Species identification of Anguilla japonica by real-time PCR based on a sequence detection system: a practical application to eggs and larvae
Y. Minegishi, J. Aoyama, and K. Tsukamoto: Ocean Research Institute, University of Tokyo
1-15-1 Minamidai, Nakano-ku, Tokyo 164-8639, Japan. 6512; fax:
To develop a practical method for identifying Japanese eel Anguilla japonica eggs and larvae to species by a sequence detection system using a real-time polymerase chain reaction (PCR), we examined (i) the sensitivity of the system using samples at various developmental stages, and (ii) influences of intra- and interspecific DNA sequence variations in the PCR target region. PCR amplifications with extracted DNA solution at 7.0 ng ml21 or lower were efficient at distinguishing A. japonica from other anguillids. A single egg at the gastrula or later developmental stages could also be identified. Two sequence variations in the PCR target region were observed in 2 out of 35 A. japonica collected from three localities, and from four year classes at a single locality. These mutations, however, did not affect the result of species identification achieved by A. japonica-specific PCR primers and probe. The accuracy of this PCR-based method of species identification will help in field surveys of the species.
The Japanese eel Anguilla japonica is commercially very important
in eastern Asia. The stock, however, has been decreasing for the
past three decades (Dekker, 2003). To compensate for the shortage
of eels, artificial fertilization has been attempted since the 1960s,
but it has not yet been achieved practically. A problem in the
current culture technique is the lack of information on the
reproductive ecology of eels. The spawning area of Japanese eels was
recently pinpointed as at the West Mariana Ridge in the western
North Pacific (Tsukamoto, 2006), but eggs have not yet been
collected, so the spawning ecology of eels remains elusive.
Field surveys of spawning eels require rapid and accurate
species identification. The eel egg, however, is difficult to identify
morphologically because of the few diagnostic characters. During a
research cruise in 1998, three morphologically eel-like eggs were
collected. However, these were found to be probably Serrivomer,
an anguilliform fish, but not Anguilla, by DNA sequences
determined in the laboratory after the cruise (Aoyama et al., 2001).
Morphological characters change with developmental stage,
so molecular characters have emerged as crucial in identifying
The basis of the genetic species identification method for
Japanese eels using a real-time polymerase chain reaction (PCR)
has been developed by Watanabe et al. (2004). A PCR-based
identification detects a species-specific DNA nucleotide sequence
with designed primers and probe (sequence detection system,
SDS). When the method is employed, however, other factors
such as template DNA concentrations and sequence variations in
the target DNA region by PCR need to be considered because
they can affect PCR amplification efficiency, so may cause errors
in species identification.
The aim of this study was to develop a practical species
identification method of A. japonica at an early life stage, e.g. eggs and
larvae. Therefore, we investigated (i) the sensitivity (a range of
template DNA concentrations) of the system, (ii) system
appropriateness for single eggs at various stages of development, and (iii)
the effect of intra- and interspecific sequence variations in the
SDS target region of the 16S ribosomal RNA (rRNA) gene of
mitochondrial DNA. Finally, we determined a practical criterion to
identify A. japonica by SDS with real-time PCR.
Material and methods
To examine the sensitivity of PCR in the system, we used DNA
samples from A. japonica, Anguilla marmorata, and Anguilla
bicolor pacifica which had been stored in TE buffer (Aoyama
et al., 1999; Minegishi et al., 2005). The last two species were
analysed because their larvae have been collected around the spawning
area of A. japonica (Aoyama et al., 1999; Miller et al., 2002; Kuroki
et al., 2006).
To test the appropriateness of PCR for a single egg, material
obtained by artificial fertilization (Irago Institute, Aichi, Japan)
was sampled at five developmental stages (unfertilized, 4, 12, 27,
and 39 h after fertilization) and immediately stored in 99%
ethanol until DNA extraction. Egg developmental stages after
fertilization roughly correspond to the morula, gastrula to embryonic
body formation, eye vesicles and heart formation, and hatching
periods, respectively (Yamamoto et al., 1975; Yamamoto, 1981).
Sequence variations in the PCR target region were investigated
as follows. A total of 15 A. japonica from three localities (Miyagi
and Ibaraki in Japan, and Taiwan, n 5 each) was analysed for
geographic variation. Temporal variations among different year
classes of A. japonica were examined using 20 eels that had
recruited to Tanegashima Island, Japan, in 1991, 1992, 1995, and
1998 (n 5 each). Interspecific variations were characterized
with 22 A. marmorata from six localities (Japan, Taiwan,
Indonesia, Fiji, Tahiti, and Madagascar), and four A. bicolor
pacifica (from the Indonesian Sea).
Sensitivity of PCR
To examine the sensitivity of PCR in the species-identification
system, a real-time PCR was performed with a wide range of
extracted DNA solutions, from 2.0 pg ml21 to 7.0 ng ml21, as
template DNA in the three species of A. japonica, A. marmorata, and A.
bicolor pacifica. Real-time PCRs were carried out on an Applied
Biosystems 7300 real-time system, with a total of 20 ml reaction
volume containing 10 ml of 2 TaqMan MasterMix (Applied
Biosystems), 900 nM each of forward (Aja16S-L3, 50-AAT CAG
TAA TAA GAG GGC CCA AGC-30) and reverse primers
(Aja16S-H3, 50-TGT TGG GTT AAC GGT TTG TGG TA-30),
200 nM of TaqMan probe (50-CAC ATG TGT AAG TCA GAA
CGG ACC GAC C-30), and 3 ml of template DNA at various
concentrations. Except for the PCR cycle number (50), amplification
parameters, primers, and probe sequences were the same as used
by Watanabe et al. (2004).
Appropriateness of PCR for a single egg
DNA extraction from a single egg was performed with three
volumes (50, 100, and 300 ml) of 5% w/v Chelex resin solution
(BioRad). After homogenization of a single egg in a certain
volume of solution, genomic DNA was prepared by incubation
at 958C for 15 min. PCRs were conducted as described above.
Intra- and interspecific variation
For a total of 35 A. japonica, 22 A. marmorata, and 4 A. bicolor
pacifica, partial DNA nucleotide sequences of 16S rRNA gene of
mitochondrial DNA were determined using a pair of PCR
primers L1854 and H3058 (Inoue et al., 2001). PCR and
sequencing methods are described by Minegishi et al. (2005). For
specimens with a mutation in the PCR target region, a real-time PCR
was carried out to evaluate the effect of such mutation on
Sensitivity of PCR
Anguilla japonica was clearly discriminated from the other
congeners at a wide range of template DNA concentrations. In A.
japonica, a PCR with 7.0 ng ml21 of extracted DNA as template DNA
yielded cycles at threshold (Ct) values ranging from 22.0 to 24.3
Figure 1. Relationship between concentrations of template DNA
and cycles at threshold (Ct) by real-time PCR. White and black
circles indicate A. japonica and other congeners, respectively.
The grey underlay shows the template DNA concentration
recommended by the manufacturer (1.33 133 ng ml21; Applied
(Figure 1). Ct values from 25.2 to 32.4 were observed when the
PCR was carried out with less template DNA, at 0.7 and
0.07 ng ml21. Even a reaction with 2.0 pg ml21 of template DNA
of A. japonica yielded 37.3 of Ct. In contrast, PCR with
7.0 ng ml21 of template DNA yielded Ct values of 40.0 for the
two anguillids other than A. japonica. Moreover, amplifications
were not detected with template DNA at 0.7 ng ml21 in either of
A. marmorata or of A. bicolor pacifica (Figure 1).
Appropriateness of PCR for a single egg
With DNA samples from an unfertilized egg and material 4 h after
fertilization, PCR amplifications were not detected regardless of
the volume of Chelex extracting solution. In contrast,
amplifications were observed in eggs 12 h or more after fertilization
(Ct 21.1 35.3), except when DNA was extracted with 100 ml
of solution for an egg 12 h after fertilization. DNA concentrations
extracted from a single egg 39 h after fertilization with 50 ml of
Chelex resin solution were 38.2 77.3 ng ml21.
Intra- and interspecific variations
About 700 bp of partial nucleotide sequences of 16S rRNA gene
were determined for the three anguillids. In all, five haplotypes
were found among 15 A. japonica collected from Miyagi, Ibaraki
(Japan), and Taiwan. Four had mutations outside the target
region of the species identification. One eel from Ibaraki had
four continuous guanines (G) from the position of the 15th
nucleotide from the 50 end of the priming site of the forward
primer, whereas the others had three guanines (Figure 2).
Among 20 glass eel stages of A. japonica collected at
Tanegashima Island in different years, five haplotypes were
found. Four contained mutations outside the target region. One
eel collected in 1998 had a nucleotide substitution from thymine
(T) to cytosine (C) at the fifth position from the 50 end of the
probe hybridize region (Figure 2). These two types of mutation
in A. japonica were not observed in the other two congeners.
Species identification of Anguilla japonica by real-time PCR
In all, nine haplotypes were found among 22 A. marmorata,
and four A. bicolor pacifica had an identical sequence.
Nucleotide variations occurred outside the target region.
Subsequently, a real-time PCR was performed using two
A. japonica from Ibaraki and Tanegashima with a mutation in
the hybridized region. With various concentrations of extracted
DNA as template DNA (1.0 57.0 ng ml21), these eels yielded
positive results, with Ct values of 18.9 27.0.
We tested experimental conditions to optimize the PCR-based
species identification protocol of Watanabe et al. (2004) for
practical application to eggs and larvae of Japanese eels. According to
the relationship between template DNA concentration and PCR
amplification efficiency, there were obvious differences in Ct
values between Japanese eels and its congeners at 7.0 ng ml21 of
extracted DNA as template DNA (Figure 1). Moreover, PCR
amplifications were detected in Japanese eels (Ct 37.3) even
with a DNA solution of 2.0 pg ml21, considerably lower than the
concentrations recommended by the manufacturer (1.33
133 ng ml21; Applied Biosystems). In contrast, amplification was
not observed in the other anguillids with DNA solutions at
0.7 ng ml21. Therefore, to distinguish A. japonica from the other
anguillids, the template DNA solution needs to be adjusted to
7.0 ng ml21.
We also confirmed the suitability of the method for single eel
eggs at the gastrula or later stages. Indeed, the DNA concentration
from a single egg was sufficiently high to be applied as template
DNA in this system (38.2 77.3 ng ml21), so, to prepare them for
optimum DNA concentration (7.0 ng ml21), the extracted DNA
solution using 50 ml of Chelex resin solution should be diluted
ten times. DNA samples from unfertilized eggs or those at the
morula stage were not feasible for use by the current method.
This may be due to the low DNA availability in the eggs at
such early developmental stages, so undeveloped eggs need to be
incubated before DNA extraction. Further, greater sensitivity
than in the current system will be necessary to identify unfertilized
Subsequently, we examined the influences of DNA nucleotide
sequence variations on the results of the PCR. Only 2 out of 35
A. japonica had mutations within the hybridized region
(Figure 2), and these did not affect species identification by
PCR. On the other hand, there was no mutation within the
hybridized region in either A. marmorata or A. bicolor pacifica.
Therefore, the nucleotide sequence of the target region in the
present protocol is likely to be conservative at an intraspecific
level, but variable at an interspecific level, suggesting that this
region is suitable for species identification of Japanese eels based
on SDS. Considering the sensitivity of this PCR-based
species-identification technique, it is unlikely that A. japonica
would not be detected (pseudo-negative) and the other congeners
incorrectly identified as A. japonica (pseudo-positive).
The method described here has, in fact, contributed to the
recent discovery of the spawning area of the Japanese eel in a
seamount area of the southern West Mariana Ridge (Tsukamoto,
2006). Additionally, the method works for boiled and baked eels
(TY, unpublished data). The decline of the Japanese eel stock
and its relatively high commercial value in Japan have incidentally
resulted in an escalation of imports of eels from elsewhere and the
camouflage of commercial products. Therefore, the method can be
used not only for ecological surveys but also to monitor products
for conservation reasons, and to identify the international source.
We thank Yoshiaki Yamada of Irago Institute for providing the
eggs. The study was partly supported by Grants-in-Aid for
Creative Scientific Research No. 12NP0201 (DOBIS) from the
Ministry of Education, Culture, Sports, Science and Technology
of Japan and the Eel Research Foundation from Nobori-kai.