Heist and Gold: Genetic identification of sharks 



55 



polyacrylamide gels and examined by autoradiogra- 

 phy. DNA sequences were read on an IBI gel reader 

 and entered directly into computer text files. Se- 

 quences were confirmed by recording a minimum of 

 two separate sequencing runs through each base. 

 Sequences were aligned by using the ESEE software 

 package (Cabot and Beckenback, 1989). Four cyt b 

 sequences — those of sandbar (C. plumbeus), lemon 

 (Negaprion brevirostris), tiger (Galeocerdo cuvier), and 

 great hammerhead sharks {Sphyrna mokarran ) — were 

 downloaded from the EMBL/NCBI Genbank database 

 (Martin and Palumbi, 1993). All other cyt b sequences 

 and all tRNA^""^ sequences were acquired from our 

 laboratory. 



Sequences from each species between primers 

 Cb3RL and Cb6H were concatenated with sequences 

 of the primers to produce a single sequence for each 

 of the eleven species (Table 2). Sequences were exam- 

 ined for predicted restriction sites with IBI MacVector 

 software (IBI Mac Vector, 1991). One hundred eigh- 

 teen restriction enzymes with unique recognition se- 

 quences were used in the search for restriction sites. 



Amplified "diagnostic" fragments were prepared 

 by amplifying genomic DNA at thirty cycles of 95°C 

 for 1 min, 48 to 52°C for 30 sec, and 72°C for 30 sec. 

 Amplified fragments were ethanol precipitated as 

 above and reconstituted in water. Generally, one 100 

 |iL PCR reaction produced enough diagnostic frag- 

 ment for all seven restriction digestions. Fragments 

 were digested by using manufacturer's buffers and 

 specifications. Most restriction patterns were scored 

 on 2% agarose gels run on IX TAE. Fragments be- 

 tween 40 and 100 bp were scored on either vertical 

 nondenaturing polyacrylamide gels or 2'7( nusieve 3:1 

 agarose gels. Fragments less that 40 bp were not 

 scored, although loss of fragments as small as 24 bp 

 could be inferred from mobility shifts in larger frag- 

 ments. All gels were stained with ethidium bromide 

 and examined under UV light (Sambrook et al., 

 1989). 



Results 



A single mtDNA fragment of 394-396 bp that con- 

 tained the 3' end of cyt b and part of the tRNATHR 

 gene was amplified in all species. Except for two ham- 

 merhead sharks, the fragment was 395 bp in length 

 (Table 2). Cyt b sequences from all eleven species 

 were identical in length and unambiguously aligned. 

 In comparison with the other nine species, the scal- 

 loped hammerhead (S. lewini) possessed a single- 

 base deletion in the tRNA^"'^ sequence, and the great 

 hammerhead possessed a single-base insertion in the 

 tRNA''""'^ sequence. 



In bignose shark (C. altimus), three fragments of 

 sizes 395 bp, 720 bp, and 1040 bp were produced re- 

 peatedly. The additional bands were more pro- 

 nounced at lower (48°C) than higher (52°C) anneal- 

 ing temperatures. Further investigation with addi- 

 tional primers revealed that bignose shark possessed 

 a much larger mitochondrial D-loop region than any 

 other shark in the study. Amplification with a light- 

 strand primer located within cyt b and a heavy- 

 strand primer located within the 12S ribosomal RNA 

 gene produced an approximately 1400-bp fragment 

 in all species except bignose shark, where a single 

 fragment of approximately 2000 bp was produced. 

 We hypothesize that the Cb6H recognition site within 

 the tRNA^"^ gene is duplicated twice within the D- 

 loop of the bignose shark, resulting in a three-banded 

 amplification product. Duplication of segments of 

 flanking regions within the mitochondrial D-loop 

 have been reported in other vertebrates (Broughton 

 and Dowling, 1994, and citations within). To our 

 knowledge, this is the first evidence of this phenom- 

 enon in elasmobranchs. To obtain the single product 

 for sequencing and restricting, the 395-bp fragment 

 was excised from an agarose gel and reamplified to 

 produce sufficient amounts of the diagnostic fragment. 



Of 1 18 restriction enzymes surveyed by MacVector 

 software, 34 were predicted to have restriction sites 

 within one or more of the eleven species. Seven re- 

 striction enzymes (Alul, Ddel. Fokl, Haelll, Hindi, 

 Hinfl, and Rsal) were chosen for screening because 

 use of these seven enzymes allowed all eleven spe- 

 cies to be distinguished (Table 3). The number of 

 sharks whose mtDNA was subjected to restriction- 

 enzyme digestion is given in Table 4. The initial 

 screen of restriction sites was undertaken with only 

 a single sequence from each species. Differences in 

 predicted and observed restriction patterns in three 

 species (sandbar, spinner, and tiger sharks) made it 

 necessary to sequence additional animals in order to 

 investigate whether differences were due to sequenc- 

 ing errors or intraspecific variation. 



Observed restriction patterns in sandbar sharks 

 from the Gulf of Mexico differed by restriction site 

 from the pattern predicted from the published se- 

 quence of Martin and Palumbi (1993) for a sandbar 

 shark (Cpl-A) from Hawaii (MartinM. Although we 

 predicted that Fokl would not cut the sandbar shark 

 fragment, Fokl digestion produced two fragments of 

 310 and 85 bp. We sequenced additional sandbar 

 sharks from the Gulf of Mexico (Cpl-B) and from 

 Hawaii (Cpl-C) and found that both sequences pos- 

 sessed theFo^I restriction site. Sequence of the speci- 



' Martin, A. P. 1997. University of Nevada-Las Vegas, Las Ve- 

 gas, NV 89154-4004. Personal commun. 



