October 07, 2011

Ancient mtDNA of Hokkaido Jomon

This is a very good paper that I don't have time right now to write a long post about; I will update this entry with some excerpts and/or additional thoughts when I can.

The gist of it is that the prehistoric Jomon people of Japan belonged to mtDNA haplogroups tying them to southeastern Siberia, but some haplogroups present there today were lacking in them. Then, the Ainu seem to have inherited the Jomon gene pool, but their major lineages tie them to the Okhotsk people. So, it seems that the deepest ancestry of Japan is not peculiar to it, but rather an extension of ancient Siberian variation with different population strata attributed to the Jomon, the Ainu, and (probably) the modern Japanese.

AJPA DOI: 10.1002/ajpa.21561

Mitochondrial DNA analysis of Hokkaido Jomon skeletons: Remnants of archaic maternal lineages at the southwestern edge of former Beringia

Noboru Adachi et al.

To clarify the colonizing process of East/Northeast Asia as well as the peopling of the Americas, identifying the genetic characteristics of Paleolithic Siberians is indispensable. However, no genetic information on the Paleolithic Siberians has hitherto been reported. In the present study, we analyzed ancient DNA recovered from Jomon skeletons excavated from the northernmost island of Japan, Hokkaido, which was connected with southern Siberia in the Paleolithic period. Both the control and coding regions of their mitochondrial DNA (mtDNA) were analyzed in detail, and we confidently assigned 54 mtDNAs to relevant haplogroups. Haplogroups N9b, D4h2, G1b, and M7a were observed in these individuals, with N9b being the predominant one. The fact that all these haplogroups, except M7a, were observed with relatively high frequencies in the southeastern Siberians, but were absent in southeastern Asian populations, implies that most of the Hokkaido Jomon people were direct descendants of Paleolithic Siberians. The coalescence time of N9b (ca. 22,000 years) was before or during the last glacial maximum, implying that the initial trigger for the Jomon migration in Hokkaido was increased glaciations during this period. Interestingly, Hokkaido Jomons lack specific haplogroups that are prevailing in present-day native Siberians, implying that diffusion of these haplogroups in Siberia might have been after the beginning of the Jomon era, about 15,000 years before present.

Link

17 comments:

eurologist said...

I can't take coalescent time seriously that end up close to LGM (likely they are off by a factor of 1.5 to 3).

At any rate, it is no surprise that extant populations in Siberia show a different composition, since much of the region underwent a twofold change first after LGM/ Younger Dryas and then after (nearby expansion pushes) of introduction of herding and of agriculture.

Andrew Oh-Willeke said...

The journal also has a wealth of free access New World genetics studies including some notable aDNA studies.

"the Dude" said...

English - water
Hittite - watar
Kusunda - wide (Nepal)
Russian - voda (wodka?)
Ainu - wakka
Quechua - yakku (Peru)
Seminole - oki (Florida)

G Horvat said...

By "South East Siberia", I think Adachi et al. was primarily referring to the Amur region:

"Recent studies have revealed that haplogroup N9b exists at a relatively high frequency in the southeastern Siberian populations (Udegey, 30.4%; Ulchi, 6.9%, Starikovskaya et al., 2005), whereas it is absent or scarce in the other East and Southeast Asian populations outside Japan...it is highly possible that this haplogroup was introduced to Japan by the Paleolithic ancestor of the Hokkaido Jomons from the lower Amur region of Siberia."

The N9b and related Y sequences appear to be concentrated in the region surrounding the Sea of Okhotsk. See map:

http://www.worldatlas.com/aatlas/infopage/okhotsk.htm

Ebizur said...

"The N9b and related Y sequences appear to be concentrated in the region surrounding the Sea of Okhotsk. See map:

http://www.worldatlas.com/aatlas/infopage/okhotsk.htm"

Please keep in mind that, according to the currently accepted phylogenetic tree of human mitochondrial DNA, haplogroup N9b is no more closely related to haplogroup Y than either is related to haplogroup N9a. Haplogroup N9a, like haplogroup Y, is actually found throughout East, Southeast, and Central Asia.

Haplogroup N9b is the only subclade of N9 that really seems to be particularly limited to the Japanese Archipelago and southeastern Siberia. Furthermore, as far as I know, haplogroup N9b has not been found in Palaeo-Siberians of Chukotka and Kamchatka, but rather among Japanese and among Tungusic peoples of Primorye (Udegey) and the lower Amur River (Ulch). Its distribution is better described as "circum-Japan Sea" than "circum-Okhotsk."

M. Smith said...

Mitochondrial haplogroup N9b is actually not just restricted to Ainu, but also found in Southeast Asians. Also Starikovskaya et al. 2005 found 30.4% N9, not N9b, among Udegeys. And the complete absence of N9b among Paleo-Siberians suggests that N9b is not a "Southeast Siberian" mtDNA lineage.

German Dziebel said...

"However, it should be noted that the haplogroups A, C, D (except for D4h2), and Y, which are widely observed in the present-day Siberian populations, were not observed in Hokkaido Jomons."

The combination of lineages found in Hokkaido Jomon people (N9b, G1, M7a, D4h2) is very different from the combination of lineages found in Amerindians. At the same time, the Amerindian combination of lineages (A, B, C, D1, X) is very different from the modern Siberian one. This suggests that Amerindians are NOT derived from either the early (as represented by Hokkaido Jomon) or the late stratum of Siberian populations.

G Horvat said...

Based upon RFLP, the Udegey and Ulchi "N9" sequences belong to N9b like the complete Nanaitci retrieved by Ingman 2007. I agree that N9b and Y sequences were brought to the Amur Region and Japan from some place else... but by which route? Up the coast from SE Asia?

Ebizur said...

Does no one else find it curious that all the haplogroups reported in this sample of human remains from a Jōmon site in Hokkaidō except N9b are the same as those that have been reported in what is (as far as I know) the only published sample of modern Ainu mtDNA?

According to A. Tajima et al. (March 2004) and
M. Tanaka, V. M. Cabrera, A. M. González et al. (October 2004), both of which analyzed the same sample of modern Ainu mtDNA, the modern Ainu belong mainly to haplogroups Y, D4(xD1), M7a, and G, all in approximately equal proportion.

The following haplogroups also have been found in this sample of modern Ainus, though in only one or two individuals each:

Ainu (Tanaka et al. 2004)
2/51 A1b
2/51 M*
2/51 M7b2
2/51 D5a
1/51 N9b
1/51 B4f
1/51 F1b
1/51 M9a
1/51 M/N (Only determined to the level of M/N? Ouch!)

Can anyone explain how the modern Ainu might have maintained the high frequency of M7a, D4(xD1), and G that has been observed in the sample of Hokkaidō Jōmon remains while almost completely switching the Hokkaidō Jōmon haplogroup N9b mtDNA for modern Ainu haplogroup Y mtDNA?

It is also notable that samples of Jōmon remains from other regions of Japan have yielded very different distributions of mtDNA haplogroups.

M. Smith said...

Can someone link the study that says the N9 sequences among the Udegey are all N9b? Also, the N9 among the Ulchi is quite negligible at only 6.9%. N9 was also found among the Tubular and Tuvan who are nearer to Central Asia or Kazakhstan than Japan.

It's interesting to see that the Ainu mtDNA haplogroups are very diverse.

Also can someone link the studies of other Jomon populations? The Ainu are just a small part of the Jomon family and it's likely that the Jomon populations were already diverse before the Yayoi invaded Japan.

Ebizur said...

It is also interesting that the four major clades (N9, D4(xD1), M7a, and G) found in this sample of Hokkaidō Jōmon remains and in the aforementioned sample of modern Ainu are also four of the five most common mtDNA clades among Japanese in general. (The remaining clade is Haplogroup B. However, as I recall, haplogroup B mtDNA has been found with fairly high frequency in a sample of Jōmon remains from the Kantō region.)

The subclades of N9 present a curious picture in the case of the general Japanese population, too. Among the subclades of N9, N9a is the most frequent, though with relatively low diversity, among mainstream Japanese. The highest reported diversity of N9b has been found in mainstream Japanese, but its frequency seems to be somewhat lower than that of N9a. Haplogroup Y also has been found in mainstream Japanese, but only in very low proportion (approx. 1% or less).

So the mtDNA pools of Hokkaidō Jōmon, modern Ainu, and modern mainstream Japanese seem to be overall quite similar, with the following discrepancies:

(1) Mainstream Japanese quite frequently exhibit haplogroup B mtDNA (especially B4, but also B5). Only a single representative of haplogroup B (in subclade B4f) has been found in the published sample of modern Hokkaidō Ainu mtDNA, and this haplogroup seems to be entirely absent from the published sample of Hokkaidō Jōmon mtDNA.

(2) In regard to haplogroup N9, mainstream Japanese examples are most frequently N9a (but with low diversity), while N9b is presented with moderate frequency but notably high diversity. N9's other subclade, Haplogroup Y, is rather rare among mainstream Japanese.

(3) The published sample of Hokkaidō Jōmon mtDNA exhibits a high frequency of the N9b subclade of haplogroup N9, but it lacks the N9a and Y subclades.

(4) The published sample of modern Hokkaidō Ainu mtDNA exhibits a high frequency of the Y subclade of haplogroup N9, but it contains only a single representative of the N9b subclade, and it seems to entirely lack the N9a subclade.

G Horvat said...

The 2 Ainu haplogroup A sequences actually belong to different subgroups: A4 & A8. A8 seems to have been carried west to east (or vice versa). In addition to the ones of northern Asia, 2 Hungarians from Transylvania carried that particular type of 'A' sequence as well as a Brazilian of Lebanese descent.

Another Ainu sequence which may have been carried from west to east is the one with the transversion at site no. 16355. As far as I can tell, it should belong to D4n. This is the clearest link with Tibetan sequences that I see. There are other possibilities.

I count 3 Ainu sequences which *could* belong to N9b. The one with 16266T could belong with the a fully sequenced "Udegei" & Japanese N9b.

Going back to haplogroup A, Tanaka et al. (2004) classified sequences as A1 (Asian) or A2 (Native American) as was common at the time. What is interesting here is that they state:

"Subhaplogroup A2 is mainly present in northeast Siberia including the Kamchatka peninsula, although a lineage has also been detected in Tibet."

I see some Tibetan haplogroup A sequences with HVR II 153G but it seems no one has bothered to fully sequence them. 153G is mainly found in New World haplogroup A sequences.

terryt said...

"By 'South East Siberia', I think Adachi et al. was primarily referring to the Amur region"

Seems likely to me.

"haplogroup N9b has not been found in Palaeo-Siberians of Chukotka and Kamchatka, but rather among Japanese and among Tungusic peoples of Primorye (Udegey) and the lower Amur River (Ulch)".

Fits the above. Did N9/Y originate there?

"Mitochondrial haplogroup N9b is actually not just restricted to Ainu, but also found in Southeast Asians".

But as Eurologist pointed out:

"according to the currently accepted phylogenetic tree of human mitochondrial DNA, haplogroup N9b is no more closely related to haplogroup Y than either is related to haplogroup N9a".

So the haplogroup as a whole is not represented in SE Asia. Neither N9a nor Y are found there. So the haplogroup is unlikely to have originated there.

"The combination of lineages found in Hokkaido Jomon people (N9b, G1, M7a, D4h2) is very different from the combination of lineages found in Amerindians".

Is that really so surprising? You mentioned yourself that Amerinds probably did not enter via Japan. It used to be generally accepted they'd entered via the tundra, which the distinctiveness between the two tends to support.

"Amerindians are NOT derived from either the early (as represented by Hokkaido Jomon) or the late stratum of Siberian populations".

Not the 'late stratum of Siberian populations' but by an earlier one. Again Eurologist mentioned:

"much of the region underwent a twofold change first after LGM/ Younger Dryas and then after (nearby expansion pushes) of introduction of herding and of agriculture".

That explains thing adequately.

"Among the subclades of N9, N9a is the most frequent, though with relatively low diversity, among mainstream Japanese. The highest reported diversity of N9b has been found in mainstream Japanese, but its frequency seems to be somewhat lower than that of N9a. Haplogroup Y also has been found in mainstream Japanese, but only in very low proportion (approx. 1% or less)".

The interchange of N9a, N9b and Y within Japan suggest to me that the haplogroup originated nearby. As you move away from NE Asia/Japan the founder effect comes in until at the geographic margins just single subclades of the three basal clades of N9/Y survive.

"Mainstream Japanese quite frequently exhibit haplogroup B mtDNA (especially B4, but also B5). Only a single representative of haplogroup B (in subclade B4f) has been found in the published sample of modern Hokkaidō Ainu mtDNA".

B is an interesting haplogroup. Its distribution is virtually circum-Pacific with the opposite extremities (America and Polynesia) containing descendants of very much downstream clades (as we would expect).

"The 2 Ainu haplogroup A sequences actually belong to different subgroups: A4 & A8. A8 seems to have been carried west to east (or vice versa). In addition to the ones of northern Asia, 2 Hungarians from Transylvania carried that particular type of 'A' sequence"

A's presence as basal haplogroups (A5, A8, A10 and all the rest as a single haplogroup) all the way from the Volga to Japan suggests it originated somewhere within that region. The SE Asian (A5b) and American (A2) versions are derived clades within A. Was a originally a neighbour to the west of N9/Y? And X even further west?

G Horvat said...

N9a and Y are found in SE Asia - maybe N9b too but N9b HVR I sequences are not as distinctive as the sequences of the other two branches. The Taiwanese aboriginal N9a sequences look very basal in the first hypervariable region but I do not know if the same applies to the rest of the genome because none have been completely sequenced (to my knowledge).

"The interchange of N9a, N9b and Y within Japan suggest to me that the haplogroup originated nearby"

'N' haplogroups had to be derived from other 'N' haplogroups...

I am inclined to place the origin of haplogroup A in Tibet or where Tibeto-Burman populations originated.

terryt said...

"I am inclined to place the origin of haplogroup A in Tibet or where Tibeto-Burman populations originated".

So am I.

"N9a and Y are found in SE Asia - maybe N9b too but N9b HVR I sequences are not as distinctive as the sequences of the other two branches. The Taiwanese aboriginal N9a sequences look very basal in the first hypervariable region"

I'm quite happy to accept an ancient move south for members of the haplogroup, probably accompanied by other basal N haplogroups.

"'N' haplogroups had to be derived from other 'N' haplogroups..."

Very much so. So if A is Tibetan, X is SW Asian and N9/Y is Northeast Asian we then find a string of basal N haplogroups down the East Eurasian coast through the Malay Peninsula and out to Australia: N8, N10, N11, R (in the form of R3, R11'B6, B4'5, R9/F, R22, R23, R12'21, R14 and P), N22, N21, O, S, N13 and N14. That leaves N1'5 and N2. Both found in South Asia, but in derived form and basically only in the northwest. Makes a coastal migration through South Asia unlikely in the case of mt-hap N.

Pam from NH said...

This is a new post on an old blog...but here goes. I am not a scientist, but am very interested I human evolution and ancient genetics. I just received the results of my mtDNA teat thru the National Geo Genome Project 2.0. The four haplogroups that were identified in my sample were L3, N, N9, and Y. I have a hard time finding much on N9 and Y. IN fact my report and map from NG were blank on these two groups. I am very confused as to what these groups represent. My mother's family was all German, with the exception of her Grandmother, who was Croatian and HUngarian. Can somebody fill me in? Thank you!!!

G Horvat said...

Pam, instead of "four haplogroups", I think it should be Y < N9 < N < L3. Therefore, your haplogroup would be Y. It is very rare in Europe but has been previously reported in a Hungarian sample. See the supplementary data of the following article:
http://onlinelibrary.wiley.com/doi/10.1111/j.1469-1809.2007.00371.x/suppinfo