mtDNA L3e2* in a Sicilian-American

[MMaddi]

I have been asked by a Sicily Project member to explain the results, both yDNA and mtDNA, for her uncle.

I don't have much problem with the yDNA results. I feel fairly certain that he is E3b-M81 (based on a comparison with 2 SNP-tested Sicily Project members who are M81+), which would indicate possible deep ancestry from the Berbers of North Africa. This makes sense given how close North Africa is to Sicily and the fact that Berbers probably made up much of the Muslim population in Sicily during the Middle Ages. She also has a very definite family story that the ancestors were "Arabs" who converted to Christianity when the Normans took over around 1100. Of course, many people might confuse Berbers with Arabs.

His mtDNA results came in yesterday with these mutations from CRS:
HVR1 - 16223T, 16320T, 16325C, 16519C
HVR2 - 73G, 150T, 195C, 263G, 309.1C, 315.1C
FTDNA gives his haplogroup as L3e2*.

I understand that this is a sub-Saharan African mtDNA group. This is actually the second sub-Saharan African mtDNA group in the Sicily Project, out of 83 results. The other Sicily Project member is an L0a1. So, any sub-Saharan African mtDNA haplogroup in the project is very unusual, about 2%.

Given the probable Berber deep ancestry, I'm thinking it's possible that L3 or L3e may be found at low levels in the North African population. This may be due to very ancient migrations or perhaps the Muslim involvement in the black African slave trade in historic times.

Does anybody know whether I'm on the right track here? Or should I look for another explanation for finding L3e2* in someone of Sicilian ancestry?

Mike Maddi

[cacio]

MMaddi:

L sequences in Italy are not that unusual, and in fact, I have been surprised by the fact that so far few L sequences have appeared in mitosearch results, relative to what found in several academic papers. 2% L in Sicily is I believe around the average of what found in papers.

L sequences are present in N Africa at frequencies around 10-15%, and they are significant (although a little less frequent) in Arabia and the Levant. Hence I would agree with your explanation: the most likely explanation is some movement from N Africa. The Arabs is a possibility, but most likely even before, eg in Roman times with slaves and traders - if not in prehistoric times. Note that L sequences are present not only in Sicily, but also, in similar frequencies, in South and Central Italy, all the way up to Tuscany and the Marche region. This confirms that it is not just about the Arabs, since they didn't come to Central Italy.

For references (feel free to contact me offlist for the papers), the Salas paper on the slave trade (The African Diaspora) briefly talks about L sequences in Eurasia, though it focuses on Spain (which had a different history). It also discusses the diffusion of L3e in Subsaharan Africa. A recent paper by Abu Amero on Saudi Arabia lists the fractions of L3e in many Arab and NA countries (it's present throughout). Right now I cannot think of the specific papers on Italian mtdna showing the percentages of L, but I know they exist, as I talked about this to another guy already (your L0a guy?). Note that also the recent much discussed paper on the Etruscan mtdna finds around 2% L sequences in the Tuscan towns considered.

cacio

[MMaddi]

Cacio,

Thanks for the help. Please e-mail me the study references you mentioned. My address is [email protected]

Mike

[F.E.C.]

In my opinion a sub-Saharan mt-lineage in Central Italy (actually if there's any, I mean above 0.5%-1%) can only be of "recent" Etruscan/Middle Eastern origin: if such mt-haplogroups had been in Italy since pre-historic times they would find a match in sub-saharan y-haplogroups which, on the contrary, seem to show a total absence in Italy.

[cacio]

F.E.C.:

potentially, a few L sequences could have been in Anatolia also in
neolithic times. Women have always been more mobile than men.

But I agree with you that most of them likely dates back to trades and contact in historic times (ie Romans). The Romans had slaves from N Africa, so these must have left a trace on the mtdna. For Sicily, one may assume as well an Arab contribution, but then it's hard to explain a movement to Central Italy, where the percentages aren't much lower than in Sicily.

cacio

[F.E.C.]

cacio, do you think L and subclades can be found around 2% throughout Central Italy? Don't you think that, maybe, the "Etruscan" towns considered have a story of their own, which makes them genetically different from the neighbouring areas?

[cacio]

F.E.C.:

2% in Central Italy is probably too much, but there are many samples (ie towns) where that is the percentage.

I do think too that a particular village in Etruria can be explained in many ways and doesn't say much per se. Note also that we're talking about one village (Murlo), as the other two Etruscan areas didn't show the Near-Eastern-ness of Murlo.

cacio

[MMaddi]

Here's another wrinkle in the Sicily Project, although it doesn't involve a sub-Saharan mtDNA haplogroup. We have a new member who joined yesterday whose haplogroup is M1. Here's the write-up about this haplogroup by FTDNA:

"The haplogroup M cluster has been characterized as generally of east Eurasia—a geographic region that includes south Asia, east Asia, and Australasia. One of the two deep roots of the mitochondrial tree of haplogroups found in Asia, haplogroup M dates to approximately 70,000 years ago. Interestingly, one of the sub-haplogroups of the M cluster, haplogroup M1, is found primarily in northern Africa, suggesting either a very early divergence near the root of haplogroup M or even migration back to Africa after the original dispersal into Eurasia. Future work will further document the historical distribution of haplogroup M1 and clarify its origin and significance for the genetic history of humans."

So, it seems that M1 probably indicates a North African deep ancestry for this new project member. Looking at a pie chart for African mtDNA haplogroups in the study "The Making of the African mtDNA Landscape" by Salas, et al. (kindly sent to me by Cacio), it indicates that M1 is about 5-10% of the North African population and about the same percentage in East Africa (Sudan, Ethiopia, Uganda).

So, adding this M1 member to the two L members, that means 3 of 84 results reflect a definite African deep ancestry, about 3.5%.

[cacio]

MMaddi:

again, no surprise here. Andrew M. could probably tell you a lot about M1 in Europe, as his wife is M1 as well.

There is a very recent paper (which I have at home) on M1 and U6. The paper argues that M1 and U6 were the first mtdna haplogroups to enter N Africa, may be 20-40K years ago. The paper also shows several M1 sequences from Italy (you can see them in Ian Logan's webpage:
http://www.ianlogan.co.uk/discussion/hap_M1.htm
they are the ones listed as Oliveri)

So M1 is nothing strange in Italy, it could have gotten there anytime from prehistoric times to the Romans and the Arabs, perhaps together with the L and the U6 sequences, but perhaps even earlier.

cacio

[F.E.C.]

Here's another wrinkle in the Sicily Project, although it doesn't involve a sub-Saharan mtDNA haplogroup. We have a new member who joined yesterday whose haplogroup is M1. Here's the write-up about this haplogroup by FTDNA:

"The haplogroup M cluster has been characterized as generally of east Eurasia—a geographic region that includes south Asia, east Asia, and Australasia. One of the two deep roots of the mitochondrial tree of haplogroups found in Asia, haplogroup M dates to approximately 70,000 years ago. Interestingly, one of the sub-haplogroups of the M cluster, haplogroup M1, is found primarily in northern Africa, suggesting either a very early divergence near the root of haplogroup M or even migration back to Africa after the original dispersal into Eurasia. Future work will further document the historical distribution of haplogroup M1 and clarify its origin and significance for the genetic history of humans."

So, it seems that M1 probably indicates a North African deep ancestry for this new project member. Looking at a pie chart for African mtDNA haplogroups in the study "The Making of the African mtDNA Landscape" by Salas, et al. (kindly sent to me by Cacio), it indicates that M1 is about 5-10% of the North African population and about the same percentage in East Africa (Sudan, Ethiopia, Uganda).

So, adding this M1 member to the two L members, that means 3 of 84 results reflect a definite African deep ancestry, about 3.5%.

Maybe I'm wrong but isn't mt-haplogroup M associated with gypsies? If it were so such a result wouldn't surprise me too much. On the contrary, I would have expected to see some more of M, considering for how long gypsies have lived in Italy.

[cacio]

F.E.C.

macrohaplogroup M is present in and east of India. It has lots of subgroups, and a couple of them are indeed found in European gypsies (M5 primarily). However, M1 is different - I was confused too, but we learnt about M1 when discussing Andrew M's wife (in another thread).

M1 is not present in India and is the only M subgroup common anywhere west of India. It has its highest percentages in the horn of Africa, and from there, it is common up north in Northern Africa and the Mediterranean. I don't think it is clear yet whether it represents the only M lineage still surviving in Africa, or rather a back-migration from Asia/Middle East into Africa - though in very early times (40-50K years ago), as the recent m1-u6 paper seems to propose.

Either way, M1 represents a Mediterranean origin, not an Indian one.

cacio

[vraatyah]

On Italian Ls. It's very unlikely that they were borrowed from claimed Asia Minor ancestors of Etruscans. Although in the modern Turks Ls are present, their total frequency is about 1% (N>700). The area, covered by Etruscans in the past, now has 2% of Ls, i.e. even more. Actually, there are some Turkish samples with more than 2% of Ls, but even in this case the admixture from the Ancient Anatolians should be about 100%. Probably, some Turkish Ls are the result of intermarriage with Arabs, due to slavery etc, because in the Eastern Europe which share with Anatolians more "neolithic" lineages than the Western part of Europe do, Ls are virtually absent. I have no idea from where Ls have come to Italy. Central Italy has nothing to do with Moorish slavery, afaik. The same in case of Roman slavery: it's very unlikely that the Romans kept Black slaves only in provinces where now Ls are found at the frequency > 1%, say. Did they really used Black slaves?

Perhaps, we are dealing with the random fluctuations, until a more detailed map of The Southern European Ls is discovered. The current sample size is not sufficient for proving of reliable difference for the Hgs of such rarity.

[MMaddi]

On Italian Ls. It's very unlikely that they were borrowed from claimed Asia Minor ancestors of Etruscans. Although in the modern Turks Ls are present, their total frequency is about 1% (N>700). The area, covered by Etruscans in the past, now has 2% of Ls, i.e. even more. Actually, there are some Turkish samples with more than 2% of Ls, but even in this case the admixture from the Ancient Anatolians should be about 100%. Probably, some Turkish Ls are the result of intermarriage with Arabs, due to slavery etc, because in the Eastern Europe which share with Anatolians more "neolithic" lineages than the Western part of Europe do, Ls are virtually absent. I have no idea from where Ls have come to Italy. Central Italy has nothing to do with Moorish slavery, afaik. The same in case of Roman slavery: it's very unlikely that the Romans kept Black slaves only in provinces where now Ls are found at the frequency > 1%, say. Did they really used Black slaves?

Perhaps, we are dealing with the random fluctuations, until a more detailed map of The Southern European Ls is discovered. The current sample size is not sufficient for proving of reliable difference for the Hgs of such rarity.

Cacio, in an earlier reply in this thread, notes that the L haplogroup and its subclades are found at a 10-15% level in North Africa and agrees that it's possible that this Sicily Project member's maternal line is from North Africa, either in the pre-historic or historic period. I had suggested this possibility in my original posting.

Do you think this is likely, given his L3e2* haplogroup and the following haplotype?
HVR1 - 16223T, 16320T, 16325C, 16519C
HVR2 - 73G, 150T, 195C, 263G, 309.1C, 315.1C
Do you know of a North African haplotype in any study that matches this?

[F.E.C.]

F.E.C.

macrohaplogroup M is present in and east of India. It has lots of subgroups, and a couple of them are indeed found in European gypsies (M5 primarily). However, M1 is different - I was confused too, but we learnt about M1 when discussing Andrew M's wife (in another thread).

M1 is not present in India and is the only M subgroup common anywhere west of India. It has its highest percentages in the horn of Africa, and from there, it is common up north in Northern Africa and the Mediterranean. I don't think it is clear yet whether it represents the only M lineage still surviving in Africa, or rather a back-migration from Asia/Middle East into Africa - though in very early times (40-50K years ago), as the recent m1-u6 paper seems to propose.

Either way, M1 represents a Mediterranean origin, not an Indian one.

cacio

I see. Thanks for the clarification cacio

[vraatyah]

Do you think this is likely, given his L3e2* haplogroup and the following haplotype?
HVR1 - 16223T, 16320T, 16325C, 16519C
HVR2 - 73G, 150T, 195C, 263G, 309.1C, 315.1C
Do you know of a North African haplotype in any study that matches this?

Hmmm.. I have only one exactly the same haplotype in the database, both hvs1 + hvs2. It's from the SWGDAM database, that portion was contributed by the Illinois State Police. American Caucasian.