Genome-wide identification and expression profile of Elovl genes in threadfin fish Eleutheronema

Abstract

Long-chain polyunsaturated fatty acids (LC-PUFA), including eicosapentaenoic acid and docosahexaenoic acid, are the essential fatty acids for organs to maintain various biological functions and processes. The threadfin fish Eleutheronema, with its rich nutritional value especially the high fatty acid contents, has become one of the promising aquaculture species in China and the potential food source of fatty acids for human consumption. However, the molecular basis underlying the biosynthesis of fatty acids in Eleutheronema species is still unknown. The elongation of the very long-chain fatty acids (Elovl) gene family in fish plays several critical roles in LC-PUFA synthesis. Therefore, in the present study, we performed genome-wide identification of the Elovl gene family to study their evolutionary relationships and expression profiles in two threadfin fish species Eleutheronema tetradactylum and Eleutheronema rhadinum, the first representatives from the family Eleutheronema. Phylogenetic analysis revealed that the Elovl genes in Eleutheronema were classified into six subfamilies (elovl1a/1b, elovl4a/4b, elovl5, elovl6/6 l, elovl7a, elovl8b). Phylogenetic, gene structure, motif, and conserved domain analysis indicated that the Elovl genes were highly conserved within the same subfamily in Eleutheronema. In addition, the Elovl genes were distributed in 7/26 chromosomes, while the duplicated gene pair, elovl4a and elovl4b, showed collinear relationships. The predicted secondary structure patterns and the 3D models revealed the highly similar functions and evolutionary conserved structure of Elovl proteins in Eleutheronema. The selection pressure analysis revealed that Elovl genes underwent strong purifying selection during evolution, suggesting that their functions might be evolutionarily conserved in Eleutheronema. Additionally, the expression patterns of Elovl genes in different tissues and species were distinct, indicating the possible functional divergence during evolution in the Eleutheronema genus. Collectively, we provided the first comprehensive genomic information on Elovl genes in threadfin fish Eleutheronema. This study enhanced the understanding of the underlying mechanisms of fatty acids biosynthesis in Eleutheronema, and provided new insights on breeding new varieties of fatty acids-enriched fish with potential benefits to farmers and the health of consumers.

Introduction

Fatty acids, the natural components of fats and oils, can be divided into saturated, monounsaturated, and polyunsaturated fatty acids of three classes based on their chemical structure. Saturated fatty acids are mainly found in animal foods such as meat. Unsaturated fatty acids, especially the omega-3 long-chain polyunsaturated fatty acids including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), are almost originated from fatty marine organisms¹. However, polyunsaturated fatty acids cannot be synthesized by humans, and must be obtained from diet or supplementation². In vertebrates, dietary long-chain polyunsaturated fatty acids (LC-PUFA) are crucial in maintaining the various biological functions and processes, including immune function, reproduction, growth, and development³. Moreover, fish are the primary source of LC-PUFA, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), for human consumption⁴.

In vertebrates, middle chain length of saturated fatty acids, mostly C16–18, are produced by fatty acid synthase, and then desaturated and elongated into long chain fatty acids (LC-FAs) and very long chain fatty acids (VLC-FAs, chain lengths > 24 carbons). The elongation process is catalyzed by fatty acyl elongases, the Elongation of Very Long-chain fatty acids (Elovl) proteins, an essential family of membrane-bound enzymes of the elongation pathway of fatty acids⁵. Seven Elovl gene family members in mammals, including ELOVL 1–7, were successfully characterized, exhibiting dramatically different expression patterns in different tissues⁶. Generally, elovl1, elovl3, elovl6, and elovl7 can elongate saturated fatty acids (SFA) and monounsaturated fatty acids (MUFA), while the preferred substrates of elovl2, elovl4, and elovl5 were PUFA⁷. In teleost, Agabaet et al.⁸ firstly reported the Elovl gene, elovol5, in zebrafish. Since then, many studies have reported the molecular and functional characterization of Elovl genes from a large variety of fish species.

Due to the teleosts-specific whole-genome duplication events, the sub-functionalized paralogues of some genes have been successfully identified in various fish species⁹. Zebrafish (D. rerio) possess two elovl1 members, elovl1a and elovl1b, which were required for kidney and swim bladder development during zebrafish embryogenesis¹⁰. The paralogues of the elovl1 gene were also characterized in Gymnocypris przewalskii and played essential roles in adaptation to cold temperature¹¹. The elovl2 was essential for the biosynthesis of docosahexaenoic acid (DHA)¹², which has been identified in various fish species, including Lepisosteus oculatus (NCBI accession: XP_015210453.1), Danio rerio (NP_001035452.1), and Salmo salar (NP_001130025.1). In addition, a previous study reported that the elovl2 activity was specific to C₂₀ and C₂₂ PUFAs¹³. However, the loss of the elovl2 gene as the consequence of the expansion of teleosts during evolution emerged in most marine species⁶. A recent study revealed that the compensation system in many marine teleosts for the absence of elovl2 might be related to the gene duplication of elovl4 (elovl4a and elovl4b), which was generally presented in most teleosts¹⁴. The elovl4 has been widely identified and investigated in marine fish, responsible for the elongation toward VLC-PUFA. Most importantly, beyond their role in VLC-PUFA biosynthesis, the elovl4 played a critical role in the biosynthesis of LC-PUFA such as DHA in organism¹⁵. The elovl5 was present in virtually all teleosts that were specific for the elongation of C₁₈ and C₂₀ PUFAs, and also had capacity towards C₂₂ PUFA in some fish, such as Pegusa lascaris¹⁶ and Scatophagus argus¹⁷. The elovl6 acted to convert C₁₆ saturated and monounsaturated fatty acids to C₁₈ fatty acids, similar to mammals, which has been reported in Larimichthys crocea¹⁸ and Oncorhynchus mykiss¹⁹. The novel elongase, elovl8, including elovl8a and elovl8b, were apparently absent in mammals but shown to have elongation activity to C₁₈ and C₂₀ PUFA in teleosts such as Clarias gariepinus and Siganus canaliculatus²⁰.

The threadfin Eleutheronema is an important marine fishery resource in Indo-Pacific regions, and has become an important and economic aquaculture species in China with high auction prices in local fish markets^{21, 22}. Moreover, the Eleutheronema fishes were considered as a promising aquaculture species in China with a worldwide market demand because of its good flavor and high nutritional value, especially the high fatty acid contents, which could be the potential main food source of fatty acids for human consumption. Thus, the exploration of the underlying molecular basis of fatty acids biosynthesis in Eleutheronema species is of great importance. The Elovl gene family, a set of crucial enzymes involved in LC-PUFA biosynthesis, has been extensively investigated in fish species in recent years. However, the evolution and function of Elovl genes in Eleutheronema species are unknown. Therefore, in the present study, we performed a genome-wide identification of Elovl genes in E. tetradactylum and E. rhadinum, characterized their gene and protein structures, chromosomal distributions, conserved motifs, evolutionary relationships, selection pressure, and expression profiles across different tissues. To our knowledge, this is the first study on the genome-wide identification, evolutionary relationship, and gene expression analysis of gene family in the Eleutheronema genus, which could provide valuable genomic resources for the future functional analysis of Elovl genes in marine fishes.

Materials and methods

Identification of Elovl gene family from E. tetradactylum and E. rhadinum

To identify the Elovl genes in Eleutheronema fish, we used the elovl amino acid sequences from zebrafish as a reference and the ELO HMM model in Pfam to detect the Elovl genes in the E. tetradactylum and E. rhadinum genome sequences. The whole genome assembly of the Eleutheronema species yielded a total genome length of 582 Mb (N50 = 18.26 Mb) in E. tetradactylum and 591 Mb (N50 = 22.59 Mb) in E. rhadinum (unpublished data). BLASTP program were used to identify the elovl like genes in E. tetradactylum and E. rhadinum protein database with the e-value of e-10 as the criterion. The hidden Markov model (HMM) files corresponding to the ELO (PF01151) domains were downloaded from the Pfam protein family database and then subjected to the E. tetradactylum and E. rhadinum protein database, respectively. The ExPASy (https://web.expasy.org/protparam/) was used to compute various physical and chemical parameters of ELOVL members. SignalP-3.0 (https://services.healthtech.dtu.dk/service.php?SignalP-3.0) and TMHMM-2.0 (https://services.healthtech.dtu.dk/service.php?TMHMM-2.0) were used to predict the presence and location of signal peptide cleavage sites and transmembrane helices in Elovl proteins. The CELLO v.2.5 (http://cello.life.nctu.edu.tw/) was utilized to analyze the subcellular localization of ELOVL members. Gene structures of ELOVL members were analyzed by the Gene Structure Display Server (GSDS, http://gsds.gao-lab.org/). The conserved motifs were identified using the MEME tool (https://meme-suite.org/meme/tools/meme) with the default parameters. All the Elovl gene sequences were subjected to the MCScanX program²³ for the chromosomal locations and collinearity analysis and then visualized using Circos software²⁴. Protein–protein interaction analysis was performed on the STRING database (http://string-db.org) using Elovl genes as the queries and the zebrafish proteins as a reference. The secondary structures were eventually predicted via the SOPMA program²⁵ and Vadar server²⁶. The three-dimensional structures were predicted through Protein Homology/analogy Recognition Engine V 2.0 (Phyre2) server²⁷. The validation of predicted protein models was assessed through Ramachandran Plot Analysis²⁶. The non-synonymous substitution (Ka) and synonymous substitution (Ks) values were estimated for Elovl genes using KaKs_Calculator software²⁸, and the selection mode of each gene pair was evaluated according to the Ka/Ks ratio.

Phylogenetic analysis of Elovl genes

To investigate the evolutionary relationships of Elovl genes in the Eleutheronema genus, the sequences of Elovl genes from E. tetradactylum and E. rhadinum were subjected to phylogenetic analysis and compared with the 106 publicly available Elovl genes of other organisms, including Homo sapiens, Mus musculus, Danio rerio, Oreochromis niloticus, Oncorhynchus mykiss, Oryzias latipes, Takifugu rubripes, Gadus morhua, Sander lucioperca, Salmo salar, Cyprinus carpio, Lepisosteus oculatus, Ictalurus punctatus, Tachysurus fulvidraco, Coregonus clupeaformis, and Ictalurus punctatus. All reference sequences were aligned by the ClustalW program in MEGAX with the default parameter²⁹, and the phylogenetic tree was constructed using the maximum like-hood method with 1000 replicates. The optimal protein substitution model for evolution of Elovl genes was the JTT model with gamma distribution, which was chosen based on the results using the option provided in MEGAX.

Expression analysis of Elovl genes across different tissues

We aimed to determine the expression patterns of Elovl genes across different tissues. In brief, adult E. tetradactylum and E. rhadinum were collected from China's coastal water. This study was reported in accordance with ARRIVE guidelines. Total RNAs of 9 different tissues, including the brain, eye, gill, heart, kidney, liver, muscle, stomach, and intestine, were isolated using RNA Extraction Kit (Takara, Japan) following the manufacturer's procedure. The RNA integrity was detected by the RNA Nano 6000 Assay Kit of the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA) and 1% agarose gels. The extracted RNAs (2 μg) of each tissue (n = 3 replicates) were respectively reverse transcribed into cDNA using the PrimeScript™ RT reagent Kit (Takara, Japan). The primers used for qPCR analysis in this study are shown in Table 1. The total qPCR mixture volume of 20 μL contained 10 μL of 2 × SYBR Green PCR buffer, 0.4 μL of forward and reverse primers, 1 μL of cDNA template, and 7.4 μL ddH₂O. The qPCR analysis was performed in triplicates on a 96-well rotor in 7300 RT-PCR system (Applied Biosystems, USA) with the following amplification program: predenaturation at 95 °C for 30 s; denaturation at 95 °C for 5 s, annealing at 58 °C for 30 s, 40 cycles. The data were analyzed statistically using the 2^−ΔΔCt method using the β-actin as the internal reference gene.

Table 1.

Gene-specific primers for qPCR.

Species	Gene name	Forward primer	Reverse primer
E. tetradactylum	elovl1a	ATGGTGAATGCAGGAGTCCA	TGTAAGCTGGATGGCAGTCA
	elovl1b	TGGATGTACGGCACCTTCTT	GATGCCGTTGCCATTCTCAT
	elovl4a	GGGCTGCAAGCTACAGTTAC	GAACATGGTGCAATGGTGGT
	elovl4b	GCAACCATCAACTCTGGCAT	ATGGTCACGTGGAACTGGAT
	elovl5	CATATGGCCGTGTGACTTCC	AAGAGAGCCATTCTGGTGCT
	elovl6	GTCCATCCCACATGCAGAAC	ACTCGCTCTTCTTGGTGACA
	elovl6l	GTCATTTGCCTTCGTCGTGA	AGCACAGTGATGTGGTGGTA
	elovl7a	GGGCCTCGGATAATGGAGAA	ATGGCCTGTGGTGAATCAGA
	elovl8b	GGCACCATGATCTTCAACTG	CACTCTGTGAACAAGTTGTAGC
	β-actin	CTGGACTTCGAGCAGGAGAT	GGATTCCGCAGGACTCCATA
E. rhadinum	elovl1a	ATGGTGAATGCAGGAGTCCA	TGTAAGCTGGATGGCAGTCA
	elovl1b	CCCTGGTGGAATGGGATCTT	ACCAGGACGAACTGGGTAAG
	elovl4a	GGGCTGCAAGCTACAGTTAC	GAACATGGTGCAATGGTGGT
	elovl4b	GCAACCATCAACTCTGGCAT	ATGGTCACGTGGAACTGGAT
	elovl5	CATATGGCCGTGTGACTTCC	AAGAGAGCCATTCTGGTGCT
	elovl6	GTCCATCCCACATGCAGAAC	ACTCGCTCTTCTTGGTGACA
	elovl6l	GTCATTTGCCTTCGTCGTGA	AGCACAGTGATGTGGTGGTA
	elovl7a	GGGCCTCGGATAATGGAGAA	ATGGCCTGTGGTGAATCAGA
	elovl8b	CCACATCAGGTGGAATGCTG	CCACATCTGGTTCACCTCCT
	β-actin	CTGGACTTCGAGCAGGAGAT	GGATTCCGCAGGACTCCATA

Ethics declarations

All experimental protocols were approved by the Research Committee of City University of Hong Kong. All methods were carried out in accordance with the relevant guidelines and regulations of the City University of Hong Kong.

Results and discussion

Identification of Elovl genes from E. tetradactylum and E. rhadinum

Totally, we successfully identified 9 Elovl genes, including elovl1a, elovl1b, elovl4a, elovl4b, elovl5, elovl6, elovl6l, elovl7a, and elovl8b, both from E. tetradactylum and E. rhadinum genome (Table 2). In E. rhadinum, the shortest and the longest putative CDS length among all Elovl genes was 810 bp and 2019 bp, respectively. Their encoded protein size ranged from 269 amino acids to 672 amino acids. The theoretical molecular weight of Elovl proteins varied from 31061.48 to 75051.42 Da, with the theoretical isoelectric points (pI) ranging from 7.86 to 9.59. Most of the Elovl proteins were characterized as stable and hydrophilic proteins. Signal peptide prediction analysis showed that the elovl1b, elovl5, and elovl6 contained signal peptide sequences. In addition to elovl8b, all Elovl proteins contained transmembrane domains ranging from 5 to 7. Almost all Elovl proteins were predicted to be endoplasmic reticulum-located except elovl8b, predominantly localized in the nucleus.

Table 2.

Basic information for the Elovl gene family members.

Species	Gene name	putative CDS (bp)	Size (aa)	Molecular weight (Da)	pI	Instability index	Aliphatic index	Hydropathicity	Transmembrane domain	Signal peptide	Subcellular localization
E. rhadinum	elovl1a	942	313	37,022.53	9.58	36.07	88.69	0.164	7	NO	Endoplasmic reticulum
	elovl1b	981	326	37,938.41	9.32	29.79	85.86	0.13	7	YES	Endoplasmic reticulum
	elovl4a	1077	358	41,720.58	9.6	37.67	83.1	− 0.1	6	NO	Endoplasmic reticulum
	elovl4b	963	320	37,086.3	9.59	32.31	85.94	0.021	6	NO	Endoplasmic reticulum
	elovl5	885	294	35,197.1	9.11	36.69	86.84	0.041	7	YES	Endoplasmic reticulum
	elovl6	810	269	31,560.24	9.38	46.26	89.55	0.234	5	YES	Endoplasmic reticulum
	elovl6l	825	274	31,061.48	8.72	50.98	94.67	0.368	5	NO	Endoplasmic reticulum
	elovl7a	885	294	34,564.52	9.14	28.66	86.5	0.241	7	NO	Endoplasmic reticulum
	elovl8b	2019	672	75,051.42	7.86	42.73	92.41	− 0.311	0	NO	Nucleus
E. tetradactylum	elovl1a	942	313	36,886.42	9.63	34.02	89.94	0.178	7	YES	Endoplasmic reticulum
	elovl1b	981	326	37,928.33	9.32	30.38	82.88	0.103	7	NO	Endoplasmic reticulum
	elovl4a	1077	358	41,763.61	9.64	37.67	82.01	− 0.123	6	NO	Endoplasmic reticulum
	elovl4b	975	324	37,605.01	9.57	33.4	88.77	0.064	7	NO	Endoplasmic reticulum
	elovl5	885	294	35,197.1	9.11	36.69	86.84	0.041	7	YES	Endoplasmic reticulum
	elovl6	810	269	31,560.24	9.38	46.26	89.55	0.234	5	YES	Endoplasmic reticulum
	elovl6l	825	274	31,049.42	8.72	50.36	93.25	0.342	5	NO	Endoplasmic reticulum
	elovl7a	885	294	34,534.49	9.14	29.08	86.84	0.252	7	NO	Endoplasmic reticulum
	elovl8b	1824	607	68,750.14	8.93	46.33	83.48	− 0.264	3	NO	Nucleus

In E. tetradactylum, the putative CDS length of Elovl genes ranged from 810 to 1824 bp, and their encoded protein size ranged from 269 amino acids to 409 amino acids. The molecular weight of Elovl proteins varied from 31049.42 to 68750.14 Da, with the pI ranging from 8.72 to 9.64. Like Elovl proteins in E. rhadinum, most elovl proteins were predicted to be stable and hydrophilic. Signal peptide prediction analysis revealed that elovl1a, elovl5, and elovl6 had signal peptide sequence, which was different from E. rhadinum that elovl1b contained signal peptide sequence, but elovl1a did not. In addition, seven members showed the same number of transmembrane structures with E. rhadinum, while the elovl8b contained three and elovl4b contained seven transmembrane structures in contrast to E. rhadinum. The elovl8b was predicted to be localized in nuclear, while other members were localized in the endoplasmic reticulum, similar to E. rhadinum.

Evolution of divergence and conservation of Elovl genes

Divergence and conservation accompany the process of species evolution. To elucidate the phylogenetic relationship of Elovl genes among different species, a maximum like-hood tree was constructed on the basis of 18 Elovl genes in E. tetradactylum and E. rhadinum and 106 publicly available Elovl protein sequences. As shown in Fig. 1, these Elovl genes can be divided into eight subfamilies, including elovl1a/1b, elovl2, elovl3, elovl4a, elovl5, elovl6/6 l, elovl7a/7b, elovl8a/8b. However, 6 subfamilies were presented in the Eleutheronema genus, and there was only one subtype for elovl7 (elovl7a) and elovl8 (elovl8b) in E. tetradactylum and E. rhadinum. The elovl3 was mainly identified in mammalians such as Homo sapiens and Mus musculus, while a recent study reported a full repertoire of Elovl genes in the Colossoma macropomum genome, including elovl3³⁰. The loss of elovl2 occurred in the vast majority of marine fish lineages, which was only presented in a few fish species, such as C. carpio, D. rerio, S. salar, and S. grahami.

Figure 1.

Phylogenetic tree for 18 Elovl proteins from E. tetradactylum and E. rhadinum, and 106 publicly available Elovl proteins from other species. All these proteins were aligned using ClustalW and then subjected to MEGAX for phylogenetic tree construction using the maximum like-hood method with 1000 replicates.

We further performed the gene structure analysis to visualize the exon–intron structure of each gene, and the results revealed that the elovl8b had the largest intron number, while the elovl6/6 l subfamily genes contained three introns (Fig. 2a). Except for elovl8, Elovl genes belonging to the same subfamily shared a similar gene structure. Additionally, we identified ten motifs in Elovl genes, and the conversed motif types, numbers, and distributions in Elovl proteins were much more similar except for the elovl8b (Fig. 2b, TableS1). Two conserved motifs were found in the Elovl gene family except for elovl8b in E. rhadinum, which were related to the ELO domain via SMART evaluation analysis (Fig. 2c and d). Gene structural variation is important for gene evolution. In E. tetradactylum and E. rhadinum, Elovl genes showed similar gene structure, and the proteins shared similar motif compositions, indicating that the Elovl genes were highly conserved in the Eleutheronema genus.

Figure 2.

Gene structure and conserved motifs diagram of Elovl genes. (a) Gene structure of Elovl genes. Exons were represented by pink boxes and introns by black lines; (b) Conserved motifs of Elovl proteins; (c and d) Logo representations of the ELO domains, motifs 1 and 2, respectively.

In the process of evolution via natural selection, adaptation to certain environmental conditions likely drove the changes in endogenous capacity for LC-PUFA biosynthesis between marine and freshwater fishes³¹. The Elovl gene family has been functionally studied and characterized in a variety of fish species, and the member of the Elovl gene family of each species varied greatly. In the present study, for a comprehensive analysis of Elovl genes in the Eleutheronema genus, the Elovl gene ortholog clusters of mammals and various teleosts with different ecological niches and habitats were collected. The results showed that only seven Elovl genes (one gene for each subtype) were observed in mammals; however, more members were variably presented in teleosts, which might be related to the teleost-specific duplication. A previous study revealed that Sinocyclocheilus graham and C. carpio possessed the highest number of Elovl genes, containing 21 members of subtypes, resulting from an extra independent 4th whole-genome duplication event^{32, 33}. Interestingly, only 9 Elovl genes were observed in Eleutheronema genus, the same as T. rubripes, possibly due to gene loss and the asymmetric acceleration of the evolutionary rate in one of the paralogs following the whole-genome duplication in some teleost fishes³⁴. Additionally, the elovl2 and elovl3 were absent, but a novel subtype, elovl8, was present in most marine fishes. The elovl8, the most recently identified and novel active member of the Elovl protein family member, has been proposed to be a fish-specific elongase with two gene paralogs (elovl8a and elovl8b) described in teleost³⁵. In Eleutheronema, we also found that the elovl8b was presented in E. tetradactylum and E. rhadinum, indicating the important roles in the LC-PUFAs biosynthesis of Eleutheronema fish. Similar results were also observed in rabbitfish and zebrafish²⁰. The Elovl gene family member number in Eleutheronema genus is the same as T. rubripes, but less than I. punctatus (10), Gadus morhua (10), D. rerio (14), S. salar (18), and C. carpio (21), which might be due to the differential expansion events during the evolutions of fish species.

Predicting the protein structure is a fundamental prerequisite for understanding the function and possible interactions of a protein. In the present study, the secondary structures as well as three-dimensional structures of Elovl proteins in both E. tetradactylum and E. rhadinum were predicted using the SOPMA and Phyre2 programs, respectively. The protein structures of all the candidate Elovl proteins were modeled at > 90% confidence. The secondary structures of these proteins in E. tetradactylum revealed 40.86–50.30% alpha helixes, 28.10–28.10% random coil, 13.75–20.67% extended strand and 2.38–4.47% beta turn, while these ratios were predicted to be 47.55–53.27, 30.00–36.01, 6.99–18.12 and 2.38–4.75%, respectively, in E. rhadinum (Table 3). High ratio of alpha helixes and random coil in the Elovl protein structure might play important roles in fatty acids biosynthesis in fish, in accordance with the literature for the order Perciformes in Perca fluviatilis³⁶. Additionally, the secondary structure pattern of Elovl proteins in the candidate E. tetradactylum and E. rhadinum species were highly similar (Fig. 3), indicating the probable similar biological functions as well as highly evolutionarily conserved Elovl genes in Eleutheronema species.

Table 3.

Properties of the secondary structures of Elovl proteins.

Species	Protein	Alpha helix (%)	Beta turn (%)	Random coil (%)	Extended strand (%)
E. tetradactylum	elovl1a	46.65	3.83	30.99	18.53
	elovl1b	46.63	2.76	36.81	13.80
	elovl4a	43.02	4.47	31.84	20.67
	elovl4b	47.22	3.40	32.10	17.28
	elovl5	47.28	2.38	34.69	15.65
	elovl6	51.30	2.60	32.34	13.75
	elovl6l	50.36	3.65	28.10	17.88
	elovl7a	48.98	2.38	31.29	17.35
	elovl8b	40.86	4.12	40.69	14.33
E. rhadinum	elovl1a	50.16	2.56	31.31	15.97
	elovl1b	47.55	2.76	33.44	16.26
	elovl4a	48.32	4.75	32.40	14.53
	elovl4b	48.75	3.12	30.00	18.12
	elovl5	47.96	2.38	33.67	15.99
	elovl6	51.30	2.60	32.34	13.75
	elovl6l	48.54	3.28	31.02	17.15
	elovl7a	47.62	3.40	30.95	18.03
	elovl8b	53.27	3.72	36.01	6.99

Figure 3.

The secondary structure pattern, including alpha helix (blue color), random coil (purple color), extended strand (red color), and beta turn (green color), of Elovl proteins in E. tetradactylum and E. rhadinum.

The 3D model results showed that all predicted Elovl proteins had complex 3D structures, composing of multiple secondary structures including alpha-helices, random coils, and others (Fig. 4). The Elovl proteins of different subfamilies showed different 3D configurations. The 3D structures of Elovl proteins also revealed the presence of the conserved domain in each Elovl protein, which showed a typical three-dimensional frame comprising of various parallel alpha-helixes. To assay the quality and accuracy of the predicted 3D model for the candidate Elovl proteins, the Ramachandran plot analysis was employed (Figure S1). In model validation, the qualities of the Elovl proteins model varied from 90 to 98% based on the Ramachandran plot analysis, suggesting the reasonably good quality and reliability of the predicted 3D models. These results indicated that the predicted 3D model of Elovl proteins could provide valuable information for the further comprehensive studies of molecular function in the fatty acids biosynthesis in Eleutheronema species. Additionally, the comparisons between these structures in E. tetradactylum and E. rhadinum suggested that the Elovl proteins encompassed the conserved structures. In addition, gene duplication resulted in obvious 3D structural variation in the duplicated genes, such as Elovl4 (elovl4a and elovl4b), Elovl6 (elovl6 and elovl6l). The ascertained variations were revealed in duplicated Elovl proteins, and the diversities in these proteins structure may reflect their different obligations in the fatty acid biosynthesis and other biological processes.

Figure 4.

Three-dimensional modeling of Elovl proteins in E. tetradactylum and E. rhadinum. All models have confidence levels above 90%.

To explore the functional selection pressures acting on Elovl gene family, Ka, Ks, and Ka/Ks ratios were calculated for each gene. Generally, Ka/Ks < 1 indicates purifying or negative selection, Ka/Ks = 1 represents neutral selection, and Ka/Ks > 1 indicates positive selection. In this study, we found that all the Ka/Ks ratios for each gene were less than 0.5, suggesting that they were subjected to strong purifying selection during evolution, and their functions might be evolutionarily conserved (Fig. 5). Therefore, theoretically, the Elovl genes in the Eleutheronema genus had eliminated deleterious mutations in the population through purification selection. Similar results were also observed in Elovl gene family of Gymnocypris przewalskii that no positive selection trace was detected in most members except elovl2¹¹. Moreover, elovl6l and elovl8b showed a higher average Ka/Ks ratio than the other seven members, indicating that the evolution of elovl6l and elovl8b might be much less conservative and thereby could provide more variants for natural selection in Eleutheronema species.

Figure 5.

The evolutionary rates of the Elovl genes in (a) E. tetradactylum and (b) E. rhadinum. The Ka, Ks, and Ka/Ks values were demonstrated in boxplots with error lines.

Chromosomal location, collinearity, and protein–protein interaction network analysis of Elovl genes

As shown in Fig. 6a and b, Elovl genes were randomly and unevenly distributed on seven chromosomes in both E. tetradactylum and E. rhadinum, including Chr5, Chr6, Chr8, Chr10, Chr11, Chr13, and Chr25. The Chr5 and Chr6 harbored two Elovl genes (elovl1b and elovl8b in Chr5, elovl5 and elovl6l in Chr6), while other chromosomes each carried a single Elovl gene. Collinearity relationship analysis was performed to further investigate the gene duplication events within the Elovl gene family. The results revealed that a pair of segmental duplication genes (elovl4a/4b) showed collinear relationships. A chromosome-wide collinearity analysis also showed that the chromosomes were highly homologous between E. tetradactylum and E. rhadinum, including the Elovl gene family (Figure S2). To infer the protein interaction within Elovl gene family, we constructed the protein–protein interaction (PPI) network of the Elovl proteins based on the interaction relationship of the homologous Elovl proteins in zebrafish. The results showed that Elovl genes had close interaction with other members except for the elovl4a/4b and elovl8b (Fig. 6c), which suggested that they might participate in diverse functions by interacting with other proteins. Thus far, elovl4a and elovl4b were widely identified in most fish, which could effectively elongate PUFA substrates³⁷. In addition, the elovl4a/4b were identified to be homologous proteins of zebrafish, indicating that the elovl4 subtype was highly conserved during evolution and played important roles in the biosynthesis of LC-PUFA in Eleutheronema.

Figure 6.

Chromosomal location and collinearity analysis of Elovl gene family in (a) E. tetradactylum and (b) E. rhadinum. Colored boxes represented chromosomes. Segmental duplication genes are connected with grey lines; (c) a protein–protein interaction network for Elovl genes based on their orthologs in zebrafish.

Expression patterns of ELOVL genes in different tissues

In the present study, we aimed to determine the expression patterns and gained insights into the potential functions of Elovl genes in the brain, eye, gill, heart, kidney, liver, muscle, stomach, and intestine. The expression patterns of Elovl genes in different tissues and species were distinct, suggesting the diverse roles during fish development (Fig. 7a and b). In our present study, the elovl1a and elvovl1b were expressed in a relatively narrow range of tissues, including the liver, stomach, and intestine. Some Elovl genes had much higher relative expression rates, e.g., elovl1a and elovl7a. The elovl4a was primarily distributed in the brain and eye, slightly expressed in gills while hardly detectable in other tissues, consistent with previous studies^{37, 38}, which might play an important role in endogenous biosynthesis of LC-PUFA in the neural system of fish. In contrast to elovl4a, elovl4b was ubiquitously, instead of tissue-specific, expressed in most tissues while hardly examined in the heart and kidney. The elovl4a and elovl4b were two commonly paralogues in evolutionarily diverged fish species, and the striking difference in expression patterns between elovl4a and elovl4b might be due to the potential functional divergence of these two paralogues. In addition, elovl8b, the novel active member of the Elovl protein family, was expressed in several tissues, suggesting the essential roles in LC-PUFAs biosynthesis of teleost as indicated by a previous study²⁰. Moreover, the differences in expression patterns among different Elovl genes indicated that these genes might possibly undergo functional divergence during evolution in the Eleutheronema genus. Overall, our present study firstly provided the preliminary organ-specific expression data of the Elovl gene family in E. tetradactylum and E. rhadinum, which could provide the foundation for further clarifying the function of these genes in the evolutionary development of the Eleutheronema genus.

Figure 7.

qPCR assessment of tissue distribution of elovl1a, elovl1b, elovl4a, elovl4b, elovl5, elovl6, elovl6l, elovl7a, and elovl8b gene expression in (a) E. tetradactylum and (b) E. rhadinum for various tissues including the brain, eye, gill, heart, kidney, liver, muscle, stomach, and intestine.

Conclusion

The Elovl gene family plays important roles in LC-PUFA biosynthesis in fish, thus the identification of the Elovl gene family in Eleutheronema species is essential to understand its underlying molecular basis of fatty acids biosynthesis. In this study, we first identified 9 Elovl genes from two Eleutheronema fish, including E. tetradactylum and E. rhadinum genome. We found that the Elovl gene family could be classified into six subfamilies in Eleutheronema species, including elovl1a/1b, elovl4a/4b, elovl5, elovl6/6 l, elovl7a, and elovl8b, which were consistently supported by their gene and protein structures and conserved motifs. The gene loss of elovl2, elovl3, elovl7b, and elovl8a was observed in Eleutheronema genus. All found genes showed highly conserved gene and protein structures and motif composition except for elovl8b. We also found that the Ka/Ks ratios for each gene were less than 0.5, suggesting that the Elovl genes in the Eleutheronema genus were subjected to strong purifying selection during evolution. The expression patterns of Elovl genes across tissues and species were distinct, suggesting possibly functional divergence during evolution in the Eleutheronema genus. Overall, our present study firstly presented comprehensive information about the sequence characteristics, phylogenetic relationships, chromosome distribution, selection pressure, and expression patterns of Elovl gene in the Eleutheronema genus. The results could provide the basis for further research on the function of the Elovl gene family in marine fish, thus facilitating the molecular breeding of fatty acids-enriched fishes with potential benefits to farmers and the health of consumers.

Author contributions

X.J: conceptualization, data analysis, writing W.WX: conceptualization, writing. All authors reviewed the manuscript.

Data availability

The sequence data obtained in this study have been deposited at GenBank of NCBI at OP391488-OP391504 (https://www.ncbi.nlm.nih.gov/).

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-023-28342-4.