College of Animal Science and Technology, Sichuan Agricultural University, Ya an, Sichuan, 0,/*+., China

http://www.jstage.jst.go.jp/browse/jpsa doi: +*.,+.+/jpsa.**3*.1 Copyright,*+*, Japan Poultry Science Association. + Bing Zhou, Shi-Yi Chen, Qing Zhu, Yong-Gang Yao and Yi-Ping Liu + +,, +, + College of Animal Science and Technology, Sichuan Agricultural University, Ya an, Sichuan, 0,/*+., China, Key Laboratory of Animal Models and Human Disease Mechanisms of Chinese Academy of Sciences & Yunnan Province, Kunming Institute of Zoology, Kunming, Yunnan, 0/*,,-, China Silkie is a famous black-bone chicken breed with beautiful silky feather. The unique medical property of this chicken was recorded in Chinese traditional medicine dictionary about 1** years ago. In this study, we analyzed the mtdna D-loop sequence variation of,0 Bairong Silkies from Fujian Province, China, together with +** reported Silkie mtdnas from China and Japan, and studied their matrilineal components and genetic relationship. A total of,+ haplotypes were detected, which could be assigned to six haplogroups (A-E, G). Among them, haplogroups D and G were exclusively presented in Japanese Silkies and Chinese Silkies, respectively. Chinese Silkies had higher frequency of lineages belonging to haplogroups A, B, and E, and lower frequency of haplogroup C than Japanese Silkies. For the four Chinese Silkie populations, most of samples of Taihe, Chengdu, and Hubei Silkies were grouped in haplogroups A, B, and C, whereas most of Bairong Silkies were grouped in haplogroup E. Five haplotypes were shared by Japanese and Chinese Silkies. The genetic diversity of each Silkie population varied, but the overall diversity of Chinese Silkies was similar to that of Japanese Silkies. Taken together, our results confirmed the genetic connection between Chinese and Japanese Silkies, but also clearly showed that the matrilineal genetic structures of Chinese and Japanese Silkies had some di# erences. Key words: d-loop, genetic diversity, matrilineal origin, mtdna, silkies J. Poult. Sci.,.1:,,,1,,*+* Introduction Although most of domesticated chickens are used for meat consumption and egg production, the ornamental The modern domestic chickens are now distributed types have been prevailed worldwide today. Among them, worldwide and have evolved along with the development the Silkie ( Gallus gallus domesticus) is well known for its of human civilization. The origin and phylogeographic unique flu# y plumage feeling like silk. In China, Silkie is history of chickens across Eurasia have been extensively specially characterized by its unique medical property, elucidated based on mitochondrial DNA (mtdna) se- which was recorded in Chinese traditional medicine dicquence variation (Fumihito et al., +33., +330; Liu et al., tionary about 1** years ago (Xie, +33/ ). There is no clear,**0a). The wildly accepted concept is that the progeni- distinction between the Silkies and other chicken breeds at tor of domestic chickens is red jungle fowls ( Gallus the matrilineal component level (Fu et al.,,**,; Liu et al., gallus), and the highly divergent mtdna haplogroups in,**0a). The exact scenario regarding where or when domestic populations would suggest for multiple domes- Silkies with the fur-like plumage were first cultivated was tications in di# erent regions, such as Yunnan, South and unknown. Most of the existing conventional literatures Southwest China and/or surrounding areas, as well as the suggested for an origin in China. Other places in South- Indian subcontinent (Liu et al.,,**0a). As the initially east Asia, such as India, have also been proposed as the domesticated chickens were adapted to new environments cultivation center for the Silkie breeds (Arisawa et al., and were subjected to di# erent artificial selection pres-,**0). The domestication history of chicken in China sures, populations with new characteristics were de- could be dated back to /.** BC (West and Zhou, +323), veloped and further formed local breeds (Zárate et al., but the first detailed description for the Silkies was,**0). recorded only 1** years ago (Xie, +33/ ). Nowadays, there are eight Chinese Silkie breeds/populations registered in the Domestic Animal Diversity Information Received: June.,,**3, Accepted: October /,,**3 Released Online Advance Publication: December +*,,**3 Correspondence: Prof. Y. Liu, College of Animal Science and Technology, System (DAD-IS,,**1) of the Food and Agriculture Sichuan Agricultural University, Ya an, Sichuan province, 0,/*+., China. (E-mail: liuyp /12@yahoo.com) Organization of the United Nations (http://www.dad. fao.org/). All these breeds/populations could be assorted

Zhou et al. : Matrilineal Components of Silkies 23 into white feather type and black feather type based on the described in our previous study (Liu et al.,,**0a) and feather color (Qiu et al., +322). Among them, the Taihe were deposited in GenBank under accession numbers GQ Silkies and Bairong Silkies, which are distributed in,,133--gq,,2*+2. Jiangxi Province and Fujian Province, respectively, are The previously reported sequences (,1 Taihe Silkies thought to be the ancestor populations of modern Silkies from Jiangxi Province, +/ Chengdu Silkies from Sichuan in China. Subsequently, these two breeds were di# used Province,,. Hubei Silkies from Hubei Province, China; into other regions (Qiu et al., +322 ; Xie, +33/ ). Unfortu- Liu et al.,,**0a) were also used for comparison in this nately, there is no strong evidence, especially for molecu- study. Thirty-four Japanese Silkies were retrieved from lar data, to support this hypothesis. GenBank. All together, +,0 mtdna D-loop partial se- Silkie is one of the most popular and ubiquitous orna- quences were analyzed in this study (Table + ). mental breeds in Japan. Silkie is called Ukokkei in Data Analysis Japanese and is mainly distributed in Tokyo, Mie, Osaka, The mtdna sequences were aligned with the reference Hiroshima, Yamaguchi and Kagawa Prefectures (Chang sequence of G. g. domesticus mitochondrial genome and Huang,,**-). There are about /* native chicken (Desjardins and Morais, +33* ). Sequence variations were breeds in Japan, which could be divided into entertain- exported using MEGA -. + (Kumar et al.,,**.). We folment, egg, and/or meat production groups. The enter- lowed the classification system in our previous report (Liu tainment group could be further classified into two sub- et al.,,**0a) to classify each mtdna into respective groups according to their introduction history, with one haplogroups/clades. An unrooted Neighbor-joining (NJ) being introduced to Japan more than,, *** years ago and tree of the haplotypes was subsequently constructed using the other one during the early Edo Era ( +0*- +201) from the Kimura two-parameter model in MEGA -. +, follow- China or India (Chang and Huang,,**-; Arisawa et al., ing the same rationale as discussed in our previous study,**0). Based on these founder breeds, other Japanese (Liu et al.,,**0a). To provide a better view of the genetic breeds were subsequently developed via artificial selection relationship of the haplotypes in these geographic populaprograms by the end of the Edo Era (Arisawa et al., tions of Silkies, a median-joining network was constructed,**0). Komiyama and colleagues (,**-,,**.) suggested using program Network -. + (http://www.fluxusthat all Japanese domesticated chickens were originated engineering.com/sharenet.htm; Bandelt et al., +333). Fiindependently from Southeast Asia and the mainland nally, the haplotype diversity (h) and nucleotide diversity China. Our later reanalysis of the gamecock mtdnas in ( p) for each Silkie population (Nei, +321) were estimated China and Japan demonstrated a genetic pattern consis- using DnaSP.. +* (Rozas et al.,,**-). tent with the proposed dual origin of Japanese gamecocks, but also left room for a single origin of Japanese gamecocks from China (Liu et al.,,**0 b). Recently, Oka et al. Results A total of,1 polymorphic sites, including,0 transitions (,**1) provided further mtdna data to support the and one transversion, were found in +,0 mtdna semultiple origins of Japanese native chickens. quences (each with a length of /+3 bp). Twenty-one hap- Mitochondrial DNA has been proved to be a very useful lotypes were defined and all of them could be found in the genetic marker in elucidating the origin of domestic data set reported by Liu et al. (,**0 a) (Fig. + ). The,+ animals, such as duck (He et al.,,**2), Cattle (Lai et al.,,**0; Chen et al.,,**2), and Goat (Liu et al.,,**3). The domestication history and phylogeographical relationship of modern chickens in China and Japan have been elaborately explored by using mtdna sequence variations haplotypes could be classified into six haplogroups (A-E, G). Briefly, six haplotypes belonged to haplogroup A, two haplotypes in haplogroup B, two haplotypes in haplogroup C, three haplotypes in D haplogroup, seven haplotypes in haplogroup E, and one haplotype in haplogroup G. (Fumihito et al., +33., +330 ; Komiyama et al.,,**-, Among them, haplotypes A,, C +, B +, E +, and B+- were,**.; Liu et al.,,**0 a, b; Oka et al.,,**1), as well as, dominant and were shared by,-,,,,,+,,*, and +, microsatellite DNA polymorphisms (Takahashi et al., samples, respectively. Nine haplotypes occurred in one +332). However, the matrilineal components of the Silkie breeds distributed in China and Japan and their genetic relationship have not been su$ ciently characterized or discussed. In this study, we compared the mtdna control sample. Each of the seven haplotypes (A 0, A ++, A +2, D 0, D +-, E 0, and E 3 ) occurred in,. samples. Haplotypes A, and B+ were shared by individuals from all four Chinese Silkie populations and one Japanese Silkie population. region (D-loop) partial sequence variation of the However, haplotypes E+ was shared by +3 individuals Silkies from China and Japan. Our results provided insightful exclusively from the Bairong Silkies and one Japanese information regarding the breed formation histo- Silkie. When we counted the number of haplotypes ac- ry and subsequent di# usion route of the Silkie. Materials and Methods Sampling and Reported Data We collected,0 Bairong Silkies from Fujian Province. These samples were sequenced according to the procedure cording to the Silkie populations, +- haplotypes were detected in Japanese Silkies, five haplotypes in Bairong Silkies, seven haplotypes in Taihe Silkies, five haplotypes in Chengdu Silkies, and five haplotypes in Hubei Silkies. The phylogenetic relationship among the six haplogroups was displayed in the NJ tree (Fig., a), and was

24 J. Poult. Sci.,.1 ( + ) Table +. Sample information and genetic diversity of Silkies GenBank Haplotype Nucleotide Breed No. Location Accession Reference diversity diversity number Bairong Silkie,0 Fujian, China GQ,,133- -GQ,,2*+2 *4.0, *4++. *4**/*3 *4**+2, This study AY.0/322 -AY.0/33+; Taihe Silkie,1 Jiangxi, China AF/+,,,+ -AF/+,,--, *42-2 *4*-. *4*++,- *4**+*3 Liu et al.,,**0a AF/+,,1- -AF/+,,2, Chengdu Silkie +/ Sichuan, China AF /+,*0*, AF/+,*0, -AF/+,*1/ *42.2 *4*0+ *4**2++ *4**,*, Liu et al.,,**0a Hubei Silkie,. Hubei, China AF/+,+23 -AF/+,,*3; AF/+,,-. -AF/+,,-0 *410. *4*., *4**2+, *4**+.0 Liu et al.,,**0a Subtotal in China 3, *42/+ *4*+/ *4*++02 *4***/, AB **11--, AB **11-/, AB **11.*, AB **11.0, Data Japanese Silkie -. Japan AB **11.1; deposited in *42.+ *4*/0 *4*+,+3 *4**+*0 AB ++.*1*, AB ++.*1+ ; GenBank by AB,0-3.1 -AB,0-31*, Miyake, T. AB,0-31. -AB,0-310 Total +,0 *422, *4*+, *4*+,32 *4***-* Fig. +. mtdna sequence variations of,+ haplotypes identified in +,0 Silkie samples collected in this study and from published sources. The haplotypes were aligned to G. g. domesticus complete mtdna sequence (Desjardins and Morais, +33* ) and were defined according the classification system in our previous study (Liu et al.,,**0a). The number of individuals sharing the same haplotype in di# erent populations is listed below the abbreviations for each population (BR-Bairong Silkie; TH-Taihe Silkie; CD-Chengdu Silkie; HB-Hubei Silkie; JP-Japanese Silkie). The column T shows the total number of individuals shared each haplotype. Dots ( ) denote the identity with the reference sequence. Short lines ( ) represent the absence of certain haplotype in the population. further demonstrated by a network graph (Fig., b). The two samples belonging to haplogroups E and G, respectopology of the tree based on the Silkie haplotypes was tively. In contrast,,, Bairong Silkies ( 2.. 0 ) from consistent with the pattern observed in a large domestic Fujian, China, were distributed in haplogroup E, while chicken data set (Liu et al.,,**0a). Among the -. four samples in haplogroups A, B, and C. Among the six Japanese Silkies, +- samples were distributed in haplo- haplogroups, D and G were exclusively found in Japanese groups D and E, whereas the remaining,+ samples could Silkies and Chinese Silkies, respectively. The frequencies be classified into haplogroups A, B, and C. For the Taihe, of haplogroups A, B, and E were higher in Chinese Silkies Chengdu, and Hubei Silkies, nearly all samples were dis- than in Japanese Silkies, but the frequency of haplogroup tributed in haplogroups A, B, and C, with the exception of C was higher in Japanese Silkies (Table,). Furthermore,

Zhou et al. : Matrilineal Components of Silkies 25 Fig.,. The unrooted neighbor-joining (NJ) tree (a) and the network graph (b) of the mtdna haplotypes in Silkie samples. The links are labeled by the nucleotide positions to designate transitions. Circle areas are proportional to haplotype frequencies. White and black colors were denoted for Chinese and Japanese Silkies, respectively. Table,. Haplogroup distribution frequency in Chinese and Japanese Silkies Haplogroup A B C D E G -+4/ -,40 342,/4* +4+ Chinese Silkies *,3 3, -* 3, 3 3,,- 3, + 3, ++42 242.+4,,*40 +140 Japanese Silkies *. -. - -. +. -. 1 -. 0 -.,04,,04, +24- /40,-4* *42 Total -- +,0 -- +,0,- +,0 1 +,0,3 +,0 + +,0 Note: Haplogroups were defined in Liu et al. (,**0a). five haplotypes were shared by Japanese Silkies and Chi- develop special and favorable strain/breed of farm aninese Silkies, whereas eight haplotypes were exclusively mals has received more attention in recent years. Due to found in Japanese Silkies. the medical property and beautiful appearance, many The mean haplotype diversity and nucleotide diversity researchers have attempted to improve the production were *. 22, *.*+, and *.*+,32 *.***-* for all chicken levels of Silkies (He,,**-). To know the genetic structure samples, respectively. For di# erent Silkie populations, and diversity of certain domestic animal is essential for both the Chengdu Silkies (*. 2.2 *.*0+) and Japanese Silkies (*. 2.+ *.*/0) presented a high value of haplotype developing rational breeding programs. mtdna sequence variations have been proved to be a useful tool in analyzdiversity. The Bairong Silkies had the lowest haplotype ing the genetic structure and tracing the matrilineal origin diversity (*..0, *. ++. ) and nucleotide diversity (*.**/*3 *.**+2,). Overall, Chinese Silkies had a relatively equal of domestic birds (Liu et al.,,**0 a; He et al.,,**2). Fu and colleagues (,**,) studied the genetic relationship diversity with Japanese Silkies (Table + ). between Silkies and other chicken breeds using mtdna Discussion sequence variations, and detected no obvious di# erentiation. This result was further confirmed in our subsequent With the improvement of living standard in China, to analysis of chicken mtdna sequences from various

26 J. Poult. Sci.,.1 ( + ) breeds (Liu et al.,,**0a). However, both studies did not high frequency of haplogroup D, which has been found to discuss the genetic relationship between the Chinese and be closely related to the distribution of the pastime of cock Japanese Silkies. fighting (Liu et al.,,**0a, b), in Japanese Silkies distin- By employing the recently described chicken mtdna guished itself clearly from the Chinese Silkies. In short, classification system (Liu et al.,,**0a) and available our results provided evidence for the genetic connection mtdna data, we compared the matrilineal components of between Chinese and Japanese Silkies, but also Chinese and Japanese Silkies, with an intention to learn demonstrated the di# erence between the matrilineal genetmore about the past history of the Silkies. In particular, ic structures of these Silkies. we collected new samples from the Bairong Silkie, which has been thought to be one of the ancestor populations of modern Silkies in China (Qiu et al., +322). Out of the nine Acknowledgments The authors thank the anonymous reviewer for helpful highly divergent mtdna haplogroups that were defined comments on the early version of the manuscript. This by Liu et al. (,**0 a), six were found in the +,0 samples work was financially supported by the earmarked fund for from Japanese and Chinese Silkies. Such a pattern further Modern Agro-industry Technology Research System, the confirmed the complexity of domestic chicken matrilineal Science Fund for Young Scholars in Sichuan Province components at the breed level. Among the four Chinese (Grant No: ZQ *,0 *+1), and the National 20- Project of Silkie populations, Bairong Silkies di# ered from the other China (No.,**2 AA +*+**+ ). three breeds/populations (Taihe, Chengdu, and Hubei Silkie) by harboring di# erent frequencies of haplogroups. This di# erence, if not formed during the later breed References Arisawa K, Yazawa S, Atsumi Y, Kagami H and Ono T. Skeletal selection and cultivation, would suggest for a di# erent analysis and characterization of gene expression related to origin of Bairong Silkies. On this point, we would think pattern formation in developing limbs of Japanese silkie fowl. Journal of Poultry Science,.-: +,0 +-..,**0. that our result supported the notion for dual ancestor Bandelt HJ, Forster P, and Röhl A. Median-joining networks for populations for modern Silkies in China (Qiu et al., +322; inferring intraspecific phylogenies. Molecular Biology and Xie, +33/ ). Normally, one would expect to detect higher Evolution, +0: -1.2. +333. genetic diversity in the founder or initial breed source Chang HL and Huang YC. The relationship between indigenous population than in derived populations. However, we animals and humans in APEC region. The Chinese Society observed a very low genetic diversity in Bairong Silkies compared with the other Silkie populations, and this result would reject the presumably ancestral status of the breed. Because of wide application of intensive artificial selection during the breeding, the rejection of null hypothesis here is of Animal Science, Taiwan.,**-. Chen SY, Liu YP, Wang W, Gao CZ, Yao YG and Lai SJ. Dissecting the matrilineal components of Tongjiang cattle from southwest China. Biochemical Genetics,.0:,*0,+/.,**2. understandable. Nonetheless, the low genetic diversity of Desjardins P and Morais R. Sequence and gene organization of the chicken mitochondrial genome. A novel gene order in Bairong Silkie would call for essential attention for conservation and sustainable usage of this breed. higher vertebrates. Journal of Molecular Biology,,+,: /33 0-.. +33*. The relationship between Chinese domestic chicken and Fu Y, Niu D, Ruan H, Luo J and Zhang YP. Studies on genetic Japanese domestic chicken has been profoundly analyzed relationship between silky chicken an other breeds. Chinese in previous reports (Komiyama et al.,,**-,,**.; Oka et Journal of Animal Science, -2: / 1.,**,. al.,,**1). All these studies suggested that the Japanese Fumihito A, Miyake T, Sumi S, Takada M, Ohno S and Kondo domestic chickens were originated independently from Southeast Asia and the mainland China (Komiyama et al., N. One subspecies of the red junglefowl ( Gallus gallus gallus) su$ ces as the matriarchic ancestor of all domestic,**-,,**.; Oka et al.,,**1). In our previous study (Liu breeds. Proceedings of the National Academy of Sciences of et al.,,**0a), we showed that the distribution of main the United States of America, 3+ : +,/*/ +,/*3. +33.. Fumihito A, Miyake T, Takada M, Shingu R, Endo T, Gojobori mtdna haplogroups in chicken presented geographic pattern: haplogroups A, B, and E were distributed ubiq- T, Kondo N and Ohno S. Monophyletic origin and unique dispersal patterns of domestic fowls. Proceedings of the uitously in Eurasia; haplogroup C was prevalent in Japan National Academy of Sciences of the United States of and Southeast China; haplogroups F and G were exclusive America, 3-: 013, 013/. +330. to Yunnan, China; haplogroup E was dominated in He DQ, Zhu Q, Chen SY, Wang HY, Liu YP and Yao YG. A Europe, the Middle East, and India. In this study, the +- haplotypes identified in -. Japanese Silkies belonged to homogenous nature of native Chinese duck matrilineal pool. BMC Evolutionary Biology, 2 :,32.,**2. five haplogroups. Specifically, the majority of Japanese He YC. Black-bone chicken and Chinese Taihe chicken and their Silkie samples ( 13.. ) could be classified into haplogroups medical property. Review of China Agricultural Science and C, D, and E. This pattern is in sharp di# erence Technology, /: 0. 00.,**-. with that of Chinese Silkies, in which haplogroups A, B, Komiyama T, Ikeo K and Gojobori T. Where is the origin of the Japanese gamecocks? Gene, -+1: +3/,*,.,**-. and E were prevalent. The high prevalence of haplogroup Komiyama T, Ikeo K, Tateno Y and Gojobori T. Japanese E in both Japanese Silkies and Chinese Silkies would point domesticated chickens have been derived from Shamo tradito their genetic connection in history. Conversely, the

Zhou et al. : Matrilineal Components of Silkies 27 tional fighting cocks. Molecular Phylogenetics and Evolu- Oka T, Ino Y, Nomura K, Kawashima S, Kuwayama T, Hanada tion, --: +0,+.,**.. H, Amano T, Takada M, Takahata N and Hayashi Y. Kumar S, Tamura K and Nei M. MEGA -: integrated software Analysis of mtdna sequences shows Japanese native chickfor molecular evolutionary genetics analysis and sequence ens have multiple origins. Animal Genetics, -2:,21,3-. alignment. Briefings in Bioinformatics, /: +/* +0-.,**.. Lai SJ, Liu YP, Liu YX, Li XW and Yao YG. Genetic diversity,**1. Qiu XP, Chen E and Chen YX. Poultry breeds in China. Shangand origin of Chinese cattle revealed by mtdna D-loop hai Scientific & Technical Publishers, Shanghai. +322. sequence variation. Molecular Phylogenetics and Evolution, Rozas J, Sanchez-DelBarrio JC, Messeguer X and Rozas R. -2: +.0 +/..,**0. DnaSP, DNA polymorphism analyses by the coalescent and Liu YP, Cao SX, Chen SY, Yao YG and Liu TZ. Genetic other methods. Bioinformatics, +3 :,.30,.31.,**-. diversity of Chinese domestic goat based on the mito- Takahashi H, Nirasawa K, Nagamine Y, Tsudzuki M and chondrial DNA sequence variation. Journal of Animal Yamamoto Y. Genetic relationships among Japanese native Breeding and Genetics, +,0: 2* 23.,**3. breeds of chicken based on microsatellite DNA polymor- Liu YP, Wu GS, Yao YG, Miao YW, Luikart G, Baig M, Beja-Pereira A, Ding ZL, Palanichamy MG and Zhang Y-P. phisms. Journal of Heredity, 23 : /.- /.0. +332. West B and Zhou BX. Did chickens go north? New evidence for Multiple maternal origins of chickens: out of the Asian domestication. World s Poultry Science Journal,./:,*/ jungles. Molecular Phylogenetics and Evolution, -2: +, +3.,**0a.,+2. +323. Xie CX. The breeding history of poultry in China. Chinese Liu YP, Zhu Q and Yao YG. Genetic relationship of Chinese and Agriculture Publishers, Beijing. +33/. Japanese gamecocks revealed by mtdna sequence varia- Zárate VA, Musavaya K and Schäfer C. Gene flow in animal tion. Biochemical Genetics,..: +3,3.,**0b. genetic resources: A study on status, impact and trends. Nei M. Molecular Evolutionary Genetics. Columbia University University of Hohenheim, Germany.,**0. Press, New York. +321.