Genetic Origins of the Japanese a Partial Support
-
Upload
cristina-i -
Category
Documents
-
view
218 -
download
0
Transcript of Genetic Origins of the Japanese a Partial Support
![Page 1: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/1.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 1/10
Genetic Origins of the Japanese: A Partial Supportfor the Dual Structure Hypothesis
KEIICHI OMOTO1* AND NARUYA SAITOU 2
1 International Research Center for Jap anese Stu dies, Kyoto, 610-11 J apan;2 National Institute of Genetics, Mishima, 411 Japan
KEY WORDS genetic dista nce; origins of J apa nese; Ainu;
Ryukyua n; Asian populations
A B S T R A C T Based on the morphological char acteristics of the skull a nd
teeth, Hanihara ([1991] Japan Review 2:1–33) proposed t he ‘‘dua l str uctu re
model’’ for th e forma tion of modern J apa nese populations. We examine this
model by dividing it into two independent hypotheses: 1) the Upper Paleo-
lithic population of Ja pan tha t gave r ise to the N eolithic J omon people was of
s ou t h e a s t As ia n or i gi n , a n d 2 ) m o de r n Ai n u a n d R yu k y u a n (O k in a w a )populations are direct descendan ts of the J omon people, while H ondo (Main
I sland) -J apanese ar e mai nly der ived fr om the migr ants fr om the nor theast
Asian cont inent after t he Aeneolithic Yayoi period. Our a im is to examine t he
extent to which the model i s suppor ted by genetic evidence f r om moder n
populations, part icularly from J apa n a nd other Asian area s. Based on genetic
dista nce an alyses u sing da ta from u p to 25 ‘‘class ic’’genet ic ma rk ers, we find
fi r s t t h a t t h e t h r e e J a p a n e s e p op u la t i on s i n cl u di n g Ai n u a n d R yu k y u a n
clear ly bel ong t o a nor theast Asian cluster gr oup. This negates the fir st
hypothesis of the model . Then, we find that Ainu and Ryukyuans shar e a
group contrasting with Hondo-Japanese and Korean, supporting the second
hypoth esis of th e model. Based on t hese r esult s, we propose a m odified version
of the dual structure model which may explain the genetic, morphological,
and archaeological evidence concerning the formation of modern Japanese
populat ions. Am J Ph ys Anth ropol 102:437–446, 1997. r 1997 Wiley-Liss, Inc.
Questions about t he or igins of moder n
Japanese have a long history of debate (for
references see Han ihara , 1991). In short, th e
cont rover sies wer e between ‘‘cont inu ity’’an d
‘‘adm ixture’’models.Th e advocat es ofth e former
model believed that the inhabitants of th e
Ja panese Islands were genetically u nchanged
from prehistoric to historic times, while their
morphology showed secular changes (e.g., Su-
zuki, 1969). The advocates of the latter model
empha sized the dra stic changes in morphologyand cultur e which took place synchronically
about 2,300 years a go, mainly in the western
part ofJ apa n, and considered these as evidence
for a dmixtu re (e.g., Kana seki et a l., 1960).
In 1991, Hanihara proposed a hypothesis
for the formation of Japanese populations,
which h e called t he ‘‘dua l st ru ctur e m odel.’’
I t i s c l e a r l y o n t h e s i d e o f t h e a d m i x t u r e
school and is the most comprehensive cur-
rent hypothesis on th e forma tion of modern
Japanese populations.
To test this hypothesis, we divide it into
t w o p a r t s t h a t a r e b r i e fl y s u m m a r i z e d a s
follows: 1) the U pper Pa leolithic populat ions
of J apa n came from somewhere in southea st
As ia a n d g a ve r is e t o J o m on e se , o r t h e
people oft he Neolithic J omon period (12,000–
2,300 year s BP) , and 2) moder n Japanesepopulations were formed by the mixture of
mainly two population groups with different
*Correspondence to: K. Omoto, Intern ational Research Cent erfor Ja pa ne se St udi e s, 3-2 Oe ya ma -cho, Goryo, Ni shi kyo-ku,Kyoto, 610-11 Japa n. Ema il: ko8516@nichibu n.ac.jp
Received 14 November 1995; accepted 2 Febr uar y 1997.
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 102:437–446 (1997)
r 1997 WILEY-LISS, INC.
![Page 2: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/2.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 2/10
a n ce s t r ie s : Ai n u a n d R yu k y u a n a r e r e la -
tively pu re descendan ts of Jomonese, while
Hondo (Main Island)-Japanese have received
strong genetic infusions from migrant popula-
tions who came to western J apan during the
Aeneolithic Yayoi period (300 BC–300 AD) and
the Proto-historic Kofun period (300–700 AD).
These migr ant s came f r om nor t heast Asia
v i a t h e K o r e a n p e n i n s u l a a n d f r o m t h e r e
spr ead to easter n and souther n Ja pan.We believe th at qu estions on th e origins of
huma n populations ar e better addr essed by
g en e t ic a p p r oa ch e s t h a n b y ot h e r a p -
pr oaches, be they mor phological or , defi-
nitely, cultural (e.g., archaeological and lin-
guistic) ones. In this paper, we present the
results of two genetic distance an alyses tha t
use ‘‘class ic’’gen etic m ar ker gene frequency
data. The fir st analysis tests whether J apa-
n e s e p op u la t i on s h a v e r oot s i n s ou t h e a s t
Asia. The second one seeks t o deter mine
w h et h er t h e re is a d u a l s t r u ct u r e a m on g
Japanese populati ons. On the basi s of our
findings, we propose that the first questioncannot but the second question can be sup-
ported. We discuss the ant hropological sig-
n ifi ca n ce of t h es e r e su lt s a n d p r es en t a
modified version of the dual st ructur e model.
MATERIALS AND METHODS
I n the fir st geneti c dist ance analysis, 26
populations of the wor ld, i ncludi ng thr ee
J a p a n e s e (Ai n u , R y u k yu a n a n d H on d o-
J apa nese), were compar ed. The populations
examined other than Japanese wer e: thr ee
Afr ican, four I ndo-Eur opean, thr ee native
American, nine east Asian, one native Aus-
t r a l ia n , on e N e w G u i n e a n (P a p u a n ), a n dtwo Pacificpopulations. Gene frequency data
on a total of 20 polymorphic loci were avail-
able for t hese p opulat ions: eight blood group
systems (ABO, MNSs, P, Fy, Rh, Jk, Di, K),
seven r ed cell enzyme systems (ACP, P GD,
PGM1, PGM2, ADA, GPT, ESD), and five
s e r u m p r ot e in s ys t e m s (H p , T f, G c, G m ,
I n v). M os t d a t a on J a p a n e se p op u la t i on s
wer e taken f r om the compilation by JI BP
Synt hesis, Volum e 2 (Wat an abe et al., 1975).
The sources of some additional da ta were as
follows: GPT (Ueda et al., 1979) and ESD
(Omoto et al ., 1975). The data on blood
genetic mar kers of Korean s were ta ken from
Yua n et a l. (1984) and Goedde et a l. (1984).
The data on other populations were obtained
from Roychoudhury and Nei (1988) and Ma-satoshi Nei (personal communication), with
the exception of our unpublished data on
two Negrito populations of the Philippines.
I n the second genetic distance analysis,
the thr ee Japanese populations mentioned
a b ov e a n d K or e a n s w er e com p a r e d u s in g
gene fr equency data fr om 25 polymor phic
loci. In addition to those mentioned above,
we were able t o use da ta from five polymor-
phic loci: pseudocholinesterase-1, cerumen,
PTC taste sensitivity, color blindness, and INH
inactivator types (Watanabe et al., 1975).
For calculat ion of genet ic dista nces, Nei’s
standar d distance ( Dst ) (Nei, 1972) and themodified Cava lli-Sforza dist an ce (DA) (Nei et
al., 1983) were used (Table 1). To cluster
each set of genetic distances, we employed
the neighbor-joining (NJ ) met hod (Saitou
and Nei, 1987). This approach to tree con-
str uction has been shown to be super ior t o
other methods such as the UPGMA method
(Sokal and Sneath, 1963) in recovering the
tr ue br anching patter n of the genetic r el a-
t ionship of populations (Saitou and Nei,
1987; Nei an d Takezaki, 1994). To evalua te
the branching pattern statistically, we used
the bootstr ap test ( Ef r on, 1982) . This t est
deter mines the pr obability of r epr oducingthe particular branching pattern in the phyloge-
netic tree by u sing ran domly recombined gene
frequency data for the given populations.
RESULTS
In the first set of our calculations, a total
of 26 populations of the wor ld wer e com-
TAB LE 1. Nei’s standa rd genetic distances (lower diagonal m atrix) and m odified Cavalli-Sforza distances(above diagonal m atrix) for the four populations
H on do-J a pa n ese Kor ea n Ain u Ryu k yu a n
H on do-J a pa n ese — 0.00354 0.00747 0.00217Kor ea n 0.00404 — 0.01155 0.00707Ain u 0.00808 0.01043 — 0.00642Ryu kyu a n 0.00336 0.00899 0.00696 —
43 8 K. OMOTO AND N. SAITOU
![Page 3: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/3.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 3/10
pared (Figs. 1, and 2). In both genetic trees,
using the genet ic dist ances D st a n d DA, t h e
African populations are very different from
non-African populations. The bootstrap value
for th is separ ation is very h igh (93–96%).
Among non-Afr icans, the I ndo-Eur opean
population group is first separated from the
r est , t hough the bootstr ap pr obability val-
ues are n ot as high as those showing separa-
tion of the African groups. Further, Native
American popula tions including Esk imos are
separa ted from Asian/Pacific populations.However, with rega rd t o clust ering of Na tive
Austr alian and New Guinean populations,
the two trees show a noteworthy difference
f r o m e a c h o t h e r . I n t h e t r e e b a s e d o n D st
(Fig. 1), th ey ar e cluster ed with the nor th-
east Asian group, which is un expected, given
t he anthr opol ogical view that t he r oots of
Austr alian and New Guinean populat ions
ar e in southeast Asia. I n the tr ee based on
DA, h owever, Austr alians and New Gui n-
eans are clearly separated from other Asian-
Pa cific populat ions (Fig. 2).
Among the rest of the Asian-Pacific popu-
lations, two cluster gr oups may be r ecog-
n i ze d , a l t h ou g h t h e b oot s t r a p v a lu e s for
their separ ation ar e low: nor theast Asian
and southeast Asian/Pacific gr oups. Thr ee
Japanese populations—Korean, Tibetan, and
Mongolian—form the northea st Asian clus-ter gr oup, while souther n Chinese, Thai,
Filipino, Indonesian, Micronesian, Polyne-
sian, an d t wo Negrito populations belong t o
the southeast Asian/Pacific cluster group.
In t he second set of phylogenetic ana lyses,
t h r ee e th n ic gr ou p s of J a p a n (Ain u ,
Ryukyuan, and Hondo- Japanese) ar e com-
Fig. 1. A neighbor-joining (NJ )genetic tree for 26 populations of th e w o rld b a se d o n sta n d a rd g e-neticdistan ce (Dst ).Allele frequencyd a ta o f 2 0 p olym o rp h ic loc i a reused. Numbers along the branchesa re b oo tstra p p ro ba b ility v a lu es
(%).
43 9GENETIC ORIGINS OF JAPANESE
![Page 4: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/4.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 4/10
par ed wit h the Kor ean sample as the con-
trol. We ha d to restr ict our a na lyses to these
samples because we wanted t o maximize the
genet ic data used in the analyses. Only 20
genetic loci are available for the other popu-
lations, but 25 loci have been examined in
t hese thr ee J apanese and the Kor ean popu-
lation samples. The NJ networks show that
Ainu and Ryukyuan samples ar e cluster ed
together in contrast to Hondo-Japanese and
Korean. The two genetic distance measures
used (Dst a n d DA) gave essentially the same
topology, a lthough the bootstr ap values for
the separa tion of the Ainu/Ryukyuan bran ch
wer e sl ightly di ffer ent : 74% f or the one
based on Dst an d 85% for DA (Fig. 3).
DISCUSSION
Background of the dual structure model
Hanihar a’s model is pr imar ily based on
findings arising from st udies on cranial an d
denta l featu res (Han ihara , 1991, 1992). The
basic observation supporting this model is
the morphological similarity ofAinus, and to
some extent of Ryukyuans, to the Neolithic
populations of th e J omon per iod, who ar e
usually r egar ded a s descendants of the Up-
per Paleoli thic population of Japan r epr e-
sented by Minatogawa Man (Howells, 1966;
S u zu k i , 1 98 2; H a n i h a r a , 1 98 4). F u r t h e r -
more, the difference in cranial morphology
between the J omon an d the AeneolithicYayoi
Fig. 2. A neighbor-joining (NJ)genetic tree for 26 populations of the world based on modified Cav-alli-Sforza distance (DA). The genefrequency data used is th e same asi n F ig . 1 . N u m be r s a lon g t h ebranches are bootstrap probability
values (%).
44 0 K. OMOTO AND N. SAITOU
![Page 5: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/5.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 5/10
populations i n t he wester n pa r t of Japa n is
so r emar kable an d occur r ed so suddenly at
about 2,300 years BP that a continuity model
provides insu fficient explan ation. H aniha ra
(1987) also used computer simulation to esti-
mate the number of migrant people who came
to the western part of Japan . He proposed that
over the 1,000 years elapsed since the begin-ning of the Yayoi period (300 BC) and the end of
the Kofun period (700 AD) an unexpectedly
l ar ge number of migr ant peopl e may have
e n t e r e d J a p a n : 1 . 5 m i l l i o n , a s s u m i n g t h e
an nu al growth ra te of th e migran t Yayoi rice
far mer s was 0.2 per cent (Hanihar a , 1987).
Although the validity of this figur e can be
examined th rough more detailed simulation
studies, we think th at H aniha ra’s dual stru c-
tur e model is su perior t o many other models
o n t h e o r i g i n s o f t h e J a p a n e s e t h a t h a v ebeen pr oposed because most of these ar e
rather nonquantitative and speculative.
The str ong advocate for the dual or i gi n
and admixtur e model f or the f or mation of
t h e J a p a n e s e p op u la t i on s , a s w el l a s t h e
idea that the ancestors of the Jomon people
came from south east Asia is Christy Turn er
(Tu r n e r , 1 9 76 , 1 9 87 , 1 9 92 ). H e d e m on -
str ated that east Asian populations can be
divided into two distinct groups on the basis
of dental pattern s: the northeast Asian group
exhibit ‘‘Sinodonty’’an d th e s outh east Asian
group sh ows ‘‘Su nd ad onty.’’Accordin g t o him ,
the Ainu and Jomon populations are sunda-
d on t s , w h er e a s t h e H on d o-J a p a n e s e a r e
sinodonts (Tur ner, 1976).
Mitochondrial DNA (mtDNA) extra cted
from an cient bones a lso seemed t o suggest a
com m on or i gi n for t h e J o m on a n d Ai n u
populations (Hor ai et al ., 1989, 1991). I n
these studies, mtDNA was extr acted f r om
the bones of five Jomonese, carbon 14–dated
a t a b ou t 6 ,0 00 ye a r s B P, a n d s ix e a r ly-
modern Ainu (seventeenth to eighteent h cen-
t u r y AD). A 190 base-pair (bp) segmen t of the
D-loop region of mt DNA was sequ enced, and
the sequences were compared with those of modern individuals from var ious regions of
the wor ld. An identical sequence was ob-
served am ong four J omonese, two Ainu, a nd
t h r e e s o u t h ea s t A s ia n a s w el l a s 1 5 n o n -
Ai n u J a p a n e s e m t D N As . A lt h ou g h H or a i
and his colleagues are cautious about their
conclusions on the origins of Jomonese a nd
Ainu (Hor ai et al ., 1989), th is r esult has
been cited as p ositive evidence for the s outh -
east Asian origin of the first population of
J apa n (Han ihar a, 1991). However, it is worth
noting tha t ther e is a big difference between
the phylogeny of molecules a nd the phylog-
eny of populations. I t is well kn own thatmolecular spli ts may f ar antedate popul a-
tion spli ts (Nei, 1987). I n the molecul ar
genetic tree of Horai et al. (1991), mt DNAs
of non-Ainu J apanese ar e scatter ed ar ound
the whole tree, among mtDNAs of Europe-
ans and Africans, an d do not cluster.
F ig . 3 . T h e N J g e n e tic tre e com p a rin g th re e Ja p a -
n e se p op u la tion s (Ain u , R yu k y u a n , a n d H o n do -Ja p a -nese) with Korean as a control, based on allele frequencydata of 25 polymorphic loci. Genetic distan ce measur esu s e d a r e DA (A) a n d Dst (B). N o te th a t th e b o o tstra pprobability values (85% for A an d 74% for B) a re muchhigher tha n t hose under th e random expectat ion (33%).
44 1GENETIC ORIGINS OF JAPANESE
![Page 6: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/6.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 6/10
The question of northernor southern origins
Contr ar y to the assumpt ions of the dual
stru cture model, classic genetic mark er da tashow that J apanese popul ations, i ncluding
Ainu, h ave definite norther n affinities. Mat-
sumoto (1984, 1988) found th at serum gam-
maglobulin (Gm) types showed a clear-cut
north/south dichotomy of east Asian popula-
t ion s a n d t h a t J a p a n e se b elon ge d t o t h e
nor ther n gr oup. He also showed t ha t J apa-
n e s e p op u l a t ion s a r e r e la t i ve ly h om og e-
neous i n the dist r ibution of Gm allele fr e-
quen cies, which contr adicts Han iha ra ’s dual
stru cture model. Matsum oto’s conclusion,
t hat the homeland of al l Japanese popula-
tions ma y have been in the Lak e Baikal area
in Siberia, h owever, met with strong objec-tion from J apa nese a nth ropologists, includ-
i ng Kazur o Ha ni har a. We also beli eve tha t
conclusions a bout t he origins ofh um an popu-
lations should be based on information from
many genetic loci r ather than a si ngle ge-
net ic locus.
Nei an d Roychoudhu ry (1993) an alyzed 26
populations fr om a r ound t he wor ld u sing
gene fr equency data fr om 29 pol ymor phic
loci of class ic genetic ma rker s an d exam ined
t h e r e s u lt i n g d e n d r og r a m s b y b oot s t r a p
t ests. Among east Asians, t hey compar ed
J apa nese (Hondo-J apa nese), Korean s, Mon-
golian s, Tibetan s, southern Chinese, Thais,
Filipinos, and I ndonesians. They did not
r e cog n iz e n or t h e r n a n d s ou t h e r n cl u st e r
gr oups and used the tr adit i onal ter m Mon-
goloid to denote all east Asian and Pacific
populations.
Recently, Nei published phylogenetic tr ees
t h a t s h ow t h e r e la t ion s h ip s of J a p a n e se
populations (Nei, 1995). This time, he com-
p a r e d Ai n u , ‘‘J a p a n e s e of O k in a w a , ’’
(Ryuk yua n) an d ‘‘J ap an ese ofTokyo’’(Hond o-
J apa nese) with neighboring populations u s-
ing dat a from 18 polymorphic loci. He sh owed
t h a t t h e t h r ee J a p a n e se p op u la t ion s a r ecl ose t o Kor ean, while souther n Chinese,
N a t iv e Ta i wa n e s e, T h a i, a n d F il ip in o
samples cluster as a separate group. On the
basis of thi s r esult , he chal lenged Han iha-
r a’s hypothesis as a whole with r espect to
t he or igi ns and the dual str uctur e of Japa-
nese populations.
Omoto (1995) examined the pr oblem of
the origins of Ainu in detail. He compared
Native Australian, New Guinean, Microne-
sian, an d Polynesian populations plus 11
east Asian populations on the basis of NJ
t r e e s w i t h DA distances, using classic ge-
netic m ar ker s fr om 23 polymor phic loci .
Ainu clear ly f ell in to the nor theast Asian
group, with Korean , Mongolian , and Tibetan
samples, while south ern Chinese, Thai, Fili-
pino, Indonesian, a nd two Negrito populations
were linked with the southeast Asian group.
The over all Hondo-J apanese admixtur e
ra te in t he original sam ple of about 500 Ainu
fr om the Distr ict of Hidaka on Hokkaido,
which is the n or ther nmost island of Japan,
was estimated to be approximately 30–40%
(Omoto, 1972). Omoto (1995), however, deter-mined the impact of admixture on the posi-
t ion of the Ainu population in the tr ee by
correcting gene frequencies for admixtur e.
No change occur r ed even when an admi x-
t u r e e s t im a t e of 4 0 % w a s e m p loy ed a n d
Ainu remained in the northeast Asian clus-
ter group. However, the bra nch length t o the
Ainu elongated considerably. When a hypo-
thetical admixtur e r ate of 60% was used t o
correct gene frequencies, the Ainu remained
in the north east Asian cluster group, but the
N a t i v e A u s t r a l ia n / N e w G u i n e a n c lu s t e r
m ov ed n e a r t h e r oot of t h e l on g b r a n ch
leading to the Ainu. The genetic r elation-ships of other populations were essentially
unchanged. Although i t is not possibl e to
draw a conclusion from such a simulation, it
lends support the view that Ainu a nd Na tive
Austr alians/New Guinean s both ar e derived
directly from the Upper Pa leolithic popula-
tions of eas t Asia (Omoto, 1995). The p ossibil-
ity tha t Ainu origins lie in a ncestral popula-
tion groups in northeast Asia has also been
shown r ecently by a DNA-based stu dy of
HLA Class II genes (Banna i et a l., 1996).
O u r s t u d y c o n fi r m s t h a t t h r e e J a p a n e s e
populations, including Ainu and Ryukyua n,belong to the northeast Asian group, along
with Kor eans, Mongolians, and Tibet ans.
T h is g r ou p a p p ea r s s e pa r a t e d fr om t h e
southeast Asian group comprising southern
Chinese, Thai, Filipino, Indonesian, Microne-
sian, Polynesian, and two Negr ito popul a-
tions, although the bootstrap probability for
44 2 K. OMOTO AND N. SAITOU
![Page 7: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/7.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 7/10
t his separ ation i s r a ther l ow and i s statist i-
cally not significant (22% in Fig. 2).
With regard to th e origins of the Mongol-
oid populations, the current ly popular view
assumes th at southeast Asia was t he centerof dispersa ls (e.g., Turn er, 1995). This view,
which is based primarily on morphological
traits, was examined by Omoto (1995). He
considered genetic evidence and favored a
model which assumes independent origins
of n or t h e a s t As ia n , s ou t h e a s t As ia n , a n d
Nat ive Amer ican population gr oups. Nei
( 1995) seems to be skept ical about the di-
chotomy of northern and southern Mongol-
oi d g r ou p s b eca u s e t h e b r a n ch i n g y ie ld s
relat ively low bootst ra p valu es. However, in
view of the genetic, archaeological, and lin-
guistic evidence for large-scale dispersals of
northern groups into south east Asia and th e
Pacific dur ing the l ast 10,000 year s (e.g.,
Bellwood, 1979, 1996), we ass um e as a work-
i ng hypothesis t h at the dichot omy in our
genetic tr ee (Fig. 2) ha s h istorical validity.
H a n i h a r a ’s v ie w t h a t t h e U p p er P a l eo-
lithic population of Japan originated some-
w h er e i n s o u t h e a st As ia m a y h a v e b e e n
based par tly on the obser vation of Suzuki
(1 98 2 ) o n M in a t o ga w a M a n (H a n i h a r a ,
1991) . Suzuki consider ed that the skull of
Mi natogawa No. 1 dated at appr oximately
18,000 years BP is morphologically similar
toJomonese. He further assumed that Mina-togawa Man is more closely related to Liu-
ja n g M a n of s ou t h e rn C h in a t h a n t o t h e
specimen No. 101 of Upper Ca ve of Zhoukou-
dian, nor ther n China. However , we doubt
that observations on a single skull can pro-
vide all that needs to be known about or i-
gins. After examining our genetic data, we
propose a count erhypothesis tha t th e Upper
Paleolithic populations of Ja pan are derived
fr om th ose of nor theast Asia and did not
necessar ily or iginat e in southeast Asia. A
critical exam ination of the t hree specimens
from Upper Cave are pa rticularly importan t
in this regard.We believe tha t our h ypothesis is more in
agreement with preh istoric evidence th an is
Han ihara ’s. According to most archa eolo-
g is t s i n J a p a n , s t on e -t ool cu l t u r es of t h e
Upper P aleolithic an d t he s uccessive Jomon
period show definite n orther n a ffinities (Cho-
suke Ser izawa, per sonal communication) .
The m icroblade tra dition accompanying th e
Ar aya type bur in was widely distr ibuted
fr om easter n Siber ia to the Ja panese ar chi-
pelago from t he postglacial period u ntil t he
beginning of the J omon per iod (appr oxi-mat ely 20,000–12,000 years BP). No stone-
tool culture of southeast Asian affinities has
been discover ed in J apan for this per iod
(Kimur a, 1993; Tan ak a et a l., 1995).
Dual structure of the Japanesepopulations
O m ot o (1 97 2, 1 99 2) a n d O m ot o a n d
Misawa ( 1976) have shown that Ainu and
Ryukyuan peoples ar e genetically r ath er
similar to each other but are different from
Hondo-Japanese. This finding seemed to fit
well with th e admixtu re model for the form a-tion of the Japanese populations. I n these
previous reports, however, the dichotomy of
Ainu/Ryukyuan and Hondo-J apanese was
emphasized, without the relationship being
tested statist ically. I n the pr esent study,
three Japanese populations are compared to
each other, with Korean as the control on the
basis of the lar gest data set ever used: 25
polymorphic loci. Also, the separa tion be-
tween the Ainu/Ryukyuan cluster and the
Hondo-J apanese/Kor ean gr oup in the NJ
network was stat istically evaluat ed by boot-
strap probabilities (Fig. 3). While two ge-
netic distance measur es, D st a n d DA, gaveessentially th e sa me topology, th e bootstrap
value is sl ightly higher (85%) in the tr ee
u s in g DA t h a n t h a t u s in g D st (74%). Al-
though this difference in bootstrap values is
statistically not significant, we have noted
t ha t DA is s u pe r ior t o Dst in pr oducing
reliable genetic relationships of populations
(Figs. 1, 2). Since th e expected value of the
bootstrap probability for four populations
un der no genet ic relationsh ip (‘‘sta r p hylog-
eny’’) is 33%, the observed valu es a re mu ch
higher than this r andom expectation. Re-
cently, theoretical studies on the bootstrap
test have been extensively conducted (e.g.,S it n i k ov a e t a l ., 1 99 5; Zh a r k i k h a n d L i,
1995; Efron et al., 1996). However, it is still
not clear how bootstr ap pr obabilit ies for
g en e t ic t r e e s s h ou l d b e t r e a t e d w i t h t h e
usual statistical tests. Therefore, we do not
consider reliance on bootstrap probabilities
as critical. In any case, we believe that our
44 3GENETIC ORIGINS OF JAPANESE
![Page 8: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/8.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 8/10
r esul ts give suppor t f or the separ ation be-
tween Ainu/Ryukyuan and Hondo-J apan ese/
Korean cluster s. If this clustering is r eal, it
i n tur n gives a par tial suppor t f or Haniha-
r a’s dual str uctur e model, in t er ms of theduality of the J apan ese populations.
It is worth n oting th at our conclusions are
different from those of Nei (1995). He pub-
lished a phylogenetic tree ba sed on 18 poly-
morphic loci th at shows th e relat ionsh ips of
Ain u , H on d o-J a p a n e s e, a n d O k in a w a n
(Ryukyuan) with neighboring populations.
He expr essed his doubt about Hanihar a ’s
dual structure model as a whole (Nei, 1995).
However , when t he hypothesi s i s br oken
into two components, the evidence, we sug-
ge st , s h ow s t h a t t h e re m a y i n fa ct b e a
duality in J apa nese populations. This differ-
e n ce i n i n t er p r e t a t ion w il l b e cl ea r e d b y
f ut ur e studi es wi th much mor e and better
d a t a . A t t h i s m o m e n t , w e t h i n k t h a t t h e
deta ils of a populat ion’s origin ar e un likely
to be revealed by using a worldwide data set
such as used by Nei (1995). To understand
local biological history, one has to examine
the genetic profiles of the populations that
have been hypothesized to be an cestra l to it
a n d t o d o t h is w it h a s la r ge a n u m be r o f
genetic loci as possible. Of course, we do not
consider our findings to be final, particularly
becau se of our relatively low bootstrap val-
ues. Pr obably, our stu dy represent s a micro-scopic rather than a macroscopic approach
in reconstructing the population history of
J a p a n .
Recently, a d eta iled genetic compa rison of
As ia n p op u l a t ion s , i n cl u d in g Ai n u a n d
R yu k y u a n , w h ich u s e d m t D N A s e qu e n ce
data was pu blished (Horai et al., 1996). The
r e s u lt s fr om t h i s s t u d y a r e i n a g r ee m en t
with ours in th at H ondo-Japa nese and Kore-
ans are genetically very close to each other
and the admixtur e model of the forma tion of
t he moder n J apanese is f avor ed. However ,
with regard to the relationship between the
Ai nu and the Ryukyuan, no evidence f or acommon ancestr y is f ound. Fut ur e studies
a r e n e ed ed t o cle a r u p t h e p os it ion of
Rykyuans.
Fina lly, when origins a re considered, ques-
tions about linguistic evidence commonly
ar ise. Our view can be summar i zed as f ol-
lows. If a lar ge number of migrant s to J apa n
actually came from northeast Asia, perhaps
v ia t h e K or e a n p e n in s u l a s t a r t in g a r ou n d
2,300 year s BP, a s Hanihar a ’s dual st r uc-
tur e model maintains, why is the language
of Hondo-Japanese (i.e., Japanese) so differ-ent from Korean or other Altaic langua ges?
In our view, language chan ges during popu-
lation mixture are complex social phenom-
ena that cannot be simply defined or mea-
sur ed chr onologically, as can biological
phenomena. Given the almost complete ab-
sence of quan titative and stat istical compari-
sons between J apa nese (including Ainu a nd
R yu k y u a n ) a n d ot h e r l a n gu a g es i n As ia ,
except for th e work of Biten Yas um oto, which
h a s n o g en e ra l s u pp or t a m on g J a p a n e se
l in g u is t s , w e h a v e a t p r e se n t n o c le a r a n -
swer, either pro or con, to our question. It is
in t er e st in g t o n ot e t h a t , on t h e b a sis of
stat istical compar isons of basic words a mong
east Asian langua ges including the Ainu, h e
r eached the conclusion that Ainu and Ko-
r ean languages may have common ancient
a n ce s t r y i n w h a t h e ca l ls t h e P a l eo-F a r -
E a s t e r n la n g u a ge g r ou p (Ya s u m ot o a n d
Honda , 1978; Yasu moto, personal commu ni-
ca t ion ). F or t h is r e a son , w e d o n ot p u t
e m p h a s is on l a n gu a g e n ow a n d h a v e f o-
cussed our a tt ent ion on genet ic, morphologi-
cal an d a rcha eological evidence.
CONCLUSIONS
We propose a modified version of the dual
str uctur e model of the or igins of Japanese
populations. Two fundam ental populat ion
g r ou p s a r e a n c es t r a l t o m od e r n J a p a n e s e
populations (Ainu, Ryukyuan, and Hondo-
Japanese) . One gr oup, r epr esented by the
J omonese wh o gave rise to the modern Ainu
a n d p r ob a bl y a l s o t h e R yu k y u a n p op u la -
tions, ha s its origin in the Upper P aleolithic
populations of northea st Asia, which were
not necessarily derived from southeast Asia.
The other group is the later migrants of the
Yayoi an d Kofun per iods who also came from
northeast Asia but were different geneticallyand mor phologically fr om the fir st gr oup.
I nter mixtur es occur r ed between these t wo
g r ou p s , b u t t h e g en e t ic i n flu e n ce of t h e
second group is pr edominan t in the m ajority
of m od e r n J a p a n e s e (H on d o-J a p a n e s e ).
Clearly, furt her stud ies are necessary which
include more informa tion, part icularly on
44 4 K. OMOTO AND N. SAITOU
![Page 9: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/9.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 9/10
a n cien t a n d m od er n DN As of Ain u ,
Ryukyuan, and other popul ations of J apan
and east Asia, to clar i f y the or i gi ns of the
Japanese people.
ACKNOWLEDGMENTS
We th an k P rofessor Masat oshi Nei of The
Pennsylvania State Univer sity ( St ate Col-
l ege, PA) for pr oviding u s with gene fr e-
quency data. We also than k two anonymous
reviewers who read the original man uscript
critically and gave us invaluable commen ts.
Thanks ar e also due to Dr . Akiko Uchida,
Chiba University, for help in preparing the
origina l man uscript.
LITERATURE CITED
B a n n a i M, T o k u n a g a K , Im a n ish i T , H a rih a ra S , F u - jisawa K, Juji T, and Omoto K (1996) HLA class IIalleles in Ainu living in Hidaka district, Hokkaido,northern J apan. Am. J. Phys. Anth ropol. 101:1–9.
Bellwood P (1979) Man’s Conques t of th e Pa cific. NewYork: Oxford U niversit y Pres s.
Bellwood P (1996) Early agriculture and the dispersal of the south ern Mongoloids. In T Akazawa an d E Szat h-mary (eds.): Prehistoric Mongoloid Dispersals. NewYork: Oxford U niversit y Pr ess, pp. 287–302.
Efron B (1982) The Jackknife, the Bootst rap and Ot herResampling Plans. Philadelphia: Society for Indus-trial a nd Applied Mathematics.
E fro n B , H a llo ra n E , a n d H o lm e s S (1 9 9 6 ) B o o tstra pconfidence levels for phylogenetic trees. Proc. Nat l.Acad. Sci. U. S. A. 93 :7085–7090.
Goedde HW, Benkmann H-G, Kriese L, Bogdanski P, DuR, Chen L, Cui M, Yuan Y, Xu J , Li S, an d Wang Y(19 8 4) P o pu la tion g en e tic stu d ie s in th re e C h in e seminorities. Am. J . Ph ys. Anth ropol. 64 :277–284.
Haniha ra, K (1984) Origins and a ffinities of Japanese a sviewed from cranial measurements. Acta Anthropo-genet. 8:149–158.
Hanihara, K (1987) Estimation of the number of earlym ig ra n ts to Ja p a n : A sim u la tiv e stu d y . J. A n th ro p .Soc. Nippon 95 :391–403.
Haniha ra, K (1991)Dua l structur e model for th e popula-tion history of the J apanese. J apan Review 2:1–33.
Hanihara K (1992) Dual structure model for the forma-tion of the Japanese population. In K Hanihara (ed.):J a p a n e s e a s a M em b er of t h e As i an a n d P a ci ficPopulations. Int ernat ional Symposium Vol. 4, Kyoto:Intern . Res. Cent. J apan. St ud., pp. 244–251.
Horai S, Hayasaka K, Murayama K, Wate N, Koike H,and Nakai N (1989) DNA amplification from ancienthuman skeletal remains and their sequence analysis.Pr oc. J pn. Acad. 65(S er. B ):229–233.
H o ra i S , K o n d o R, Mu ra y a m a K , H a y a sh i S , K o ik e H ,and Nakai N (1991) Phylogenetic affiliation of ancienta n d co n tem p o ra ry h u m a n s in fe rre d from m ito ch o n -drial DNA. Philos. Trans. R. Soc. Lond. [Biol.] 333:409–417.
H o ra i S , Mu ra y a m a K , H a y a sa k a K , Ma tsu b a y a sh i S ,Hat tori Y, Fucharoen G, Hariha ra S, Pa rk K-S, OmotoK, and Pan I-H (1996) mtDNA polymorphism in eastAsian populations, with special reference to the peo-pling of Japa n. Am. J . Hum. Genet. 59:579–590.
Howells, WW (1966) The J omon population of J apan: Astudy by discriminant analysis of Japanese and Ainu
crania. Pa p. Peabody Mus. Archeol. Eth nol., Har vardUniv., 57:1–4.
Kanaseki T, Na gai M, a nd Sano H (1960) Craniologicalstudies of the Yayoi-period a ncients, excavated at theDoigahama site, Yamaguchi P refectur e. Jinr uigaku
Kenkyu, 7(Suppl.):1 – 36 . (in Ja p a n e se w ith E n g lishsu m m a ry )
K im u ra H (ed . ) (1 9 93 ) T h e O rig in a n d D ispe rsa l o f Microblade Industry in Norther n Eu rasia. Pr oc. Inter-national Conference, Sapporo Un iv. Archeol. Mu s,Sapporo.
Matsum oto H (1984) On t he origin of the J apanese r ace.Studies of genetic markers of the immunoglobulins.Pr oc. J pn. Acad. 60:211–216.
Matsumoto H (1988) Characteristics of Mongoloid andneighboring populations based on the genetic mar kersof h u m a n im m u n o glob u lin s. H u m . G e n et. 80 :207–218.
Nei M (1972) Genetic distances between populations.Am. Nat. 106:283–292.
Nei M (1987) Molecular Evolutionary Genetics. NewYork: Columbia Un iversity Pr ess.
N e i M (19 9 5) T h e orig in s o f h u m a n p op u la tion s: G e -netic, linguistic, and a rcheological dat a. In S Brenn er
and K H anihar a (eds.): The Origin an d Past of ModernH u m a n s a s Vie we d from D N A. S in ga p ore : WorldScientific, pp. 71–91.
N e i M, a n d R oy ch o u dh u ry A K (19 9 3) E v olu tion a ryrelationships of human populations on a global scale.Mol. Biol. Evol. 10:927–943.
N e i M, a n d Ta k e z a k i N (19 9 4) E st im a tion of g e n e ticdistances and phylogenetic trees from DNA analysis.In: Proceedings of the 5th World Congress on GeneticsApplied to Livest ock P roduction, Vol. 21. pp. 405–412.
N e i M, T a jim a F , a n d T a te n o Y (1 9 83 ) A ccu ra c y of estimat ed phylogenetic trees from molecular dat a. II.Gene frequency data. J. Mol. Evol. 19 :153–170.
Omoto, K (1972) Polymorph isms and gen etic affinities of the Ainu of Hokkaido. Hum. Biol. Oceania 1:278–288.
Omoto, K (1992) Some as pects of the genetic composi-tion of the Ja panese. In K Haniha ra (ed.): Japan ese asa Member of Asian and Pacific Populations. Interna-tional Symposium Vol. 4, Kyoto: Intern. Res. Cent.
Ja pan. Stu d., pp. 138–145.Omoto K (1995) Genetic diversity a nd t he origins of the‘‘Mongoloids.’’ In S Brenner and K Haniha ra (eds.):T h e O rig in a n d P a st o f Mo d e rn H u m a n s a s V ie w e dfrom DNA. Singap ore: World Scientific, pp. 92–109.
Omoto K, and Misawa S (1976) The genetic relations of th e Ainu. In RL Kirk an d AG Thorne (eds.): The Originof th e A u stra lia n s. C a n b e rra : T h e A u stra lia n In st.Aborigina l Stu d., pp. 365–376.
Omoto K, Aoki K, and Harada S (1975) Polymorphism of esterase D in some population groups in Japan . Hum.Hered. 25:378–381.
Roychoudhury AK, and Nei M (1988) Huma n Polymor-phic Genes: A World Distribution. Oxford: OxfordUniversity Press.
Saitou N, and Nei M (1987) The neighbor-joining meth -od: A new method for constructing phylogenetic trees.Mol. Biol. Evol. 4:406–425.
Sitnikova TL, Rzhetsky A, and Nei M (1995) Interior-branch and the bootstrap tests of phylogenetic trees.Mol. Biol. Evol. 12:319–333.
Sokal R, and Sneath PAP (1963) Principles of NumericalTaxonomy. San Francisco: WH Freeman.
Suzuki H (1969) Microevolutionary changes in the Japa-nese population from the prehistoricage to the present-day. J. F ac. Sci. Un iv. Tokyo, Sec. V3, 279–308.
Suzuki H (1982) Skulls of the Minatogawa Man. In HSuzuki and K Ha nihar a (eds.): The Minatogawa Man:The Upper Pleistocene Man from the Island of Oki-nawa . Bull. Un iv. Mus., U niv. Tokyo, 19, pp. 7–49.
44 5GENETIC ORIGINS OF JAPANESE
![Page 10: Genetic Origins of the Japanese a Partial Support](https://reader031.fdocuments.net/reader031/viewer/2022021200/577d22591a28ab4e1e972438/html5/thumbnails/10.jpg)
8/3/2019 Genetic Origins of the Japanese a Partial Support
http://slidepdf.com/reader/full/genetic-origins-of-the-japanese-a-partial-support 10/10
T a n a k a M, S a sa k i K , a n d S a g a w a M (1 9 9 5 ) Ja p a n e searchaeology. In: An Intr oductory Bibliograph y forJa panese Stu dies, Vol. 9, Part 2: Human ities 1991–92.Tokyo: Japan Foundation, pp. 1–30.
Turner CG II (1976) Dental evidence on the origins of
the Ainu and Ja panese. Science 193:911–913.Tu rn e r C G II (19 8 7) L a te P le isto ce n e a n d H o loc en epopulation h istory of East Asia. Am. J . Phys. Anthr o-pol. 73:305–321.
Tu rn e r C G II (19 9 2) S u n d a d o n ty a n d S in od on ty inJa p a n : T h e d e n ta l b a sis for a d u a l o rigin h y p oth e sisfor th e p e op lin g o f th e J a p a n e se Isla n d s. In K H a n i-h a ra (e d . ): Ja p a n e se a s a Me m b e r o f th e A sia n a n dPacific Populations. In terna tional Symposium Vol. 4.Kyoto: Intern. Res. Cent. J apan. Stud., pp. 96–112.
Turn er CG II (1995) Shifting continuity: Modern huma nor i gi n . I n S B r en n e r a n d K H a n i h a r a (e ds .): T h eOrigin and Past of Modern Humans as Viewed fromDNA. Singapore: World Scientific, pp. 216–243.
Ueda S, Omoto K, Park KS, and Kudo T (1979) Polymor-phism of red cell glutamic-pyruvic transaminase in
Ja panese: Gene frequencies in sam ples from norther nJa p a n . H u m . H e re d. 29 :208–212.
Watanabe S, Kondo S, and Matsunaga E (eds.) (1975)
Anthropological an d Genetic Studies on t he J apanese,JIBP Synthesis 2. Tokyo: University of Tokyo Press.
Yasum oto B, and Honda M (1978) Nihongo no Tanjo
[The Birth ofth e J apanese Languages]. Tokyo:Daishu-kan (in Japanese).
Yua n Y, Du R, Ch en L, Xu J , Cui M, Wang Y, Li S, GoeddeH W, B e n k m a n n H -G , K riese L , a n d B og d a n sk i P
(1984) Distribution of eight blood group systems andAB H se cre tion in Mon g olia n , K ore a n , a n d Z h u a n gnationalities in Ch ina. Ann. H um. Biol. 11:377–388.
Zhark ikh A, and Li W-H (1995) Estima tion of confidencein phylogeny: Complete-and-partial bootstra p tech-nique. Mol. Phyl. E vol. 4:44–63.
44 6 K. OMOTO AND N. SAITOU