.

DNA sequence information has thus provided excellent confirmation of

what we already knew. That is not a trivial contribution; confirmation of

findings from disparate sources and methodologies lies at the heart of science.

We know a great deal about relationships based on phenotypic characters.

Likewise, we know a great deal about these relationships based on genetic data.

And although these data sets occasionally conflict, much more often than not

they reinforce each other.

In general, genetic information provides some advantages over the data

based on morphological or other phenotypic characters in determining the

phylogenies of close relatives. Genetic data easily rejected the hypothesis that

humans are more closely related to orangutans than chimpanzees (a position

held by a minority of anthropologists during the middle of the twentieth

century). After a considerable effort, genetic researchers were able to show

that chimpanzees and humans are indeed more closely related to each other

than either species is to gorillas.

Unlike the information gathered from morphological and other phenotypic

traits, the genetic information also seems to provide us with a definite

number; DNA sequence data inform us that humans are . diverged from

chimps. Case closed? In actuality, as we discussed above, the situation is more

complex. Genetic information provides us with several different numbers,

because that . figure only applies to nucleotide sequence substitutions of

genes that code for proteins. If you include insertions and deletions, we are

 diverged. Moreover, there is substantial variation among genes; many

genes haven’t changed at all between humans and chimpanzees, and others

Are We the Third Chimpanzee? 

have changed many fold times the average rate. So, what exactly does that

number (.) mean?

In my opinion, this number based on DNA sequence data does provide us

with some vital information about our relationship with chimpanzees and the

other apes. The number, however, informs us more about our differences

from, rather than our identity with, chimpanzees. Although much has been

made of our genetic similarity, the . genetic difference in single nucleotide

changes alone represents considerable genetic distance. Our genetic difference

with both the common chimpanzee and the bonobo is at least ten-fold higher

than the average genetic difference found between any two people within our

species. Moreover, recall that the vast majority of the differences within

Homo sapiens are among individuals of the same ethnicity and not between

what have been designed as “racial groups.” Given DNA sequences, a geneti-

cist could easily distinguish humans from chimpanzees (common or bonobo)

almost regardless of what genes were studied—with the exception of where

persistent and strong balancing selection has maintained genetic variants

since before the species diverged (chapter ).

Caught Between Scylla and Charybdis?

In Greek mythology, Scylla and Charybdis are two sea monsters that live on

opposite sides of a narrow strait of water. Sailors who tried to avoid Scylla

would often pass disastrously close to Charybdis and vice versa. If you will

permit me to mix metaphors, the chasm Huxley had mentioned regarding

humans and nonhuman animals presents its own Scylla and Charybdis. One

potential danger is looking at chimpanzees and the other great apes and exag-

gerating their differences with humans. Avoidance of that danger, however,

can lead us toward a different risk: exaggerating their similarities with us.

The dangers sea voyagers faced from the two mythological sea monsters

were not equal; Charybdis, but not Scylla, would create great whirlpools

thrice daily by swallowing vast quantities of water and then spouting it up

again. In fact, according to Homer’s Odyssey, Odysseus was advised to steer

closer to Scylla because the whirlpools of Charybdis had the ability to destroy

his whole ship. Odysseus, following that advice, avoids Charybdis, though at

the expense of losing six of his crew to Scylla.

So, which is the greater danger—overemphasizing the similarities between

humans and chimpanzees (and bonobos) or overemphasizing the differences?

I agree with Marks that questions about sameness and otherness are philo-

sophical, not scientific, in nature. I also think that the answer to this question

of the greater danger depends on what the answer will be used for. When

contemplating questions of ethics, it behooves us to err on the side of our

similarity with chimpanzee and other apes. A greater damage can arise if we

treat them as though they are more different from us than if we treat them as

more similar to us. Although I certainly do not believe great apes should be

DARWINIAN DETECTIVES

accorded the full moral standing of human beings, their phylogenetic and

phenotypic similarities do impose on us the need for special consideration in

how we treat them.

Yet similarity, not difference, is likely to be the Charybdis in anthropolog-

ical study. I believe that overemphasis on the similarities poses a greater risk

than overemphasis on the differences when contemplating our human

natures. Apes are not disabled people, nor are they subhuman. They are not

human. We are chimpanzees only in the most strict phylogenetic sense that

our closest relatives are chimpanzees. We are sufficiently different from

chimpanzees in so many other ways. The history of anthropology has been

rife with scientists projecting our natures on those apes, and vice versa. Take

the following question: Is the “warfare” seen in chimpanzee “societies” an

indicator of our real nature that has been hidden beneath the cloak of culture

and civilization? The first problem is that both warfare and societies need to

be set in quotation marks, because the chimp versions of these abstractions

need not be identical to what we humans consider warfare and society. We

view both “warfare” and “society” with the biases and baggage of our own

culture.

The other problem with looking to the chimps as mirrors to our nature is

that we have two equally foggy mirrors that give us different views. The

common chimpanzee and the bonobo, which are far closer to one other than

either is to us, are worlds apart in terms of their behaviors—as different as

Genghis Khan and Dr. Ruth. A pop psychologist could write a book called

Chimps Are from Mars and Bonobos Are from Venus. We are just as close in

genetic terms to the bonobo as we are the chimp. We share just as many

genes with each species. But just as neither Venus nor Mars is planet Earth,

neither bonobos nor chimps are human.

Are We the Third Chimpanzee? 

This page intentionally left blank



What Are the Genetic Differences

That Made Us Human?

Blasé, matter-of-fact, encumbered by the infant (who, face to

her chest, clutches her fur), she carefully positions the hard-

shelled fruit on the log and smashes it open—using a stone

tool procured for the purpose. Hammer and anvil. No light

bulb goes off above her head. There’s no chin to fist, no hint of

insight struggling to emerge, no moment of revelation, no

strains from Also Sprach Zarathustra. It’s just another routine,

humdrum thing that chimps do. Only humans, who know

where tools can lead, find it remarkable.

—Carl Sagan and Ann Druyan (Sagan and Druyan,

, p. )



The Needles in the Haystack

As we saw in the previous chapter, the extent to which we are different or

similar to chimpanzees and other apes—or for that matter other animals in

general—depends upon one’s point of view and the types of questions that

one asks. Although genetics can inform debates revolving around the same-

ness-otherness dichotomy, such questions are philosophical in nature, as

Jonathan Marks stated.

Regardless of how close to or how far apart we consider ourselves from

chimps, we certainly differ substantially from our relatives among the apes

with respect to brain size and capacity for language. The vocabulary of

four-year-old human children outstrips that of apes—even those apes that

researchers assiduously trained in language. No other animal has composed

sonnets or, on the flip side, a Mein Kampf. Although several of the great apes

use tools, no other animal has built anything resembling a go-cart, much less

a microprocessor.

As exemplified by the quote from Sagan and Druyan, observations of other

species of apes often leads to contemplation of our origins. What made us

different from them? Is it a case of “there but for the grace of different selec-

tive pressures . . .”? Can we pinpoint the genetic changes that made us

human? The answer to the last question depends upon what one means

by human. Such a question, laden with philosophical overtones, may never be

answered. Evolutionary geneticists are, however, becoming better able to