Vandersea et at: Identification of larval Centropristis spp. using ribosomal DNA-specific molecular assasy 
189 
1.00 
1.00 
1.00 
Diplectrum formosum 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis philadelphica 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis ocyurus 
Centropristis striata 
Centropristis striata 
Centropristis striata 
1 00 |L Centropristis striata 
Centropristis striata 
Centropristis striata 
Centropristis striata 
Centropristis striata 
Centropristis striata 
0.1 Substitutions per site 
Figure 3 
Phylogeny showing the relationship between black sea bass ( Centropristis 
striata), bank sea bass ( C . ocyurus), and rock sea bass ( C . philadelphica) 
based on ITS1, 5.8S, and ITS2 rDNA sequences. Sand perch ( Diplectrum 
formosum) served as the outgroup. Branch support for each species was 
100% (1.0). The scale at the bottom of the phylogeny is proportional to 
the branch lengths in the phylogenetic tree that correspond to a diver- 
gence of 0.1 nucleotide substitutions per base pair in the DNA sequences. 
The within-species sequence variation was consistently less than that 
observed between species. 
species and they did not amplify non- 
target species (Fig. 4). The accuracy of 
the species-specific assays was checked 
by PCR assaying 15 adult and juvenile 
sea basses (5 of each species). The PCR 
assays were successful in identifying the 
target species and did not cross-react 
among the three Centropristis species. 
We validated the PCR assays by using 
them to assess the identity of 76 Centro- 
pristis larvae (Fig. 5). Sixty-nine of the 
76 larvae (91%) were identified success- 
fully: 33 C. striata, 32 C. ocyurus, and 
four C. philadelphica. DNA from four of 
the larvae failed to PCR amplify and in 
three cases the PCR assay results were 
inconclusive. The four DNA samples that 
failed to amplify were tested for PCR 
inhibition by spiking a small amount of 
target Centropristis plasmid DNA into 
PCR reaction mixes containing DNA 
from the larvae. These spiked controls 
all amplified, indicating that PCR in- 
hibition was not a factor. DNA was re- 
extracted from the larvae and attempts 
were made to re-amplify it with the spe- 
cies-specific primers and the universal 
primers. Neither primer sets produced a 
PCR product, indicating that DNA deg- 
radation had occurred to the extent that 
amplifiable DNA could not be isolated 
from these four specimens. 
The three inconclusive results were 
symptomatic of DNA cross-contamina- 
tion between larval samples. Both the 
C. striata and C. ocyurus specific primer 
sets produced correct size PCR products 
for these samples. DNA from these lar- 
vae was re-extracted in an attempt to 
eliminate the cross-contamination, but 
multiple PCR assay attempts yielded 
the same ambiguous results. The issues 
of cross-contamination and methods to 
avoid it will be discussed below. 
RFLP assays were also developed for 
each species as an alternative to the 
PCR assays. Genus-specific PCR primers 
(Table 3) that flanked the ITS regions 
were used to amplify DNA from identi- 
fied adult sea basses. The genus-specific 
primer sites were located very near the universal prim- 
er sites but were designed to amplify only Centropris- 
tis DNA. The primers were tested for cross-reactivity 
against the same related reef species that were used 
in the PCR assays. Cross-reactivity did not occur (Fig. 
6A). The PCR products were digested with the Alu 
I restriction enzyme. Electrophoresis yielded unique 
banding patterns (Fig. 6B). Digests of C. striata yielded 
fragments of 446, 418, 218, 125, and 4 base pairs. The 
restriction fragment sizes for C. ocyurus were 449, 362, 
245, 120, 51 bp, and 4 base pairs. For C. philadelphica 
the fragments were 481, 261, 187, 185, 121, 58, and 4 
base pairs. The advantages and disadvantages of this 
diagnostic approach are discussed below. 
Discussion 
Ribosomal DNA-specific assays for Centropristis larvae 
have proven to be robust and accurate. The difficul- 
