188 
Fishery Bulletin 106(2) 
an extension of 45 sec. at 72°C. This was followed by a 
final extension of 5 min. at 72°C. The cycling conditions 
for the C. ocyurus assay were the following: 2 min. at 
95°C, 35 cycles each consisting of 30 sec. denaturation 
at 95°C, 40 sec. initial annealing temperature at 64°C 
which decreased by 0.5°C per cycle for four cycles and 
62°C thereafter and an extension of 45 sec. at 72°C. 
This was followed by a final extension of 5 min. at 72°C. 
Lastly, the cycling conditions for the C. philadelphica 
assay were: 2 min. at 95°C, 35 cycles each consisting of 
30 sec. denaturation at 95°C, 40 sec. initial annealing 
temperature at 60°C which decreased by 0.5°C per cycle 
for four cycles and 58°C thereafter and an extension of 
45 sec. at 72°C. This was followed by a final extension 
of 5 min. at 72°C. An aliquot (4 pL) from each amplifi- 
cation was analyzed by agarose gel electrophoresis (3% 
agarose TAE gel, 75V). The sizes of the PCR products 
were estimated with either a 123-bp or 50-bp ladder 
(Promega, Madison, WI). 
Validation of the PCR assays for identifying larval fish 
Following assay development, the primer pairs were 
tested for cross-reactivity with a panel of DNAs that 
included the 10 reef fish species listed previously (20-50 
ng of DNA per PCR reaction). DNA from 76 Centropi'istis 
larvae was extracted using the methods described above. 
The larvae were assayed by using 20-50 ng of DNA 
from each specimen. Each PCR assay included a posi- 
tive control, a negative control, a blank DNA extraction 
control, and two PCR inhibition controls. The positive 
control incorporated 20 ng of appropriate Centropris- 
tis DNA in the reaction mixture. The negative control 
substituted lx PCR buffer for DNA to confirm that the 
reagents were not contaminated with target DNA. The 
blank extraction control was included to assess pos- 
sible cross-contamination during the extraction proce- 
dures. The inhibition control consisted of spiking 20 ng 
of appropriate Centropristis DNA into two arbitrarily 
chosen larval fish DNA samples. The PCR inhibition 
controls confirmed that negative PCR reactions were 
due to the absence of appropriate Centropristis DNA and 
not from PCR inhibition. The C. striata and C. ocyurus 
assays were tested on all of the larvae in the collection. 
The C. philadelphica assay was applied by process of 
elimination to larvae that did not yield results from the 
C. striata or C. ocyurus assays. 
Development of RFLP assays for identifying larval fish 
Forward and reverse genus-specific primers for three 
species of Centropristis were designed from the 3' SSU- 
5' LSU sequence alignment described above. The forward 
primer, CentropFWl (Table 3), overlapped the boundary 
of the 3' end of the SSU rRNA gene and the 5' end of the 
ITS1 spacer (Fig. 2A). The reverse primer, CentropRevl 
(Table 3), overlapped the boundary of the 3' end of the 
ITS2 spacer and the 5' end of the LSU rRNA gene (Fig. 
2A). The primer pair yielded an approximate 1200-bp 
PCR product and was tested for cross-reactivity against 
the 10 non -Centropristis reef fishes listed above. The 
RFLP-PCR reaction mix is listed in Table 2. Twenty to 
50 ng of genomic DNA was used per PCR reaction. PCR 
amplifications were conducted with an MJ Research 
PTC-150 MiniCycler under the following cycling condi- 
tions: 2 min. at 95°C, 40 cycles each consisting of 30 
sec. denaturation at 95°C, 40 sec. initial annealing 
temperature at 63°C which decreased by 0.5°C per cycle 
for six cycles and 60°C thereafter, and an extension of 
1 min. at 72°C. This was followed by a final extension 
of 5 min. at 72°C. A 4-pL aliquot of each PCR reaction 
was checked for the presence of a specific amplification 
product by agarose gel electrophoresis (2% agarose TAE 
gel, 50 V) and ethidium bromide staining. The sizes of 
the PCR products were estimated by using the DNA 
molecular weight marker IX (Roche Diagnostics, India- 
napolis, IN). 
RFLP analysis of Centropristis PCR products was 
simulated with the RFLP analysis tool in the Vector 
NTI program (Informax Inc., Bethesda, MD) to identify 
species-specific RFLP patterns. The restriction enzyme 
Alu I (New England Biolabs, Beverly, MA) was selected 
on the basis of the distinct restriction patterns pre- 
dicted for each Centropristis species. Restriction digests 
were performed in 30-pL reactions containing 3 uL of 
New England Biolabs lOx reaction buffer #3, 25 pL of 
PCR product, 2000 units of Alu I enzyme, and were 
incubated at 37°C for 2 hours. Restriction fragments 
were separated by gel electrophoresis on 3% agarose, 
Tris-Borate EDTA NuSieve 3:1 gels (Cambrex Bio Sci- 
ence Rockland, Inc., Rockland, ME) and fragments sizes 
were estimated by using a 50-bp DNA ladder (Promega, 
Madison, WI). 
Results 
Species-specific molecular assays based on differences in 
the ITS rDNA regions were developed successfully for the 
identification of Centropristis larvae. The approach we 
employed in developing the molecular assays was 1) to 
identify universal PCR primer sites that would amplify 
the ITS regions of many fish species; 2) to sequence the 
ITS regions from three Centropristis species and other 
co-occurring reef species; 3) to develop species-specific 
PCR and RFLP assays for the Centropristis species 
based on alignments of the resulting sequences; and 
4) to use the assays to identify field-collected larval 
Centropristis. 
The universal PCR primers (Table 1) were successful 
in amplifying the ITS regions of many reef fish species. 
The primers generated an approximate 1200-1300 base 
pair product that varied with the species. A phyloge- 
netic analysis of the Centropristis ITS sequence align- 
ments showed that between-species sequence divergence 
in this region was much greater than within-species 
divergence (Fig. 3). This variation made it possible to 
develop species-specific PCR assays for C. striata, C. 
ocyurus, and C. philadelphica. The primers were tested 
for cross-reactivity against ten other related reef fish 
