Considering mtDNA testing for genetic genealogy?
Here’s our next post in our tour through the use of DNA testing for genealogy. This time the focus is on matrilineal researches using mitochondrial DNA aka M-DNA aka mtDNA. As we mentioned before, the mitochondria is a so-called “organelle” that resides within the our cells, and those of other animals. It is the engine that provides us all with our cells’ metabolic energy. There is a theory that these mitochondria were once independent creatures, but that somewhere in the mists of time, a symbiotic relationship developed between early animals and the mitochondria. One argument in favor of this proposition is that the mitochondria have their own relatively simple DNA structure. Unlike the pairs of chromosomes within the cell nucleus, the mitochondria has its DNA looped into a circle, with a certain region having an overlap in a kind of “tail-biting” structure. This is so different from any other DNA structure as to almost cry out for a separate evolutionary history.
Whatever the actual origin of the mitochondria and its DNA, it exists within every cell including the male sperm cells, where the mitochondria exists at the base of the tail or flagella, and in the female egg. However, during the process of fertilization, the mitochondria from the sperm cell does not enter the egg, which has its mitochondria from the mother. This is simply because the egg cells were actually part of the mother at the time of her birth, having developed in utero. By extension, the egg cell of the mother’s mother were developed when she was in utero and received from her mother, and so on back in time. As the newly fertilized egg divides repeatedly during the development of the growing fetus, each cell copy receives its mitochondria and its mtDNA by a process of replication from that of the original unfertilized egg cell.
While this process is intellectually fascinating in its own right, the focus in this post is on the testing of mtDNA for genealogy; and as we mentioned in an earlier post, the interest in mtDNA is due to this property of transmission along a matrilineal descent. It is not that males do not have their mtDNA from their mothers; they do. But a male cannot transmit his mtDNA to either son or daughter since that within his sperm is discarded at the outer wall of the egg.
Well, that’s one theory anyway. It is the theory that you will find in most stories on the value of mtDNA for genealogy. Unfortunately, it is not the whole story, as this Wikipedia article on paternal mtDNA transmission tells. Here the story unfolds that mtDNA from the sperm can and does enter the egg, although in significantly smaller proportions to that contained there from the mother. There is also evidence that some individuals can wind up with a mosaic of mtDNA in different parts of their tissue; say, most from their mother but a scattering from their father. And so on. Hence, the story of mtDNA remains incomplete.
However, since the crux of the matter is the source of mtDNA in the egg cell from which any daughter will develop upon fertilization, the question is whether or not the transmission of maternal mtDNA is “a lock”. I mean, we don’t develop from our mother’s muscle tissue, after all. Even a 1:1,000,000 chance could disturb some scientific theories in this regard, although that would probably not overturn the value of mtDNA for genealogy. However, there is practically no information available on the likelihood of paternal mtDNA transmission in humans outside of some references to what might be termed freak cases.
A related issue is whether or not maternal and paternal mtDNA might recombine in some way during fertilization; that is, whether mtDNA in the fetus might in some way mix the mtDNA of both parents. This is also a controversial statement at present with few scientific results in its favor in in vivo human cases, which is to say, in real life. This too would overturn an apple cart or two.
For the time being, I’m going to ignore some of these potential cracks in the theory, and continue to tell the story as you will find it in most of the literature. I’ll come back to what this might mean for your researches at the end of the post.
One of the primary foci for study of mtDNA are two so-called hypervariable regions in the control region or “D loop”. Consider the image below:
mitochondrial DNA
The control region is shown at the top center position of the circle. Within it are two hypervariable regions called HVR1 & HVR2, simply enough. HVR1 spans base pair locations numbered from 16001 to 16568. HVR2 spans locations 001-574. HVR1 is considered to have low resolution, while HVR2 has high resolution, meaning just that the nucleotides in HVR1 mutate less often.
The mechanism of mutation in the control region is single nucleotide polymorphism, just as we considered previously in Y-DNA studies. So, what we will be looking at will be substitutions of one nucleotide for another, say, C by T. Just in the same way as we considered haplogroups that arise because of these SNP mutations in the Y-chromosome, we’ll also be considering female haplogroups associated with specific SNPs in mtDNA.
In 1981, researchers at Cambridge University sequenced the entire mitochondrial genome of a human donor, showing it to have 16,569 base pairs and 37 genes. This mtDNA sequence has become known as the Cambridge Reference Sequence (CRS). Some errors were discovered in the original publication and a revised CRS or rCRS was published. Now, most mtDNA results are expressed in terms of differences between the individual being tested and the rCRS. Again, in the interests of providing examples, here are my results for HVR1 & HVR2, given with respect to the rCRS from Family Tree DNA.
So, the rCRS has adenine (A) at position 16066 and I have guanine (G), and so on. All of this data goes to show that I am in the U5a mtDNA haplogroup. This is presumably also true of my mother, my grandmother, my aunts, my sisters, my sisters’ children, my first cousins on my mother’s side, and et cetera. Should there be some question about such a maternal relationship between myself and some other person, then a match for this mtDNA pattern could go a long way in proving the relationship, or disproving it if a match failed.
Just as in the case of Y-DNA haplogroups, mtDNA haplogroups have been carefully analyzed. And just as Bryan Sykes wrote a popular version of the analysis of Y-DNA haplotypes in the history of Britain, he also wrote a similar book on mtDNA haplogroups and the populations of Europe titled The Seven Daughters of Eve, and subtitled The science that reveals our genetic ancestry. In that book, Sykes considers seven major mtDNA haplogroups, gives them names instead of initials, and places them in their own ancestral “Gardens of Eden”. So, Ursula is in Greece 45,000 years ago. Xenia is in the Caucasus 17,000 years ago. Jasmine is in the middle East 10,000 years ago. Katrine is in Italy near what is now Venice 15,000 years ago. Tara is on the western coast of Italy 17,000 years ago. Helena is in France 20,000 years ago. Velda is in Spain 17,000 years ago. These are the seven daughters of mtDNA-Eve. Of course, this is all quite romantic; and worth the read. I recommend the book.
However, as in the case of the basic Y-DNA haplotypes that may be revealed with a 12 STR panel test, many people begin with the mtDNA test for HVR1 and learn little more than they are in one of these basic mtDNA haplogroups, which originated anywhere from 10,000 to 45,000 years ago. By now, you’ll have lots of relatives.
Including an HVR2 test will provide much more resolving power than just doing HRV1. That’s the good news. The bad news is that relatively few people have tested HVR2 and published their results. In the same way that FTDNA has put a large database of Y-DNA results online at ysearch.org, they have also posted results of mtDNA tests at Mitosearch. However, the size of the database and its “resolving power” is much more limited. In part to augment the limitations of just testing HVR1 and HVR2, it has become possible to test the entire mtDNA sequence. Together with HVR1 and HVR2, such a “mega” test covers all base pairs from 00001 to 16569 inclusive, not just the 500 – 600 or so pairs that tie together to form the D-loop. By the way, it’s called this since it forms a sort of “D” shape with three strands of DNA in this area.
This is a some good some bad proposition. The good side is that you are more likely to find high quality matches in the future. The bad news is that this entire region includes some coding DNA elements in which certain mutations are associated with genetically transmitted disease. So, in publishing such information about yourself, you can run the risk that some health insurance organization might declare you a risk and deny you coverage for the problem. This is a problem in the US where the health care system is designed around private insurers finding ways to deny coverage to their clients, IMHO. It’s less of a problem elsewhere in the developed world. For such reasons, testing labs like Family Tree DNA provide secure methods of matching your full mtDNA sequence to other users without making the results public. You can go ahead and make your mtDNA sequence public; but this is not required in order to find matches. Publicly sharing your HVR1 and HVR2 tests do not have this stigma since they are not coding DNA.
Here is a chart showing parts of the mtDNA loop that are associated with genetic disease:
So, unless you do at least and HVR2 test, and possibly a complete mtDNA mega-test, your results will provide you with little resolution in discovering your relatives. Even if you do a complete mega-test, as I have done, you’re not likely to obtain much immediate value from your results until others follow suit. If you have the money to do this test, and if you’re interested in tracking these results, then I recommend doing the full mtDNA sequence test and being patient. However, unlike Y-DNA testing, do not expect a rapid return on your investment. In subsequent posts, I’ll talk about bang for your buck in testing.
Earlier, I made a reference to an mtDNA-Eve. This idea is consistent with what’s called the “Recent single origin of humans” theory. The essence of the idea is that modern Homo sapiens evolved in Africa and eventually spread out over the world from there. A completely contradictory model would be that modern humans evolved uniformly from previous species all over the planet, arguably into different races. This last theory has been discredited, but more recently a middle ground has been proposed, based on very recent evidence. This theory argues that modern humans of African origin interbred with Neanderthals in Europe and with the recently discovered Denisovans in Asia. These hypotheses are consistent with, among other things, tests of mtDNA and autosomal DNA from modern humans, Neanderthal remains and Denisovan remains. However, they are not conclusive within the scientific community. This is to be expected since the first Denisova hominin was discovered only in 2010.
So, in the mtDNA-Eve theory, all living humans have their mtDNA from this single woman and along the way, there have been certain mutations that have accumulated leading to new haplogroups. In this competing model, there is an allowance that some modern Europeans have mtDNA in part from Neanderthal admixture and some Melanesians and Asians have mtDNA in part from Denisovan admixture. From the point of view of using mtDNA sequencing for genealogical research, the truth or falsehood of either theory is really irrelevant. The events in question date back tens of thousands of years; and they wouldn’t make a difference to whether some living person is a cousin on your mother’s side as might be proven by an mtDNA test.
To summarize, mtDNA testing is the standard approach to tracing your matrilineal ancestors, and therefore, identifying relationships with living persons who have the same descent. Along the way, you may also find some interesting, if “romantic” information about the very ancient history of your ancestors going back tens of thousands of years.
Good luck!