Molecular Cloning and Bioinformatics Analysis of DQA Gene from Mink (Neovison vison)

In the present study, we cloned, sequenced, and explored the structural and functional characteristics of the major histocompatibility complex (MHC)-DQA gene from mink (Neovison vison) for the first time. The full-length sequence of DQA gene was 1147-bp-long, contained a coding region of 768-bp, which was predicted to encoding 255 amino acid residues. The comparison between DQA from mink (Neovison vison) and other MHC-DQA molecules from different animal species showed that nucleotide and encoded amino acid sequences of the mink DQA gene exhibited high similarity with the ferret (Mustela pulourius furo). Phylogenetic analysis revealed that mink (Neovison vison) DQA is grouped with that of ferret (Mustela pulourius furo). The cloned sequence contained a 23-amino acid NH2-terminal signal sequence with the signal peptide cutting site located in amino acids 23–24, and had three Asn-Xaa-Ser/Thr sequons. Three cysteine residues were also identified (Cys-85, Cys-121, and Cys-138). The 218 to 240 amino acids were predicted to be the transmembrane domains. The prediction of the secondary structure revealed three α-helixes and fourteen β-sheets in Neovison vison DQA protein, while random coil was a major pattern. In this study, the whole CDS sequence of Neovison vison DQA gene was successfully cloned, which was valuable for exploring the function and antiviral molecular mechanisms underlying the molecule. The findings of the present study have laid the foundation for the disease resistance and breeding of mink.


Introduction
The major histocompatibility complex (MHC) is a chromosomal region consisting of a group of closely linked loci. It binds to specific epitopes, which is the key to triggering the immune system to attack the virus-infected cell through the proteins involved in the immune response of the body [1][2][3]. Due to the polygenic characteristics, MHC was divided into MHC class I, II, and III genes according to the differences in the structure, tissue distribution, and function of the coding molecules [4]. The class II genes of MHC encoded for α and β chains of DR and DQ dimer molecules, which present antigenic peptides to the helper T cells and was considered as a candidate gene for animal disease-resistant breeding [5][6][7]. A striking characteristic of MHC II-DQA genes is their extreme polymorphism [8][9][10]. Due to the central role of DQA gene in the vertebrate immune system, it is presumed that one of the main selective pressures promoting MHC diversity gives rise to the susceptibility or resistance against infectious diseases and autoimmune disorders [11,12].

Molecular Cloning of DQA from Neovison vison
The NCBI DQA sequence of Mustela putorius furo and Enhydra lutris kenyoni were used for molecular cloning of DQA from Neovison vison. The amplified cDNA was separated by electrophoresis on 1.5% agarose gel, which showed the expected band size as compared to the standard molecular weight ladder ( Figure 1). This fragment was excised from the agarose gel, purified, ligated in the TA vector pMD19-T (TaKaRa Biotechnology Co., Ltd., Dalian, China), and cloned in E. coli. The positive clones were confirmed by colony PCR. The amplicons were obtained by PCR amplification of the cDNA with gene-specific primers encoding the DQA gene ( Figure 1). The amplicons were purified, cloned, and characterized by sequencing. This sequence represented the first cloned DQA from Neovison vison, which encompasses the full coding region in comparison with the corresponding regions from different organisms. Sequence analysis revealed that the cDNA encoding the   The deduced amino acid sequence is reported in one-letter code. The start is in boldface type, and the underlined and stop codon is showed with asterisks. The three cysteine residues (Cys-133, Cys-189, and Cys-221) are highlighted in gray. Two potential Asn-Xaa-Ser/Thr sequons were also observed in the sequence (underlined in boldface type).  The deduced amino acid sequence is reported in one-letter code. The start is in boldface type, and the underlined and stop codon is showed with asterisks. The three cysteine residues (Cys-133, Cys-189, and Cys-221) are highlighted in gray. Two potential Asn-Xaa-Ser/Thr sequons were also observed in the sequence (underlined in boldface type).

Figure 2.
The nucleotide and deduced amino acid sequences of Neovison vison DQA. The deduced amino acid sequence is reported in one-letter code. The start is in boldface type, and the underlined and stop codon is showed with asterisks. The three cysteine residues (Cys-133, Cys-189, and Cys-221) are highlighted in gray. Two potential Asn-Xaa-Ser/Thr sequons were also observed in the sequence (underlined in boldface type). Both sequences were aligned with the retrieved sequences using Clustal (W) method from Mega 5.2 software, and percent identity and divergence were determined. Several clades were formed during the phylogenetic analysis of the broad class of DQA gene. Neovison vison and Mustela pulourius furo formed one monophyletic clade, whereas that from other species except Equus caballus formed a separate clade. The Neomonachus schauinslandi DQA was grouped closely with that of Odobenus rosmarus divergens and Zalophus ( Figure 3). walrus (Odobenus rosmarus divergens), 90.5% Sea lions (Zalophus), 90.5% Hawaiian monk seal (Neomonachus schauinslandi), 89.6% wolf (Canis lupus), and 86.6% horse (Equus caballus). The deduced amino acid sequences showed a maximum divergence of 23.3%, with sequence identities ranging from 80.0 to 96.9% as compared to other species. Furthermore, DQA proteins shared 96.9% and 89.4% identity with ferret (Mustela pulourius furo) and sea otters (Enhydra lutris), respectively. Also, 88.2% identity was detected with sea bear (Ursus maritimus) and giant panda (Ailuropoda melanoleuca), 86.7% with walrus (Odobenus rosmarus divergens), 86.3% with sea lions (Zalophus) and wolf (Canis lupus), 85.1% with Hawaiian monk seal (Neomonachus schauinslandi), and 80.0% with horse (Equus caballus) proteins.
Both sequences were aligned with the retrieved sequences using Clustal (W) method from Mega 5.2 software, and percent identity and divergence were determined. Several clades were formed during the phylogenetic analysis of the broad class of DQA gene. Neovison vison and Mustela pulourius furo formed one monophyletic clade, whereas that from other species except Equus caballus formed a separate clade. The Neomonachus schauinslandi DQA was grouped closely with that of Odobenus rosmarus divergens and Zalophus ( Figure 3).

Hydrophilic and Hydrophobic Analysis of Neovison vison DQA Protein
ProtScale was used to analyze the hydrophilicity and hydrophobicity of DQA protein, and Hphob./Kyte & Doolittle were selected as the prediction standards. Figure 4 presented three distinct hydrophobic regions, with the highest score 3.089 at the position 236Thr (T) of the polypeptide chain, which represented the strongest hydrophobicity, while the lowest score −2.144, at position 53Asp (D), represented maximal hydrophilicity. Since more than half of the amino acid sequences are hydrophobic residues, the entire polypeptide chain is deemed as such with an average hydrophobic index of 0.300.

Hydrophilic and Hydrophobic Analysis of Neovison vison DQA Protein
ProtScale was used to analyze the hydrophilicity and hydrophobicity of DQA protein, and Hphob./Kyte & Doolittle were selected as the prediction standards. Figure 4 presented three distinct hydrophobic regions, with the highest score 3.089 at the position 236Thr (T) of the polypeptide chain, which represented the strongest hydrophobicity, while the lowest score −2.144, at position 53Asp (D), represented maximal hydrophilicity. Since more than half of the amino acid sequences are hydrophobic residues, the entire polypeptide chain is deemed as such with an average hydrophobic index of 0.300.   Figure 5). This result was in agreement with the predicted hydrophobic peak (Figure 4), which further verifies that this area might be a transmembrane region.   Figure 5). This result was in agreement with the predicted hydrophobic peak (Figure 4), which further verifies that this area might be a transmembrane region.

Subcellular Localization and Domain Prediction of Neovison vison DQA Protein
The subcellular localization of DQA protein was predicted using Target P1.1 server, which indicated a secretory channel signaling peptide (SP) in the protein (Figure 9a). SMART program was used to predict the structural domain of DQA protein (Figure 9b). Amino acids 30-110 aa are MHC II, 128-199 aa is the IGc1 domain, and 218-240 aa is the transmembrane region ( Figure 5).

Function Prediction of DQA Protein
In this study, ProtFun 2.2 Server analysis predicted the functional domain and classification of the DQA protein. Table 1 lists that the highest probability is voltage-gated ion channel with the probability value 12.682. However, additional tests are required to determine whether it is a Na+, K+, or Ca2+ ion channel. The secondary structure of Neovison vison DQA using the PredictProtein software predicted that the DQA protein was composed of 19.2% alpha helix, 34.5% beta fold, and 46.3% random coils. Figure 10 displayed the results of PSIPRED analysis, wherein the addition of three alpha helices and fourteen beta folding regions rendered a majority of the amino acids in a state of irregular curls in DQA protein. The homologous modeling of Neovison vison DQA protein was carried out using the SWISS-MODEL Server. The tertiary structure of the DQA protein primarily in the form of an alpha helix, beta folding, and random crimp was in agreement with the predicted secondary structure of DQA protein (Figure 11).

Function Prediction of DQA Protein
In this study, ProtFun 2.2 Server analysis predicted the functional domain and classification of the DQA protein. Table 1 lists that the highest probability is voltage-gated ion channel with the probability value 12.682. However, additional tests are required to determine whether it is a Na + , K + , or Ca 2+ ion channel. The secondary structure of Neovison vison DQA using the PredictProtein software predicted that the DQA protein was composed of 19.2% alpha helix, 34.5% beta fold, and 46.3% random coils. Figure 10 displayed the results of PSIPRED analysis, wherein the addition of three alpha helices and fourteen beta folding regions rendered a majority of the amino acids in a state of irregular curls in DQA protein. The homologous modeling of Neovison vison DQA protein was carried out using the SWISS-MODEL Server. The tertiary structure of the DQA protein primarily in the form of an alpha helix, beta folding, and random crimp was in agreement with the predicted secondary structure of DQA protein (Figure 11).

Discussion
The DQA gene presents foreign peptide antigens from infectious agents, including bacteria, viruses, and autoimmune antigens [21], providing a valuable source of genetic markers for examining the correlations between host variants and disease resistance or susceptibility [22]. As a functional unit, exon 3 of the swine leukocyte antigen DQA (SLA-DQA) gene is pivotal to both diarrhea susceptibility and resistance based on the effects on diarrheal infection in 425 suckling piglets [23,24]. The correlation between the polymorphisms of SLA-DQA exon 2 and piglet diarrhea in three Chinese native pig breeds indicated that the amino acid variations in the pepetide binding pockets play a critical role in the antigen-binding groove in piglet diarrhea resistance [25]. Feng et al. demonstrated significant differences in the white blood cell count, red blood cell count, neutrophil ratio, and other immune indicators of GOLA-DQA genotypes, which provide a basis for breeding disease-resistant species of goats [26]. Bao et al. reported that the resistance to E. coli F18 infection in weaned piglets is related to the mRNA expression of SLA-DQA [12]. However, whether the DQA gene can be considered as a candidate gene for mink disease-resistant breeding necessitates further

Discussion
The DQA gene presents foreign peptide antigens from infectious agents, including bacteria, viruses, and autoimmune antigens [21], providing a valuable source of genetic markers for examining the correlations between host variants and disease resistance or susceptibility [22]. As a functional unit, exon 3 of the swine leukocyte antigen DQA (SLA-DQA) gene is pivotal to both diarrhea susceptibility and resistance based on the effects on diarrheal infection in 425 suckling piglets [23,24]. The correlation between the polymorphisms of SLA-DQA exon 2 and piglet diarrhea in three Chinese native pig breeds indicated that the amino acid variations in the pepetide binding pockets play a critical role in the antigen-binding groove in piglet diarrhea resistance [25]. Feng et al. demonstrated significant differences in the white blood cell count, red blood cell count, neutrophil ratio, and other immune indicators of GOLA-DQA genotypes, which provide a basis for breeding disease-resistant species of goats [26]. Bao et al. reported that the resistance to E. coli F18 infection in weaned piglets is related to the mRNA expression of SLA-DQA [12]. However, whether the DQA gene can be considered as a candidate gene for mink disease-resistant breeding necessitates further Figure 11. Prediction of the tertiary structure of DQA protein.

Discussion
The DQA gene presents foreign peptide antigens from infectious agents, including bacteria, viruses, and autoimmune antigens [21], providing a valuable source of genetic markers for examining the correlations between host variants and disease resistance or susceptibility [22]. As a functional unit, exon 3 of the swine leukocyte antigen DQA (SLA-DQA) gene is pivotal to both diarrhea susceptibility and resistance based on the effects on diarrheal infection in 425 suckling piglets [23,24]. The correlation between the polymorphisms of SLA-DQA exon 2 and piglet diarrhea in three Chinese native pig breeds indicated that the amino acid variations in the pepetide binding pockets play a critical role in the antigen-binding groove in piglet diarrhea resistance [25]. Feng et al. demonstrated significant differences in the white blood cell count, red blood cell count, neutrophil ratio, and other immune indicators of GOLA-DQA genotypes, which provide a basis for breeding disease-resistant species of goats [26]. Bao et al. reported that the resistance to E. coli F18 infection in weaned piglets is related to the mRNA expression of SLA-DQA [12]. However, whether the DQA gene can be considered as a candidate gene for mink disease-resistant breeding necessitates further researches that focus on the correlation between the polymorphism of DQA gene in mink and susceptibility to disease and disease phenotype.
The phylogenetic tree based on the nucleotide coding sequence of Neovison vison DQA gene was divided into three major subgroups as follows. Neovison vison had the shortest genetic distance from Mustela pulourius furo, followed by Enhydra lutris, Ursus maritimus, Ailuropoda melanoleuca, and Canis lupus, and the furthest from Equus caballus. The results accorded with the rule of genetic evolution of species, and showed that polymorphism of DQA gene in Neovison vison harbored cross-species maintenance. The DQA gene of mink contained an open reading frame (ORF) of 768 nucleotides, encoding a polypeptide of 255 amino acids, which was similar to ferret, gayal, Bubalus bubalis, and Bamei pigs [27][28][29]; however, some differences were observed from that of Wuzhishan pig [30]. Moreover, 19 nucleotide and eight amino acid substitutions were detected in the DQA gene coding regions as compared to the ferret, which indicated relatively abundant polymorphisms. Based on the theories that mutations in different varieties might affect the spatial structure and function of proteins, it can be speculated that DQA gene is a potential genetic marker.
The secretion of proteins depends on the signal peptide located to the N-terminal of the protein precursor. The DQA protein of mink consisted of a signal peptide with 23 amino acid residues, indicating it to be a secreted protein with transmembrane transport function. This phenomenon was consistent with the prediction that DQA protein harbored a transmembrane domain at 218-240 amino acids. It is proposed that the longer the half-life, the more stable is the protein [31]. In this study, the DQA protein was an unstable entity with up to 30 h half-life in mammalian reticulocytes, which might be related to its different physiological functions in different subcellular organelles [32].
N-glycosylation is one of the major modifications after translation of eukaryotic proteins, which can affect the antigen-determining clusters, charge properties, and thermal stability of the proteins. The degree of glycosylation and abnormalities in the structure of sugar chains are the markers of cancer, which aid in the early diagnosis and timely treatment of diseases [33,34]. The mink DQA protein consists of two putative N-glycosylation sites (N104-Y105-T106 and N144-V145-T146), which are consistent with those in Bamei pigs [29]. The reliability of the predicted glycosylation site of DQA protein in mink and its physiological function needs to be investigated further. In addition, the secondary structure prediction for DQA protein in mink indicated that the percentages of α-helix, β-sheet, and random coil are 19.2%, 34.5%, and 46.3%, respectively. This result was in agreement with that of Bamei pigs, canine, and Yantai black pig [35,36]. The SWISS-MODEL server was used to predict the three-dimensional structure of 25-208 amino acids of mink DQA protein using HLA-DQA1 protein as the homologous modeling tool. The similarity and coverage between mink-DQA and HLA-DQA1 sequences was 78.87% and 76%, respectively. Nevertheless, further exploration of the DQA protein structure would clarify its biological functions and immunological mechanisms.

Experimental Mink
The liver tissue of mink was provided by the Special Animal Genetic Resources Key Laboratory of the Institute of Special Products, Chinese Academy of Agricultural Sciences.

RNA Extraction, cDNA Synthesis, and Reverse Transcription PCR
Tissue samples (heart, liver, spleen, lung, kidney, intestines, skin, and cureus) were excised, snap-frozen in liquid nitrogen, and stored at −80 • C until further use. Total RNA was extracted from mink tissues using RNAiso Plus (TaKaRa) and quantified by spectrophotometry on the NanoDrop (Thermo Fisher Scientific, Waltham, MA, USA). Then, formaldehyde agarose gel (1%) electrophoresis was used to assess the integrity of the RNA samples. From each tissue sample, 1 µg of total RNA was reverse transcribed into complementary DNA (cDNA) by RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, K1622), following the manufacturer's instructions.

Oligonucleotide Design and Molecular Cloning of DQA
Three overlapping PCR primers were designed according to the DQA genes available in the GenBank (XM_004782395.2, XM_022510513.1) for Neovison vison DQA gene amplification (Table 2). These primers were used in RT-PCR for the amplification of DQA-cDNA fragment. Genomic DNA PCR amplification was performed using ExTaq Kit (TaKaRa, RR01AM). The reaction volume of 25 µL comprises of the proviral genomic DNA template, 0.125 µL of Ex Taq, 2.5 µL of Ex Taq Buffer, 2 µL of dNTP mixture, 0.5 µL of both forward and reverse primers (20 µmol/L), and 1 µL cDNA template. The PCR reaction included an initial denaturation cycle of 5 min at 94 • C, followed by 35 cycles of denaturation for 30 s at 94 • C, annealing for 30 s (the optimum annealing temperatures are listed in Table 1), and an extension for 1 min at 72 • C, with a final extension of 10 min at 72 • C. The PCR products were excised from a 1.0% agarose gel, purified using GeneJET Gel Extraction Kit (Thermo Fisher Scientific, K0692), ligated into the pMD19-T vector (TaKaRa, 6013), and transformed into Escherichia coli DH5α competent cells. Then, the plasmid DNA was extracted from the transformed cells using a TIANprep Mini Plasmid Kit (Tiangen, Beijing, China). Three independent clones were sequenced by the Beijing Genomics Institute (Beijing, China).

Multiple Sequence Alignment and Phylogenetic Analysis
Multiple sequence alignment was carried out using the sequence analysis software Lasergene 1 (DNASTAR Inc., Madison, WI, USA) and Clustal X program, version 1.83. A neighbor-joining tree was drawn with MEGA5.0 software (http://www.megasoftware.net), with confidence levels assessed using 1000 bootstrap replications. The sequences of reference DQA genes concluded in the multiple sequence alignment were obtained from GenBank and summarized in Table 3.

Conclusions
The DQA gene of Neovison vison was amplified for the first time. The results of this study lay a foundation for further study on the antiviral molecular mechanism of DQA.