IFN-β is a unique member of type I IFN in humans and contains four positive regulatory domains (PRDs), I-II-III-IV, in its promoter, which are docking sites for transcription factors IFN regulatory factor (IRF) 3/7, NF-κB, IRF3/7, and activating transcription factor 2/Jun proto-oncogene, respectively. In chicken IFN-β and zebrafish IFNφ1 promoters, a conserved PRD or PRD-like sequences have been reported. In this study, a type I IFN gene, named as xl-IFN1 in the amphibian model Xenopus laevis, was found to contain similar PRD-like sites, IV-III/I-II, in its promoter, and these PRD-like sites were proved to be functionally responsive to activating transcription factor 2/Jun proto-oncogene, IRF3/IRF7, and p65, respectively. The xl-IFN1, as IFNφ1 in zebrafish, was transcribed into a long and a short transcript, with the long transcript containing all of the transcriptional elements, including PRD-like sites and TATA box in its proximal promoter. A retroposition model was then proposed to explain the transcriptional conservation of IFNφ1, xl-IFN1, and IFN-β in chicken and humans.
In mammals, the production of type I IFNs is initiated from the recognition of pathogen-associated molecular patterns by pattern-recognition receptors, including TLRs, nucleotide-binding oligomerization domain-like receptors, RIG-I–like receptors, and cytosolic DNA sensors (1, 2), followed by the interaction with critical adaptors, such as Toll/IL-1R domain-containing adapter inducing IFN-β (also known as TRIF), mitochondrial antiviral signaling protein and mediator of IFN regulatory factor (IRF) 3 activation/stimulator of IFN genes protein (MITA/STING), and then the recruitment of kinases (e.g., TANK binding kinase 1), the activation of transcription factors, such as the AP-1 family, IFN regulatory factors (IRFs), and NF-κB (3–5). IFN-β, as a unique member of the type I IFN family, plays an important role in antiviral immunity (6–8). The promoter of IFN-β contains four positive regulatory domains (PRDs) I–IV (GAGAAGTGAAAG, GGGAAATCC, GAAAACTGAAAG, and TGACATAG), which provide the binding sites for IRF3/7, NF-κB, IRF3/7, and the heterodimer of two AP-1 family members, activating transcription factor 2/Jun proto-oncogene and AP-1 transcription factor subunit (ATF2/JUN), respectively (9). Importantly, the arrangement and orientation of IFN-β PRDs (IV-III-I-II) may influence the assembly and function of IFN-β enhanceosome (9, 10). Furthermore, the PRDs and transcriptional regulation of chicken IFN-β are conserved as human IFN-β, although IRF7 is employed to reconstitute IFN signaling due to the loss of IRF3 in chicken, suggesting the conserved regulation of amniote IFN-β gene in response to various immune signaling pathways (11, 12).
To date, it has been revealed that teleost fish possess conservatively a variety of pattern-recognition receptors and signaling molecules involved in the production of type I IFNs as reported in mammals, such as TLRs, nucleotide-binding oligomerization domain-like receptors, retinoic acid-inducible gene I–like receptors, cytosolic DNA sensors, TIR domain-containing adapter molecule 1, mitochondrial antiviral signaling protein, mediator of IRF3 activation/stimulator of IFN genes protein/STING, TANK binding kinase 1, AP-1, NF-κB, and IRFs (13–20). As an important model species, zebrafish (Danio rerio) is found to contain four copies of type I IFNs, named as IFNφ1–4, and only IFNφ1 has similar transcriptional elements as human IFN-β (18, 21). In fact, conserved PRD-like sequences, IV-I/III-II, are found in zebrafish IFNφ1 promoter, which were reported to be associated with JUN (AP-1 family), IRF3/7, and p65 (NF-κB family), respectively (17, 18).
However, zebrafish IFNφ1 shows low amino acid identity, and differs obviously in gene organization, to human IFN-β (21). Zebrafish IFNφ1 gene has five coding exons, four introns, and two transcriptional start sites (TSSs), whereas human IFN-β is an intronless gene with one TSS (22, 23). In fact, type I IFN loci, the IFN–PLEKHM1/ARHGAP27 gene cluster, from teleost fish, including zebrafish IFNφ1, have conserved colinearity with the multiexon type I IFN locus in amphibian (e.g., Xenopus), but not with IFN-β gene located on the intronless type I IFN locus (PTPLAD2–IFN–MTAP gene cluster) in amniotes (22, 24–26). Simultaneously, amniote multiexon type I IFN genes were found to be lost in the syntenic region (RDM1–PLEKHM1/ARHGAP27 locus) (24, 26). A hypothesis to explain how introns were lost from type I IFNs in amniotes is that intronless type I IFNs might have originated from intron-containing type I IFN transcripts via evolutionary retroposition event in a transition period when vertebrates migrated from an aquatic environment to land (27). Indeed, the retrotransposed intronless type I IFN genes were replicated and clustered to form a novel locus in amniotes (24, 26). However, an interesting question arises as to how the conserved transcriptional mechanism is maintained in amniote IFN-β gene in response to immune signaling molecules as mentioned above. Amphibians as an important vertebrate lineage have diverged from the amniotes some 360 million years ago (28), and the analysis of proximal promoters of multiexon type I IFNs in the amphibian model species, the African clawed frog Xenopus laevis may provide evolutionary clues for answering the question. In this study, the promoter of a type I IFN gene, xl-IFN1 gene in X. laevis, was found to be similar to human IFN-β gene with conserved PRD-like sites (IV-III/I-II) in consideration of sequences, arrangement, and orientation. Furthermore, these PRD-like sites in xl-IFN1 promoter were proved to be associated with the functional transcription factors ATF2/JUN, IRF3/IRF7, and p65, respectively. Additionally, like zebrafish IFNφ1, xl-IFN1 gene has two TSSs to generate two transcripts, named as long and short transcripts, respectively. The cDNA sequence of the long transcript contains all transcriptional elements, including PRD-like sites (IV-III/I-II) and TATA box, in the proximal promoter of xl-IFN1. Therefore, we propose, at least hypothetically, a retroposition model for explaining the transcriptional conservation of IFNφ1, xl-IFN1, and IFN-β in chicken and humans that retroposition of the long transcript might have resulted in the origin of human intronless IFN-β transcriptional elements in response to various immune signaling.
Materials and Methods
Cells and virus
The X. laevis A6 cells (CCL-102) were purchased from American Type Culture Collection and maintained in 75% NCTC-109 medium (Life Technologies) supplemented with 15% distilled water and 10% FBS (Life Technologies) at 26°C in a CO2 (5%) incubator (Thermo Fisher Scientific). Epithelioma papulosum cyprini (GDC174) cells from the China Center for Type Culture Collection were cultured in medium 199 (Life Technologies) with 10% FBS and 5% CO2 at 26°C. FV3 (VR-567) was purchased from American Type Culture Collection and propagated in Epithelioma papulosum cyprini cells as previously described (25).
RNA extraction, gene cloning, and plasmid construction
Total RNA was extracted from A6 cells by using TRIzol reagent (Ambion) according to the standard protocol. RNA quality control was performed through agarose electrophoresis and absorbance measurement. The full-length 5′- and 3′-end sequences were amplified by RNA ligase-mediated RACE PCR using the GeneRacer Kit with SuperScript III RT kit (Invitrogen), following the manufacturer’s instruction. The full-length open reading frames (ORFs) and promoter fragments of target genes were amplified with the specific primers and inserted into an p3XFLAG-CMV-14 expression vector (Sigma-Aldrich) at KpnI/BamHI sites and pGL3-basic vector (Promega) at XhoI/KpnI sites by using ClonExpress II One Step Cloning Kit (Vazyme Biotech), respectively.
The translation of nucleotides to protein sequences was analyzed using the online tool (Translate, https://web.expasy.org/translate/). The protein homology was detected by BLASTP program (https://blast.ncbi.nlm.nih.gov/). Putative conserved domains were analyzed based on the feedback searching from the Conserved Domain Database (https://www.ncbi.nlm.nih.gov/cdd/). Multiple sequence alignments were performed by using the Clustal X program. Phylogenetic analyses were carried out according to the construction of neighbor-joining tree with 1000 time repeat of bootstrap in MEGA4 package.
Northern blotting assay was performed using the DIG Northern Starter Kit (Roche) following the manufacturer’s instruction with modification. Briefly, DIG-labeled probe was generated with the primers (Supplemental Table I) by using a PCR DIG Probe Synthesis Kit (Roche). The A6 cells (∼4 × 106 cells/well) were seeded into six-well plates and then treated with synthetic polyinosinic-polycytidylic acid [poly(I:C)]; Sigma-Aldrich) at a final concentration of 25 μg/ml for 3 or 6 h. The total RNA was collected and separated by formaldehyde denaturing 1% agarose gel electrophoresis at 4°C overnight. After gel rinse, the RNA was transferred to positively charged nylon membranes (0.45 μm, HyBond N+; Amersham) by capillary blotting method. Next, the membranes were washed and baked at 80°C for 2 h to fix the RNA. After prehybridization, the membranes were hybridized in DIG Easy Hyb solution (Roche) containing denatured DIG-labeled probe at 50°C overnight with gentle agitation. Subsequently, 2× SSC buffer (0.3 M NaCl and 30 mM sodium citrate [pH 6.9–7.1]) and 0.1× SSC buffer (0.015 M NaCl and 1.5 mM sodium citrate [pH 6.9–7.1]) with 0.1% SDS was used to wash the membranes at room temperature and 68°C, respectively. The membranes were then blocked and incubated with anti–digoxigenin-AP Ab (Roche) for immunological detection using the Gel Doc XR System (Bio-Rad Laboratories).
Luciferase activity assay
Measurements of luciferase activity were conducted as described in a previous report (29). Briefly, A6 cells (∼1 × 105 cells/well) were seeded into 24-well plates and cotransfected separately with ATF2/JUN, IRF7, and p65 expression plasmids (including wild-type [wt] and mutants [mut]), pRL-TK plasmid (Promega), and luciferase luciferase-reporter vectors of IFN promoter (including wt and mut) using FuGENE HD Transfection Reagent (Promega). At 24 h posttransfection, cells were collected and detected by using Dual-Luciferase Reporter Assay System (Promega) and GloMax-Multi Jr Detection System (Promega) following the standard protocol.
The method for prokaryotic recombinant protein production was described previously (30, 31). Briefly, the synthetic coding sequences from IRF7 (N-terminal region, Met1-Glu143), p65 (N-terminal region, Met1-Leu194), ATF2 (C-terminal region, Gln317-Ser486), and JUN (C-terminal region, Met230-Phe314) were inserted into the pET-32a (+) expression vector, and the constructed plasmids were transformed into Escherichia coli BL21 (DE3) strain to induce recombinant protein expression by isopropyl β-d-thiogalactoside. After filtration, the soluble fusion proteins were harvested for purification by using Ni-NTA agarose (Millipore) affinity chromatography, and the protein concentration was determined by Bradford’s method. EMSA was performed as reported (32, 33), with modifications. Briefly, wt and mut oligonucleotides (Supplemental Table I) were labeled by using the EMSA Probe Biotin Labeling Kit (Beyotime) following the manufacturer’s instructions. Binding reactions of probe and recombinant protein were performed by using the EMSA/Gel-Shift Kit (Beyotime) according to the standard protocols. Probe/protein complex and free probes were separated by nondenaturing 6% PAGE and were then transferred to the nylon membrane. Next, the probes were detected by autoradiography using the LightShift Chemiluminescent EMSA Kit (Pierce) and ChemiDoc MP imaging system (Bio-Rad Laboratories), following the user guide.
The cDNA synthesis and quantitative real-time PCR
Genomic DNA remnants in total RNA were removed by digestion with DNase I (Thermo Fisher Scientific) at 37°C. The first-strand cDNA was synthesized by using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific) following the manufacturer’s instructions. Quantitative real-time PCR was performed by using iQ SYBR Green Supermix (Bio-Rad Laboratories) in the CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories), and the standard curves were established as described previously (29). The reaction protocol was as follows: one cycle of 95°C for 3 min, followed by 45 cycles of 95°C for 10 s, 60°C for 15 s, and 72°C for 15 s. Gene expression was normalized against β-actin, and data analyses were carried out using the 2−ΔΔCT method (34).
The A6 cells (∼1 × 105/well) were seeded into 24-well plates and transfected with ATF2/JUN, IRF7, p65-expressing plasmids, and empty vector (as control) using FuGENE HD Transfection Reagent (Promega) for 24 h, respectively. The transfected cells were then infected with FV3 for 72 h at 0.3 multiplicity of infection. Viral titers were detected by the observation of cytopathic effect method as previously reported (25).
Data were analyzed statistically with the Student t test in SPSS 16.0 software. Significant difference was indicated by *p < 0.05 and **p < 0.01, respectively. The data are presented as mean ± SE.
Discovery of a short and a long xl-IFN1 transcript
The proximal promoters from the described multiexon type I IFNs in X. laevis were analyzed, and some IFN-stimulated response elements (ISREs) or NF-κB sites were found in the proximal promoters of these African clawed frog (xl)-IFNs, but only xl-IFN1 has all PRD-like sites, including PRD-like IV (5′-TGAGATCT-3′), PRD-like III/I (5′-GGAAAGTGAAAATGAAAC-3′), and PRD-like II (5′-GGAAAATCC-3′) (Supplemental Fig. 1). Multiple sequence alignments showed that the xl-IFN1 promoter elements, including PRD-like (IV-III/I-II) and TATA box, in consideration of sequences, arrangement and orientation are conserved when compared with zebrafish IFNφ1, chicken IFN-β, and human IFN-β (Fig. 1A).
Moreover, RACE PCR was performed to investigate whether xl-IFN1 contains two TSSs, like zebrafish IFNφ1. As shown in (Fig. 1B and 1C, xl-IFN1 has two TSSs, the first and the second TSS, which generate two transcripts, designed as long and short transcripts, respectively, without introns based on the PCR verification. The relative short 5′-untranslated region (UTR) without PRD-like (IV-III/I-II) and TATA box was observed in the cDNA sequences from chicken IFN-β, human IFN-β transcript, transcript 2 of zebrafish IFNφ1, and short transcript of xl-IFN1 (Fig. 1B). In fact, the 5′-UTR lengths of these transcripts are ∼20–75 bp, being 42, 20, 70, and 75 bp for IFNφ1, xl-IFN1, chicken IFN-β, and human IFN-β, respectively. But, the cDNA sequence of xl-IFN1 long transcript possesses longer 5′-UTR (818 bp), which contains the second TSS and the proximal promoter, including PRD-like (IV-III/I-II), TATA box and flanking sequences, of the short transcript (Fig. 1B). The zebrafish IFNφ1 long transcript has 162 bp in the 5′-UTR. The existence of xl-IFN1 two transcripts was also proved by Northern blotting using the probe of xl-IFN1 full-length ORF (Fig. 2A), and the two transcripts were induced by poly(I:C) (Fig. 2). It was suggested that the two TSSs from xl-IFN1 promote two transcripts, and PRD-like (IV-III/I-II) sites are included in the cDNA sequences of long transcript but not the short transcript.
Identification of ATF2, JUN, IRF7, and p65 in X. laevis
The full-length ORF of X. laevis ATF2 (1461 nt), JUN (945 nt), IRF7 (1827 nt), and p65 (1584 nt) genes were amplified form A6 cells using specific PCR primers and deposited into the GenBank database under accession numbers MZ603622 (https://www.ncbi.nlm.nih.gov/nuccore/MZ603622), MZ603624 (https://www.ncbi.nlm.nih.gov/nuccore/MZ603624), MZ603623 (https://www.ncbi.nlm.nih.gov/nuccore/MZ603623), and MZ603625 (https://www.ncbi.nlm.nih.gov/nuccore/MZ603625), respectively. The ATF2, JUN, IRF7, and p65 cDNA sequences were predicted to encode 486, 314, 608, and 527 aa (Fig. 3A–D). According to the domain prediction from the Conserved Domain Database, it was found that both ATF2 and JUN have conserved basic leucine zipper domain and are identified as members of the AP-1 family (Fig. 3A, 3B). IRF7 contains a DNA-binding domain (DBD) and an IRF association domain (Fig. 3C), and p65 possesses an N-terminal subdomain of the Rel homology domain (RHD-n) and an Ig-like fold, plexins, transcription factor (IPT) domain (Fig. 3D). Thus, it was believed that IRF7 and p65 belong to IRF and NF-κB family, respectively. Phylogenetic tree showed that X. laevis ATF2, JUN, IRF7, and p65 were clustered with their corresponding homologous genes from other vertebrates (Fig. 3E), indicating the conservation of X. laevis ATF2, JUN, IRF7, and p65 sequences with their counterparts in other vertebrates.
Antiviral function of X. laevis ATF2, JUN, IRF7, and p65
To understand whether X. laevis ATF2, JUN, IRF7, and p65 are functional genes, the ORFs of these genes were cloned into expressional vectors. Because the heterodimer of ATF2/JUN, IRF7 homodimer, and p65 homodimer have positive regulation activity, ATF2 and JUN were cotransfected, and IRF7 and p65 were separately transfected, into A6 cells to detect the antiviral function. It was shown that ATF2/JUN, IRF7, and p65 were able to induce IFN-stimulated genes (ISGs), including Mx and viperin, and to decrease significantly the FV3 titers in A6 cells (Fig. 4). These results indicated that X. laevis ATF2, JUN, IRF7, and p65 are functional genes.
Functional PRD-like sites in xl-IFN1 promoter
To verify that PRD-like sequences in xl-IFN1 promoter are functional sites, luciferase reporter assay was performed to detect the activation of xl-IFN1 proximal promoter with wt or mut PRD-like sites induced by ATF2/JUN, IRF7, and p65 in A6 cells. Because xl-IFN1 possesses two TSSs, the proximal promoters from upstream of the first and second TSSs were designated as xl-IFN1 promoter-1 and -2, respectively (Fig. 5A). As shown in (Fig. 5B–E, the wt xl-IFN1 promoter-2 was significantly induced by IRF7-wt, but not the DBD-deletion IRF7. The IRF7-induced promoter activity increased further to a higher level when stimulated with poly(I:C) (Fig. 5F). The luciferase activity of xl-IFN1 promoter-2 reporter plasmid enhanced by IRF7 was significantly reduced by the mutation of PRD-like III/I site, which had a negative effect on the binding activity of xl-IFN1 promoter-2 and recombinant IRF7 DBD (Fig. 5C, 5G). Similarly, xl-IFN1 promoter-1 reporter plasmids were significantly induced by IRF7-wt through the ISRE and further activated in response to poly(I:C) (Fig. 5H–K). In addition, IRF7 knockdown in A6 cells decreased significantly the expression of xl-IFN1 as well as other multiexon type I IFN genes (xl-IFN2, IFN3, IFN4, IFN5, IFN6, and IFN7), which were strongly induced by poly(I:C) (Supplemental Fig. 2). Moreover, the conserved IRF3 gene (GenBank accession number MZ603621; https://www.ncbi.nlm.nih.gov/nuccore/MZ603621) was cloned from A6 cells (Supplemental Fig. 3A, 3B), a constitutively active form of IRF3 (IRF3-5D) promoted significantly the luciferase activity of xl-IFN1 promoter-1 and -2 reporter plasmids through ISRE and the PRD-like III/I sites, respectively (Supplemental Fig. 3C–E).
It was also observed that the luciferase activity of xl-IFN1 promoter-2 reporter vector was significantly induced by p65 and further increased by the stimulation of poly(I:C); however, the mutant with the deletion of p65 RHD-n and IPT domains had no such effect (Fig. 6A–E). Mutant of PRD-like II site led to the reduction of xl-IFN1 promoter-2 activity induced by p65 and of binding activity to recombinant p65 RHD-n and IPT domains (Fig. 6). Furthermore, the wt ATF2/JUN but not ATF2/JUN mutant (without basic leucine zipper domain) significantly activated the xl-IFN1 promoter-2 (wt), and PRD-like IV mutant significantly decreased the ATF2/JUN induced xl-IFN1 promoter activity. Recombinant ATF2/JUN showed binding ability to xl-IFN1 promoter-2 based on EMSA (Fig. 7). These results suggested that all PRD-like IV, III/I, and II in xl-IFN1 promoter-2 are functional transcription sites and associated with the induction of xl-IFN1 (Fig. 8).
In this study, a conserved presence of PRD-like sites, IV-III/I-II, was found in the proximal promoter of xl-IFN1 in the amphibian model species X. laevis, as in its homologous type I IFNs (i.e., IFNφ1 in zebrafish and IFN-β in chicken and humans) (9–12, 17, 18). Promoter activity analyses showed that these transcriptional elements, PRD-like III/I, II, and IV, were significantly responsive to their specific corresponding transcription factors, revealing an evolutionarily conserved transcriptional mechanism at least in the expression of this unique type I IFN in vertebrates.
IFN-β gene in humans is a critical antiviral type I IFN and is regulated by a variety of immunological signaling pathways, depending mainly on transcriptional sites, including PRD IV-III-I-II, in its proximal promoter (3, 4, 9). Interestingly, the pattern of transcriptional regulation of zebrafish IFNφ1 is similar to the IFN-β gene because of conserved transcriptional elements, PRD-like sites, IV-III/I-II, in the proximal promoter of IFNφ1, which respond to a set of conserved immune signals (17, 18). However, the gene organization of multiexon type I IFN in fish and amphibians is distinct from the intronless type I IFN genes in mammals, and the retroposition event has been proposed to explain the loss of introns from multiexon type I IFN (24, 27). In fact, the duplicated intronless type I IFN genes are clustered to form a new locus, with the disappearance of evolutionarily primitive multiexon type I IFN locus in amniotes (24, 26). With the presence of multiexon and intronless type I IFNs, amphibians are considered valuable for finding further evidence to explain the evolution from multiexon type I IFNs in fish primitive vertebrates to intronless type I IFNs in higher vertebrates. The finding of PRD-like sites (IV-III/I-II) in xl-IFN1 provides the possible evolutionary congruence of type I IFN or at least the IFN-β gene in vertebrate lineages in respect to its transcriptive signaling and its evolutionary selective pressure from pathogens.
Although zebrafish IFNφ1 has two TSSs, the 5′-UTR of short transcript promoted by the second TSS contains relatively short sequences without PRD-like sites and TATA box, which are similar to the short 5′-UTRs of chicken and human IFN-β from the single TSS. Due to the lack of introns, the IFN transcript with short 5′-UTR may become possibly retroposed to new locus on chromosome, but promoter elements of retroposed IFN might have re-evolved under the evolutionary pressure from conserved pathogenic stimuli, such as pathogen-associated molecular patterns. However, the arrangement and orientation of PRDs (IV-III/I-II) are necessary for the assembly and function of IFN-β enhanceosome (9, 10). Thus, it may be impossible that the entire composition of all conserved elements can be maintained in the proximal promoter of retroposed short transcript IFN.
The xl-IFN1 gene possesses two transcripts, and the regulation of short transcript is conserved from zebrafish to human due to the presence of functional PRD-like sites (IV-III/I-II), which with TATA box and the flanking sequences are contained in the cDNA sequence of long transcript. Therefore, it is hypothesized that retroposition of the long transcript from IFN-β homologs in ancestry vertebrates might have given rise to the intronless IFN-β transcription elements in higher vertebrates (Fig. 8). To a large extent, retroposition of the long transcript may have resulted in the consistency between the proximal promoter of intronless IFN-β gene and of short transcript from multiexon ancestry IFN-β homologs. This may at least partially account for the evolutionary occurrence of intronless IFN-β and its transcriptional mechanism in mammals, but may also shed light on understanding the evolution of multiexon type I IFNs to intronless type I IFNs in vertebrates in response conservatively to pathogens.
Type I IFNs in mammals and in lower vertebrates are currently classified into different subtypes or subgroups, which are phylogenetically clustered in different clades (15, 25). In mammals, 11 subtypes, including IFN-α, IFN-β, IFN-δ, IFN-ε, IFN-ζ, IFN-κ, IFN-μ, IFN-ν, IFN-τ, IFN-ω, and IFN-αω, have been reported (6, 35, 36). But in fish, seven subgroups of type I IFNs are reported according to current nomenclature, including a, d, e, and h in group I and b, c, and f in group II (15, 37, 38). However, these subtypes or subgroups are not all existent in a single species of mammal, nor in a species of fish. In mammals, some type I IFN subtypes have a species- or lineage-specific distribution, such as IFN-δ, IFN-μ, IFN-ν, IFN-τ, IFN-ζ, and IFN-αω (6, 35, 36), and they may also show different expression patterns (4, 39–41). It appears possible that most fish species have only three subgroups of type I IFNs (i.e., a, c, and d) (22, 26, 42–44), and IFN-h is only reported in perciform fish so far (45–47). In amphibian lineage, independent bifurcation of type I IFNs has been observed, which may have resulted in the occurrence of intronless and intron-containing type I IFNs (25). It is thus considered that type I IFN may have undergone independent evolution in different lineages of vertebrates from fish to mammals (25, 37, 38, 47). With the increasing availability of vertebrate genome sequences and analyzing practicability of transcriptional elements, the members of type I IFNs and their transcription mechanisms can then be understood from an evolutionary point of view to reveal the compositional and functional conservation and variation of type I IFNs in vertebrate lineages.
We thank the laboratory members for their views and discussion in the research.
This work was supported by the National Natural Science Foundation of China (Grant 31320103913) and China Agriculture Research System (CARS-46). P.N. received funding from a special top talent plan “One Thing One Decision (Yishi Yiyi)” [(2018)27] and “First Class Fishery Discipline” [(2018)8] program, both in Shandong Province, China.
The online version of this article contains supplemental material.
The sequences presented in this article have been submitted to GenBank under accession numbers MZ603622, MZ603624, MZ603623, and MZ603625.
Abbreviations used in this article
activating transcription factor 2/Jun proto-oncogene
Ig-like fold, plexins, transcription factor
IFN regulatory factor
IFN-stimulated response element
open reading frame
positive regulatory domain
N-terminal subdomain of the Rel homology domain
transcriptional start site
The authors have no financial conflicts of interest.