## Introduction

Shark fins are primarily consumed for their translucent collagen fibres served in Asian luxurious shark fin soup1,2. The global shark fin market was estimated to be worth USD $400–550 million annually and Hong Kong is one of the largest shark fin trade markets in the world, accounting for more than half of global trade volume3,4,5. Although the trade volume in Hong Kong has dropped by 30–50% since peaked at 2003 due to public campaign of shark population awareness, Hong Kong is still remaining a major market for fin trading with about 80 nations annually and the import volume of shark fins into Hong Kong was still over 4,500,000 kg in 2013, amounting USD$170 million6,7,8.

Sharks grow slowly and reach their sexual maturity late and they have a low fecundity and reproductive rate. Growing demand of shark fins and meat has led to overexploitation of shark, resulting in serious declines in some of the shark populations9,10. In order to avoid some shark species from overexploitation, the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES) has listed twelve shark species in Appendix II, including, since 2000, great white shark (Carcharodon carcharias), basking shark (Cetorhinus maximus), and whale shark (Rhincodon typus); since 2013, oceanic whitetip shark (Carcharhinus longimanus), porbeagle shark (Lamna nasus), and three species of hammerhead sharks (Sphyrna lewini, Sphyrna mokarran, Sphyrna zygaena); since 2016, all three species of thresher sharks (Alopias pelagicus, Alopias superciliosus, Alopias vulpinus) and the silky shark (Carcharhinus falciformis). International trade of any part of the listed species must be accompanied by permits or licenses to certify that each specimen has been harvested sustainably and legally11. Shark species which were added to the list of CITES Appendix II since 2013 are especially commercially valuable and traded in large numbers3,12.

Species-level identification of shark fin products is critical for effective law enforcement and fisheries management. In order to enforce the CITES obligations and identify illegal shark fin trade, morphological identification or polymerase chain reaction (PCR)-based method is employed for identification of shark fin products traditionally3,13. Regular practices of molecular means for shark species identification include DNA barcoding3,14,15, mini-DNA barcoding for highly degraded DNA16, and species-specific amplification with PCR-based method17,18,19,20,21,22. These genetic approaches for identifying the species-of-origin of shark fins have been carried out since 1990s and they are effective for shark fin samples with different levels of processing. However, these current amplification approaches are time-consuming. The requirement of molecular expertise and sequencing facilities would also be a big hindrance for customs officials. There have also been no species-specific primers targeting the oceanic whitetip shark and the whale shark.

An easy-to-use, rapid, and effective diagnostic method is needed for species-level identification of all twelve CITES-listed shark species. Recently, loop-mediated isothermal amplification (LAMP), a rapid amplification technique using DNA polymerase with strand displacement activity at a single temperature, has been developed23,24. In LAMP reactions, two to three pairs of specific primers are designed to target four to six distinct regions of target DNA and generate loop-formed DNA, which increases sensitivity and specificity when compared to the conventional PCR methods25. Since it does not require a thermal cycler and can be detected by visual observation of fluorescence, it is favourable for field detection26. Recent studies have applied the LAMP technique on species-specific detection of animals27,28 and medicinal material29,30.

Most shark fins in Hong Kong are processed and classified in Chinese-name categories instead of species origin3,12,31. This presents difficulty in species-level identification for the law enforcement as well as for research and development of sustainable shark fisheries. Here, we present a rapid molecular method for on-site identification of shark fins based on species-specific amplification and instant detection for all twelve CITES-listed shark species. We have designed specific primers for both conventional species-specific PCR and LAMP methods based on cytochrome oxidase I (COI) and NADH dehydrogenase subunit 2 (NADH2) sequences. After sample preparation and DNA extraction, we are able to detect all twelve CITES-listed shark species individually within an hour at constant temperatures using these assays. Target shark species could be differentiated from non-target species within three hours from DNA extraction to LAMP assay.

## Materials and Methods

### Sources of samples and genetic identification of samples

In this project, 291 samples of 93 shark and related species from seven orders and 26 families were collected from the collection of The Chinese University of Hong Kong; and donations of Agriculture, Fisheries and Conservation Department of the HKSAR Government; Biodiversity Research Museum of Academia Sinica, Taiwan; Institute of Marine Biology of National Taiwan Ocean University; Kadoorie Farm & Botanic Garden Hong Kong; Ocean Park Hong Kong, and Ocean Park Conservation Foundation Hong Kong (Table 1). Among 291 samples, there were 94 dried processed fin samples, 174 frozen tissues, and 22 genomic DNA. For the 12 CITES-listed shark species, at least two samples were collected for each species. No live specimens were involved in the sample collection and the samples were either dry products from the market or donated by the various institutions. All samples, including dried processed fin samples, were stored at -80 °C until DNA extraction and downstream experiments.

In order to ensure the species identity of the samples and avoid misidentification of samples due to other parties, DNA of all samples were extracted and amplified by PCR amplification. PCR products were sequenced, and the sequencing results were compared to the NCBI database using BLAST analysis. DNA was extracted from the samples using Biomed Genomic DNA/Tissue Extraction Kit (Biomed, Beijing, China) according to the manufacturer’s instructions (~60 minutes). PCR was performed to amplify COI gene using universal fish primers, FISHCOILBC_ts (5′-CACGACGTTGTAAAACGACTCAACYAATCAYAAAGATATYGGCAC-3′), FISHCOIHBC_ts (5′-GGATAACAATTTCACACAGGACTTCYGGGTGRCCRAARAATC-3′)32, FishF1 (5′- TCAACCAACCACAAAGACATTGGCAC-3′) and FishR1 (5′-TAGACTTCTGGGTGGCCAAAGAATCA-3′)33 for all the samples with an expected product size of ~700 bp. Sanger sequencing of PCR products purified with gel extraction kit (Biomed, Beijing, China) was performed by Tech Dragon Ltd., Hong Kong. Samples were identified using BLAST search against nucleotide sequences available on GenBank. The top-match results with a sequence similarity of at least 99% were used for species identification34. Sequences of COI, NADH2 and 12 S rRNA regions of sharks (12265, 1761, and 327 sequences respectively) were downloaded from the NCBI database for alignment. Species-specific primers were designed based on nucleotides unique for each of the 12 CITES-listed species.

### Sensitivity and specificity of the LAMP and PCR assays

The sensitivity of the LAMP and PCR assays was determined using genomic DNA of target species that was adjusted to a concentration of 10.0 ng/μL and diluted to 5.0 ng/μL, 1.0 ng/μL, 0.4 ng/μL, 0.2 ng/μL, and 0.1 ng/μL. The specificity of the LAMP and PCR assay was determined by amplification of the 93 shark and related species with the species-specific primers. All amplifications were performed in triplicates.

## Results

To evaluate the specificity of LAMP and PCR assays for the CITES-listed shark species, all species-specific primers (Table 2) were tested on their target species and against all non-target species as listed in Table 1. All samples were successfully amplified using universal COI primers32,33 and their identity confirmed using BLAST analysis. Most external primers (F3 and B3) of the LAMP primer sets were the same as their PCR forward and reverse primers (LP and RP) except for silky shark, porbeagle shark and internal control, as their LAMP and PCR primer sets were designed based on different regions. The expected size of target amplicon of each primer set is listed in Table 2.

### Species-specific LAMP assays

The 12 CITES-listed shark species LAMP assays specifically amplified their target species only and non-target species were not detected (Fig. 1). Three independent experiments had been performed for each assay. The melting temperatures of LAMP products of the 12 LAMP assays were between 81.1 and 86.6 °C. The melting curves of target species with their corresponding LAMP assays were identical and no melting curve was found for non-target species showing the specificity of these LAMP assays (Fig. 2). For the sensitivity, the detection limit of the LAMP assay was 0.2 ng/μL for pelagic thresher shark, bigeye thresher shark, oceanic whitetip shark, porbeagle shark, scalloped hammerhead shark and smooth hammerhead shark; 0.4 ng/μL for great white shark, silky shark, and great hammerhead shark and 5.0 ng/μL for common thresher shark, basking shark and whale shark (Fig. 3). For the internal control of the LAMP assay, positive result was found with all samples and no amplification was found with the negative control. The LAMP assay for internal control was able to detect the 12 CITES-listed shark species at the concentration equivalent to the detection limit of their corresponding species-specific LAMP assays (Supporting Information). Apart from the real-time amplification curve and melting curve obtained from Genie II, positive LAMP signals could be observed by naked eye after addition of SYBR Green I dye. As shown in Fig. 4, under ambient white light, all positive LAMP reactions turned to yellow colour and negative reactions remained in brownish orange colour (Fig. 4a,c). The lower panel (Fig. 4b) shows the limit of detection of each LAMP assay under visual detection. These features allowed rapid on-site screening.

### Species-specific PCR assays

All CITES-listed shark species were specifically amplified by their corresponding PCR assays with expected sizes (Table 2) and no amplification was found for non-target species, showing the specificity of each of the PCR assays (Supporting Information). The limit of detection of each PCR assay was 0.2 ng/μL for porbeagle shark and bigeye thresher shark; 0.4 ng/μL for common thresher shark, oceanic whitetip shark, and smooth hammerhead shark; 5.0 ng/μL for pelagic thresher shark, great white shark, silky shark, basking shark, scalloped hammerhead shark and great hammerhead shark; 10.0 ng/μL for whale shark (Supporting Information). All samples showed positive results in the PCR assay for internal control and no amplification was found in the negative control. Internal control could be successfully amplified from the 12 CITES-listed shark species at the DNA concentration equivalent to the limit of detection of their corresponding species-specific PCR assay (Supporting Information).

## Discussion

In this work, we have developed sensitive, simple to use, and to our knowledge, the fastest species-specific assays using LAMP technique, as well as simple and low-cost PCR assays to authenticate all 12 CITES-listed shark species. The LAMP and PCR assays we developed can successfully discriminate their target species from the other 92 shark and related species, which has included species commonly found in the Hong Kong shark fin market12.

We have designed our primers based on COI and NADH2 loci for the CITES-listed shark species and on 12 s rRNA locus for internal control. Previous studies about species-specific primers on shark species identification are mostly focused on the ITS2 locus17,18,19,20. The COI locus, a standard marker for fish and shark species DNA barcoding16,32,33,35,36, is highly species specific and the number of nucleotide difference of congeneric species is sufficient for designing species-specific primers. For common thresher shark, silky shark, oceanic whitetip shark and porbeagle shark, NADH2 locus is a better region for designing species-specific primer, due to high nucleotide sequence differences between the target species and the corresponding congeneric species37. High degree of nucleotide sequence difference allows us to design LAMP primers targeting four to six species-specific regions instead of only two for conventional PCR. Our LAMP assays are highly specific, especially for species with many congeneric species such as the silky shark and the ocean whitetip shark. Furthermore, care was taken to avoid sites of known intraspecific variation among sharks from different regions or ocean basins during primer design, so that amplification success would not be affected by intraspecific nucleotide variation. In contrast, 12 s rRNA locus is a suitable target for designing primers for the internal control as it has high sequence homology among different shark species. This allows us to design six LAMP primers that can amplify all of the shark species in this work.

Samples of 91 shark species and 2 Chimaera species were collected for testing the specificity of our developed LAMP and PCR assays. These species covered 56 out of 61 species found in the Hong Kong fin market in 2014–201512 The other five species were rarely found in the fin trade market. For LAMP and PCR assays targeting the silky shark and the oceanic whitetip shark, we have tested 21 out of 31 congeneric species from Carcharhinus and a total of 76 Carcharhinus samples were tested. Although we were not able to test the assays on all 31 recognised Carcharhinus species37, we have confidence on the specificity of our assays for three reasons. First, we have tested our assays against most species which are phylogenetically close to the silky shark and the oceanic whitetip shark38,39 and no amplification was found with non-target species. Second, those nine untested species are phylogenetically more distant from the silky shark and the oceanic whitetip shark39 and they are rarely found in the Hong Kong fin market12. Third, we have compared nucleotide sequences of those nine species on NADH2 region available on NCBI GenBank except for C. hemiodon, which has no sequences available on any public database, for checking of primer mismatching with Primer-BLAST to ensure the specificity of our primers. For the oceanic whitetip shark, previous studies have found difficulties on designing species-specific primers targeting the oceanic whitetip shark due to its high phylogenetically similarity with C. obscurus and C. galapagensis17,22,39. We are the first to successfully develop species-specific assays for the oceanic whitetip shark. We have found that these three species share less similarity among their sequences on the NADH2 region and hence easier to design species-specific primers discriminating the oceanic whitetip shark from the other two species. Furthermore, we have used DNA extracted from processed shark fin for testing to ensure that the LAMP assays could provide accurate result when performed on-field.

Besides oceanic whitetip shark, we are also the first to present species-specific assays for the whale shark. Whale shark, being the only species in the family Rhincodontidae, is phylogenetically more distant from other species in Orectolobiformes37,39,40. We have examined the COI sequences in Orectolobiformes that are available on the NCBI database and performed the Primer-BLAST analysis to ensure the specificity of our assays. For the thresher sharks, basking shark, great white shark and porbeagle shark, they are classified in the mackerel shark order Lamniformes. For the LAMP and PCR assays targeting these six species, we have tested them with 12 out of 15 recognised species of the order Lamniformes including all species in the family Alopiidae and the family Lamnidae to ensure the specificity of our primers. For the three CITES-listed hammerhead sharks, we have tested our assays with six out of nine species in the family Sphyrnidae. The other three species (S. corona, S. gilberti, and S. media) are phylogenetically more distant from the CITES-listed hammerhead sharks and not commonly found in the Hong Kong shark fin market3,12,19,41,42.

LAMP allows rapid amplification of DNA using Bst polymerase with high strand displacement activity at isothermal temperature within an hour. This facilitates diagnostic on-site detection with limited equipment and provides species-level identification in a short period of time, increasing efficiency for law enforcement. To our knowledge, this is the first report on applying LAMP on shark species identification. Although the optimal temperature for Bst polymerase is between 60 °C and 65 °C43, the optimal reaction temperatures for most of the LAMP assays developed were found to be 68 °C or 70 °C, except for those targeting bigeye thresher shark, common thresher shark and porbeagle shark. Increase in LAMP reaction temperature allows a better specificity especially for eliminating congeneric species. Apart from increasing reaction temperature, another common approach of optimizing LAMP reaction is the addition of loop primers, which can increase the sensitivity of LAMP reaction. The LAMP assays targeting great white shark, porbeagle shark, great hammerhead shark, and that for the internal control include loop primers in their primer sets. Loop primers are designed to bind to the additional sites that are not accessed by the internal primers and accelerate the rate of LAMP reaction and provide higher sensitivity24. This allows the LAMP assay for the internal control to amplify all shark species in Table 1 at a relatively low reaction temperature, which also favours primer annealing and amplification. The presence of loop primers also allows the LAMP assays developed to give amplification with detectable threshold within 60 minutes, reducing the reaction time required for the assays. For LAMP and PCR assays amplifying the same region of a species, the species-specific LAMP assays generally have a better sensitivity than the species-specific PCR assays, except for the LAMP assays targeting common thresher shark and basking shark. Looped DNA product increases the sensitivity of the LAMP assay and hence facilitates detection with limited sample amount from dried shark fins. In addition to better sensitivity, since LAMP produces more DNA products than conventional PCR method, it allows easy visual detection with naked eye to tell between positive and negative results without fluorescence detection equipment. On-site detection of CITES-listed shark species can be achieved. On the other hand, there are also limitations on LAMP assays. Since LAMP assay is highly sensitive, it is important to perform them at their optimal reaction temperature. Four different temperatures, according to conditions listed on Table 2 should be used to achieve repeatable result. Furthermore, there may be a chance of cross contamination leading to false positive results. Therefore, it is important to clean the equipment before test and keep good laboratory practice. Since there are only 16 wells available on Genie II, using of multi-block thermal cycler may be considered.

For on-field screening and monitoring of shark fin products with unknown identity, use of multi-block thermal cycler is suggested for performing LAMP reactions, at four different temperatures within 60 minutes. Sample DNA, in different forms including dried processed shark fin, dried and frozen flesh tissues and inner organs, can be extracted using (1) Biomed Genomic DNA/Tissue Extraction Kit in 30–60 minutes depending on the rate of sample dissolution in lysis buffer; or (2) Kaneka Easy DNA Extraction Kit (Version 2) (Funakoshi, Osaka, Japan) in 15 minutes. After sample extraction, 1 μL of the extracted DNA is amplified by LAMP (50–60 minutes). At the end of the reaction, 2 μL of 1000X SYBR Green dye is added for visual detection. If there is positive LAMP reaction, the colour would change simultaneously with the addition of SYBR Green dye. For a 96-well multi-block thermal cycler with temperature control in each column, eight samples, including positive and negative control, with 12 LAMP assays could be tested per hour. To obtain the best result, portable device for DNA concentration measurement could also be used. In our experience, DNA with concentration of at least 10.0 ng/µL was able to be extracted from all unprocessed and processed samples using the Biomed Genomic DNA/Tissue Extraction Kit.

There are several on-field detection protocols available for CITES-listed shark species identification. They include multiplex real-time PCR assay by Cardeñosa et al.22 and genome skimming using MinION hand-held sequencer by Johri et al.44. Our protocol has good potential to be used on field. As shown in Table 3, our LAMP assays have covered all twelve CITES-listed shark species and it is cheaper in the aspects of cost and equipment compared to the other two methods. It is less time consuming for test on a small scale and allows simple result analysis which is more favourable to customs officials without scientific background. Multiplex real-time PCR assay allows test on 94 samples within 4 hours, which is the fastest one among these three methods and data analysis is also straightforward. This could be favourable for on-field test on a large scale. However, the multiplex real-time PCR developed only covers 9 out of 12 CITES-listed shark species. For the other three shark species, use of morphological guide is needed22. For genome skimming using MinION hand-held sequencer, the major limiting factors are the high cost of MinION flow cell compared to the other two methods. Although this protocol could provide additional genetic information, result has to be analysed by bioinformatic expertise, which is less favourable for customs use. Furthermore, this approach has not been tested with highly processed shark fin products which are generally skinned and bleached. In short, for on-field detection by customs, LAMP assays are more favourable for small-scaled quick test while multiplex real-time PCR assay is more favourable for large-scaled test.

Our rapid on-site detection assays have shown good specificity and can be used for on-site law enforcement of all CITES-listed shark species which are commonly found in the fin market3,12. It will be useful for screening of processed shark fin products which is difficult to be identified using morphological identification guide and shark products such as shark meat as they are hardly identified based on morphology22,45. The result has also shown that COI and NADH2 regions are suitable for species-level identification, especially for species with a high nucleotide similarity with congeneric species.

Our LAMP assays allow easy inspection of imported shark fin products at the border for all 12 CITES-listed shark species and facilitate monitoring of international trade46. Same model using our LAMP assays could also be applied for law enforcement in other area. For example, as the 4th largest shark-catching area47, Taiwan has recently passed the “Regulations for Import of Shark Fins” of which catching of whale shark and import of whale shark products are forbidden48. Our protocol for whale shark could provide a reliable and efficient test for monitoring the trade of whale shark products. With the exact species identity of shark fin products, it would provide a better picture on the CITES-listed shark products in Hong Kong and elsewhere. The data could be important for better monitoring of international shark fin trade and the conservation of shark by CITES members49.

Besides shark species currently listed on CITES appendix II, there is a need for development of rapid identification assay for other endangered shark species in the future CITES appendix and the list of IUCN to ensure effective detection of illegal trade. In fact, two more shark species, Isurus oxyrinchus and I. paucus, have been recently included in CITES Appendix II at the 18th meeting of the Conference of the Parties of the CITES in 2019 and will enter into effect on 26 November 201911. Our innovation can be extended to these new CITES-listed shark species and processed products and for further development of multiplex LAMP assays. We also expect that our work will serve as a model for the rapid identification assay for other endangered species.