Suitability of AFLP markers for the study of Genetic relationships among Korean native dogs

Genes Genet. Syst. (2001) 76, p. 243 250 Suitability of AFLP markers for the study of Genetic relationships among Korean native dogs Kyung Seok Kim 1, Ho Won Jeong 1, Chan Kyu Park 2, Ji Hong Ha 1 * 1 Department of Genetic Engineering, Kyungpook National University, Taegu 702-701, Korea 2 National Creative Research Initiative Center for Behavioral Genetics, Department of Biological Sciences, Korea Advanced Institute of Science and Technology, Yusong-Ku, Taejon 305-701, Korea (Received 16 April 2001, accepted 20 August 2001) To determine the genetic relationships among domestic dog breeds, we performed both a sequence comparison of mitochondrial DNA (mtdna) and an amplified fragment length polymorphisms (AFLP) analysis. Three of four regions of mtdna, cytochrome b, cytochrome oxidase subunit II, and 16S rrna genes were highly homogeneous among dog breeds, whereas the other region, the control region, showed relatively high polymorphisms with a maximum percentage difference of 3.18%. However, the control region showed extensive polymorphism even within breeds, and the relationship tree derived from the data could not clearly delimit distinct breeds. 19 EcoRI/MseI primer combinations were used to generate AFLP markers among 25 dogs from 11 breeds including three Korean native dogs. These amplification reactions allowed the detection of more than 1900 amplification products of which 408 were identified as polymorphic bands. Unrooted neighbor-joining tree based on dissimilarity values showed that the Korean native dogs were clustered together with the Asian dogs and that the Asian originated dogs were clustered separately from Western originated dogs. A consensus tree using parsimony method also showed Korean native dogs were grouped separately from the other dogs with moderate bootstrap values. Taken together, it is concluded that AFLP analysis is a more informative tool for revealing genetic relationships among dog breeds than mtdna sequence comparison. INTRODUCTION The Dog, Canis familiaris, is one of the oldest domestic animals, originally domesticated in the pre-agricultural age between 14,000 and 12,000 years ago (Turnbell and Reed 1974; Tanabe et al. 1991). Currently, there are more than 400 dog breeds with various morphological and behavioral traits that are used for various purposes including hunting, guarding, guides for the blind, pets, etc. Although dogs have been a common focus for research concerning their evolution, origin, and variation (Tanabe et al. 1991; Roy et al. 1994; Vila et al. 1997; Tsuda et al. 1997), the genetic backgrounds and genetic polymorphisms among the various breeds are still not completely understood. Mitochondrial DNA (mtdna) has become widely utilized for evolutionary studies due to rapid nucleotide substitution, rare recombination, strict maternal inheritance, conservation of the size, gene content, and gene Edited by Naoyuki Takahata * Corresponding author. E-mail: jhha@bh.kyungpook.ac.kr order. Especially, partial comparisons of mtdna sequences have been used to investigate the phylogenetic relationships among closely related taxa and populations of the same animal species (Aquadro and Greenberg 1983; Ramharack and Deeley 1987; Ruvulo et al. 1993; Brown et al. 1993; Janczewski et al. 1995). MtDNA sequence comparisons have also been widely utilized in studies concerning the origin and evolution of domestic dogs and wolves (Wayne and Jenks 1991; Girman et al. 1993; Vila et al. 1997). However, there is another opinion regarding the usefulness of mtdna for dog phylogeny. Some researchers have suggested that mtdna sequence comparisons are not appropriate for studying phylogenetic relationships between dog breeds (Okumura et al. 1996; Tsuda et al. 1997), due to extensive polymorphism in the control region. It should be noted, though, that these researchers did not investigate any other regions of mtdna that have been used effectively in other species. It is also important to understand the genetic relationships among different dog breeds using nuclear DNA, but it is very difficult to find nuclear DNA genes suitable for evolutionary

244 K. S. KIM et al. study. Compared to sequence comparison of single copy genes, profiling of the genome by DNA fingerprinting can yield more information. Recently, various systems of markers that are representative of the whole genome have been developed. AFLP-PCR technology is a DNA fingerprinting technique that combines both a classical, hybridization-based fingerprinting method and a PCR-based fingerprinting method (Vos et al. 1995). The DNA polymorphisms identified by AFLP-PCR are typically inherited in Mendelian fashion and may, therefore, be used for typing, for the identification of molecular markers, for mapping of genetic loci, and for establishing relatedness between DNA samples. AFLP-PCR has been used for the relationship analysis mainly of plants or bacteria, and as a genetic marker for the construction of a genetic map (Janssen et al. 1997; Simons et al. 1997; Van der Lee et al. 1997; Qi and Lindhout 1997). Recently, AFLP markers have been used successfully for DNA fingerprinting and for studies of genetic relationships in closely related animals (Ajmone-Marsan et al. 1997; Ovilo et al. 2000). Resolving the genetic relationships of domestic dog breeds is very difficult, given the varied origins of and substantial gene flow among breeds. In Korea, there are three native dogs: the Jindo dog, the Sapsaree, and the Chejudo dog. Of these, the Jindo dog and the Sapsaree are registered as Korean native breeds, originating from Jindo Island and Kyungpook Provinces, respectively. The Chejudo dog is indigenous to Cheju Island. The three Korean native dogs have not been interbred, and have maintained morphologically distinct traits. Therefore, the use of these dog breeds will highlight the applicability of mtdna sequence and AFLP analysis for the documentation of genetic relationships between dog breeds. In this paper, we performed a mtdna sequence comparison and an AFLP analysis to clarify the genetic relationships among the different dog breeds and compared the results obtained from both analyses. We concluded that AFLP analysis is the more valuable tool for revealing genetic relationships of local dog breeds where the hypervariable region of mtdna fails. MATERIALS AND METHODS Sample collection and description of Korean native dog breeds A total of 25 subjects from eleven dog breeds were analyzed in this study. Subjects utilized in this study comprised 14 Korean native dogs, five Asian (non- Korean) dogs, and six Western dogs. The 14 Korean native dogs included seven Sapsaree (KS1~7), five Jindo dogs (KJ1~5), and two Chejudo dogs (KC1~2). The five Asian dogs included one Akita dog (AA), one Shiba dog (AS), one Tosa inu (AT), one Pekingese (Apk), and one Pug (Apu). The six Western dogs included four Great Danes (WGD1~4), one Collie (WC), and one German Shepherd (WGS). Fresh blood was taken from subjects from well-defined geographical areas and chosen at random without consideration of the relationship within and between breeds except for the Sapsaree, two of which were bred by the author. The Jindo dog is a short-haired, medium-sized breed, with a body height of 47 55 cm, originating from Jindo Island but now found all over Korea. We took blood samples from pedigreed dogs on Jindo Island. The Sapsaree is a long-haired and medium-sized breed, body height 49 55 cm, originating from Kyungpook Province. Blood samples from the Sapsaree (pedigreed) were taken in Kongsan, Kyungpook Province, from which the Sapsaree has been conserved exempting it from endangered crisis. The Chejudo dog is a short-haired, medium-sized breed, body height 40 46 cm, originating from Cheju Island. We took blood samples from pedigreed Chejudo dogs on Cheju Island. Blood samples from the other dogs, including Western originated dogs were taken from several institutions and veterinary hospitals in Korea. Genomic DNA preparation, PCR reaction, and sequencing of mtdna Genomic DNA was extracted from the blood samples collected from the foreleg vein of each dog as described by Maniatis et al. (1982). Based on the complete nucleotide sequence (GenBank accession no.: U96639) of the domestic dog mtdna (Kim et al. 1998), four primer sets were synthesized. The primer sequences for two protein-coding genes, one ribosomal RNA gene, and the control region are as follows: cytochrome b (L14,252 L14,631) F: ACTCATTCATTGACCTCCCAGCG R: AGT- TCCGATATAAGGGATGGCAGAG; COII (L7,054 - L7,46-5) F: ATGGCGTACCCATTTCAACT R: GGATGGTTATT- TCTATTGG; 16S rrna (L2,033 L2,472) F: GCAAAG- GTAGCATAATCAT R: AGGACTTTAATCGTTGAAC; Control region (L15,622 L16,030) F: CATAGGACATA- TTAACTCAATC R: AAGTCCAGCTACAAGTTATTTG. 20 60 ng of genomic DNA was used for PCR. Thermal cycle amplifications were performed in a 50 µl final volume with 5 µl of a 10 X reaction buffer (10mM Tris-HCl (ph 8.3), 40mM KCl, 1.5mM MgCl 2, 1mM DTT, 50 µg/ml BSA), dntps at 200 mm, primers at 1 µm, and 2 units of Taq DNA polymerase with 50 µl of mineral oil overlaid. The reaction cycle consisted of denaturation for 1 min at 94 C primer annealing for 1 min at 58 C, and an extension for 2 mins at 72 C; the cycle was repeated 35 times. The amplified DNA was electrophoresed in 1.2% agarose gels and extracted from a gel slice using a Gene Clean Kit II (Bio101 Inc.). The purified DNA was then ligated into a pgem T-vector and transformed into E. coli DH- 5α. The mtdna sequence was determined manually by 6% polyacrylamide gel electrophoresis using the dideoxynucleotide chain termination method (Sanger et al. 1977) with [ 35 S]-dATP using a commercial DNA sequencing kit (Sequenase Version 2.0, United States Bio-

Genetic relationships among domestic dogs using AFLP markers 245 chemical). The sequencing was carried out from both strands using universal T7 and SP6 primers. Amplified Fragment Length Polymorphisms (AF- LP)-PCR AFLP analysis was performed as described in the manufacturer s protocols (AFLP TM Analysis System I, AFLP Starter Kit, Gibco-BRL). For each genomic DNA sample, approximately 500ng were prepared by the double digestion with EcoRI and MseI, and subsequent ligation of asymmetric quantities of enzyme specific adapters. Selective nucleotides extending into the restriction fragments were added to the 3' ends of the PCR primers such that only a subset of the restriction fragments was recognized. Total numbers of 19 primer combinations were employed to detect AFLP polymorphism between the different genotypes. Selective nucleotide combinations tested with EcoRI/MseI primers were as follows: AGG/ CTG, ACA/CAG, ACA/CAC, ACA/CTC, ACA/CAA, ACA/ CTA, ACA/CTT, ACT/CAG, ACT/CAC, ACT/CAA, ACG/ CAA, ACG/CAT, AGC/CTA, AGC/CTT, AGC/CAT, ACC/ CAA, ACC/CTA, ACC/CTT, ACC/CAT. The subset of amplified fragments was analyzed by a 6% denaturing polyacrylamide gel electrophoresis to generate the fingerprint. Electrophoresis was performed at constant power (55W) and AFLP patterns were visualized by autoradiography. Data analysis Calculations of the distance matrix and a cluster analysis were carried out using the PHYLIP software package (Ver.3.5c; Felsenstein 1993). The mtdna sequence data were analyzed using distances corrected by the two-parameter method of Kimura (1980) using the DNADIST program of PHYLIP package. The distance values were used to construct a neighbour-joining (NJ) tree (Saitou and Nei 1987), and bootstrap resampling (n = 100) was performed to test the robustness of the dendrogram topology. For AFLP analysis, each AFLP band was scored as either present (1) or absent (0) across all genotypes to create a binary matrix. The results were converted into a similarity matrix based on the Jaccard s coefficient (Sj) for qualitative data (Jackson et al. 1989); Sj(ij) = W/(W+X+Y) Where, W is the number of polymorphic DNA fragments common to both individuals, X is the number of fragments present in i and absent in j, and Y is the number of fragments present in j and absent in i. Jaccard s similarity algorithm was used due to the dominant feature of AFLP bands and the fact that this algorithm ignores 0/0 matches that theoretically provide less evidence of relationship (Tosto and Hopp 2000). Sj was then converted to a dissimilarity (calculated as 1-Sj) to construct a relationship tree between the individual dogs. The data were used to construct a NJ tree using the NEIGHBOR module of the PHYLIP package. A phylogenetic tree using the polymorphism parsimony method was constructed using DOLLOP module and strict consensus tree program with bootstrap resampling (n = 100) of a binary matrix. All trees were graphically represented with Tree-view (Page 1996). RESULTS MtDNA sequence analysis We determined partial sequences of four regions of dog mtdna, including cyt b, COII, a 16S rrna gene of ten subjects and the control region of 25 subjects (Table 1). From the sequence alignments, we found that the cyt b, COII, and 16S rrna gene did not show a significant difference among the dog breeds. The 16S rrna gene showed only two haplotypes with a low percentage difference of less than 0.23%, differing in only one base pair among the dog breeds. In addition, the cyt b and COII gene showed the same sequences in all of the dogs utilized. Although the number of dogs analyzed was small, we speculate that the protein coding genes and rrna genes of the domestic dog mtdna are very homogeneous within the dog breeds. However, the mtdna control region showed a highly polymorphic sequence among dog breeds with a maximum polymorphic value of 3.18%. The results of pairwise alignments of 409 base pairs (bps) of the control region sequenced are shown in Table 2. In total, 16 haplotypes with 18 nucleotide substitutions were observed. Eighteen transitions and one transversion (position 18) were detected in the polymorphic sites, showing high transition/transversion ratio. The three Korean native dog breeds had at least nine Table 1. The results of partial sequencing of mtdna among several dog breeds Target gene Length of No. of examined No. of observed Sequence sequences (bp) Individuals haplotypes variation (%) Cyt b 380 10 1 0.00 CO II 412 10 1 0.00 16S rrna 440 10 2 0.00 ~ 0.23 Control region 409 25 16 0.00 ~ 3.18 The partial nucleotide sequences of domestic dog mtdna reported in this paper are available in the GenBank database under Accession No. AF064567 AF064588.

246 K. S. KIM et al. Table 2. The nucleotide substitutions within 409 bps of control regions among 25 different dogs Nucleotide Positions Dog Breed 1 1 1 1 1 1 2 3 3 3 3 4 1 1 2 3 4 0 0 7 9 9 9 9 2 3 3 8 0 6 1 8 2 1 4 1 5 9 0 3 4 1 1 4 8 2 4 Sap. A (KS1) G C A A G T A C T A T T C C C T A C Sap. B (KS2), Jindo C (KJ3) C T Sap. C (KS3), Sap. E (KS5), Jindo D (KJ4) A C C T Sap. D (KS4), German Shepherd (WGS) A T C C T Jindo A (KJ1) A T A CG C G T Jindo B (KJ2) A A G C T Chejudo A (KC1), Sap. G (KS7) A T C T Chejudo B (KC2), Great Dane D (WGD4) A T G G A C C T T C G T Shiba (AS) A T C Great Dane C (WGD3) A C T T C T Tosa (AT), Sap. F (KS6), Jindo E (KJ5) T C T Pug (Apu) T T C T Akita (AA) A T G G A T C C T T C G T Collie (WC) T C T Great Dane A (WGD1) T C G T Pekingese (Apk), Great Dane B (WGD2) A C Dots indicate matching with the reference sequence of Sapsaree A (KS1) and the numbers of nucleotide position on the top were labeled in accordance with the first base of the amplified control region as for position 1. The partial nucleotide sequences reported in this paper are available in the GenBank database under Accession No. AF064569-AF064585. haplotypes, indicating that Korean native dogs originated from a matriarchal lineage of more than nine. The Sapsaree and Jindo dogs had six and five haplotypes respectively, and three of these were common to both. Additionally, seven haplotypes of the 16 were found in two or more breeds. The relationship tree (Figure 1) showed no clade specific for any dog breeds including Korean native dogs. In addition, Western originated dogs were formed in all clades, with no discrimination of Asian dogs from Western originated dogs. These results showed that the sequence comparisons of dog mtdna could not contribute to the documentation of a reliable relationship among dog breeds. Fig. 1. Neighbor-joining tree of 25 dogs from 11 breeds based on the indices of the nucleotide substitution per site calculated by Kimura s two-parameter method in 409 bps of a control region. The numbers at the nodes are the percentage bootstrap values from 100 replications of sequences. Symbols in the figure are the same as those of the breeds given in Table 2. Genetic relationship between dog breeds using AFLP-PCR In order to evaluate the applicability of AFLP markers for the clarification of genetic relationships among dog breeds, a set of 19 pairs of EcoRI and MseI primer combinations was analyzed. As shown in Table 3, the number of AFLP markers generated by each primer combination was variable, ranging approximately from 60 to 160 bands. From 19 primer combinations, a total of 1991 bands were produced for 25 dogs of 11 breeds investigated in this study. Of these, 408 polymorphic bands were produced with a polymorphic ratio of 20.5%, but no breed specific markers were detected. As expected, a much lower polymorphism level was detected within the same breeds (10.7%). A lower polymorphism level was detected within Asian dogs (14.3%) than within the total sample of dogs analyzed.

Genetic relationships among domestic dogs using AFLP markers 247 Table 3. The number of AFLP markers obtained with the 19 primer combinations Primer combination Total no. No. of of bands polymorphic bands E+AGG / M+CTG E+ACA / M+CAG E+ACA / M+CAC E+ACA / M+CTC E+ACA / M+CAA E+ACA / M+CTA E+ACA / M+CTT E+ACT / M+CAG E+ACT / M+CAC E+ACT / M+CAA E+ACG / M+CAA E+ACG / M+CAT E+AGC / M+CTA E+AGC / M+CTT E+AGC / M+CAT E+ACC / M+CAA E+ACC / M+CTA E+ACC / M+CTT E+ACC / M+CAT Total 80 89 125 109 123 119 138 120 105 156 106 59 101 92 110 112 91 107 108 1991 16 7 23 28 20 19 24 21 38 38 28 7 18 16 27 20 9 26 23 408 The 408 polymorphic bands detected were used for the analysis of the genetic relationships between dog breeds. Dissimilarity coefficients were calculated as 1-Sj based on the existence of AFLP bands and the values are shown in Table 4. The lowest dissimilarity values were detected between individuals of the same breed (with the values ranging from 0.07 to 0.27) as compared with those between individuals from different breeds (ranging from 0.18 to 0.43). It is interesting that the lowest dissimilarity value was obtained within the Sapsaree, a Korean native dog. In fact, some breeders have inbred the dog extensively over the past three decades, reflecting a similar genomic background within the breed. To understand genetic relationships among dog breeds using AFLP markers, we constructed phylogenetic trees by the neighbor-joining (N-J) and character based parsimony methods. Figure 2 depicts an unrooted radial N-J tree obtained from the pair-wise genetic distances based on the dissimilarity values. Figure 3 is a strict consensus tree from 100 replications of a binary matrix. The N-J tree showed a grouping of breeds on the basis of their locality of domestication. The tree split the entire group of Asian dogs with respect to Western originated dogs. In addition, the three Korean native dogs, the Jindo dog (KJ1~5), the Sapsaree (KS1~7), and the Chejudo dog (KC1~2) were grouped separately from other Asian native dogs. Moreover, the seven Sapsaree were sub-grouped within other Korean native dogs. The four individuals of the Great Dane (WGD), one of the Western originated breeds, showed a close relationship with each other. A consensus tree using the polymorphism parsimony method, in spite of a low bootstrap value, is also similar to the N-J tree with a reasonable relationship on the basis of local breeds. Apart from two Asian native dogs (AS and AA), which formed the basal branch to all other dogs, the tree split the Korean native dogs from other dogs and clustered the Western originated dogs as a single group. Some relationship trees using other methods (such as UPGMA and character-based parsimony methods) showed similar topologies to the above trees. It should be noted that two individuals (KS3 and KS5) of the Sapsaree, which Table 4. Dissimilarity coefficient matrix among 25 dogs from 11 breeds based on existence of AFLP markers Breeds KS1 KS2 KS3 KS4 KS5 KS6 KS7 KJ1 KJ2 KJ3 KJ4 KJ5 KC1 KC2 AA AS AT Apk Apu WC WGS WGD1 WGD2 WGD3 WGD4 KS1 0.00 KS2 0.07 0.00 KS3 0.11 0.10 0.00 KS4 0.13 0.13 0.14 0.00 KS5 0.08 0.10 0.08 0.12 0.00 KS6 0.17 0.20 0.18 0.13 0.17 0.00 KS7 0.16 0.19 0.18 0.16 0.16 0.15 0.00 KJ1 0.18 0.18 0.20 0.23 0.18 0.23 0.27 0.00 KJ2 0.20 0.20 0.20 0.18 0.20 0.19 0.21 0.21 0.00 KJ3 0.24 0.26 0.26 0.26 0.25 0.26 0.24 0.23 0.18 0.00 KJ4 0.22 0.22 0.23 0.22 0.25 0.24 0.25 0.22 0.22 0.27 0.00 KJ5 0.22 0.24 0.24 0.20 0.24 0.20 0.19 0.26 0.22 0.23 0.22 0.00 KC1 0.20 0.22 0.21 0.20 0.19 0.19 0.24 0.23 0.21 0.23 0.21 0.23 0.00 KC2 0.21 0.19 0.20 0.18 0.19 0.21 0.24 0.21 0.20 0.23 0.20 0.20 0.15 0.00 AA 0.26 0.27 0.25 0.23 0.27 0.26 0.24 0.26 0.23 0.21 0.26 0.21 0.28 0.26 0.00 AS 0.31 0.29 0.28 0.26 0.29 0.26 0.28 0.31 0.30 0.32 0.29 0.29 0.30 0.29 0.22 0.00 AT 0.25 0.23 0.24 0.22 0.26 0.24 0.24 0.29 0.23 0.26 0.25 0.30 0.28 0.26 0.24 0.22 0.00 Apk 0.23 0.23 0.25 0.24 0.23 0.26 0.22 0.28 0.27 0.28 0.28 0.29 0.31 0.30 0.27 0.23 0.21 0.00 Apu 0.26 0.25 0.26 0.25 0.27 0.27 0.28 0.27 0.27 0.29 0.28 0.26 0.26 0.28 0.28 0.26 0.21 0.20 0.00 WC 0.36 0.34 0.37 0.33 0.35 0.31 0.30 0.35 0.35 0.35 0.32 0.33 0.31 0.32 0.32 0.33 0.26 0.31 0.26 0.00 WGS 0.31 0.29 0.27 0.31 0.30 0.29 0.27 0.28 0.28 0.29 0.27 0.29 0.30 0.25 0.31 0.35 0.25 0.27 0.25 0.27 0.00 WGD1 0.31 0.31 0.34 0.31 0.43 0.29 0.28 0.30 0.32 0.31 0.28 0.28 0.28 0.31 0.26 0.28 0.25 0.29 0.26 0.26 0.24 0.00 WGD2 0.31 0.29 0.31 0.28 0.32 0.27 0.30 0.31 0.27 0.29 0.31 0.27 0.26 0.28 0.28 0.29 0.24 0.29 0.20 0.28 0.20 0.17 0.00 WGD3 0.29 0.26 0.27 0.26 0.28 0.29 0.29 0.30 0.31 0.34 0.32 0.29 0.27 0.29 0.31 0.34 0.30 0.32 0.25 0.30 0.22 0.23 0.20 0.00 WGD4 0.34 0.33 0.33 0.30 0.32 0.30 0.31 0.30 0.30 0.35 0.30 0.30 0.29 0.29 0.29 0.31 0.29 0.34 0.27 0.26 0.24 0.17 0.20 0.14 0.00

248 K. S. KIM et al. Fig. 2. Unrooted neighbor-joining tree from Korean, Asian, and Western originated dogs. The tree was constructed using a dissimilarity matrix of 1-Jaccard estimates based on AFLP analysis with 19 primer combinations. Symbols in the figure are the same as those of the breeds given in Table 2. is brood kinship, clustered with each other with the relatively high bootstrap value (76%) with respect to other clade. Although more survey for relationships within kinship should be performed, it seems that closer kin have a similar genetic relationship with each other. DISCUSSION In this study, to explain the genetic relationships among domestic dog breeds, a sequence comparison of mtdna and an amplified fragment length polymorphisms (AFLP) analysis were compared. The four portions of dog mitochondrial DNA (mtdna), cyt b, COII, 16S rrna genes, and the control region were compared among 25 dogs of 11 breeds including three Korean native dogs. Two proteincoding genes and one rrna gene showed a relatively high similarity among different dogs. Since these genes were very conserved and hardly changed, the dogs could not have had sufficient time to accumulate any nucleotide mutations. In contrast, the control region exhibited a high degree of nucleotide substitutions within dog breeds as well as between dog breeds. The sequence variations (0.00~3.18%) within the domestic dog were similar to those (0.00~3.19%) of Tsuda et al. (1997), showing similar values irrespective of the breeds studied and localities of domestication. However, no sequences of the control region that were specific for any particular dog breed were Fig. 3. A strict consensus tree from Korean, Asian, and Western originated dogs using polymorphism parsimony method from AFLP data. The numbers at the nodes are the percentage bootstrap values from 100 replications of a binary matrix. identified. Instead, extensive polymorphisms, even within breeds, were found. The relationship tree based on sequence divergences did not show any clade or haplotype specific to a particular dog breed. These findings indicate that although Korean native dog breeds have been maintained with a geographical separation for a long time, they could not be delimited as distinct dog breeds from the sequence comparisons of the hypervariable region of mtdna. Recently, from studies of the mtdna, it has been concluded that dogs have multiple origins and that extensive crossbreeding has occurred since the beginning of domestication (Okumura et al. 1996; Vila et al. 1997; Tsuda et al. 1997; Vila et al. 1999). Tsuda et al. (1997) proposed that the existing breeds of domestic dogs were developed from an ancestral wolf population and that extensive interbreeding has occurred among the multiple matriarchal origins. Additionally, Okumura et al. (1996) stated that Japanese native dog breeds could not be clearly delimited as distinct dog breeds due to extensive polymorphisms of the mtdna control region. Taken together, the present results as well as previous studies show that the mtdna control region of the domestic dog is not suitable for docu-

Genetic relationships among domestic dogs using AFLP markers 249 menting the relationships among different dog breeds. In our study, the problem will not be helped by additional sequence information from other mtdna genes. Even breeds with ancient histories, such as the dingo, had only a few substitutions different from recent dog breeds. It is likely that ancient polymorphisms have been retained in dogs from a wolf ancestor and unique mutations that have occurred since their founding have been widely spread among breeds by interbreeding. Consequently, phylogenetic trees of dog sequences will be reticulated with respect to breeds. Therefore, the relationships between dog breeds should be surveyed in nuclear genomic DNA levels. In this study, relationship trees based on AFLP markers accorded with the grouping of dog breeds on the basis of the areas where they are currently living, or originated from. In spite of the drawback with low bootstrap support, AFLP markers can be used as a possible tool to understand relationships of local breeds. In fact, the low bootstrapping support value is not surprising considering the complex and recent origin of dogs. The AFLP analysis showed different aspects of the genetic relationships compared with the molecular analysis of the mtdna. Discrepancies between the AFLP and the mtdna analysis can be explained by the difference in the inheritance mode between nuclear- and mtdna-coded genes. Domestic dogs have been well characterized by various morphological and behavioral traits. These characteristics may be reflected in nuclear genomic DNA, which is transmitted biparentally, rather than mtdna, which is maternally inherited. Some differences between breeds can be explained in gene recombination or mutation that has occurred in the nuclear genome. Because the closer kinship seems to have a higher genetic similarity, the relationships may be reflected in the nuclear genome. The most common method to improve the genetic status of domesticated animals by breeding is to introduce useful genes from males. Since domestic dogs have been improved in this way, the results may be reflected in the nuclear genome of the male. In contrast, the strictness of the maternal inheritance of mtdna and the lack of genetic recombination in mtdna make mtdna analysis impossible for the explanation of indigenous morphological and behavioral traits in domestic dogs. In addition, nucleotide sequences of a single gene such as specific mtdna gene may follow an independent evolutionary process compared to those of the nuclear genome. Recently, several studies concerning the phylogenetic relationship among local dog breeds have been performed in nuclear genomic levels. Tanabe et al. (1991) analyzed 25 blood protein loci in about 3000 specimens of 40 breeds, which included Japanese, Korean, other Asian, and European dog breeds. The authors suggested that most extant Japanese native dogs had been affected by two major genetic components and hypothesized influxes from two distinct routes in ancient times. Additionally, Jeong et al. (1997) reported that RAPD analysis might be a valuable tool for the genetic relationship analysis of dog breeds. Frequency differences in nuclear markers using microsatellites have been successfully utilized in the population genetics of Canidae family (Roy et al. 1994; Zajc et al. 1997; Kim et al. in press). Accordingly, a study based on the nuclear genome may be the best choice for resolving relationships of dog breeds with indigenous morphological and behavioral traits. We would like to thank to anonymous revisers for helpful discussions and comments on the manuscript. This work was supported by a grant from the Creative Research Initiative Program and by grants from the Cultural Properties Administration of Korea. REFERENCES Ajmone-Marsan, P., Valentini, A., Cassandro, M., Vecchiotti- Antaldi, G., Bertoni, G., and Kuiper, M. (1997) AFLP TM markers for DNA fingerprinting in cattle. Anim. Gen. 28, 418 426. Aquadro, C. F., and Greenberg, B. D. (1983) Human mitochondrial DNA variation and evolution: Analysis of nucleotide sequences from seven individuals. Genetics 103, 287 312. Brown, J. R., Beckenbach, A. T., and Smith, M. J. (1993) Intraspecific DNA sequence variation of the mitochondrial control region of White Sturgeon (Acipenser transmontanus). Mol. Biol. Evol. 10, 326 341. Felsenstein J, 1993. PHYLIP-phylogenetic inference package, version 3.5c. University of Washington, Seattle. Girman, D. J., Kat, P. W., Mills, M. G., Ginsberg, J. R., Borner, M., Wilson, V., Fanshawe, J. H., Fitzgibbon, C., Lau, L. M., and Wayne, R. K. (1993) Molecular Genetic and Morphological Analyses of the African Wild Dog (Lycaon pictus). Journal of Heredity 84, 450 459. Jackson, D. A., Somers, K. M., and Harvey, H. H. (1989) Similarity coefficients: measures of co-occurrence and association or simply measures of occurrence? Am. Nat. 133, 436. Janczewski, D. N., Modi, W. S., Stephens, J. C., and O Brien, S. J. (1995) Molecular evolution of mitochondrial 12S RNA and cytochrome b sequences in the pantherine lineage of Felidae. Mol. Biol. Evol. 12, 690 707. Janssen, P., Maquelin, K., Coopman, R., Tjernberg, I., Bouvet, P., Kersters, K., and Dijkshoorn, L. (1997) Discrimination of Acinetobacter genomic species by AFLP fingerprinting. Int. J. Syst. Bacteriol. 47(4), 1179 1187. Jeong, H. W., Kim, K. S., and Ha, J. H. (1997) Analysis of Phylogenic Relationships among the Asian 8 Dog Breeds (Canis familiaris) through Randomly Amplidied Polymorphic DNA. Korean J. Genetics 19, 143 149. Kaneda, H., Hayashi, J., Takahama, S., Taya, C., Lindahl, K. F., & Yonegawa, H. (1995) Elimination of paternal mitochondrial DNA in intraspecific crosses during early mouse embryogenesis. Proc. Natl. Acad. Sci. USA 92, 4542 4546. Kim, K. S., Lee, S. E., Jeong, H, W., and Ha, J. H. (1998) The Complete Nucleotide Sequence of the Domestic Dog (Canis familiaris) Mitochondrial Genome. Mol. Phyl. Evol. 10(2), 210 220. Kim, K. S., Tanabe, Y., Park, C. K., and Ha, J. H. (2001) Genetic variability in East Asian dogs using microsatellite loci analysis. J. Heredity in press. Kimura, M. (1980) A simple method for estimating evolutionary

250 K. S. KIM et al. rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol., 16, 111 120. Maniatis, T., Sambrook, J., and Fritsh, E. F. (1982) Molecular Cloning; A labolatory Manual. 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Okumura, N., Ishiguro, N., Nakano, M., Matsui, A., and Sahara, M. (1996) Intra- and interbreed genetic variatins of mitochondrial DNA major non-coding regions in Japanese native dog breeds (Canis familiaris). Anim. Genet. 27, 397 405. Ovilo, C, Cervera, M. T., Castellanos, C., and Martinez-Zapater, J.M. (2000) Characterization of Iberian pig genotypes using AFLP markers. Anim. Genet. 31, 117 122. Page, R. D. (1996) TREEVIEW: an application to display phylogenetic trees on personal computers. Computer Application in the Biosciences 12, 357 358. Qi, X., and Lindhout P. (1997) Development of AFLP markers in barley. Mol. Gen. Genet. 254, 330 336. Ramharack, R., and Deeley, R. G. (1987) Structure and evolution of primate cytochrome c oxidase subunit II gene. J. Biol. Chem. 262, 14014 14021. Roy, M. S., Geffen, E., Smith, D., Ostrander, E. A., and Wayne, R. K. (1994) Patterns of differentiation and hybridization in north American wolflike canids, revealed by analysis of microsatellite loci. Mol. Biol. Evol. 11, 553 570. Ruvulo, M., Zehr, S., Dornum, M. V., Pan. D., Chang. B., and Lin, J. (1993) Mitochondrial COII sequences and modern human origins. Mol. Biol. Evol. 10, 1115 1135. Sanger, F., Nichlen, S., and Coulson, A. R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA. 74, 5463 5467. Simons, G., van der Lee, T., Diergaarde, P., et al. (1997) AFLPbased fine mapping of the Mlo gene to a 30-kb DNA segment of the barley genome. Genomics 44(1), 61 70. Saitou, N., and Nei, M. (1987) The neighbor-joining method: a new method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4, 406 425. Sneath, P. H. A., and Sokal, R. R. (1973) Numerical Taxonomy. W. H. Freedman and Co., San Francisco. Tanabe, Y., Ota, K., Ito, S., Hashimoto, Y., Sung, Y. Y., and Faruque, M. O. (1991) Biochemical genetic relationships among Asian and European dogs and the ancestry of the Japanese dog. J. Anim. Breed. Genet. 108, 455 478. Tosto, D. S., and Hopp, H. E. (2000) Suitability of AFLP markers for the study of genomic relationships within the Oxalis tuberosa alliance. Plant Syst. Evol. 223, 201 209. Tsuda, K., Kikkawa, Y., Yonekawa, H., and Tanabe, Y. (1997) Extensive interbreeding occurred among multiple matriarchal ancestors during the domestication of dogs: Evidence from inter- and intraspecies polymorphisms in the D-loop region of mitochondrial DNA between dogs and wolves. Genes Genet. Syst. 72, 229 238. Turnbell, P. F. and Reed, C. A. (1974) The fauna from the terminal Pleistocene of Pale gawra Cave, a Zarzian occupation site in northeastern Iraq. Fieldiana Anthropol. 63, 81 146. Van der Lee, T., De Witte, I., Drenth, A., Alfonso, C., and Govers, F. (1997) AFLP Linkage Map of the Oomycete Phytophthora infestans. Fungal Genet. Biol. 21(3), 278 291. Vila, C., Savolainen, P., Maldonado, J. E., Amorim, I. R., Rice, J. E., Honeycutt, R. L., Crandall, K. A., Lundeberg, J., and Wayne, R. K. (1997) Multiple and ancient origins of the domestic dog. Science 276, 1687 1689. Vila, C., Maldonado, J. E., and Wayne, R. K. (1999) Phylogenetic relationships, evolution, and genetic diversity of the domestic dog. J Hered 90(1):71 77. Vos, P., Hogers, R., Bleeker, M. et al. (1995) AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res. 23(21), 4407 4414. Wayne, R. K., and Jenks, S. M. (1991) Mitochondrial DNA analysis implying extensive hybridization of the endangered red wolf, Canis rufus. Nature 351, 565 568. Zajc, I., Mellersh, C. S., and Samson, J. (1997) Variability of canine microsatellites within and between different dog breeds. Mammalian Genome 8, 182 185.