This blog is currently on hiatus owing to work commitments. Whilst I still keep an eye on the goings on at RiAus, and contribute to the work of the good folks at eLife, little will be added to this blog for the foreseeable future. Simon Says remains open for business, albeit at a reduced capacity. Thanks for stopping by, and I hope the archive of content found here will prove to be of interest.

Saturday, 17 November 2012

But How Do They Know? | The Evolution Enigma

Badongo, Durrell Wildlife Conservation Trust,
September 2011
Source: me
Earlier this year the genetic sequence of the gorilla was revealed for the first time. It led to a flurry of news articles, flagging new findings that parts of the gorilla genetic sequence are more similar to humans or chimpanzees than those species are to each other — which defies the traditionally accepted order of species evolution — and other tantalizing titbits of information on our handsome cousins. But punctuating these articles was a statement that slipped by without explanation, to be accepted as written and memorized by fact fans everywhere. Thus: the last common ancestor of gorillas and humans lived 10 million years ago.

But this is a big statement.

We frequently read and hear phrases such as “scientists tell us”, followed often by superfluous details of a larger tale. But rarely is an attempt made to explain how such knowledge is acquired. You know the type: “humans and fruit flies are 60% genetically identical” or “sea levels will rise X metres by the year Y.” But how do we know these statements are true? Do we know they are true? If no explanation is given, should we be surprised if many refuse to accept scientific consensus?

This, then, is for everyone who has heard something and sought to know more, who has shouted in frustration: “BUT HOW DO THEY KNOW THAT?”

So: 10 million years?

You'll have heard that DNA is a code. This can be difficult to understand given that it's a physical chemical inside the body. How can a chemical provide information? The key is that while DNA is a chemical, there's actually four different forms of it, and these string together to form enormous chains. The four are nicknamed with the letters A, C, T and G, so the DNA sequence in the body can be written down as pages and pages of these four letters in all sorts of combinations. Mechanisms and machinery in the body then use these strings of chemicals in a specialized way to make the body function.

There's another level of complexity on top of this - the famed 'double helix'. DNA is in fact not one enormous chain but two: these line up alongside each other in parallel, with the As always lining up opposite the Ts, the Cs always with the Gs. The parallel strands then twist around each other to form the helix, which saves space, much like a twisted piece of string is shorter than an unwound piece of the same starting length. These strict pairings are crucial to the way DNA works: when the strands are pulled apart during growth, free ‘letters’ align automatically with their counterparts in the exposed strands (which are now acting as templates) and are then stuck together, and so it is that the entire sequence is replicated quickly, easily and efficiently. This versatility is exploited in 'sequencing' reactions in laboratories, the aim of which being to deduce a particular section of the genetic code. Here, the individual loose DNA letters are given unique coloured dyes that, as they align with the template strand, are observed by a machine in the order in which they are added. So red-red-green-blue-yellow-blue, for example, would correspond to T-T-A-C-G-C.

In this new study, published in the journal Nature, an international team of scientists took the genome of a western lowland gorilla and compared 2 billion of the individual letters of DNA with sequences from humans, chimpanzees, orang-utans and macaques. When you try to line those sequences up alongside each other, you can see how different the genomes — and therefore the species — are.

One of the first molecular attempts to date the great apes was by Vincent Sarich and Allan Wilson in 1967, comparing certain blood proteins. Their estimate of a 5 million year old common ancestor of humans and gorillas flew in the face of what was then known from the fossil record. Modern DNA sequencing now allows us to retest this number by looking not at a few proteins but the entire genome.

The problem is that this gives only a score of how different the species are, not how long ago they separated in evolution. We need to know how quickly the genome changes, or mutates, to turn this score into an age. This rate is not so easy to work out, as it is controlled by an enormous range of variables controlled by the very processes of evolution.

Mutations are the agent of change throughout evolution. They are the culprit behind the variation within species and the changes that have, over time, led to new species coming into existence. Yet the historical rate of mutation cannot be observed directly: we don’t have tissue from our ancestors to compare, and we clearly haven’t been around for millions of years, watching ourselves evolve, so we have to patch together clues in our role as genetic detectives (deerstalkers optional).

Our first clue is the rate of mutation in modern human genetic diseases — that is, how quickly genetic diseases have changed over generations and in different populations — which has been observed to be 1 mutation per billion DNA letters every two years. That means that in those 2 billion letters of gorilla genome, you would expect 1 spelling error cropping up every year. Now, the rules of nature are such that this rate changes depending on where it happens in the body, and the rules of evolution mean that changes need to occur in eggs or sperm and to not cause any harmful consequences in order for it to be passed on, but, needs must, we are dealing with averages, so 1 per billion letters every two years it is.

Our second clue comes from the fossil record. To achieve the number of mutations seen in macaques in the time the fossil record suggests, you need a rate twice as fast as that seen in modern human genetic diseases. So which rate is closer to the truth?

To answer this, I'm afraid I'm now going to have to draw a graph. This is, after all, a jigsaw of evidence, and it is on the graph that the clues shall fit.

Consider the relationship between mutation rate and the time it would take, at that rate, to allow for the number of genetic changes that exist between humans and chimps. It would look something like this:

Now let’s add in the other Great Apes: here is that same graph with the times required to allow for the number of changes between humans, chimps and gorillas (HCG), and humans and orang-utans (HO).
This shows that, for example, if a mutation rate of 1 mutation per billion DNA letters per year is correct, the last common ancestor of humans and chimpanzees lived ~4-5 million years ago (mya); the HCG split occurred ~8 mya and the HO split occurred ~15 mya. This varies dramatically with other mutation rates.

Now, a number of ape ancestor fossils have been found, each with estimates of their age, as based on archaeological evidence, and of their position in the puzzle of ape evolution, as based on clues from the bones themselves. This might include the orientation of the hips, or the position of the foramen magnum, the hole through which the spinal cord passes into the skull on its way to the brain. Both of these details might indicate the form of walking that the species adopted — the more human it is, the greater its ability to consistently walk on two legs only. How does this evidence tally with our graph?

Take this chap.

His name is Geoff. It’s not really. It’s Orrorin tugenensis, but Geoff is easier. Geoff was a bipedal hominin* and, of the species we’ve found evidence of to date, is the second-oldest in the human lineage, living around 5.6-6.2 mya. Geoff was found in the Lukeino Formation in western Kenya, thought at that time to be a dry, evergreen environment. That Geoff lived here at around 6 mya fitted well, at the time of his discovery, with the East Side story proposed by Dr Yves Coppens. This, sadly, was not a musical on the origins of mankind, but the theory that a change in environment in East Africa from woodland to savannah was the trigger for hominins to come down from the trees and eventually become bipedal and... human.

If evolution were a musical, our protagonist’s first appearance as Geoff comes part way into the show, after the first of many costume changes. The first appearance would be as Brenda, the Sahelanthropus tchadensis:

Brenda** lived around 7 mya, was found hiding in the central African nation Chad at the turn of the millennium, and is thought to be the oldest human ancestor (that is, the oldest in the line up to the human-chimpanzee split). She is unusual in that from the back her skull looks like a chimpanzee, but from the front she is somewhat more human.

The beauty of the graph we drew above is that there is a third, hidden dimension to it. As we move in a straight line up and down between each curve we are not just travelling through time. As we move away from the human-chimpanzee split towards the present day, we are progressively looking at a more human-like species, thus it is not just an axis of time but of development too.

Because we know by fossil evidence that Brenda and Geoff are the two oldest human ancestors, and not very different to whatever the human-chimp ancestor was, we know they have to be pretty close to the HC curve. At the same time, because they are in the human lineage, and not that of the chimp, they need to be placed just below the HC curve, because anywhere above it would place them further back in human evolution than the palaeoanthropological evidence would suggest. Add on top of this the archaeological dates calculated for when Brenda and Geoff lived and you suddenly have a very clear picture of where to insert their Pushpins of Evolution on the Graph of Answers:
Of course, our musical has many costume changes, many separate story threads and, given that this is a theatre production, an element of variation in each actor’s performance from day to day; that is to say, there are other species to add and a degree of uncertainty in where to place them. But it all comes together to produce a surprisingly clear picture:
The archaeological evidence all seems to fit around the slower rate of 1 mutation per billion DNA letters ever 2 years (0.5x10-9yr-1), and from this we reach the conclusion that humans and gorillas started their separate evolutionary adventures around 10 million years ago. Mystery solved.

But there’s a problem.

In the jazz-swing-opera fusion that is the orang-utan evolutionary story, there’s a character, let’s call him Archibald, that defies our model.

Archibald, the Sivapithecus indicus, was a chimpanzee-like ape with the face of an orang-utan. A wise and scholarly fellow (possibly), Archie lived, so say the archaeologists, 10.5-12.5 million years ago in and around Pakistan and India. But this simply doesn’t sit right with our predicted mutation rate. Indeed, if we apply this same rate to the human–orang-utan split, a figure of 36 million years arises, which disagrees with the fossil record by a whopping 16 million years. Something is either wrong with our working, or a more intriguing possibility arises.

Throughout school we are told to show our working. Yet in an age of short, snappy news, science reporting has come to steadfastly refuse to show the working behind its stories. But why should we believe these fanciful, fantastic stories, if we are not allowed to know how they were discovered?

So: 10 million years.

The main findings of the gorilla genome study were reported widely. But in failing to explain its methods a bigger story, the one to which Archie belongs, has slipped through the net: in the Great Apes, evolution seems to be slowing down.

*Definitions have changed over time but, at present, hominins are defined as humans and the immediate ancestors of humans since their split from what became the chimpanzees, whereas hominids are all modern and extinct members of the Great Apes, therefore also including gorillas, chimpanzees, bonobos and orang-utans, as well as their extinct ancestors.

**Clearly there were male and female Orrorin and Sahelanthropus, I merely use Geoff and Brenda as names to make the text easier to read and, selfishly, to type. To my knowledge the gender of the specimens shown here are unknown.

Geoff: Orrorin tugenensis
Brenda: Sahelanthropus tchadensis
Edgar: Ardipithecus ramidus 
Wilhelmina: Chororapithecus abyssinicus 
Archibald: Sivapithecus indicus 

This article was initially submitted for the Wellcome Trust Science Writing Prize 2012. It didn't win because, in it's original form, it was a big pile of rubbish. This expanded form, hopefully, is better, but has not been peer reviewed. Please do inform me of any inaccuracies or misunderstandings. 

21/11/12 I corrected 'last shared a common ancestor 10 million years ago' to 'the last common ancestor of ... and ... lived 10 million years ago' at the wise suggestion of @Sarah_May1


  1. You spell memorised like an American but apart from that, an enjoyable read!

  2. Thanks!
    I was using Oxford English spelling ;-p