“It is possible to travel back in time” is a bold way to begin a scientific paper by any standard. This promising first sentence appeared in a the respectable journal Nature Structural and Molecular Biology, earlier this year. The words that follow reveal why: “at the molecular level by reconstructing proteins from extinct organisms.”
When I think of ‘time travel’, I think of 1985 DeLoreans and flaming skid marks. But the reconstruction of an extinct protein is time travel of a different kind: it provides a glimpse of a distant past, just like a reconstruction of an extinct creature’s skeleton does. In this metaphor, amino acids are the molecular biologist’s bones. Instead of finding out which bone goes where she reassembles protein skeletons by lining up the correct sequence of amino acids.
But how do you bring back an extinct protein from the dead? The earth contains no fossil proteins. They are long gone, as are the ancient strands of DNA that coded for them. Yet modern DNA still contain clues as to what they looked like. How so?
Suppose a linguist wants to reconstruct the Germanic word for apple as it was spoken by Germanic tribes thousands of years ago. This word is just as exitnct as the proteins of our distant ancestors: it has not been written down anywhere, nor has it been spoken by anyone in millennia. The only way our linguist could learn something about its original form is to compare the current incarnations of the word apple in modern Germanic languages.
Appel, apple, æble, Apfel, äpple, eple or eplið
Here’s a list of Germanic apples in a row: apple (English), appel (Dutch), Apfel (German), æble (Danish), äpple, (Swedish), eple (Norse) and eplið (Icelandic). It is obvious that all these apples have some common features, such as the a or e followed by a p. This ubiquity suggests that this letter combination is inherited from the ancestral Germanic word for apple. Other combinations, such as pf in the German Apfel, are unique and derived: pf only ‘evolved’ in the German lineage of apples.
It is via such comparisons that linguists can reverse-engineer extinct words, always with some degree of uncertainty (the proto-Germanic word for apple was eventually reconstructed as apla(z)).
Modern proteins also contain traces of their ancestral shape and form. By comparing the amino acid sequences of a sufficient number of modern proteins, the ancestral protein sequence can be predicted with some confidence.
The authors of the NSMB paper set out to resurrect the ancestor of a a protein family known as the thioredoxins. These small proteins are like a pair of scissors that cut with molecular precision: they can cleave the tight bonds between two sulfur atoms. Such sulfur bridges stabilize a protein’s structure and can be found in all sorts of proteins.
A fossil apple?
The ancestral thioredoxin makes a great target for a reconstruction because all organisms on earth carry the code for one or more thioredoxins in their genome. The advantages of such an ubiqutous distribution are twofold. First of all, the more modern descendants are known, the easier it is to rebuild their common ancestor. If all linguists could work with were Äpfel and epli, reconstructing the proto-Germanic apple would have been more difficult. Second, since the last common ancestor of all life on earth must already have carried a thioredoxin, the trail that these proteins have left behind is billions of years old. It almost leads to the earliest origins of life itself.
For every amino acid that makes up a thioredoxin, the team calculated which of the twenty amino acids was most likely to have occupied this specific spot in ancestral proteins. They did this for a variety of different ancestors, such as the ancestor of all bacteria, the ancestor of eukaryotes, and so on. From these most likely sequences the researchers then produced the most likely proteins, which they subjected to a battery of different tests.
A reconstructed Tyrannosaurus remains dead forever, but a resurrected thioredoxin is just as alive as its modern descendants. They still fold into their proper shape and can still cut sulfur bonds. Not every thioredoxin slices up sulfur-sulfur bonds in the same way, however.
Bacterial thioredoxins have a different cut than the thioredoxins of eukaryotes, the branch of life that includes animals, fungi, and various amoeba-like organisms. Eukaryotic thioredoxins have a deep binding groove that rotate and force the sulfur-sulfur links into a specific position, like a restraint that prevents a patient from moving during surgery. Once the link is in the correct position, the thioredoxin makes the cut. Bacterial thioredoxins also have such a groove, but it is more shallow and bacteria don’t rely on it to break the sulfur bond.
To test the different cutting styles of ancient thioredoxins, the team devised an experiment in which they stretch out a single protein from end to end, like an unwinded ball of yarn.
When the thioredoxin (gray) cuts the sulfur bond (yellow), the protein suddenly increases in length.
The team engineered one sulfur bridge into this thin protein string, which is placed in such a way that part of the thread is locked up in a side loop. When the sulfur link is broken, thanks to a thioredoxin for example, the entire protein strand suddenly increases in length, about 14 nanometers. It is this small jolt that the scientists measured to determine a thioredoxin’s cutting style.
By pulling on one end of the string, it becomes more difficult for thioredoxins with a deep binding groove to break the sulfur bond. The thioredoxin needs to align itself at 180 degrees with the sulfur link, but this becomes harder to do when the protein string is stretched out. Thioredoxins with a more shallow groove can also break sulfur links outside of the groove, and so are less hindered by the pulling force. The researchers suspected that more ancient thioredoxins would have a binding groove that is not as deep and complex as that of modern eukaryotes.
But contrary to this expectation, the experiments revealed that even the most ancient eukaryotic thioredoxins already cut sulfur links in the same way as modern thioredoxins with a deep binding groove. Similarly, ancient bacterial thioredoxins have the same cutting style as modern bacteria. This suggests that thioredoxin chemistry was already established early in evolution and has conserved ever since. Making and breaking sulfur bonds has become such an important foundation of life’s biochemistry, that modern life couldn’t afford to change its nature.
At first glance, a paleoenzymologist looking to reconstruct the distant past might be disappointed to find that today’s protein doesn’t differ from yesterday. But compare his find to that of the explorers who discovered the cave paintings of Chauvet. The realization that early humans already represented animals and humans in an abstract way tells us something about what it means to be human. Extinct enzymes whose inner workings have been preserved teach us what it means to be alive.
Detail of the Chauvet cave paintings
Thanks to @onetruecathal for tweeting the paper
Image credits
Apples on a tree by Imaffo.
Fossil carbon pulp from PaleoSearch.com
Thioredoxin cutting a sulfur bond in a protein string is modified from reference.
Chauvet cave paintings from Wikimedia.
References
Perez-Jimenez R, Inglés-Prieto A, Zhao ZM, Sanchez-Romero I, Alegre-Cebollada J, Kosuri P, Garcia-Manyes S, Kappock TJ, Tanokura M, Holmgren A, Sanchez-Ruiz JM, Gaucher EA, & Fernandez JM (2011). Single-molecule paleoenzymology probes the chemistry of resurrected enzymes. Nature structural & molecular biology, 18 (5), 592-6 PMID: 21460845
You might also like:
The language analogy works amazingly well. I think quite a few taxonomists already do reconstruct ancient DNA (at least in their head) as a thought experiment while trying to work out relatedness.
(Back to the Future is an awesome film :D )