4274 Electrophoresis 2007, 28, 4274–4281

James Chun-I Lee1 Research Article

Li-Chin Tsai2
Yuan-Yang Kuan3
Wen-Hsien Chien3 Racing pigeon identification using STR and
Kai-Tai Chang4
Cheng-Hsien Wu4 chromo-helicase DNA binding gene markers
Adrian Linacre5
Hsing-Mei Hsieh2 Pigeon racing appeals to many in Taiwan, due in part to the potential large financial gains
1
based on illegal betting. The races are unregulated with frequent examples of fraud, such as
Department of Forensic Medicine,
College of Medicine, substitution of one bird for a substandard one. There is no test available to reliably verify the
National Taiwan University, bloodline of pigeons and thus help to resolve such disputes. In this study, we describe a
Taipei, Taiwan multiplex PCR amplification system combining 7 STR loci and a chromo-helicase DNA
2
Department of Forensic Science, binding gene (CHD) marker for the identification of individual pigeons. The cumulative
Central Police University,
Kwei-San, Taoyuan, Taiwan power of discrimination (CPd) of the 7 STR loci was 0.99999234 based upon our population
3
Forensic Biology Office, study. The cumulative probability of paternity (CPP) when used in paternity testing of 17
Criminal Investigation Bureau, pigeon families ranged from 97.36 to 99.99% and the combined probability of exclusion
Taipei, Taiwan
4
Forensic Science Section,
(CPE) was 0.9325 for these seven STR markers. The statistical data illustrates the potential
Taipei County Government of this system to be used in pigeon individualization and paternity testing. Furthermore, the
Police Bureau, established STR system could be also used in the other areas, such as ecology, population
Banciao City, Taipei County, genetics, and avian breeding programs.
Taiwan
5
Centre for Forensic Science,
Department of Pure & Applied Keywords:
Chemistry, Chromo-helicase DNA binding gene marker / DNA identification / Multiplex PCR /
University of Strathclyde,
Glasgow, UK STR DOI 10.1002/elps.200700063

Received January 31, 2007

Revised April 23, 2007
Accepted April 23, 2007

1 Introduction puter-coded leg bands have been adopted to avoid mis-

Although pigeon racing is an illegal form of gambling in
Taiwan, it appeals to a large number of the population due to
the large financial gains that are possible. Since it is illegal
and unregulated, cheating occurs frequently. This may be the
substitution of a superior racer for a substandard bird. Iden-
tification of an individual pigeon is necessary to verify win-
ners of races. Blood line and pedigree factors make a partic-
ular pigeon highly valued. The offspring of a valued pigeon
will also be prized. Since there is no reliable method to con-
firm a genuine bloodline of pigeons, trade disputes occur if a
pigeon thought to be of a successful bloodline fails to have a
successful racing career. If such disputes are to be resolved a
reliable method of pigeon identification is necessary. Com-
Correspondence: Dr. Hsing-Mei Hsieh, Department of Forensic
Science, Central Police University, 56 Shu-Jen Road, Kwei-San,
Taoyuan 33334, Taiwan
E-mail: mei@mail.cpu.edu.tw
Fax: 1886-3-3275907
Abbreviations: CHD, chromo-helicase DNA binding gene; CPd,
cumulative power of discrimination; CPE, combined probability
of exclusion; CPP, cumulative probability of paternity
identification, but a more powerful identification method
will be required when the bands are suspected of being fab-
ricated or tampered. DNA has been used in similar circum-
stances for the identification of pedigree animals [1–4].
Microsatellite DNA or STR typing is one of the most power-
ful tools in individualization and parentage testing not only
in humans but animals [5] and even plants [6]. Micro-
satellites have been studied in a number of avian species,
including a high level of kinship and mixed paternity in a
population of Australian magpies (Gymnorhina tibicen) [7],
microsatellite variation in Red-Winged Blackbirds (Agelaius
phoeniceus) [8], and mutation rates of microsatellite loci of
barn swallows [9]. Characterization of 35 novel microsatellite
DNA markers from the duck (Anasplatyrhynchos) genome
and cross-amplification in other birds, such as chickens,
geese, and peacocks, were reported by Huang et al. [10].
Seven polymorphic STR markers have been isolated from
the domestic pigeon (Columba livia, var. domestica) [11]. Of
these seven STR markers, two had a tetranucleotide repeat
motif and the other five were dinucleotide repeats. A total of
18 STR markers from C. livia, including the above-men-
tioned seven STR loci, are registered in GenBank, although
12 of these are dinucleotide repeat sequences. Dinucleotide
repeat sequences are used less preferentially for genetic

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2007, 28, 4274–4281
identification due to the large stutter products of PCR
amplification and the associated resolution problems.
We report newly isolated STR markers with tri- and tet-
ranucleotide repeat unit for pigeon (C. livia) identification.
These loci are demonstrated to provide a high degree of
confidence in pigeon identification. In addition to the STR
markers of pigeons, the avian sex marker chromo-helicase
DNA binding gene (CHD) [12] was combined into the test.
The multiplex PCR amplification [13] combining the STR
and CHD markers was designed for rapid identification to
resolve identification and paternity disputes arising from
racing pigeons. Identification of individual pigeons will have
implications in legal proceedings relating to antigambling
legislation. Furthermore, these sets of genetic markers can
be used in the other areas, such as ecology to analyze the
genetic distribution, population genetics, and avian breeding
programs.
2 Materials and methods
2.1 Sample collections and DNA extraction
Blood samples from pigeons were collected from the veter-
inary hospital. Buccal swabs were collected from 96 unrelated
pigeons and a further 51 buccal swab samples of 17 pigeon
families with known parents and offspring were collected
from five pigeon farms. DNA from the blood samples was
extracted by the salt/chloroform method [14, 15]. DNA from
the buccal swab samples was extracted with a commercial kit
(Blood and Tissue Genomic Mini Kit, Viogene, Taiwan).
2.2 STR isolation and polymorphic study
STR loci were isolated by the hybrid-capture enrichment
procedure using a modification of the method reported by
Ciofi and Bruford [6, 16]. About 10 mg of DNA from one of
the blood samples was digested completely with the restric-
tion enzyme MboI (GeneMark, Taipei, Taiwan) and the DNA
fragments between 200 bp and 1.2 kb were eluted with 2.5%
agarose gel electrophoresis. Adapters constructed with
SAULA and SAULB oligonucleotides were then ligated to the
ends of the eluted fragments. After PCR amplification, the
products were hybridized with 43 probes which included ten
trinucleotide repeats and 33 sequences of tetranucleotide
repeats. The hybridized DNA fragments were then amplified
and cloned into a vector pOSI-T (GeneMark) from which
colonies were screened by PCR amplification and sequenc-
ing. Sequencing was performed using the BigDye™ Termi-
nator Kit (ABI PRISM™ BigDye Terminator Cycle Sequenc-
ing Ready Reaction Kit). The cycle sequencing products were
detected using an ABI 3730 DNA analyzer.
Ninety-six samples were screened by analysis of their
fragment lengths using a fluorescent dye attached to one
primer of each primer set. LIZ500 (Applied Biosystems) was
used as the internal size standard. The fragment was ana-
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Nucleic Acids 4275
lyzed with an ABI 3100 DNA analyzer. The sequenced sam-
ples were also used as the reference sample for further com-
parison.
2.3 Linkage analysis
Linkage disequilibrium of the STR markers adopted in the
multiplex amplification were analyzed with PowerMarker
(http://statgen.ncsu.edu/powermarker/) and Genepop v3.4
softwares (http://wbiomed.curtin.edu.au/genepop/index.html)
[17].
2.4 Multiplex PCR amplification
Multiplex PCR amplification was performed using multiple
STR markers including the four newly isolated STR markers
in this study, three polymorphic STR markers extracted from
the GenBank database (accession nos. G73190, G73196, and
AF188632) [11], and the bird sex marker CHD [12]. The
primer sequences used and their concentrations are shown
in Table 1. CHD-W gene is on chromosome W and is female
specific, and CHD-Z gene is on chromosome Z and present
in both sexes for most bird species. Amplification was per-
formed in a volume of 10 mL and under the following condi-
tions: 957C for 10 min, 28 cycles of 947C for 1 min, 587C for
1 min and 727C for 1 min, and extension for 45 min at 607C
in a 2400 Applied Biosystems thermal cycler. The PCR prod-
ucts were analyzed using an ABI 3100 DNA analyzer.
2.5 Sensitivity test
The DNA isolated from blood was quantified with absorption
260 nm by spectrophotometer and serially diluted as 4, 1,
0.25, 0.0625, 0.0156, and 0.0040 ng/mL. DNA (1 mL) of each
concentration was used to perform the multiplex amplifica-
tion in a total volume of 10 mL.
2.6 Paternity testing
The buccal swab samples of 17 different pigeon families
were used to test the power of the established system for
paternity testing.
2.7 Statistical analysis
The probability of a match Pm, power of discrimination Pd,
and cumulative power of discrimination (CPd) were calcu-
lated using the following equations [18]:
X
n
Pm ¼ ðP k Þ2 (1)
k¼1
where Pk is the genotype frequency.
Pd ¼ 1
Pm (2)
www.electrophoresis-journal.com
4276 J. C.-I. Lee et al. Electrophoresis 2007, 28, 4274–4281

Table 1. Primer sequences and concentrations for amplification of each locus

Locus Repeat Primera) Sequence (50 –30 ) Concentration

unit (mM)

PG1 (G73190) TATC PG1-F ATGTGTGTTTGTGCATGAAG 0.15

PG1-R ATGAAAGCCTGTTAGTGGAA 0.15
PG2 (G73196) ATGG PG2-F CCTTCCAACCCACATTATT 0.5
PG2-R CCAGCCTAAGTGAAACTGTC 0.5
PG3 (AF188632) GGAT PG3-F ATGGGTTTGGAGATGTTTTG 0.5
PG3-R GATGGAGTTGCTATTTTGCT 0.5
PG4 TCCA PG4-F CCCATCTCCTTGCCTGATGC 0.3
PG4-R CACAGCAGGATGCTGCCTGC 0.3
PG5 TTTG PG5-F GTTCTTGGTGTTGCATGGATGC 0.25
PG5-R AGTTACGAAATGATTGCCAGAAG 0.25
PG6 AAAC PG6-F AAGCAATCAGAACAGTGCTTCG 0.125
PG6-R GTCCCTATGTTGCCTTCCCTC 0.125
PG7 TTG PG7-F CATTGGTCAGGAGGTGGTGGG 0.5
PG7-R TCTGCCACTCACTCGCCCTC 0.5
CHD P8 CTCCCAAGGATGAGRAAYTG 0.5
CHD-R ATGGAGTCACTATCAGATCCAG 0.5

a) The primers of PG1-F, PG6-F, and CHD-R were labeled with a FAM dye; the primers of PG2-F and PG7-F were
labeled with a VIC dye; and the primers of PG3-F, PG4-F, and PG5-F were labeled with a NED dye.

n
CPd ¼ 1
P Pm i (3) 3 Results and discussion
i¼1

3.1 STR isolation and polymorphic study

For paternity testing, the paternity index (PI), combined
paternity index (CPI), cumulative probability of paternity
(CPP), probability of exclusion (PE), and combined prob-
ability of exclusion (CPE) were calculated as follows:
PI ¼ X=Y
where X represents the probability that the putative parents
are the real biological parents based upon the matching
alleles, and Y represents the probability that random
pigeons are the biological parents and share the alleles by
chance. CPI was calculated as the product of each PI.
CPP ¼ CPI=ðCPI þ 1Þ
PE was obtained from the formula


PE ¼ H2 1
ð1
H ÞH2
where H represents the heterozygosity [19]. CPE was calcu-
lated using the equation:
n loci obtained in this study were consistent with the report by
CPE ¼ 1
P ð1
PEi Þ
i¼1
The allele structures from isolated STR loci were verified by
DNA sequencing. From 352 colonies, only four novel STR
loci with a repeat motif not less than trinucleotides were
obtained. The other colonies contained either dinucleotide
repeat loci or no repeat sequences. These four loci were dif-
(4) ferent from those reported by Traxler et al. [11] and registered
in GenBank. The repeat units for the four STR loci are
TCCA, TTTG, AAAC, and TTG. To determine the degree of
polymorphism at these loci, 96 samples were screened by
analyzing the length of each PCR fragment. The locus name,
respective repeat unit, and primer sequences are shown in
Table 1. The allele frequencies and Pd of each locus in pop-
ulation are listed in Table 2. The results showed that these
four markers (PG4, PG5, PG6, and PG7) were sufficiently
(5)
polymorphic for individualization and paternity testing. In
order to amplify these STR markers simultaneously and
raise the CPd value of this system, the above four markers
and three polymorphic STR markers extracted from Gen-
Bank database (accession nos. G73190, G73196, and
(6) AF188632) were all included in a multiplex PCR amplifica-
tion system. The allele frequencies and Pd of these three STR
markers (PG1, PG2, and PG3) are also shown in Table 2.
By combing the newly isolated STR loci with the three
described previously, the CPd was increased significantly as
expected. The results of only the four tri- and tetrameric STR
(7)
Primmer et al. [20], where it was reported that the frequency

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2007, 28, 4274–4281 Nucleic Acids 4277

Table 2. Allele frequencies and Pd of each STR locus of microsatellites in the avian genome was low. Due to the
large stutter products of PCR amplification and the asso-
Allele Locus ciated resolution problems of SSR (simple sequence repeat)
PG1 PG2 PG3 PG4 PG5 PG6 PG7 with dinucleotide repeat sequences, STR markers with
repeat sequences less than trinucleotides were ignored in
3 0.173 this study.
5 0.540 0.122
7 0.065 0.797 0.622
8 0.050 0.038 0.203 0.256 3.2 Linkage analysis
9 0.005
10 0.185 0.020 Linkage disequilibrium of the seven STR markers was ana-
11 0.310 0.082 0.155 lyzed by PowerMarker and Genepop v3.4 softwares. The
12 0.488 0.006 0.090 0.005 results showed that none of the loci were associated with
13 0.339 0.024 0.170 0.356 0.149 any of the other loci using the PowerMarker software, but
14 0.161 0.060 0.505 0.048 the PG2 and PG7 loci exhibited association using Genepop
15 0.012 0.143 0.014 0.452 v3.4. When applying the Bonferroni correction [21] and
16 0.155 0.024
checking the genotypes of PG2 and PG7 for all samples
17 0.149
manually, it was noted that these two loci showed no asso-
18 0.030
Pda) 0.78 0.93 0.86 0.78 0.49 0.72 0.88 ciation. The Pd of PG2 (0.93) was larger than that of PG7
(0.88) (Table 2). Based upon the samples in this study, the
A total of 96 pigeon considered to be unrelated were used to
CPd of the seven STR loci was 0.99999234 and the CPd was
develop an allele database using the seven loci. 0.99993399 when excluding PG7. The loci were shown to be
a) The row of Pd represented the power of discrimination of in Hardy–Weinberg equilibrium based upon the same soft-
each locus. ware.

Figure 1. Example of genotypes obtained from the multiplex amplification system. The PCR products were separated on an ABI 3100 and
alleles designated by use of an allelic ladder and panel developed for these loci.

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

4278 J. C.-I. Lee et al.
3.3 Multiplex PCR amplification
A multiplex PCR amplification system was performed using
the bird sex marker CHD and the seven STR markers. The
primers (CHD-R and P8) for pigeon CHD gene amplifica-
tions were based on the sequences of PCR products from P2
and P8 [12]. The sizes of pigeon CHD-W and CHD-Z genes
were 270 and 289 bp, respectively. The primer sequences and
concentrations are shown in Table 1. Figure 1 is an electro-
pherogram representing an example using this system. The
genotype was confirmed when compared to an allelic ladder
prepared from all alleles analyzed in this study. Genotypes of
Figure 2. Different concentrations of DNA were used to test the sensitivity of the multiplex system. The template ranged from 4 ng (top
panel) through 1, 0.25, 0.0625, 0.0156, and 0.004 ng.
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2007, 28, 4274–4281
the eight loci for this sample using the multiplex system
were consistent with the results predicted from the poly-
morphic study, in which all the STR markers were amplified
in a separate reaction and typed independently.
Initially another STR marker (accession no. AF188631)
was also included in the multiplex system, but it would not
amplify simultaneously with the other eight loci. It may be
due to the interferences between multiple primers, or
unsuitable conditions, highlighting the challenge associated
with a multiplex amplification. More STR markers could be
incorporated into this system to raise the power of dis-
crimination after further optimization.
www.electrophoresis-journal.com
Electrophoresis 2007, 28, 4274–4281 Nucleic Acids 4279

© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

4280 J. C.-I. Lee et al.
Figure 3. Example of the multiplex to illustrate the use in paternity testing. At each of the eight loci there is a matching allele shared by the
daughter when compared to the father and mother.
3.4 Sensitivity test
DNA from the blood samples using serial dilutions were
used to detect the sensitivity of the multiplex system and the
results are shown in Fig. 2. All the eight loci were success-
fully and correctly amplified when the DNA template was not
less than 0.0156 ng. The optimal quantity of DNA was about
0.25 ng when considering the peak intensity and resolution.
Locus drop out was noted when the DNA template was
0.0040 ng or less and as expected the larger loci failed to
amplify compared to the smaller.
The optimal quantity of DNA for human identification is
usually about 1–5 ng, but only about 0.25 ng of pigeon DNA
was needed for the system established in this study, indicat-
ing the sensitive nature of this PCR test.
3.5 Paternity testing
DNA from the buccal swab of 51 samples from 17 families
was typed by this multiplex system. Genotypes of one of the
families are shown in Fig. 3. Matching alleles are shown be-
tween the parents and female offspring. The alleles followed
the rules of inheritance with the alleles transmitted from one
generation to the next. The typing results of the other 16
families also followed these rules. The CPP ranged from
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Electrophoresis 2007, 28, 4274–4281
97.36 to 99.99%. The CPE was 0.9325 for these seven STR
markers, and the CPE was 0.8734 when excluding PG7.
These data confirmed the paternity of these families and
provide a high degree of confidence in resolving paternity
disputes of pigeons.
4 Concluding remarks
This study describes a multiplex PCR amplification to iden-
tify pigeons of C. livia and illustrates its application for for-
ensic science purposes. Eight loci, including seven STR
markers and the sex marker CHD, were successfully ampli-
fied and analyzed simultaneously. The CPd of the seven STR
loci was 0.99999234. When used in paternity testing of 17
pigeon families, the CPP ranged from 97.36 to 99.99%, and
the CPE was 0.9325 for these seven STR markers. The sta-
tistical data show the potential of this system to be used in
pigeon individualization, paternity testing, and in similar
disputes in a forensic context. This system could be a model
for the study of other species of pigeons or even other avian
species. Furthermore, the robust system could be also used
in the other areas, such as ecology, population genetics, and
avian breeding programs.
www.electrophoresis-journal.com
Electrophoresis 2007, 28, 4274–4281
This study was supported in part by the Department of Med-
ical Research in NTUH, Taiwan, ROC.
5 References
[1] Lee, S. Y., Cho, G. J., J. Vet. Sci. 2006, 7, 63–67.
[2] Harrison, S. P., Turrion-Gomez, J. L., Mitochondrion 2006, 6,
53–63.
[3] Kanthaswamy, S., von Dollen, A., Kurushima, J. D., Alminas,
O. et al., Am. J. Primatol. 2006, 68, 73–95.
[4] Ichikawa, Y., Takagi, K., Tsumagari, S., Ishihama, K. et al., J.
Vet. Med. Sci. 2001, 63, 1209–1213.
[5] Dimsoski, P., Croat. Med. J. 2003, 44, 332–335.
[6] Hsieh, H. M., Hou, R. J., Tsai, L. C., Wei, C. S. et al., Forensic
Sci. Int. 2003, 131, 53–58.
[7] Hughes, J. M., Mather, P. B., Toon, A., Ma, J. et al., Mol. Ecol.
2003, 12, 3441–3450.
[8] Williams, L. C., Homan, H. J., Johnston, J. J., Linz, G. M.,
Biochem. Genet. 2004, 42, 35–41.
[9] Brohede, J., Moller, A. P., Ellegren, H., Mutat. Res. 2004, 545,
73–80.
© 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Nucleic Acids 4281
[10] Huang, Y., Tu, J., Cheng, X., Tang, B. et al., Genet. Sel. Evol.
2005, 37, 455–472.
[11] Traxler, B., Brem, G., Muller, M., Achmann, R., Mol. Ecol.
2000, 9, 366–368.
[12] Griffiths, R., Double, M. C., Orr, K., Dawson, R. J. G., Mol.
Ecol. 1998, 7, 1071–1075.
[13] Butler, J. M., Ruitberg, C. M., Vallone, P. M., Fresenius’ J.
Anal. Chem. 2001, 369, 200–205.
[14] Mullenbach, R., Lagoda, P. J., Welter, C., Trends Genet. 1989,
5, 391.
[15] Lee, J. C., Tsai, L. C., Yang, C. Y., Liu, C. L. et al., Electropho-
resis 2006, 27, 3359–3362.
[16] Ciofi, C., Bruford, M. W., Mol. Ecol. 1998, 7, 133–140.
[17] Raymond, M., Rousset, F., J. Hered. 1995, 86, 248–249.
[18] Nei, M., Molecular Evolutionary Genetics, Columbia Uni-
versity Press, New York 1987.
[19] Fischer, R. A., Heredity 1951, 5, 95–102.
[20] Primmer, C. R., Raudsepp, T., Chowdhary, B. P., Moller, A. P.,
Genome Res. 1997, 7, 471–482.
[21] Sankoh, A. J., Huque, M. F., Dubey, S. D., Stat. Med. 1997, 16,
2529–2542.
www.electrophoresis-journal.com