23
would be hard to create a new universe.
24
But it’s not impossible. (At least for evolution; we still don’t know how
25
to create new universes.) And that’s exactly what Lenski set out to do.
26
His basic setup was— and is, as the experiment is still ongoing— a sim-
27
ple one. He started with twelve flasks containing growth medium: a liquid
28
with a specific mixture of chemicals, including a bit of glucose to provide
29
energy. He then introduced a population of identical E. coli bacteria into
30
each of them. Every day, each flask goes from a few million to a few hun-
31
dred million cells. One percent of the surviving bacteria are extracted and
32
moved to new flasks with the same growth medium as before. The remain-
33
ing bacteria are mostly disposed of, although every so often a sample is fro-
34
zen for future examination, creating an experimental “fossil record.”
S35
(Unlike human beings, live bacteria can easily be frozen and revived at a
N36
273
Big Picture - UK final proofs.indd 273
20/07/2016 10:02:50
T H E B IG PIC T U R E
01
later date using current technology.) The total population growth amounts
02
to about six and a half generations in a day; the limiting resource is nutri-
03
tion, not time (it takes less than an hour for a cell to divide). As of late 2015,
04
this added up to more than 60,000 generations of bacteria— enough for
05
some interesting evolutionary wrinkles to develop.
06
Confined to this extremely specific and stable environment, the evolved
07
bacteria are by now quite well adapted to their surroundings. They are now
08
over twice the size of the individuals in the original population, and they
09
reproduce more rapidly than before. They have become very good at me-
10
tabolizing glucose, while generally decaying in their ability to thrive in
11
more diverse nutrient environments.
12
Most impressively, there have been qualitative as well as quantitative
13
changes in the E. coli. Among the ingredients in the initial growth medium
14
was citrate, an acid made of carbon, hydrogen, and oxygen. The original
15
bacteria had no ability to use this compound. But around generation
16
31,000, Lenski and his collaborators noticed that the population in one
17
particular flask had grown larger than the others. Looking more closely,
18
they realized that some of the bacteria in that flask had developed the abil-
19
ity to metabolize citrate, rather than just glucose.
20
Citrate is not as good an energy source as glucose is. But if you’re a bac-
21
terium in a flask full of other bacteria that are competing for a fixed amount
22
of glucose, the ability to live off of this other energy source is very useful.
23
Without having any particular goal to work toward, without the benefit of
24
any external prompting or instruction, evolution had come up with a clever
25
new way of allowing organisms to flourish in their particular environment.
26
•
27
28
The origin of life was the mother of all phase transitions. Like other chem-
29
ical reactions or combinations thereof, life proceeds by converting free en-
30
ergy into disordered energy. The aspect that makes life special among
31
chemical reactions is that it carries with it a set of instructions. Like the
32
tape in one of John von Neumann’s Universal Constructors, the genetic
33
information contained in DNA regulates and guides the interconnected
34
dance of reactions that defines a living organism. Those instructions can
35S
change as they are passed down from generation to generation. That ability
36N
is what gets natural selection off the ground.
274
Big Picture - UK final proofs.indd 274
20/07/2016 10:02:50
E v O l u t I O n ’ S b O O t S t R A P S
We’ve speculated that DNA came from RNA, which in turn may have
01
self- catalyzed its own production under the right circumstances. It’s possi-
02
ble that the creation of the first RNA molecule involved random fluctua-
03
tions at critical points along the way. Boltzmann taught us that entropy
04
usual y increases, but there is always some probability that it will occasion-
05
ally move downward. The more moving parts a system has, the more rare
06
such fluctuations will be; at macroscopic scales, the number of atoms in-
07
volved is so large that it’s not worth worrying about. But at the level of in-
08
dividual molecules, rare fluctuations are frequent enough to be important.
09
The appearance of the first self- replicating RNA molecule might just have
10
been a matter of good luck.
11
We sometimes think of natural selection as “survival of the fittest.” But
12
even before evolution in Darwin’s sense officially kicked in, there was a
13
competition of sorts going on for the available free energy. Some of it would
14
have been readily accessible, but some— similarly to that locked up in the
15
citrate in Richard Lenski’s flasks of bacteria— would have required more
16
ingenuity to unlock. An intricate network of reactions, directed by proteins
17
created by a sequence of nucleotides in RNA, could have prospered where
18
simpler processes would have flickered out. Once heritable genetic informa-
19
tion starts playing a role, all of the ingredients are in place for natural selec-
20
tion to commence.
21
22
•
23
From a certain perspective, Darwin’s theory is sufficiently commonsensical
24
that it seems almost inevitable. Upon first reading Origin, Thomas Henry 25
Huxley, Darwin’s contemporary and vocal supporter, exclaimed, “How ex-
26
tremely stupid not to have thought of that!” But natural selection is a very
27
specific process, and by no means inevitable or obvious. It’s not simply “spe-
28
cies gradually change over time,” or “ well- adapted organisms are more likely
29
to reproduce.”
30
Organisms reproduce, and they hand down their genetic information to
&
nbsp; 31
the next generation. That information is largely stable— children resemble
32
their parents— but it’s not absolutely fixed. Small, random variations can be
33
introduced at every step. The variations do not strive to reach any future
34
goals, and neither can individual organisms influence them by their actions.
S35
(Your offspring don’t become more muscular just because you work out.) If
N36
275
Big Picture - UK final proofs.indd 275
20/07/2016 10:02:50
T H E B IG PIC T U R E
01
we have descent with inheritance, and there is slight, random variation in
02
the genetic information that can affect the likelihood of reproduction,
03
natural selection can occur. Variations that fortuitously improve an organ-
04
ism’s chances of handing down its genetic heritage will be more likely to
05
persist than those that are harmful or neutral.
06
These ingredients shouldn’t be taken for granted. This is why biologists
07
highlight the difference between “evolution” and “natural selection.” The
08
former is the change of the genome (complete set of genetic information)
09
over time; the latter refers to the specific case where changes in the genome
10
are driven by different amounts of reproductive success.
11
Darwin didn’t know about DNA or RNA, or even of genes, discrete
12
units of inherited information. It was the Augustinian monk Gregor Men-
13
del who established the basic rules of heredity, through a set of now- famous
14
experiments crossing different varieties of pea plants. In the 1930s and ’40s,
15
biologists developed the modern synthesis, combining natural selection with
16
Mendelian genetics. The paradigm continues to be elaborated upon as we
17
learn more and more about biology and inheritance, but the basic picture
18
remains enormously successful.
19
The reality of biology here on Earth is, unsurprisingly, more compli-
20
cated than the simplest statement of natural selection. Like any way of talk-
21
ing about the world, Darwin’s theory works only within its domain of
22
applicability.
23
There are forces at work in the history of life other than organisms
24
adapting to their environments. This is completely compatible with Dar-
25
win’s conception; natural selection happens, but it happens within the
26
messiness of the real world, and it’s not the only thing happening. Many
27
features of the genome of any individual species are going to be the results
28
of accidents rather than any particular adaptation. This is known as genetic
29
drift. Sometimes there will be mutations that neither increase nor decrease
30
the fitness of an organism; other times, the randomness inherent in sexual
31
reproduction or unpredictable features of the environment will cause some
32
traits to become common while others die off. Biologists debate the relative
33
importance of adaptation and genetic drift, but there is little doubt that
34
both are important.
35S
In Lenski’s long- term evolution experiment, the mutation that allowed
36N
some of the bacteria to metabolize citrate occurred around generation
276
Big Picture - UK final proofs.indd 276
20/07/2016 10:02:50
E v O l u t I O n ’ S b O O t S t R A P S
31,000. When the researchers unfroze some of the earlier generations to see
01
if they would evolve this ability again, they found that the answer was yes—
02
but only when they started with cells from generation 20,000 or later.
03
Around generation 20,000, one or more mutations must have occurred that
04
did not themselves allow the bacteria to metabolize citrate, but set the stage
05
for a later mutation that would do so. A single trait can be brought to life
06
by multiple, separate mutations, which may not individually have much
07
noticeable impact at all.
08
Selection pressures work on traits, while genetic information is passed
09
down through DNA, and the map from one to the other isn’t a simple one.
10
Something as basic as how tall a person is won’t typically be fixed by one
11
particular string of nucleotides, but instead will depend on an interplay
12
between different factors working simultaneously. As a result, selection
13
pressure acting on one trait may end up affecting another one, if they de-
14
pend on common sets of DNA sequences. Evolutionary history is replete
15
with “spandrels,” as was famously emphasized by biologists Stephen Jay
16
Gould and Richard Lewontin. These are traits that arise for one reason and
17
then end up being used for something quite different. By-products of the
18
evolutionary process, rather than aspects that are directly selected for.
19
Gould and Lewontin imagine that many features of the human brain fall
20
under this category.
21
To make matters worse, inheritance can be more than simply a matter
22
of passing down DNA from one generation to the next. There is horizontal
23
gene transfer, in which genes are passed from one organism to another in a
24
way other than reproduction. It is relatively common in bacteria, and oc-
25
casionally happens in multicellular species. There are epigenetic phenomena,
26
in which the chemical structure of inherited DNA is modified during de-
27
velopment by influences such as the nutritional intake of an organism and
28
the maternal environment in which an embryo develops. It is currently
29
unclear how much such changes can be inherited by subsequent genera-
30
tions, but to the extent that they are, natural selection will act upon them
31
as usual.
32
So the real world is a beautiful mess. Is this kind of undirected
33
mechanism— just what we would expect in a universe governed by un-
34
thinking underlying laws and with a strong arrow of time— sufficient to
S35
account for all the spectacular intricacy of our planet’s biosphere? “There is
N36
277
Big Picture - UK final proofs.indd 277
20/07/2016 10:02:50
&nb
sp; T H E B IG PIC T U R E
01
grandeur in this view of life,” Darwin writes in On the Origin of Species. But 02
is his simple mechanism really enough to make dolphins and butterflies and
03
rain forests from a meager collection of organic molecules fighting for free
04
energy? Can the wonders of efficiency and ingenuity we see in biological
05
organisms really come about from random variation plus time? (Hint: yes.)
06
07
08
09
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35S
36N
278
Big Picture - UK final proofs.indd 278
20/07/2016 10:02:50
01
02
34
03
04
Searching through the Landscape
05
06
07
08
09
10
11
12
I
13
n computer science, as in life, we are often faced with the simple prob-
14
lem of finding some particular item in a long list of possibilities. Con-
15
sider the traveling- salesman problem: given a list of cities and the
16
distances between them, what is the shortest route that visits each city ex-
17
actly once? That can be rephrased in the following way. Take a list of cities
18
and the distances between them. Now make another list, consisting of ev-
19
ery possible route that goes through each city at least once. (It will be an
20
enormously longer list, but it is still finite.) Which route is the shortest?
21
A search algorithm is a precisely stated procedure for finding what you
22
are looking for in a list of objects. Of course you could trudge through every
23
element of the list, asking, “Is this the one?” That can be hard, since quite
24
reasonable- sounding questions can involve very unreasonably sized lists to
25
sort through. For the traveling- salesman problem, the number of possible
26
routes grows roughly as the factorial of the number of cities involved. The
27
factorial of a number n is equal to 1 times 2 times 3 times 4 . . . times ( n – 1) 28
times n. For twenty- seven cities, that’s about 1028 routes to search through.
The Big Picture Page 47