A case study for application of DNA barcoding in identifying species of some imported frozen fish fillets in Egypt

Aquaculture plays a vital role in producing aquatic animals to meet the rising demand for protein^1,2. The production of aquaculture worldwide hit a record 130.9 million tonnes in 2022 and expected to reach 205 million tonnes in 2032³ to meet rising demands. More than 90% of aquaculture production takes place in low- and middle-income nations, where it reduces food insecurity and provides a vital source of income⁴.

Catfish (order Siluriformes) are a diverse group of fish representing more than 3,000 species, 478 genera and 36 families⁵. By the end of 2030, the number of catfish species could reach 5,000⁶. Catfish are high tolerance to low dissolved oxygen, which may have been brought on by pollutants like heavy metals⁷. Because of their high market value⁸, rapid growth, and resistance to a wide range of environmental conditions^{9,10,11,12,13,14,15}, catfish have been cultivated in nearly 90 countries worldwide⁶, such as (i) North African catfish Clarias gariepinus which is widely farmed in South Africa, Kenya, Mali, Nigeria, and Brazil, (ii) channel catfish Ictalurus punctatus primarily cultured in USA, Mexico, Cuba, Russia, and China, (iii) striped catfish Pangasianodon hypophthalmus which is primarily raised in Vietnam and Thailand.

Pangasius fish, mainly Swai/Sutchi Pangasianodon hypophthalmus and basa Pangasius bocourti belonging to the family Pangasiidae in the order Siluriformes of catfish¹⁶. They are presently mostly raised in intensive farming ponds, although Vietnamese and Cambodian farmers throughout the Mekong basin have been cage-raising them for decades¹⁷. Although basa grows much more slowly than Swai/Sutchi and takes twice as long to become commercially available, it has a better texture, taste and flavour. Because it cannot withstand contaminated water settings as well as Swai/Sutchi, basa is also less able to adapt to intensive farming practices¹⁸. Pangasius fish have a great deal of promise for the production of convenient fish products, including fillets, which are easy to fillet because they don’t have intramuscular pin bones¹⁹. It is usually prepared into frozen fillets for both domestic use and export to the United States, Middle East, and Europe.

Vietnam is currently the world’s third-largest producer of fisheries products²⁰. Vietnamese catfish products have exported to 140 markets worldwide²¹. Vietnam’s exports of pangasius to various markets reached 191 million USD in 2024²² and reached 1.023 billion USD²³. Egypt, United Arab Emirates, and Saudi Arabia are the top three markets for Vietnamese pangasius enterprises. By the end of 2021, Vietnam’s pangasius exports to the Middle East market had a total value of 56.32 million USD. Pangasius exports to Egypt reached 18.7 million USD, which was more than 32% of the total export value to the whole region²⁴.

Commercial fish product mislabeling is a widespread issue that occurs all around the world. The majority of seafood consumers are not aware that this problem directly affects them and may potentially have detrimental effects on their health. When a product’s label doesn’t match its contents, it’s called mislabeling²⁵. While intentional mislabeling is a more prevalent practice to boost profits and/or circumvent fishing regulations.

Import restrictions and labelling laws have had an impact on the species and origins of imported catfish, but they haven’t substantially reduced the volume of imports. Numerous anecdotal and proven cases exist when catfish species are mistakenly classified as either another catfish species or a higher value species. To improve the examination of imported fish products such as fish fillets, including verifying accurate species labelling, the government should create new laws²⁶. Accurate identification of fish species in the markets is a growing concern due to the high incidence of species substitution at international level²⁷.

DNA barcoding using COI gene sequences offer a reliable method for identifying known species and finding new ones by comparing sequence variation^28,29. The COI gene’s extensive taxonomic coverage in validated databases, such as National Center for Biotechnology Information GenBank (NCBI GenBank) and Barcode of Life Data System (BOLD), makes it a common marker in animal metabarcoding³⁰. DNA-based methodologies have revealed seafood mislabeling, checking for fraud across the worldwide supply chain and confirming the legitimacy of species^31,32,33,34. In Egypt, commercial seafood mislabeling is a serious problem, especially in processed products like fish fillets where visual identification is challenging. Studies using DNA barcoding have shown high rates of species substitution, which are frequently motivated by financial incentives to market less valuable fish as premium kinds³³.

The current study aimed to conduct an initial survey to identify fish species and detect mislabeling of some imported frozen fish fillets purchased from local grocery markets in Alexandria and Cairo, Egypt, by comparing the cytochrome oxidase subunit I (COI) gene sequences of frozen fish fillets against global genetic databases.

Sample collection

Twelve samples of frozen fish fillets labeled as Pangasius were purchased, between May, 2024 and May, 2025, from local grocery markets in Alexandria and Cairo, Egypt. Details about each sample were recorded, including purchase location, name on the label, and country of origin labeling. Samples details are shown in Table 1.

DNA extraction

The genomic DNA extractions from the muscle tissues were performed using the G-spinTM total DNA kit (Cat NO. 17045, iNtRON Biotechnology, Inc., Korea) following the manufacturer’s protocol, and stored at − 20 °C prior to further analysis. The quantity and quality of extracted DNA was measured using a NanoDrop 2000/2000c spectrophotometer (Manufactured by ThermoScientific™, USA) and 1.2% agarose gel electrophoresis. The A260/280 ratios for all isolated genomic DNA were ≥ 1.8. All isolated DNA were diluted to 200 ng.

PCR amplification and sequencing

The mitochondrial COI gene was amplified using primer of FishF1 (5’ TCA ACC AAC CAC AAA GAC ATT GGC AC 3’) and FishR1 (5’ TAG ACT TCT GGG TGG CCA AAG AAT CA 3’), described by Ward et al.³⁵. The amplification reactions were performed as described in the protocols of Ashour et al.²⁹. PCR products were visualized on a 1.5% agarose gels, containing ethidium bromide (10 mg/ml) and the most intense products were selected and purified using the Qiagen gel extraction kit for sequencing. All sequencing was done by the Macrogen sequencing lab, Korea using the Life Technologies POP7 technology. The Abass et al.^10,13 techniques were followed in the purification and sequencing of the amplified PCR products. Sequences were submitted to the GenBank database with accession numbers PX795194- PX795205 (Table 2).

Molecular and bioinformatics data analysis

All sequences were aligned using the Clustal Omega website (https://www.ebi.ac.uk/jdispatcher/msa/clustalo). Sample identification based on the COI barcode sequences similarity approach was carried out using two databases; GenBank and the Barcode of Life Data System (BOLD), https://v3.boldsystems.org. The highest percent pairwise identity of the consensus sequence from each sample blasted (BLASTN) against NCBI were compared to the percent specimen similarity scores of the consensus sequence from each species using FASTA sequences within the BOLD, accessed on Jan. 10^th, 2026. The BLASTN software available from NCBI (http://www.ncbi.nlm.nih.gov) was used to analyze DNA sequences. Sequence similarity > 99% and query coverage = 100% were used as thresholds to identify fish species. The multiple sequence alignments were processed using the Clustal Omega website to illustrate the homologous relationship of each sample. Accession numbers (GenBank) for voucher species which were used to construct multiple sequence alignments are listed in Table 2. 21 sequences of COI gene aligned with one out-group as a root, Devario aequipinnatus to determine evolutionary relationships and identification of fish species. The MEGA version 12.0³⁶ was used to construct the phylogenetic tree. Phylogenetic analysis was established using the Neighbor-Joining (NJ) and maximum parsimony (MP) methods³⁷. The robustness of the tree was assessed by performing bootstrapping analysis with 1000 replicates³⁸. The distance of the COI nucleotide bases between and within the fish species was estimated using the Kimura-2-parameter distance model (K2P)³⁹.

Identification of fish species using DNA barcoding

For the information stated on the labels, all of them stated the common names, and the country of origin labeling. However, no trade market on the label (Table 1). The partial fragment (~ 650 bp) of mitochondrial cytochrome oxidase subunit I (COI) gene was successfully amplified using PCR from all the fish samples (Fig. 1). Table 3 shows the comprehensive barcoding identification results based on GenBank and the BOLD databases. The reference sequences of GenBank for the identification of all imported fish fillets were fish fillet_1 (JF292415), fish fillet_2 (JF292411), Fish fillet_3 (JF292419), fish fillet_4 (JF292425), fish fillet_5, and fish fillet_6 (PX312189), fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, and fish fillet_12 (PP396129), and fish fillet_11 (MW829646) (Table 3). The BIN number of BOLD for the identification of fish fillets were fish fillet_1, fish fillet_2, fish fillet_3, and fish fillet_4 (BOLD: AAB7484), fish fillet_5, fish fillet_6, fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, fish fillet_11, and fish fillet_12 (BOLD: AAE3237) (Table 3). The GenBank database revealed definitive identity matches in the range of 99.85%–100% for all imported frozen fish fillets and BOLD database revealed 100% identities at the species level. GenBank-based identification for all fish species yielded an alignment E-value of 0.0. At the time of analysis, GenBank results were in agreement with BOLD results in identification of all imported fish fillets. The GenBank and BOLD database hit using our fish fillet_1, fish fillet_2, fish fillet_3, and fish fillet_4 sequences was a single Pangasius bocourti sequence, while fish fillet_5, fish fillet_6, fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, fish fillet_11, and fish fillet_12 was a single Pangasianodon hypophthalmus sequence. Fish fillet_1, fish fillet_5, fish fillet_6, fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, fish fillet_11, and fish fillet_12 had 100% maximum identity in GenBank and BOLD database. Fish fillet_2, fish fillet_3, and fish fillet_4 had 99.85% maximum identity in GenBank and 100% maximum identity in BOLD database. Mislabeling was detected in the imported frozen fish fillets analyzed since all samples were genetically identified as Pangasius bocourti and Pangasianodon hypophthalmus. The results showed that 66.7% of the basa fish, Pangasius bocourti had been replaced by Swai/Sutchi fish, Pangasianodon hypophthalmus.

Phylogenetic analysis

The COI gene sequences of the imported frozen fish fillets in the Egyptian markets were aligned using the Clustal Omega website (Supplementary, Fig. S1). Multiple sequence alignment of imported frozen fish fillets in Egyptian markets against reference sequences from GenBank, generated by Clustal Omega website are shown in (Supplementary, Fig. S2). From the COI gene sequence alignments, there were genetic variations at a nucleotide level (A, T, C, or G) as determined at different positions of the representative sequences. The current study’s nucleotide sequence data from the COI gene of different fish samples generated an alignment of ~ 650 bp. The fish species nucleotide discrimination revealed varied AT (adenine + thymine) and GC (guanine + cytosine) contents. Among the twelve representative fish samples, the observed nucleotide base composition of all analyzed sequences was 56.84% AT (range: 357–386) and 43.16% GC (range: 275–294) (Supplementary, Table S1).

Nine catfish species namely walking catfish Clarias batrachus, North African catfish Clarias gariepinus, bighead catfish Clarias macrocephalus, helmet catfish Cranoglanis bouderius, long-barbel catfish Hemibagrus macropteru, blue catfish Ictalurus furcatus, channel catfish Ictalurus punctatus, swai or sutchi catfish Pangasianodon hypophthalmu, and basa catfish Pangasius bocourti were downloaded from NCBI GenBank (Table 2) to perform the phylogenetic analysis. The classification of the fish species used in this study is shown in Table 4. Classification, and scientific and common names of fish species were obtained from BOLD and Fishbase databases. The phylogenetic tree (Fig. 2) was constructed from the combined dataset of COI gene consisting of 21 sequences aligned with one out-group as a root, Devario aequipinnatus (Supplementary, Fig. S3).

From dendrogram it is clear that all genera of catfish were differentiated into 4 clusters. In cluster 1 the one species of the genus Pangasianodon was clustered with 96%, 81%, 73%, and 62% bootstrap values. In cluster 2 the one species of the genus Pangasius was clustered with 99%, 90%, 75%, and 71% bootstrap values. In cluster 3 the three species of the genus Clarias, Cranoglanis, and Hemibagrus were clustered with 99%, 97, 61%, and 39% bootstraps. In cluster 4 the one species of the genus Ictalurus was clustered with 84% bootstrap values. Cluster 1 and Cluster 2 consist of fishes of interest which shows that these fishes are from same lineage and must share a common ancestor. Genetic diversity among the twelve frozen fish fillets shows that fish fillet_1, fish fillet_2, fish fillet_3, and fish fillet_4 are closely related with Pangasius bocourti whereas fish fillet_5, fish fillet_6, fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, fish fillet_11, and fish fillet_12 are related with Pangasianodon hypophthalmus (Fig. 2). The evolution of phylogenetic tree for all the fish species combined showed a clear distinction of the most fish families with most species of the same family were clustered together with observation of misplacement of species (Fig. 2).

The intraspecific and interspecies genetic distances were calculated with The Kimura-2-parameter model (K2P) to trace the evolutionary relationship between fish fillets samples, species, and families are summarized in Tables 5, 6, 7, 8 and 9. The minimum genetic distance between fish fillets samples (intraspecific genetic distances) was 0.00% and the maximum distance (interspecific genetic distances) was 0.12% (Table 5). The minimum genetic distance between fish species was 0.12 (between Pangasianodon hypophthalmus and Pangasius bocourti) and the maximum distance was 0.24 (between Ictalurus furcatus and Clarias batrachus) (Tables 6 and 7, and 9). The minimum value of genetic distance between fish families was 0.18 and the maximum value was 0.22 (Tables 8 and 9).

Molecular species identification, especially DNA barcoding, offers a powerful solution for the expanding issue of fisheries and aquaculture mislabeling by providing accurate, rapid genetic “fingerprints” to verify the actual species of food products, helping to protect consumers, stop economic fraud, and aid in the conservation of threatened species²⁶. DNA barcoding is a powerful, standardized technique for species identification that excels in speed and application throughout life stages. However, its limitations appear from its dependence on extensive reference databases and difficulties with specific taxonomic complexity, such as hybrids⁴⁰. DNA barcoding is superior to other molecular markers in terms of universality, although it occasionally lacks the discriminating capacity of genome-wide approaches. In the current study, we successfully sequenced ~ 650 bp barcodes from the COI gene from all imported frozen fish fillets purchased from Alexandria and Cairo, Egypt. Since Alexandria and Cairo are Egypt’s biggest cities and main economic markets, they constitute an important cross-section of the country’s consumer market, making the choice of local grocery stores as the sampling regions extremely relevant. To cover a wider geographic area, it is strongly recommended that larger samples be used in future studies.

Comparison of COI gene sequences with sequences published in public databases (GenBank and BOLD) revealed 99.85–100% identities at the species level. In the present investigation, we were able to successfully sequence around 650 bp barcodes from the COI gene of imported frozen fish fillets that were bought from Egyptian marketplaces. Consequently, all of the imported fish fillets were correctly identified by the COI barcodes. According to the COI gene sequence, the fish species can be identified with the help of the DNA barcode^29,41,42,43. According to our results, all sample of imported frozen fish fillets matched the reference sequences of COI gene in both databases (GenBank and the BOLD) and every species under investigation had a unique sequence that set them apart from one another.

Our results revealed that the average nucleotide base content was 56.84% AT and 43.16% GC. This result is consistent with previous studies that reported higher AT content than GC content with COI gene amplification in freshwater fish species^{29,44,45,46,47}. The average GC% content for COI sequences gene for 30 freshwater fish sampled obtained from Lake Nasser and the River Nile was 46.44%²⁹ and the average GC% content of 44 freshwater fish sampled obtained from Enugu and Anambra States was 45.78%⁴⁸.

The COI gene sequences revealed the intraspecific genetic distances was 0.00% and the interspecific genetic distances was 0.12%. This finding is consistent with previous research showing that genetic distance within a species is smaller than that between species^29,49. In Discogobio species, the average interspecific genetic distances were approximately 18.8 times greater than the average intraspecific genetic distances²⁸. Our results showed that average genetic distance based on the K2P within species, and families were 0.20. Ashour et al.²⁹ reported that the K2P values for 8 freshwater fish species collected from Lake Nasser and the River Nile were 0.227%, and 0.249% between species, and families, respectively. A more comprehensive understanding of the evolutionary history of catfish species can be obtained through taxonomic analysis and population genetics. A maximum genetic distance of 3% is adequate to distinguish the catfish species26. Species belonging to the same genera were, as expected, closely clustered into a single clade with a well-supported bootstrap proportion⁵⁰. According to the Neighbour-joining tree, the dendrogram of the studied fish species and some imported frozen fish fillets samples revealed that the same cluster contained fishes of interest, indicating that these fishes must have a common ancestor and are from the same lineage. In a phylogenetic tree, low bootstrap values (< 70%) indicate uncertainty or weak signal. These are frequently caused by rapid evolution (short branches), contradicting data (incomplete lineage sorting, horizontal gene transfer), a lack of informative characters (short sequences, high similarity), or methodological problems (long-branch attraction), which indicate that the data does not strongly support that particular grouping over alternatives. It implies that, particularly if other branches have significant support, the node’s placement may be the result of artifacts or chance rather than compelling evolutionary evidence. The genetic diversity of the five fish fillet samples reveals that whereas fish fillet_5, fish fillet_6, fish fillet_7, fish fillet_8, fish fillet_9, fish fillet_10, fish fillet_11, and fish fillet_12 are related with Swai or Sutchi catfish Pangasianodon hypophthalmus, fish fillet_1, fish fillet_2, fish fillet_3, and fish fillet_4 are closely related to basa catfish Pangasius bocourti. The phylogenetic analysis showed the sequenced DNA similar to the NCBI database reference sequence. According to this study, the labeling on the imported catfish fillets that were sold in Egyptian local grocery stores was incorrect. This is consistent with the findings by Galal-Khallaf et al.³³. The sequencing of COI gene of ninety commercial samples from Egyptian markets showed 33.3% species substitution in the fish fillets examined, 50% Nile perch Lates niloticus and 50% basa fish Pangasius bocourti being replaced by imported Vietnamese fish Pangasianodon hypophthalmu³³. Similar to other global markets where seafood mislabeling has been exposed using DNA-based technology^32,51,52. All imported fish must have an accurate product description and the country of origin written on the label in Arabic. According to Egyptian standards, such as those established by the Egyptian Organization for Standardization (EOS), using a “grade” or “type” that does not correspond with the biological species is deemed non-compliant. Mislabeling food items, such selling Swai as Basa, is considered a criminal violation under Egyptian law that involves “commercial fraud” and “misleading consumers.” The Consumer Protection Agency (CPA) and the National Food Safety Authority (NFSA) are responsible for enforcing penalties.

In contrast, Wong et al.²⁶ reported that the labeling on the catfish fillets that were bought from the local grocery and oriental markets in USA was accurate. Pangasiidae was observed as sister group to Ictaluridae at a relatively low bootstrap percentage of 55%, whereas Cranoglanididae represented by Cranoglanis bouderius and Bagridae represented by Hemibagrus macropterus were the most diverged families from the rest of the groups and the mean genetic distance between these two families is 0.194%. Our findings are in line with those of Wong et al.²⁶ has found that Pangasiidae was observed as sister group to Ictaluridae at a comparativaly low bootstrap percentage of 56%. In comparison to the other groups, Bagridae represented by H. macropterus was the most divergent⁵³ and Funk and Omland⁵⁴ also found that the clustering of C. batrachus and C. macrocephalus in one lineage and C. gariepinus in another lineage caused them to become geographically separated during the early phases of their evolution; with the former two catfish species being native Asian catfish and the latter of African origin. Ictaluridae are closely related to Pangasiidae, according to mitochondrial data⁵⁵.

In conclusion, DNA barcoding and phylogenetic analysis provide Egypt with a reliable and rapid technique to identify commercial fish products, helping to prevent seafood fraud, where high-value fish species are illegally replaced with cheaper, lower-quality fish species for economic gain and safeguarding public health. DNA barcoding can accurately distinguish poisonous and harmful fish species and remove them from the market. By analyzing a short fragment of mitochondrial COI gene, regulatory authorities can confirm that the species on the label matches the product inside the packaging, even after processing. This molecular technique represents a significant improvement over traditional visual checks, which are often useless for processed fish such as fillets, fish balls, and canned fish. Economic incentives, morphological similarity in processed forms, market demand, and regulatory gaps all contribute to fish species substitution in imported frozen fish fillets in Egypt, especially within catfish species. Pangasianodon hypophthalmus, also known as swai or Sutchi catfish, is sometimes employed in substitute of more expensive basa, Pangasius bocourti, due to its quick growth, higher meat yield, and much cheaper market price.