GENETICS

Microsatellite Marker Analysis for the Genetic Relationships Among Japanese Long-Tailed Chicken Breeds

R. Tadano*, M. Sekino , M. Nishibori* and M. Tsudzuki*,1

   ABSTRACT

 The present study was conducted to evaluate the genetic diversity and relationships of 9 native Japanese long-tailed chicken breeds (Shoukoku, Koeyoshi, Kurokashiwa, Minohiki, Ohiki, Onagadori, Satsumadori, Toumaru, and Toutenkou) together with 2 commercial breeds (White Leghorn and White Plymouth Rock), using 40 polymorphic microsatellite markers covering 23 linkage groups. The 8 breeds mentioned, except for Shoukoku and 2 commercial breeds, were believed to be descendants derived from crossings of the ancestor of Shoukoku and some other breeds. Three to 14 alleles per locus were detected across all the breeds. The mean number of alleles per locus, the mean unbiased expected heterozygosity, and the mean polymorphic information content ranged from 2.60 (Minohiki) to 4.07 (Shoukoku), from 0.293 (Koeyoshi) to 0.545 (Satsumadori), and from 0.250 (Koeyoshi) to 0.478 (Satsumadori), respectively. The mean fixation coefficient of subpopulation within the total population of 9 Japanese long-tailed breeds showed that approximately 38% of the genetic variation was caused by breed differences and 62% was due to differences among individuals. Toumaru had the largest number of breed-specific alleles with relatively high (>20%) frequency. In the phylogenetic tree of 11 breeds constructed by the neighbor-joining method from modified Cavalli-Sforza chord genetic distance measure, White Leghorn and White Plymouth Rock clustered together apart from the Japanese breeds. Among the Japanese long-tailed breeds, Toumaru, Kurokashiwa, and Koeyoshi showed relatively far distance from the other breeds. The Ohiki, Onagadori, Shoukoku, and Toutenkou were grouped into the same branch. Minohiki and Satsumadori were also clustered together. Kurokashiwa was not genetically close to Shoukoku, differing from a traditional hypothsis. It was confirmed in the present study that the microsatellite is a suitable tool to evaluate genetic diversity and relationships in chicken breeds.

Key Words: genetic diversity • genetic relationship • Japanese long-tailed chicken breed • microsatellite marker • Shoukoku

   INTRODUCTION

 The importance of keeping genetic diversity in domestic livestock is advocated worldwide by the Food and Agriculture Organization (FAO; http://dad.fao.org/en/Home.htm). In recent years in the poultry industry, only limited kinds of breeds, which match economic demands, are reared on a large scale. This limitation of the number of breeds probably results in a decrease of genetic diversity in domestic chickens, because these commercial breeds or their hybrids are generated during a considerably short time to satisfy economic benefit. On the other hand, native breeds, which are not used for industrial purposes, seem to have more genetic diversity as compared with recent commercial breeds, because they have been improved and established in their long breeding history through processes greatly differing from those used with commercial breeds. Therefore, conservation of native chicken breeds as a genetic resource is important to fill unanticipated breeding demands in the future.

Native Japanese chicken breeds were mainly improved as ornamental breeds before the opening of Meiji era (A.D. 1868). About 30 unique and artistic breeds were established up to the time (Mitsui, 1979; Tsudzuki , 2003 ). Seventeen types of them have been designated as "natural monuments" by the Japanese Government (Kuroda et al., 1987). Today, most of these breeds are conserved by fanciers on a small scale, and some breeds face the brink of extinction.

Oana (1951) studied relationships among native Japanese chicken breeds based on ancient documents, pictures, and morphological characters such as body shape, comb, and tail structure. In his report, native Japanese breeds, except for some breeds, were largely classified into 3 types according to their characters; that is, Jidori (similar to Leghorn body shape), Shoukoku (having a larger number of tail feathers than usual), and Shamo (having an erect body shape). Also, he believed that the Shoukoku was one of the oldest breeds, and its ancestor had been introduced into Japan from China in the Heian era (A.D. 794 to 1192) and that several breeds were thereafter generated based on the Shoukoku. They are the Koeyoshi, Kurokashiwa, Minohiki, Ohiki, Onagadori, Satsumadori, Toumaru, and Toutenkou (Figure 1 ). Among them, Minohiki, Ohiki, Onagadori, and Toutenkou have long and thick tail feathers and saddle hackles as in Shoukoku. Although Kurokashiwa has long and thick tail feathers, its saddle hackles are not as long. Koeyoshi, Satsumadori, and Toumaru have thick tailfeathers; however, these tail feathers are not as long as that of Shoukoku. Saddle hackles are also not long in these breeds.

View larger version (334K):[in this window]

[in a new window]

 

Figure 1. Males of the 9 Japanese breeds used in the present study: Shoukoku (panel A), Kurokashiwa (panel B), Minohiki (panel C), Koeyoshi (panel D), Ohiki (panel E), Onagadori (panel F), Satsumadori (panel G), Toumaru (panel H), and Toutenkou (panel I).

 After the work of Oana (1951), several studies to investigate genetic relationships among native Japanese chicken breeds were conducted from biochemical markers, such as blood group systems and blood protein polymorphisms (Tanabe and Mizutani, 1980; Okada et al., 1980, 1984, 1989; Hashiguchi et al., 1981; Tanabe et al., 1991; Okabayashi et al., 1998). However, these genetic markers generally show low levels of polymorphisms, in other words, a limited number of polymorphic loci and a small number of alleles per locus. Therefore, results of these studies were not sensitive enough to reveal genetic diversity in and genetic relationshipsamong native Japanese chicken breeds.

The microsatellite marker is a valuable tool to evaluate genetic diversity and relationships of domestic livestock including chickens, because it shows a higher degree of polymorphisms and ease of identification than other markers, such as allozyme assay or random amplified polymorphic DNA analysis (Zhang et al., 2002a). Up to now, several studies using microsatellite markers have been conducted to evaluate genetic diversity and relationships among populations of various domestic

chicken breeds and jungle fowls (Vanhala et al., 1998 ; Wimmers et al., 2000; Romanov and Weigend, 2001; Zhang et al., 2002b; Hillel et al., 2003). Some studies havebeen conducted (Takahashi et al., 1998; Osman et al., 2004 , 2005, 2006) for native Japanese chicken breeds as well. However, these studies were generally performed from a relatively small number of microsatellite loci and individuals per breed or population. Takezaki and Nei (1996)  suggested that at least 30 microsatellite markers should be used to obtain exact genetic information about phylogenetic relationships in a case of a lower level of diversity among populations or species. Furthermore, FAO has recommended that 25 or more individuals should be used in this kind of study (FAO, 1998). In the present study, genetic diversity was measured and relationships evaluated for 9 native Japanese chicken breeds, which have Shoukokutype tail morphology, and 2 commercial breeds from a larger number of microsatellite loci and individuals per breed than earlier studies.

   MATERIALS AND METHODS

 Sample Collection and DNA Isolation

In total, 480 birds from 9 native Japanese long-tailed chicken breeds (Koeyoshi, Kurokashiwa, Minohiki, Ohiki, Onagadori, Satsumadori, Shoukoku, Toumaru, and Toutenkou) and 2 commercial breeds (White Leghorn and White Plymouth Rock) were investigated. Details of the source of samples, number of individuals investigated per breed, and specific traits for each breed are shown in Table 1 . Samples (varying from 35 to 48) were collected from several different fanciers and livestock experiment stations, because native Japanese chickens are generally maintained as a small population composed of a small number of individuals by each fancier or station.

View this table:[in this window]

[in a new window]

 

Table 1. List of native Japanese and commercial breeds used in the present study

 Blood samples were taken from the ulnar vein and stored at –20°C before examination. Genomic DNA was extracted from the whole blood sample by the following methods, basically the same as Sambrook and Russell (2001). About 10 µL of whole blood was dissolved in 400 µL of TNES-urea buffer (1.0 M Tris-HCl, pH 7.5; 0.5 M NaCl; 0.5 M EDTA; 10% SDS; and 4 M urea) with proteinase K (10 mg/mL). After mixing the buffer and blood sample with overnight incubation at 37°C, proteins were removed from the samples by phenol and chloroform-isoamyl alcohol extractions, and genomic DNA was precipitated by 3 M CH3COOH (pH 5.2) andethanol.

PCR Amplification and Microsatellite Genotyping

Forty microsatellite markers (Table 2 ) were chosen from the US Poultry Genome Project Web site (http://poultry.-mph.msu.edu/) and the ArkDB database Web site by the Roslin Bioinformatics Group (http://www.thearkdb.org/), among which 14 markers were the recommended markers by a joint International Society of Animal Genetics-FAO working group for biodiversity study (http://dad.fao.org/en/Home.htm). The 40 microsatellite markers covered 23 linkage groups.

View this table:[in this

window][in a new window]

 

Table 2. Microsatellite used in the present study and F-statistics per each locus for 9 native Japanese long-tailed chicken breeds

 Each PCR amplification was conducted in a 20-µL reaction mixture, which included 10 pmol of each primer, 100 µM deoxynucleoside triphosphate, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 0.001% gelatin, 0.5 U of AmpliTaq Gold (Applied Biosystems, Foster City, CA), and approximately 20 ng of genomic DNA as a template. The PCR procedure, using the GeneAmp PCR System 9700 (Applied Biosystems), was as follows: first denaturation step at 95°C for 10 min, 43 cycles of denaturation at 95°C for 1 min, annealing at primer-specific temperature (50 or 55°C) for 1 min, extension at 72°C for 1 min, followed by final extension at 72°C for 10 min.

The DNA fragments produced by the PCR amplification were electrophoresed with a size standard Genescan-350 TAMRA (Applied Biosystems), using an ABI 310 automated DNA sequencer (Applied Biosystems).Fragment length was determined with GeneScan software version 3.1 (Applied Biosystems), and microsatellite genotypes were assigned to the birds using Genotyper software version 2.5 (Applied Biosystems).

Statistical Analysis

The genetic diversity of each breed was assessed by calculating the number of alleles per locus and its mean (MNA), observed heterozygosity (HO), unbiased expected heterozygosity (HE; Nei , 1987 ), and polymorphic information content (PIC; Botstein et al., 1980), using the CERVUS version 2.0 software package (Marshall et al., 1998). Additionally, F-statistics [fixation coefficient of an individual within a subpopulation (FIS), fixation coefficient of an individual within the total population (FIT), and fixation coefficient of a subpopulation within the total population (FST)] per locus across 9 native Japanese breeds (Weir and Cockerham, 1984) werecalculated using the GENEPOP version 3.4 program (Raymond and Rousset, 1995).

Genetic distances among 11 breeds were evaluated by modified Cavalli-Sforza chord distance (DA; Nei et al., 1983 ). A phylogenetic tree was constructed based on the DA genetic distance by using the neighbor-joining (NJ) method (Saitou and Nei, 1987). The robustness of tree topologies was evaluated with a bootstrap test of 1,000 resampling across loci. These processes were conducted using the DISPAN computer program (Ota, 1993). The phylogenetic tree was edited using TreeView program (Page, 1996).

   RESULTS

 Genetic Diversity

In total, 286 alleles were detected at the 40 microsatellite loci distributed on 23 autosomes in 480 birds representing the 9 native Japanese long-tailed chicken breeds and 2 commercialbreeds. Across all the breeds, the average MNA was 7.15 (286/40), with the range from 3 (MCW0037) to 14 (MCW0287 and MCW0301). Allele size difference was smallest at MCW0037 and MCW0165 (4 bp: 149 to 153 bp, and 112 to 116 bp, respectively) and largest at LEI0196 and MCW0301 (42 bp: 170 bp to 212 bp and 260 bp to 302 bp, respectively). Table 2  shows marker information and F-statistics for the 9 Japanese long-tailed breeds. In the Japanese breeds, 272 alleles were detected, and the average MNA was 6.80 (272/40), with a range from 3 (ADL0106, ADL0278, MCW0037, MCW0078, MCW0081, and MCW0165) to 14 (MCW0287). The FIS, FIT, and FST

 values ranged from –0.0245 (ADL0147) to 0.5113 (LEI0099), from 0.2767 (MCW0080) to 0.7237 (LEI0099), and from 0.2210 (ADL0190) to 0.5085 (MCW0248) with the mean values of 0.109, 0.450, and 0.383, respectively.

Table 3  shows measures of genetic diversity in 11 chicken breeds for the 40 microsatellite markers, such as MNA, HO, HE, and PIC. The MNA ranged from 2.60 (Minohiki) to 4.07 (Shoukoku). The HO and HE ranged from 0.273 (Koeyoshi) to 0.523 (Shoukoku) and from 0.293 (Koeyoshi) to 0.545 (Satsumadori), respectively. The PIC ranged from 0.250 (Koeyoshi) to 0.478 (Satsumadori). In particular, Koeyoshi exhibited a lower degree of genetic diversity than all the other breeds in all measures of genetic diversity (MNA = 2.75, HO = 0.273, HE = 0.293, and PIC = 0.250). Conversely, a high degree of diversity was observed in Satsumadori and Shoukoku. (Satsumadori: MNA = 3.90, HO = 0.496, HE = 0.545, PIC = 0.478; Shoukoku: MNA = 4.07, HO = 0.523, HE = 0.527, and PIC = 0.464).

View this table:[in this

window][in a new window]

 

Table 3. Genetic diversity parameters1 estimated for 40 microsatellite markers in 9 native Japanese and 2 commercial chicken breeds

 As shown in Table 3 , 61 breed-specific alleles were detected across the 11 breeds. The number of breed-specific alleles per breed ranged from 2 (Toutenkou) to 9 (Shoukoku, Toumaru, and White Leghorn). The frequency of breed-specific alleles ranged from 0.010 (Toutenkou: 310-bp allele at MCW0193 and 252-bp allele at MCW0287; White Plymouth Rock: 280-bp allele at MCW0287) to 0.969 (Ohiki: 163-bp allele at MCW0069). Approximately one-third of the 61 breed-specific alleles showed frequency exceeding 20%, details of which are summarized in Table 4 .

View this table:[in this

window][in a new window]

 

Table 4. Breed-specific alleles with frequency >20% detected across 9 native Japanese and 2 commercial chicken breeds

 Genetic Distance and Relationships Among Breeds

Table 5  shows DA between each pair for all 11 chicken breeds, based on 40 microsatellite loci genotypes. The DA ranged from 0.2411 (between Onagadori and Toutenkou) to 0.5988 (between Onagadori and White Leghorn). Figure 2  shows a phylogenetic tree of 9 native Japanese long-tailed chicken breeds and 2 commercial breeds that was constructed from DA by using the NJ method. Two commercial breeds, White Leghorn and White Plymouth Rock, clustered together apart from Japanese breeds. Among the Japanese breeds, Kurokashiwa, Toumaru, and Koeyoshi formed separated branches from the other long-tailed breeds. Shoukoku, Ohiki,Onagadori, and Toutenkou were grouped into the same branch. Minohiki and Satsumadori were grouped together.

View this table:[in this window]

[in a new window]

 

Table 5. Modified Cavalli-Sforza chord distance (DA) between each pair of 11 chicken breeds

 

View larger version (11K):[in this window]

[in a new window]

 

Figure 2. A neighbor-joining tree of 9 Japanese long-tailed fowl breeds and 2 commercial breeds based on the modified Cavalli-Sforza chord distance. The number at each branch represents the percentage of bootstrap values from 1,000 replications of resampled loci.

 

   DISCUSSION Genetic Diversity

The FIS represents a degree of nonrandom mating (deviation from Hardy-Weinberg equilibrium). A positive number for FIS means deviation from Hardy-Weinberg equilibrium. Only ADL0147 showed a negative number; however, all the others showed a positive number. This result indicated that nonrandom mating was performed in the native Japanese breeds studied here. It is thought that these Japanese breeds have been more or less inbred to improve breed-specific characteristics in each breed, such as beautiful figures and long crowing. The mean FST value of 0.383

indicates that approximately 38.3% of the total genetic variation is caused by breed differences, whereas the remaining 61.7% is due to differences among individuals within breeds. Vanhala et al. (1998) reported the mean FST value of 0.303 from 8 Finnish chicken lines using 9 microsatellite markers. As compared with other domestic animals, the mean FST value in native Japanese long-tailed breeds was relatively high. For instance, Kim et al. (2001, 2005) reported the mean FST value of 0.261 and 0.154 for pig breeds and East Asian native dog breeds, respectively. This fact might suggest that the Japanese chicken breeds in the present study are genetically subdivided to a higher extent than foreign chicken lines and other livestock breeds because of inbreeding between related individuals or more intensive selection to fix desirable traits.

In native Japanese long-tailed chicken breeds, the lowest diversity was observed in Koeyoshi (MNA = 2.75, HO = 0.273, HE = 0.293, PIC = 0.250). Among the measures, heterozygosity was extremely lower than in the other breeds. This result seems to reflect the present breeding state of Koeyoshi, in which inbreeding is generally carried out among closely related birds in narrow geographical areas in north Japan to improve the long crowing trait. Low genetic diversity of Koeyoshi has also been reported by Osman et al. (2004)  using a set of 20 microsatellite loci and 24 birds, in which MNA and HE were 1.65 and 0.206, respectively.

In contrast, high diversity was observed in Satsumadori (MNA = 3.90, HO = 0.496, HE = 0.545, PIC = 0.478) and Shoukoku (MNA = 4.07, HO = 0.523, HE = 0.527, PIC = 0.464). Today, Satsumadori and Shoukoku are kept by fanciers as ornamental breeds in large areas of Japan, which leads to breeding from many individuals. This condition seems to result in a higher degree of diversity than other breeds. Osman et al. (2004) also reported that Satsumadori showed a high degree of diversity (MNA = 4.70 and HE = 0.670).

The MNA in White Leghorn and White Plymouth Rock was not as high when compared with native Japanese long-tailed breeds that are generally reared in a small-size population. The HO:HE in these commercial breeds was apparently larger than that of the Japanese breeds. These results suggested that the population size was relatively small at the starting point of these 2 commercial breed populations; thereafter, sufficient random matings were performed for the breeds with a large population size.

The HO and HE observed in native Japanese chicken breeds in the present study were similar to or slightly lower than those reported for European or American breeds or lines (Vanhala et al., 1998 ). On the other hand, native Chinese breeds possess higher HO. According to Zhang et al. (2002b), HO ranged from 0.707 to 0.861. It was believed that the population size was large in each Chinese breed, and no intensive control has been performed for these Chinese breeds.

Genetic Relationships

In a NJ tree, Shoukoku, Ohiki, Onagadori, and Toutenkou showed a close relationship. This result supports the supposition of Oana (1951)  based on literature and morphological studies. Osman et al. (2005)  also found the same result from 20 microsatellite loci analyses.

Kurokashiwa and Toumaru showed no close relationship to each other in the present study. On the contrary, Osman et al. (2006)  reported a relatively close genetic relationship between the 2 breeds. Although Kurokashiwa samples used in the present study were partly the same as those used by Osman et al. (2006) , the Toumaru samples were collected from a different population. There is a possibility that some genetic differentiation occurredamong different Toumaru populations. Furthermore, studies will be necessary to reveal a genetic relationship between Kurokashiwa and Toumaru.

The NJ tree showed that Toumaru is genetically distant from other Japanese breeds. Also, 9 breed-specific alleles, the largest number of breed-specific alleles, were detected in Toumaru. Although breed-specific alleles were generally detected with very low frequency, 5 of the 9 breed-specific alleles detected in Toumaru showed relatively high frequency (>0.200). Takahashi et al. (1998) also reported that DA between Toumaru and other native Japanese breeds are relatively far. Oana (1951) supposed that Toumaru originated from Oh-Toumaru that had been imported from China in the 16th or 17th century and became extinct by the opening of Meiji era (A.D. 1868). There is a possibility that the breeding history of Toumaru is different from other Japanese breeds. In the present study, Koeyoshi, Toumaru, and Toutenkou did not exhibit close DA between each pair. Conversely, Komiyama et al. (2004) reported that these 3 breeds are genetically close to each other, based on the polymorphism analysis for the mitochondrial D-loop region. Further studies will be necessary in the future to reveal genetic relationships among these 3 breeds.

Oana (1951) mentioned that Kurokashiwa was a direct descendant of Shoukoku or a variety of Shoukoku. In the present study, however, Kurokashiwa showed no close relationship to Shoukoku. So far, Osman et al. (2006)  also obtained the same result. Thus, the hypothesis of Oana (1951)  should be rejected. It was concluded that Shoukoku and Kurokashiwa are not genetically close.

Minohiki and Satsumadori showed a close relationship. Osman et al. (2006)  also reported that these 2 breeds are genetically close. Moreover, Komiyama et al. (2004) also reported a close relationship between the 2 breeds based on mitochondrial DNA polymorphisms. Thus, the hypothesis of Oana (1951) , in which he supposed that these 2 breeds have a common ancestor, is thought to be supported by 2 kinds of DNA analyses (i.e., nuclear and mitochondrial DNA analyses).

In the present study, a precise study on genetic diversity and relationships was performed in and among chicken breeds from microsatellite markers and approximately 48 birds per breed. Previously, Osman et al. (2004 ,2005, 2006) performed similar studies based on 20 microsatellite markers and approximately 24 birds per breed, the results of which are the same in almost all points as those of the present study. Thus, it is thought that the use of 20 microsatellite markers and approximately 24 birds might be adequate to obtain accurate results.

Evaluating genetic diversity might be useful for conservation of native Japanese chicken breeds as a genetic resource and natural monument. For instance, judging from the result in the present study, conservation of the Koeyoshi breed should be an emergency. It was confirmed through the present study that the microsatellite is a valuable tool to evaluate genetic diversity of chicken breeds.