Authors: Bryan Sykes
Mr Robinson had also given us the contact details of Syrian hamster breeders' and owners' clubs throughout the world, and Chris was about to write to them asking for hair samples when it occurred to us that this might not go down very well. We had already discovered that you needed quite a number of hairs to get out the DNA. Hamster hairs were very fine and tended to break off above the root. Although the animals didn't mind a few hairs being plucked, they were likely to feel a little uncomfortable, and so were their owners, if we asked for substantial tufts. That's when we realized we needed another source of DNA. We hit on what seemed at first a completely wild idea. We knew the DNA amplification reaction was exquisitely sensitive, which is why it had worked with the ancient DNA from the archaeological bones. Would there be enough hamster cells shed from the walls of the large intestine to survive in their droppings? Surely, not even the most devoted owner would begrudge parting with a few droppings for the cause of science. But would it work? There was only one way to find out â so next day Martin appeared with a fresh crop from his house guests. They were dried and shrivelled, rather like mouse droppings, and totally inoffensive. Even so, Chris used tweezers to pick them up and put them into a test tube. He boiled the droppings for a few minutes, spun down the sediment in a centrifuge and took a drop of the clear liquid into the DNA amplification reaction. It worked a treat.
For the rest of the summer small packets arrived from hamster enthusiasts all over the world. With their characteristic rattle, we knew immediately what they were. We eventually got DNA from thirty-five hamsters, and it wasn't long before Chris had sequenced the mitochondrial control region in all of them. They were all absolutely identical. So the story was true after all. All the pet hamsters in the world really do come from a single female. But more importantly for us, the control region had remained completely stable. From that very first hamster captured in the Syrian desert to its millions of great-great-greatâ¦great-grandchildren from every corner of the world, the control region DNA had been copied absolutely faithfully with not even a single mistake.
It was an amazing thought. Going flat out, hamsters can manage four or five generations a year. At that rate there would have been time for at least two hundred and fifty hamster generations since 1930. Even though all thirty-five of our hamsters would not have traced independent maternal lines all the way back to 1930, the fact that there were absolutely no DNA sequence differences between any of them had to mean that the anxiety I had that mutations in the control region might be happening too quickly was unfounded. Quite the reverse, in fact: this was a very reliable region of DNA after all, not given to fickle fits of mutation that would make it impossible to trace over the hundreds of generations we wanted to explore in our own human ancestors. Of course, there was a chance that even though the control region was stable in hamsters, it might not be in humans. I didn't think this was very likely, given the very fundamental nature of mitochondria, and I was prepared to take that risk.
I was not alone in my interest. Before very long it was plain that other scientists were thinking along the same lines and had realized the potential of this very special piece of DNA to illuminate not only the grand schemes of human evolution but much more recent mysteries as well.
In July 1991 the remains of nine bodies were dug out of a shallow grave in birch woods just outside Ekaterinburg, formerly Sverdlovsk, in the Russian Urals. This exhumation was the culmination of years of research and persistence by the Russian geologist Aleksandr Avdonin, who thought he had located the resting place of the last of the Romanovs, the imperial Russian royal family. The last Tsar, Nicholas II, his wife, the Tsarina Alexandra, and their five children had been executed, or murdered â depending on your point of view â along with their doctor and three servants in the basement of the house in which they had been kept prisoner by the Bolsheviks. This was the night of 16 July 1918, in the turmoil of revolutionary Russia, and rather than risk the royal captives being released by White Russians who were then closing on the town, the decision was made, at the highest level, to kill them.
According to a contemporary account, the family were already in bed when the final elements of the plan were put into action. A telegram was sent to Lenin in Moscow asking him to sanction the execution. Delays on the way meant that it did not reach the Kremlin until after eleven o'clock at night. The reply which gave the green light arrived at one o'clock the following morning. At half past one, a truck drew up at the house ready to take away the bodies. The family were roused and informed that, because of the military action in the town, they must spend the rest of the night in the basement where they would be safer. The Romanovs had heard the distant sound of artillery every night for the past fortnight, and saw nothing particularly sinister in this request, so they all made their way quietly down the stairs.
When they got down to the cellar, they were still not alarmed to find several guards had joined them. Even when they were asked to line up in a group, they were not suspicious. Then the leader of the execution squad approached the Tsar and took a piece of paper out of his pocket with one hand while his other rested on a revolver inside his jacket. Hastily he read the notice which condemned them to death. The Tsar was confused. He turned to his family, then to the guards, who drew their weapons. The girls started to scream. The firing began. First to be hit was the Tsar; he slumped to the floor. The cellar echoed with the screams of the victims mixed with the sound of gunfire and bullets as they ricocheted around the room. It was pandemonium, and the room soon filled with smoke, making it even harder for the squad to pick out their targets who were rushing to and fro in a blind panic. The order to cease firing was given and the victims were finished off with bayonets and rifle butts. It had taken less than three minutes to put an end to a dynasty that had ruled Russia for three hundred years.
The house is no longer there. It was demolished in 1977 on the orders of the First Secretary of the Sverdlovsk Region, the young Boris Yeltsin. But the fate of the Romanovs themselves remained something of a mystery. In the atmosphere of uncertainty and disinformation that prevailed in Bolshevik Russia, just because there were official reports of events, even events as historic and infamous as the âexecution' of the Romanovs, this did not mean that the events described had actually taken place. There were persistent rumours, actively encouraged by Soviet propaganda at the time, that the Romanovs had been sent to a safe place for their own protection. Another rumour circulated that the Tsarina and the children had been smuggled out to Germany. Yet another had the Tsar in the Kremlin, where Lenin was preparing to reinstate the monarchy as soon as the bourgeoisie had been eliminated and the Tsar was âreconnected to the people'.
The discovery of the skeletons at least promised to introduce some objective evidence into the debate. The proof of the execution story depended entirely on showing that the remains taken from the pit really were those of the Romanovs. The location at least tallied with some contemporary accounts that told of the bodies being loaded on to a truck and driven into the woods on the outskirts of the town. According to these accounts, the executioners panicked when their truck became stuck in the mud, and they threw the bodies into a hastily dug pit before dousing them with sulphuric acid in a vain attempt to remove all features which could be used for identification.
When all the recovered bones were assembled, it soon became clear that these were the remains of only nine bodies, two fewer than there should have been if all the victims of the massacre had been buried in the same grave. After the long and painstaking process of refitting more than eight hundred bones and rebuilding the shattered skulls that had been crushed by the rifle butts of the burial detachment, it was concluded from the skeletons that the nine bodies were those of the Tsar and Tsarina; three of their five children â Maria, Tatiana and Olga; their physician, Dr Eugeny Botkin; and three servants, Alexei Trupp the valet, Ivan Kharitonov the cook and Anna Demidova the Tsarina's maid. There was no sign of the bodies of the youngest daughter Anastasia, nor of the Crown Prince, the Tsarevich Alexei. Other than these reconstructions, what further tests could be done on the remains to confirm their identity?
We had already published a paper in 1989 showing that DNA could be extracted from much older bones than these, so it was only natural to try to get DNA from the Ekaterinburg remains in the hope of confirming that these were the Romanovs. The work was carried out by the Russian Academy of Sciences and the British Forensic Science Service. First they used conventional forensic genetic fingerprints to identify the sex of the skeletons and to confirm that they did indeed include a family group of two parents and three children. DNA from the remains presumed to be those of Dr Botkin and the servants showed that they were unrelated to the family group or to each other. So far, everything fitted in well with the conclusions of the bone experts.
These scientists also succeeded in recovering mitochondrial DNA from the bones, and came up with two different sets of sequences from the family group. The female adult, the presumed Tsarina, and all three children had an identical mitochondrial DNA sequence. The male adult in the family group, the presumed Tsar, had a different sequence. This was exactly what you would expect from a family. All three children had inherited their mother's mitochondrial DNA sequence while the father, who had got his from his own mother, had not passed it on to any of his children. However, on its own, extracting the mitochondrial DNA and sequencing it did not identify this family as the Romanovs â any family would show the same pattern of identity between mother and offspring, with the father showing a different sequence. The only way of proving
family this was was to locate living relatives of the Tsar and Tsarina who were connected to the dead Russians through a series of entirely maternal links. They didn't have to be especially close relatives; the real power of mitochondrial DNA is that it is not diluted by distance. So long as the connections are exclusively maternal and not disrupted by a fatherâchild link, then the mitochondrial DNA will be identical.
Fortunately, it was possible to trace living direct maternal relatives of both the Tsar and the Tsarina. The Tsar had an unbroken maternal connection through his grandmother Louise of Hesse-Cassel, the Queen of Denmark, to a Count Nicolai Trubetskoy, seventy years old and living in peaceful retirement on the CÃ´te d'Azur after a lifetime as a merchant banker. The Tsarina could trace a direct maternal link through her sister Princess Victoria of Hesse to His Royal Highness Prince Philip, the Duke of Edinburgh, the husband of Queen Elizabeth II. After several rounds of discreet negotiation both men agreed to provide a small blood sample from which their DNA could be extracted. What would they show?
The notation everyone uses to compare mitochondrial DNA sequences involves quoting differences from a set reference sequence, in fact the very first mitochondrial DNA to be entirely sequenced, by a team from Cambridge in 1981. In this notation, a DNA sequence which differs from the reference sequence at the fifteenth and one hundredth positions in the 500 base control region segment is abbreviated to 15, 100. The sequence from the Duke of Edinburgh was 111, 357 using this notation. At all the other 498 positions along the 500 base stretch, the Duke's sequence was exactly the same as the reference sequence.
It is always much harder to get a complete sequence in one go from ancient DNA than from a modern sample. The strands are fragmented by the ageing process, so even the relatively short 500 base segment of the control region has to be built up in overlapping stages of a hundred bases or so. This is a laborious process, but eventually the sequences of the presumed Tsarina and her three children were typed. They all had exactly the same sequence of 111, 357. They were all an exact match with the Duke of Edinburgh.
The same, however, was not true for the adult male, the presumed Tsar. He was not an exact match with Count Trubetskoy. Whereas the Trubetskoy sequence was 126, 169, 294, 296, the presumed Tsar's DNA had mutations at only 126, 294 and 296 â very similar but not identical. This was a definite setback. There was so much circumstantial evidence connecting the bodies with the Romanovs, and there was the exact match of the females with the Duke of Edinburgh. But there is no point in doing a genetic test if you don't take notice of the result. A close match is not an exact match. And if the maternal connection over six generations with Count Trubetskoy was unbroken, the match would be exact.
Was there a chance that the Count was not really a relative of the Tsar, even though the family tree had recorded him as such? If so, there would have to have been a break somewhere along the line going back from the Tsar to Louise of Hesse-Cassel and then down to Count Trubetskoy. It would mean, in fact, that one of the people on this line had a different mother from the woman recorded on the pedigree. This is always a possibility â there could have been an adoption or a mix-up at the birth â but these are only remote possibilities. If it were a paternal line that was being followed it would be different. A child can easily have a different biological father from the man married to his mother; but such mis-identification is much more unlikely down the maternal line. After all, both mother and baby have to be present at the birth. The only formal conclusion that could be reached was that this was not the Tsar; and so, that since the conventional genetic fingerprints had already identified him as the father of the three children found in the grave, this was not the grave of the Romanovs after all.
But even though the mitochondrial DNA sequences of Trubetskoy and the male skeleton were not exactly the same, they were very close; and so near a miss invited further thought. They both shared three mutations at positions 126, 294 and 296. Trubetskoy had another one at position 169. Was it possible that there had been an error in reading the sequence of the âTsar's' mitochondrial DNA? The team went back to the original trace from the sequencing machine and looked very closely at the readings at position 169 for the âTsar's' sample. The trace itself looks like four superimposed lines of different colours, representing the readout from four separate channels which detect the four DNA bases: red for T, black for G, blue for C and green for A. While Trubetskoy's trace showed a clear red peak at position 169 corresponding to the mutation T, the âTsar's' trace at the same position showed the blue peak for C, the same as the reference sequence. But underneath the blue peak was a small red blip. Could it be that the âTsar's' DNA was a mixture of two mitochondrial DNA sequences, the main one with the sequence 126, 294, 296 and another, much smaller, with the same sequence plus the mutation at position 169? There was one way to find out, and that was to clone it.
Cloning is the only way to separate the different DNA molecules in a mixture. Briefly, it involves tricking bacteria into accepting just a single molecule of DNA and then copying it as if it were their own. Getting DNA into bacteria is a very inefficient process; only one in a million accepts it. Still, if just a couple of dozen bacteria can be persuaded to take in the DNA, they can be treated in such a way that the only bacteria to survive and grow as colonies on a culture dish are the ones with the extra DNA. They can then be picked off and the DNA sequenced. Within each colony, all the DNA will be copies of the original molecule that was accepted. If there is a mixture of two different DNA molecules to start with, some of the colonies will have one type and some will have the other. The scientists managed to create twenty-eight clones containing mitochondrial DNA from the âTsar'. When each of these was individually sequenced, twenty-one contained the main sequence 126, 294, 296 read from the original trace, without the mutation at 169. But the DNA from seven clones did contain the additional 169 mutation, making it absolutely identical to Count Trubetskoy's.
What the researchers had stumbled across was the very rare state where a new mutation, in this case at position 169, is part way to becoming established. This state, formally called
, had scarcely ever been observed before and was very little understood. As we will see in a later chapter, we know a lot more about heteroplasmy now; in 1994, when the paper on the âRomanov' remains was published, it was a novelty. But it did get the researchers off the hook. Here was the evidence they needed that there was indeed a continuous maternal link between the bones of the Ekaterinburg âTsar' and a living relative of Tsar Nicholas II.
The mitochondrial DNA matches were certainly good evidence to support the case that the Ekaterinburg bones were the remains of the Romanovs. But was it proof? Proof can never be absolute. It is always relative. In the case of the Romanovs the degree of certainty could be given a mathematical form depending on how common these mitochondrial sequences are in Europe. In those early days of the research we didn't know many European sequences, so it was hard to know how strong the evidence was. Now we have far more sequences to compare, and we know that the Duke of Edinburgh's sequence (111, 357) is actually extremely rare: it has not been found again in over six thousand Europeans. Since it has not been seen elsewhere, we cannot accurately estimate its frequency, but it is very unlikely to be higher than one in a thousand. This means there is, at most, a one in a thousand chance that the mitochondrial DNA sequence from a European picked at random would match the Duke of Edinburgh. So there was still a very small chance that the female Ekaterinburg bones did not belong to the Tsarina and her children at all, but to another family who just happened to have the same mitochondrial DNA as the Duke of Edinburgh. The Trubetskoy sequence (126, 169, 294, 296) is again very rare and has not been seen in six thousand modern Europeans. However, the Tsar's main sequence (126, 294, 296) is much more frequent, with just under one in a hundred Europeans matching it exactly. So, once again there was a small but finite chance that the bones of the adult male were not the Tsar's but those of someone else who just happened to match.