Activity of the porcine type I interferon family

TYPE I IFNS are a family of cytokines prominent in antiviral responses. In contrast to a single type II IFN, i.e., IFN-γ, type I IFNs consist of multiple subclasses including IFN-α, IFN-β, IFN-ε, IFN-ω, and IFN-κ (16, 30). Humans have multiple IFN-α subtypes and single-subtype subclasses of IFN-β, IFN-ε, IFN-ω, and IFN-κ (42). Type I IFNs also include other subclasses with a more limited species expression, including IFN-αω (6), IFN-δ (6, 20), IFN-τ, and IFN-ζ (limitin) (27). Ubiquitously expressed IFN-α and IFN-β are among the most studied subclasses for antiviral responses. Although less extensively studied, tissue- and cell-specific expressed subclasses, such as IFN-ω in various leukocytes, IFN-δ and IFN-ε in female reproductive tissues (6, 7, 20), and IFN-κ in epidermal keratinocytes, are potently induced by viral infection and confer an antiviral state on uninfected cells (30, 42). In pigs type I IFNs are known to consist of several multiple-subtype subclasses (IFN-α, IFN-ω, and IFN-δ), and they also possess a single gene locus encoding IFN-β (3, 4, 6, 18, 25). Although quite a few porcine type I IFN genes have been identified, information about the complete repertoire, gene structure, expression profiles, and comparative antiviral activity of most porcine type I IFNs is not available.

Type I IFNs, particularly IFN-α and IFN-β, are central to antiviral innate immunity. Through a feedback loop of production and action, type I IFNs not only activate antiviral responses by autocrine means but also function systemically to induce an antiviral state in surrounding and distal cells. Induction of the antiviral state, which includes suppression of cellular metabolic processes and profound expression of antiviral IFN-stimulated genes (ISGs) (10, 45), is critical for developing effective immune protection against viral infections. Hundreds of ISGs have been identified with various functions such as direct virus targeting, amplification of antiviral resistance, and sequestration of cellular metabolic processes to repress virus replication (35, 40). The efficacy of induction of antiviral responses appears different between subclasses and even between subtypes belonging to the same subclass. For example, human IFN-α subtypes vary in their ability to activate human natural killer cells, and IFN-β shows more potency than IFN-α2 in inhibition of monocyte proliferation (8, 41, 43). Therefore, to elucidate the antiviral potency of a type I IFN family, it is essential to systematically analyze their antiviral activity among members of the same and different subclasses.

Here we report, based on the current understanding of the porcine genome, a near-complete analysis of the porcine type I IFN family consisting of 39 functional genes, classified into 7 subclasses distributed along draft genomic sequences of chromosomes 1 and 10. Several IFNs have been identified for the first time in this study, and others, which have appeared in the sequence literature but have not been functionally characterized, were evaluated for gene expression and antiviral activity. We show that members of different or the same porcine type I IFN subclasses have diverse expression profiles and antiviral activities against porcine reproductive and respiratory syndrome virus (PRRSV) and vesicular stomatitis virus (VSV) and that some newly identified IFNs have potent anti-PRRSV activity. 

All virus procedures were approved by the Kansas State University (KSU) Biosafety committee. The PRRSV isolate was a cell culture-adapted strain, SDSU-23983-P140 (13, 34). VSV (Indiana 1 serotype) was a natural isolate standardized for VSV diagnostic evaluations in the Kansas State Veterinary Diagnostic Laboratory (26). MARC-145 cells, an African green monkey kidney cell line sensitive to PRRSV infection, PK-15 cells, a porcine circovirus-1-free subclone of the porcine kidney PK-15 cell line (American Type Culture Collection, Manassas, VA), and porcine primary alveolar macrophages (AMs), isolated and cultured as previously described (38), were used as target cells to test IFN antiviral activity. Procedures for cell culture, viral infection, and virus detection were conducted as previously described (37–39).

With the web-based algorithms tblastn, protein blast(http://blast.ncbi.nlm.nih.gov/Blast.cgi), and Genomview(http://www.ncbi.nlm.nih.gov/Tour/3a.html), putative porcine IFN sequences were extracted from working draft sequences deposited by the Swine Genome Project (GenBank accession nos. AC127471, AC130792, AC135219, AC138785,CU463237, CU469469, CU633414) and later refined with reference genome assemblies of Sscrofa5 (http://www.ncbi.nlm.nih.gov/projects/genome/guide/pig/index.html) and Sscrofa9 (http://www.sanger.ac.uk/Projects/S_scrofa/). Further confirmation of coding regions, sequence alignments, and phylogenic analysis was conducted by combined use of bioinformatic tools in EBI (http://www.ebi.ac.uk/services/) and ExPASy (http://us.expasy.org/) as previously described (29, 36).

Porcine cDNA was reverse transcribed from a pool of total RNA extracted from intestine, skin, and mesenteric lymph nodes with a SuperScript III First-Strand Synthesis System and random primers (Invitrogen, Carlsbad, CA) (37). Coding regions of porcine IFNs were amplified from this cDNA pool for transcription confirmation and isolation of open reading frames (ORFs). Whole ORFs of some IFNAs and IFNK were isolated from porcine genomic DNA (Novagen, Madison, WI). A high-fidelity PCR AccuPrime Taq system (Invitrogen) with gene-specific or subtype-common cloning primers (Supplemental Table S1) was used.¹ PCR fragments of ORFs were purified and cloned into a pcDNA 3.3-TOPO TA cloning/expression vector (Invitrogen). An Applied Biosystems 3730 DNA Analyzer was used for sequencing IFN genes cloned in a pcDNA 3.3-TOPO TA vector with the vector sequencing primers (Invitrogen). At least five clones for each IFN gene were selected by gene-specific PCR and sequenced from both forward and reverse directions. Two or three clones with the highest sequence identity (98–100%) to the template sequence extracted from working draft sequences deposited in the Swine Genome Project were selected for protein overexpression and antiviral assays. Plasmids for cell transfection were purified from individual bacteria cultures with an Endo-free plasmid purification kit (Sciresys, Cherry Hill, NJ).

Gene-specific or subtype-common primers were designed based on multiple alignments of related IFN sequences (Supplemental Table S1). Some primers and PCR conditions for IFN-α subtypes were derived from previous publications with modifications (4, 5). Primers and PCR conditions were optimized and validated by using confirmed IFN plasmids to show specific amplification only with templates containing confirmed IFN clone(s). Tissues were obtained from healthy 5-wk-old crossbred pigs as previously described (36). All collection procedures were approved by the Kansas State University Institutional Animal Care and Use Committee. Real-time RT-PCR arrays in a 96-well microplate format (iCycler5.0, Bio-Rad, Hercules, CA) were performed with the validated primers. Reactions were conducted with a SYBR Green RT-PCR system (Qiagen, Valencia, CA) with 150 ng of total RNA in a 25-μl reaction mixture and RT-PCR conditions as described previously (36). Specific optic detection was set at 78°C for 15 s after each amplification cycle of 95°C for 15 s, 56–59°C for 30 s. and 72°C for 40 s. Critical threshold (C_t) values and melt curves were monitored and collected with iCycler 5.0 software, and final products after 40 PCR cycles were analyzed on agarose gels. Relative gene expression data in different tissues were normalized against C_tvalues of the housekeeping gene (GAPDH), and the relative expression index (2^−ΔΔC_t) was determined compared with the average expression levels of control samples, with the index defined as 1.0 (36).

MARC-145 cells were added to 96-well culture plates at 1.0 × 10⁴ cells/well; 24 h later at ∼80% confluence, cells were transfected by the addition of 10 μl of a mixture containing 0.1 μg of plasmid DNA plus 0.3 μl of FuGENE transfection reagent (Roche, Indianapolis, IN) in OptiMEM medium (Invitrogen). Twenty-four hours after transfection, cells were infected with virus as described previously (38). Protection afforded by IFN expression constructs was evaluated at 48–72 h with crystal violet staining (PRRSV) or detection of immunofluorescence in AMs with an antibody (SDOW17, Rural Technologies, Brookings, SD) against PRRSV nucleocapsid protein (38). To collect IFN-containing medium, individual IFN expression constructs were used to transfect HEK293F cells (Invitrogen) seeded in 12-well plates at 2.5 × 10⁵ cells/well in serum-free medium. The transfection procedure was the same as above but with a proportionally (15×) increased transfection mixture for each well. IFN-containing supernatants and media from mock-transfected cells were collected separately at 72 h after transfection. Expressed IFNs were the major protein band in supernatants with protein concentrations of 5–10 μg/ml (micro BCA protein assay kit, Pierce, Rockford, IL). IFN peptides were partially purified (Centricon centrifugal filters, 10k and 50k NMWL, Millipore, Billerica, MA), subjected to gel electrophoresis (NuPAGE Bis-Tris gel system, Invitrogen), and stained (Coomassie blue G-250 solution, Bio-Rad). Before use, protein concentrations in supernatants were adjusted to 2 μg/ml (the lowest concentrations obtained in some subtypes) with serum-free medium. Supernatants were filtered (0.2 μm) and stored at −135°C until use. In addition, total RNA was extracted from transfected HEK293F cells to examine the intensity and uniformity of the expression of transfected IFN genes by real-time RT-PCR as described above.

Antiviral activity of IFN-containing supernatants was assayed as the inhibition of cytopathic effect (CPE) of PRRSV and VSV on MARC-145 and PK-15 cells, respectively, as previously described (4). In brief, cells were cultured in flat-bottom 96-well plates to 95% confluence, and 10 μl of IFN-containing medium was added to each well with predispensed 90 μl of fresh medium containing virus. Virus concentrations for infections were at a multiplicity of infection (MOI) of 0.1 and 0.2 50% tissue culture infectious dose (TCID₅₀)/ml for PRRSV and VSV, respectively. Controls included medium collected from mock-transfected cells (MEM); medium collected from cells transfected with an antisense IFNA1 cloned in pcDNA3.3 vector; incubation of 50 μl of IFN-α1- or IFN-β1-containing medium with 10 μg of antibodies against porcine IFN-α (K9 clone, R&D Systems, Minneapolis, MN) or human IFN-β (R&D Systems), respectively (anti-α and anti-β); and a standard curve (0–250 U/100 μl) of recombinant porcine IFN-α (3.8 × 10⁷ U/mg, catalog no. 17100-1, R & D Systems). At 48, 60, or 72 h after infection, protective activity of IFNs on virus infections was determined with crystal violet staining (PRRSV- and VSV-MARC-145 and VSV-PK-15) or immunofluorescence staining of PRRSV (PRRSV-AM) (38). Images were collected with an AlphaImager (Alpha Innotech, San Leandro, CA) for crystal violet-stained plates. Quantitative data were obtained by measuring absorbance at 570 nm for dye extracts from crystal violet-stained wells (4) with an ELISA microplate reader or calculating percentages of virus-positive cells for immunofluorescence-labeled viruses. Percent protection of an IFN was calculated with the formula 100 × (V_t − V_i)/(V_t − V₀), where V_t represents the value of total/highest occurrence of a viral infection in untreated cells, V_i is the value obtained from IFN-treated cells, and V₀ is the value from cells without addition of virus. All V_t or V₀ values were averages of data from 8 or 16 replicate wells transformed with one or two IFN polymorphic clones. To titrate antiviral activity, 10-fold serial dilutions of IFN-containing medium were assayed with the PRRSV-MARC-145, VSV-MARC-145, and VSV-PK-15 virus-target cell systems. Antiviral activity was calculated by the Reed-Muench method (31), with 1 unit as the highest dilution that reduced cell number by 50%.

The antiproliferative assay was performed by adding serial dilutions of IFN to MARC-145 cells in flat-bottom microtiter plates and monitoring cell density after 72 h by crystal violet staining. The 50% activity concentrations (EC₅₀) of IFN were determined from an IFN dose-response curve (SigmaPlot11.0, Systat, Chicago, IL) as described by Roisman et al. (32).

Promoter regions corresponding to human IFN regulatory factor-3 (IRF-3, −779/+1) (21), IRF-7 (−1123/+575) (22), and MxA (−530/+1) (33) were retrieved from human genomic DNA (Invitrogen) with a high-fidelity PCR kit (Invitrogen). Promoter regions were cloned into luciferase reporter vector pGL4.14 [luc2/Hygro] (Promega, Madison, WI) with the KpnI and HindIII cloning sites (37). Sequence-confirmed constructs were used to transfect MARC-145 cells and to establish stable reporter cell lines after selection of surviving cell colonies after 10 passages in medium containing 500 μg/ml hygromycin B (Invitrogen). All three promoter-reporter MARC-145 cell lines used were tested numerous times for their responses to virus-related stimuli (37) and recombinant IFN-α (R&D Systems). Cells were grown to confluence on 96-well plates and then incubated with IFN-containing media for 16 h. Cells were lysed, and released luciferase was detected simultaneously by addition of a Steady-Glo luciferase assay reagent at 100 μl/well (Promega, Madison, WI). After incubation for 10 min at 22°C, luciferase activity was measured with a luminometer (Fluoroskan, Ascent, FL). Statistical analyses were done with SigmaPlot 11.0 software (Systat).

Nucleotide sequences for the new identifications, characterizations, and sequences in this article have been deposited in GenBank. Accession numbers are listed in Table 1.

Table 1. Porcine type I interferons: annotation of functional genes

To investigate the repertoire of porcine type I IFNs, we used bioinformatic tools to scan swine genome databases (http://www.ncbi.nlm.nih.gov/genome/guide/pig/). At least 38 porcine type I IFN gene loci were clustered densely in 6 regions spanning ∼10–20 kb each along Sus scrofa autosome 1 (SSC1). These regions of SSC1 are covered by working draft sequences with GenBank accession numbers AC127471, AC130792, AC135219,AC138785, CU469469, and CU633414 deposited by the Swine Genome Project. These 38 loci likely encode most of the porcine type I IFNs and include genes of 17 IFN-α subtypes, 11 IFN-δ subtypes, 7 IFN-ω subtypes, and single-subtype subclasses of IFN-αω, IFN-β, and IFN-ε. In addition, we correlated most of these type I IFN loci with putative annotations on pig reference genome assembly Sscrofa5 (Table 1). The gene of another single-subtype subclass, IFN-κ, is located on porcine chromosome 10 (SSC10) spanned by the working draft sequence with accession number CU463237. Additional evidence of the IFNK location was found by searching the draft assembly of SSC10 at Pre-Ensembl (http://www.ensembl.org/Sus_scrofa/index.html). The location of the IFNK gene was defined in a region of SSC10 between 37,587,861 and 37,588,487 bp on the reverse strand with a transcript ID of ENSSSCT00000012053. Gene loci of the three single-subtype IFN subclasses are generally separated from other IFN genes by greater than ∼30 kb. In contrast, each IFN locus of multiple-subtype subclasses (i.e., IFNA, IFND, and IFNW genes) was within ∼20 kb distance from other closest IFN gene loci (Supplemental Fig. S1).

Among all identified porcine type I IFNs, the IFNK gene has the longest ORF (627 bp) encoding a 208-amino acid peptide. All other genes of porcine type I IFNs distributed on SSC1 have ORFs of 462–582 bp, encoding prepeptides of 153–193 amino acids (Supplemental Fig. S1). Examination of predicted peptides of all porcine type IFNs against the conserved domain database (CDD;http://www.ncbi.nlm.nih.gov/Structure/cdd/cdd.shtml) indicates that they all contain an IFabd functional domain, which is a type I IFN signature (24). Multiple putative binding sites for type I IFN receptor (IFNAR) subunits and one putativeN-glycosylation site (Asn/Asp-X-Ser/Thr), which are important for IFN signaling and IFN peptide stability, were also detected in all porcine IFN peptides. All 39 porcine IFNs bear NH₂-terminal 20- to 30-amino acid signal peptides. Except for IFN-δ7, all porcine IFN peptides have two conserved cysteine residues, the second Cys⁵² and fourth Cys¹⁶² relative to IFN-α1 (Fig. 1). However, the other two cysteine residues (the first Cys²⁴ and third Cys¹²² relative to IFN-α1) are only conserved in IFN-α, IFN-ω, and IFN-κ subclasses. All 11 IFN-δ isoforms have two other cysteine residues at positions Cys⁷⁷ and Cys¹²⁸ relative to IFN-δ1, and, except for IFN-δ2 and IFN-δ7, they also have an additional COOH-terminal cysteine residue (Cys¹⁶⁶ relative to IFN-δ1).

Fig. 1. Multiple sequence alignment of porcine type I interferons (IFNs) with IFN-α1, IFN-α4, and IFN-β. Residues with identities >50% among all compared sequences are shaded (light shading for similar residues and dark shading for identical residues). All porcine type I IFNs contain NH₂-terminal 20- to 30-amino acid signal peptides. The COOH-terminal hydrophobic or positively charged tails of some IFNs are boxed. Horizontal lines identify the 5 α-helical regions (A–E) as predicted from human IFN-α2a and IFN-β structures. Numbered black arrows indicate conserved cysteine residues among all or some (nos. in parentheses) subtypes of aligned IFNs.

Within the IFabd domain of each IFN isoform, five regions that may conform into five α-helixes were found, except for IFN-δ7, which only has the first four helical regions (Fig. 1). The three multiple-subtype subclasses (IFN-α, IFN-δ, and IFN-ω) can be further subclassified into short and long members within each subclass. The major difference is that the long members have 6–16 extended residues forming a COOH-terminal positively charged or hydrophobic tail. For example, IFN-α4, IFN-α5, IFN-α6, IFN-α8, IFN-α9, IFN-α12, IFN-α13, and IFN-α14 all have eight more residues of QDRFRKKE compared with other IFN-α subtypes; IFN-δ5, IFN-δ6, IFN-δ9, and IFN-δ11 have a 15-residue tail of QQSSTKSQERKKAHL compared with other IFN-δ subtypes; and all IFN-ω subtypes, except IFN-ω1, have 14 or 16 extended residues (Fig. 1). The four single-subtype subclasses (IFN-αω, IFN-β, IFN-ε, and IFN-κ) have a COOH-terminal positively charged or hydrophobic tail of 6, 6, 11, and 7 residues, respectively. In the IFN-α subclass, this COOH-terminal positively charged or hydrophobic tail is thought to contribute to higher antiviral activity in the long isoforms (41).

Phylogenic classification of porcine type I IFNs in this study was based on the following criteria. First, sequence similarity of subtypes within a subclass is near or higher than 80% nucleic acid or 60% amino acid, and sequence similarity of subtypes between subclasses is lower than 70% nucleic acid or 50% amino acid. Second, multiple subtypes of each subclass generally share the same cysteine distribution and likely intramolecular disulfide bond patterns, even though some distribution patterns may be conserved among subclasses. For example, subtypes of IFN-α and IFN-ω have four cysteine residues likely forming two intramolecular disulfide bonds as in human IFN-α2a (14), and IFN-β has two cysteine residues likely forming one disulfide bond as in human IFN-β (12). Third, names of each subclass were based on their closest homologs within or across species in previous references. To this end, porcine IFN-β (3) and IFN-α (mostly IFN-α1) (18) were previously isolated, and IFN-α1 to IFN-α14 and several IFN-δ subtypes were recently defined by bioinformatic analyses (4, 6); IFN-ε, IFN-κ, and some isoforms of IFN-α (IFN-α15, IFN-α16, IFN-α17) were characterized for the first time in this study. Two to three subtypes of IFN-δ or IFN-ω were originally reported as short type I porcine interferon (SPI-IFN) or porcine IFN-αII (PoIFN-αII) (18–20, 25). Recently, eight IFN-ω were defined in the porcine genome with a similar bioinformatic approach. However, identical sequences were shared in pairs of IFN-ω2/6 and IFN-ω3/5; thus those eight IFN-ω reflect six of the IFN-ω designated here (44). IFN-αω, first coined by Krause and Pestka (15) during an evolutionary analysis of IFNs, is a unique isoform that has only been identified in a few species such as pigs and cattle. Relative to other IFNs, IFN-αω shows only 65% nucleic acid identity and 50% amino acid similarity to the closest subclass. Similar to IFN-β, it has a six-amino acid tail.

Fig. 2. Phylogenic relationship of porcine type I interferons (poIFNs) and identified human (hu) and bovine (bo) orthologs. The analysis is based on predicted amino acid sequences. GenBank accession numbers for poIFNs are listed in Table 1. GenBank accession numbers for huIFNs and boIFN are huIFN-α1, NP_076918; huIFN-α6, NP_066282; huIFN-α17, NP_067091; huIFN-β,NP_002167; huIFN-ε,NP_795372; human-κ, NP_064509; huIFN-ω, NP_002168; and boIFN-αω, XP_001254773. 

Although sharing high sequence similarity in coding ORFs, the putative promoter regions among most members of porcine type I IFNs are quite different (Y. Sang, unpublished observation). This suggests that expression of type I IFN genes even within a subclass is different in respect to tissue/cell types or different stimuli. We thus devised an assay to detect the differential expression of most porcine type I IFN genes with a real-time RT-PCR array. As shown in Table 2, significant expression of multiple-type IFN genes was detected in skin, intestine, lymph nodes, spleen, and testis. For example, in skin, 15 IFN genes belonging to all subclasses were highly expressed, with a relative expression index higher than 1.00. IFN genes were highly expressed in intestine (10 genes) and lymph nodes (9 genes); however, most of them were from subclasses other than IFN-α. In spleen and testis two to four IFN genes from each multiple-subtype subclass (i.e., IFN-α, IFN-δ, or IFN-ω) were highly expressed. In contrast, bone marrow cells and liver showed a relatively weaker expression pattern of type I IFNs. Relative to subtype differences, IFN-α1, IFN-α8, and IFN-α12 and IFN-β were detected in all tested tissues. Although shown to have tissue and cell preference in some of their orthologs, porcine IFN-δ8 and IFN-δ5/6; IFN-ω1, IFN-ω2, and IFN-ω3; and single-subtype subclasses IFN-ε and IFN-κ were detectable in most tested tissues and highly expressed in three or four assayed tissues. IFN-αω was highly expressed only in skin, although detectable in intestine, lymph nodes, and spleen.

Table 2. Tissue expression profile of porcine type I interferon genes

For our initial evaluation of antiviral activity of porcine IFNs two approaches were used: 1) direct transfection of MARC-145 cells with IFN expression constructs and 2) treatment of MARC-145 cells with IFN-containing medium collected from transfected HEK293A cells (Supplemental Fig. 2A). Antiviral protection of expressed IFN genes or IFN-containing medium was shown by decreasing the ability of PRRSV to produce CPE on cell monolayers. Alignment of data from both expression vector transfection and IFN-containing medium treatment showed similar characteristics between the two activity curves (r² = 0.95; Supplemental Fig. 2A). The protective effect of exogenous IFN was specific, because media from mock transfection of cells with an antisense IFNA1-pcDNA construct did not protect against viral infection. In addition, incubation of IFN-containing medium from either IFN-α1 or IFN-β with antibodies to porcine IFN-α or IFN-β (anti-α or anti-β, respectively) eliminated the anti-PRRSV activity (Supplemental Fig. 2B). For the purpose of comparison, incubation with commercial recombinant porcine IFN-α showed that concentrations >31 U/100 μl of IFN-α were sufficient to fully protect MARC-145 cells from PRRSV infection (Supplemental Fig. 2B). Collectively, these findings showed that IFN peptides expressed by transfected porcine IFN genes conferred antiviral activity in the cells, and that IFN-containing medium collected from transformants could be used to determine the antiviral activity of transformed porcine IFN genes. Therefore, we further purified and concentrated IFN peptides from the IFN-containing medium (Supplemental Fig. S3), and protein concentrations of all IFN supernatants were measured for further evaluation of IFN activity.

The protective role of IFN-containing medium against viral infection was evaluated in both porcine cells and MARC-145 cells. In both porcine AMs and MARC-145 cells, all IFN-α subtypes, except IFN-α7 and IFN-α11μ (a guanosine-deletion mutant of IFN-α7/11), provided nearly full protection against PRRSV infection (10³ dilution, ∼2 ng/ml; Fig. 3A). In contrast, other subtypes, including IFN-β and most IFN-δ and IFN-ω, showed effective protection (60–80%) against PRRSV in porcine AMs but not in MARC-145 cells (Fig. 3A). Porcine type I IFN antiviral activity against VSV in PK-15 and MARC-145 cells displayed similar activity profiles, i.e., most IFN-α subtypes exerted antiviral activity in both porcine and MARC-145 cells, but other subtypes were selectively active in porcine cells (Fig. 3B). Interestingly, IFN-αω, a unique porcine IFN subtype, exerted higher antiviral protection in MARC-145 cells than in porcine cells (Fig. 3). In all evaluations, IFN-α7, IFN-α11μ, IFN-δ2, IFN-δ7, IFN-ε, IFN-κ, and IFN-ω4 showed little antiviral activity against PRRSV or VSV in either porcine cells or MARC-145 cells (Fig. 3).

Fig. 3. Antiviral activities of porcine type I IFNs. Serial dilutions (1:10) of IFN-containing medium with initial protein concentration at 2 μg/ml were added into wells in 96-well plates. Porcine cells [alveolar macrophages (AM); porcine kidney-15 (PK-15)] or MARC-145 cells were infected with either porcine reproductive and respiratory syndrome virus (PRRSV; A) or vesicular stomatitis virus (VSV; B). The protective role of IFNs was evaluated by calculating % reduction of virus-infected cells compared with positive [mock infected, minimum essential medium (MEM)] and negative (mock treated, Aα1, antisense IFNA1) controls. Virus-infected cells were determined by crystal violet staining in MARC-145 and PK-15 cells or immunostaining of PRRSV-positive cells in porcine alveolar macrophages. Data (1:1,000 dilution) are means ± SE; n = 8. *Different from MARC-145 cells (P < 0.05).

 

 

On the basis of further analysis of antiviral activity against PRRSV and VSV, porcine type I IFNs were categorized into high- and low-activity groups (Fig. 4). Members in the high-activity groups were those having average activity higher than 5 × 10³ U·ng⁻¹·ml⁻¹. In MARC-145 cells, the high-activity group exhibited a range of anti-PRRSV activity of 10⁴ to >10⁵ U·ng⁻¹·ml⁻¹ and included most members of the IFN-α subclass tested (except IFN-α7 and IFN-α11μ) and IFN-αω. The highest-activity subtype, IFN-α6, provided full antiviral protection in the PRRSV-MARC-145 system at ∼0.02 ng/ml. Members in the low-activity groups had activity near or at 0 U·ng⁻¹·ml⁻¹. The low-anti-PRRSV activity group included the three single-subtype subclasses (IFN-β, IFN-ε, and IFN-κ), all members of tested IFN-δ and IFN-ω, and IFN-α7 and IFN-α11μ (Fig. 4B). The anti-PRRSV activity of porcine IFNs in MARC-145 cells was comparable to the anti-VSV activity in both MARC-145 and PK-15 cells (Fig. 4, C–F). Although most active IFNs showed higher anti-VSV activity than anti-PRRSV activity in MARC-145 cells, the profiles of the high- and low-activity groups were similar to that categorized above (Fig. 4, C and D). Exceptions included IFN-δ8, which was the lowest active member in the high-anti-VSV activity group (Fig. 4C), and IFN-δ1 and IFN-β in the low-activity group that had some activity, but not different from the negative controls (Fig. 4D). Anti-VSV activity in porcine PK-15 cells was unlike that obtained in MARC-145 cells (Fig. 4, E and F). The high-anti-VSV activity group in PK-15 cells not only included all IFN-α subtypes but also IFN-β, four IFN-δ (δ1, δ4, δ5, and δ6), and IFN-ω2 (Fig. 4E). IFN-αω, which had high activity in MARC-145 cells, had low anti-VSV activity in PK-15 cells and was categorized in the low-activity group along with 11 other porcine IFNs (Fig. 4F). Gel electrophoresis of IFN-containing samples confirmed the presence of correct proteins of both high- and low-activity IFNs, such as IFN-α7 and other IFN-α subtypes (Supplemental Fig. S3). In addition, some IFNs that are minimally active in MARC-145 exerted antiviral activity in porcine cells (Fig. 3), indicating that the absence of activity of an IFN was not caused by the protein in an inactive form. IFN alone was not responsible for cell death because antiproliferative activity of porcine IFNs required much higher concentrations than that required for antiviral activity (Supplemental Table S2).

Fig. 4. High and low antiviral activity of porcine type I IFNs. MARC-145 (A–D) or PK-15 (E and F) cell monolayers were infected with PRRSV (MARC-145 only) or VSV in the presence of IFN-containing medium (1:10 serial dilutions). Protection by IFN was assayed by crystal violet staining, and the dye was extracted and quantified with a microplate reader at absorbance of 570 nm. Antiviral activity of each IFN was calculated by using the Reed-Muench method (31) to define 1 unit as the highest dilution that reduced the cell number by 50%. Porcine type I IFNs were categorized into high (>5 × 10³ U·ng⁻¹·ml⁻¹; A, C, and E)- and low (<5 × 10³ U·ng⁻¹·ml⁻¹; B, D, and F)-activity groups. Data are means ± SE; n = 8.

 

IRFs and the MxGTPase pathway are main effector pathways for IFN-meditated antiviral responses. To investigate the signaling pathways of porcine type I IFNs, promoters for IRF-3, IRF-7, and MxA were transfected into MARC-145 cells. Comparison of anti-PRRSV activity of porcine type I IFNs against activation of IFN-stimulated genes showed that antiviral activity was highly correlated (r² = 0.97) with an IFN's ability to induce MxA promoter activity (Fig. 5A). A lower correlation was found between IFN antiviral potency and stimulation of IRF-7 promoter activity (Fig. 5B). In contrast, almost no correlation existed between anti-PRRSV activity and the ability of porcine type I IFNs to stimulate the IRF-3 promoter (Fig. 5C).

Fig. 5. Anti-PRRSV activity of porcine type I IFNs is highly correlated with IFN induction of the MxA gene (A) and less correlated with the induction of interferon regulatory factor (IRF)-7 (B) and IRF-3 (C). Data represent 2 independent experiments with similar results.

 

Continuing progress of animal genome projects and the use of functional genomic approaches have expanded and given definition to several mammalian gene families (29, 36). This study provides experimental support for three new findings concerning the porcine type I IFN family. First, it shows that the porcine type I IFN family consists of 39 functional genes, classified into 7 subclasses. Second, it identifies for the first time seven new porcine IFNs and provides new functional analyses for many others. Third, it shows that some newly identified and functionally characterized porcine IFNs are differentially expressed and have potent anti-PRRSV activity.

Porcine type I IFN gene loci consist of at least 39 single-exon (intronless) functional genes, spanning an ∼500-kb region of SSC1 with an IFN-κ positioned on SSC10. The finding of a porcine IFNK locus on a different chromosome than other type I IFNs was unexpected because its human ortholog is located on the same chromosome (chromosome 9) as other type I IFN genes (42). Further searching of the draft assembly of SSC10 refined the annotated location of this gene. In addition, a radiation hybrid map of SSC10 refined a syntenic region to HSA9 at 27–99 Mb, which might contain the human IFNK gene locus, indicating a possible chromosome relocation of porcine IFNK compared with its human ortholog (23).

Of the known porcine IFNs, IFN-α1 (presumably) and IFN-β are among the most characterized subtypes, including their molecular identification and antiviral responses related to multiple viral infections (3, 4). Other subtypes have received less, but some, characterization. For example, porcine IFN-δ subtypes were originally designated as small porcine IFNs, and IFN-δ1 was well defined for its expression in trophoblasts (19). In addition, three functional porcine IFN-ω genes that were originally designated as IFN-α-II4, -II5, and -II3 and characterized by Mege et al. in 1991 (25) were matched to our IFNW1, IFNW3, and IFNW7, respectively. In this study, seven porcine IFNs have been identified for the first time (IFN-α15, IFN-α16, and IFN-α17, IFN-ε, IFN-κ, and IFN-ω2 and IFN-ω7), and others, which have appeared in the sequence literature but have not been characterized (6, 15, 19), were evaluated for antiviral activity. Each porcine IFN gene has a 500- to 600-bp ORF, which encode a peptide containing a type I IFN functional domain (IFabd) but are diverse from each other in residue composition. On the basis of their phylogenic sequence relationship, porcine type I IFNs can be classified into seven subclasses consisting of IFN-α, IFN-β, IFN-δ, IFN-ε, IFN-κ, IFN-ω, and IFN-αω. IFN-α, IFN-δ, and IFN-ω are subclasses consisting of multiple subtypes. Furthermore, gene loci of porcine type I IFNs are not clustered together by similar subclasses but are interspersed with members from different subclasses. This likely indicates that mechanisms of gene duplication are involved in the evolution of type I IFN genes (11). In addition, gene duplication appears more apparent in loci of subclasses with multiple subtypes such as IFN-α, IFN-δ, and IFN-ω; loci of single-member subtypes are typically separated from other IFN gene loci and latent in evolving novel isoforms. Functional domain analysis reveals that all porcine type I IFN proteins contain an IFabd domain, a signature of the type I IFN family, and putative binding sites for IFN-α receptor (IFAR) subunits, which indicates functional conservation in IFN-stimulated antiviral signaling. Members of different IFN subclasses generally share <50% identity in peptide sequences and are likely different in their structural conformation resulting from variable intramolecular disulfide connections. Even among subtypes of the same subclass, e.g., IFN-α, IFN-δ, and IFN-ω, differences in COOH-terminal residue composition, especially the positively charged or hydrophobic tails, likely differentiate their antiviral potencies and regulation of other physiological processes. The COOH-terminal positively charged or hydrophobic tail has been proposed to contribute to higher antiviral activity in the IFN-α subtype (4). However, our data indicate that longer IFN subtypes in the multiple-subtype subclasses do not always have higher activity than the shorter subtypes. For example, IFN-α1 and IFN-α2 exhibited greater anti-PRRSV activity than longer subtypes, such as IFN-α4, IFN-α8, and IFN-α13 in MARC-145 cells.

We examined differential expression of most type I IFN genes in several porcine tissues. Because these tissues were from healthy pigs without symptoms of disease, we reasoned that the expression data represent basal levels of these innate immune antiviral effectors in these tissues. A preliminary expression analysis in PRRSV-positive porcine tissues has shown differences from basal expression in control tissues (Y. Sang, unpublished observation).

Different IFNs possessed diverse activities when tested in different target cell-virus systems; therefore, porcine IFNs were categorized into high- and low-activity groups according to their anti-PRRSV or anti-VSV activities in MARC-145 cells (Fig. 4, A–D). Generally, porcine IFNs were more active in porcine cells (Figs. 3 and 4). An intriguing exception is IFN-αω, which was more active in MARC-145 cells than in porcine cells. Recombinant porcine IFN-α (presumably IFN-α1) and IFN-β have been shown to suppress PRRSV infectivity in AMs (2, 28). Our data further indicate that most subtypes of the IFN-α subclass are highly active against PRRSV in both porcine AMs and MARC-145 cells; however, other subclasses including IFN-β, IFN-δ, IFN-ε, and IFN-κ are minimally active in MARC-145 cells. This likely reflects a difference between homologous and heterologous IFN ligand-receptor interaction. Work by others indicates that IFN-α is downregulated in response to PRRSV infection (1, 2, 9). This suggests that IFN-α subtypes should be emphasized in modulation of porcine anti-PRRSV innate immunity. This point is supported by findings showing that PRRSV isolates differ in their sensitivity to IFN-α suppression (17). Therefore, extensive investigation of effective anti-PRRSV IFN scenarios may uncover critical aspects that shape efficient immune protection to PRRSV infection.

Antiviral activity of porcine type I IFNs in MARC-145 cells is correlated with their ability to induce expression of MxA rather than IRF-3 and IRF-7. This indicates that the efficacy of an IFN to signal downstream ISG expression may contribute more to anti-PRRSV activity than to autoregulating IFN production. Alternatively, some ISGs, like MxA, may have special PRRSV-targeting activity (40).

We have shown that most subtypes of the porcine IFN-α subclass were very active against PRRSV in in vitro systems. Why then are these IFNs not effective in mitigating PRRSV infection? Answers to this question are not known; however, two reasonable answers are possible: 1) endogenous IFN concentrations are lower than the in vitro overexpressed concentration and 2) stimulation of systemic antiviral immunity in vivo may require multiple type I IFNs. More defined analyses of overexpressed and endogenous IFNs at the protein level should help to answer these questions and to elucidate porcine IFN antiviral innate immunity.

In summary, we have identified an expanded repertoire of porcine type I IFNs through bioinformatic annotation and extensive comparative antiviral assays. Elucidation of most of the active IFN molecules and downstream ISG candidate(s) against PRRSV will help clarify porcine anti-PRRSV immunity and may suggest novel strategies to limit costly pandemic viral disease. 

This work was supported in part by U.S. Department of Agriculture National Research Initiative Competitive Grants Program (NRICGP) Grants 2006-35204-17337 and 2003-35204v13704. 

No conflicts of interest, financial or otherwise, are declared by the author(s). 

Data for the IFNK genomic location were provided by the Sanger Institute Porcine Project Team at the Wellcome Trust Sanger Institute and can be obtained from http://www.ensembl.org/Sus_scrofa/index.html. We thank Danielle Goodband for her expert technical assistance and Joe Anderson for his help with the VSV experiments. 

1. Ait-Ali T, Wilson AD, Westcott DG, Frossard JP, Mellencamp MA, Drew TW, Bishop SC, Archibald AL. Dynamic differential regulation of innate immune transcripts during the infection of alveolar macrophages by the porcine reproductive and respiratory syndrome virus. Dev Biol 132: 239–245, 2008.

2. Albina E, Carrat C, Charley B. Interferon-alpha response to swine arterivirus (PoAV), the porcine reproductive and respiratory syndrome virus. J Interferon Cytokine Res 18: 485–490, 1998.

3. Artursson K, Gobl A, Lindersson M, Johansson M, Alm G. Molecular cloning of a gene encoding porcine interferon-beta. J Interferon Res 12: 153–160, 1992.

4. Cheng G, Chen W, Li Z, Yan W, Zhao X, Xie J, Liu M, Zhang H, Zhong Y, Zheng Z. Characterization of the porcine alpha interferon multigene family. Gene 382: 28–38, 2006.

5. Cheng G, Zhao X, Chen W, Yan W, Liu M, Chen J, Zheng Z. Detection of differential expression of porcine IFN-alpha subtypes by reverse transcription polymerase chain reaction. J Interferon Cytokine Res 27: 579–87, 2007.

6. Cochet M, Vaiman D, Lefèvre F. Novel interferon delta genes in mammals: cloning of one gene from sheep, two genes expressed by the horse conceptus and discovery of related sequences in several taxa by genomic database screening. Gene 433: 88–99, 2009.

7. Demmers KJ, Derecka K, Flint A. Trophoblast interferon and pregnancy. Reproduction 121: 41–49, 2001.

8. García-Sastre A, Biron CA. Type 1 interferons and the virus-host relationship: a lesson in détente. Science 312: 879–882, 2006.

9. Genini S, Delputte PL, Malinverni R, Cecere M, Stella A, Nauwynck HJ, Giuffra E. Genome-wide transcriptional response of primary alveolar macrophages following infection with porcine reproductive and respiratory syndrome virus. J Gen Virol 89: 2550–2564, 2008.

10. Haller O, Weber F. Pathogenic viruses: smart manipulators of the interferon system. Curr Top Microbiol Immunol 316: 315–334, 2007.

11. Kaessmann H, Vinckenbosch N, Long M. RNA-based gene duplication: mechanistic and evolutionary insights. Nat Rev Genet 10: 19–31, 2009.

12. Karpusas M, Nolte M, Benton CB, Meier W, Lipscomb WN, Goelz S. The crystal structure of human interferon beta at 2.2-Å resolution. Proc Natl Acad Sci USA 94: 11813–11818, 1997.

13. Kim TS, Benfield DA, Rowland RR. Porcine reproductive and respiratory syndrome virus-induced cell death exhibits features consistent with a nontypical form of apoptosis. Virus Res 85: 133–140, 2002.

14. Klaus W, Gsell B, Labhardt AM, Wipf B, Senn H. The three-dimensional high resolution structure of human interferon alpha-2a determined by heteronuclear NMR spectroscopy in solution. J Mol Biol 274: 661–675, 1997.

15. Krause CD, Pestka S. Evolution of the Class 2 cytokines and receptors, and discovery of new friends and relatives. Pharmacol Ther 106: 299–346, 2005.

16. LaFleur DW, Nardelli B, Tsareva T, Mather D, Feng P, Semenuk M, Taylor K, Buergin M, Chinchilla D, Roshke V, Chen G, Ruben SM, Pitha PM, Coleman TA, Moore PA. Interferon-kappa, a novel type I interferon expressed in human keratinocytes. J Biol Chem 276: 39765–39771, 2001.

17. Lee SM, Schommer SK, Kleiboeker SB. Porcine reproductive and respiratory syndrome virus field isolates differ in in vitro interferon phenotypes. Vet Immunol Immunopathol 102: 217–231, 2004.

18. Lefèvre F, La Bonnardière C. Molecular cloning and sequencing of a gene encoding biologically active porcine alpha-interferon. J Interferon Res 6: 349–360, 1986.

19. Lefèvre F, Boulay V. A novel and atypical type one interferon gene expressed by trophoblast during early pregnancy. J Biol Chem 268: 19760–19768, 1993.

20. Lefèvre F, Guillomot M, D'Andréa S, Battegay S, La Bonnardière C. Interferon-delta: the first member of a novel type I interferon family. Biochimie 80: 779–788, 1998.

21. Lowther WJ, Moore PA, Carter KC, Pitha PM. Cloning and functional analysis of the human IRF-3 promoter. DNA Cell Biol 18: 685–692, 1992.

22. Lu R, Pitha PM. Monocyte differentiation to macrophage requires interferon regulatory factor 7. J Biol Chem 276: 45491–45496, 2001.

23. Ma JG, Yasue H, Eyer KE, Hiraiwa H, Shimogiri T, Meyers SN, Beever JE, Schook LB, Beattie CW, Liu WS. An integrated RH map of porcine chromosome 10. BMC Genomics 10: 211, 2009.

24. Marchler-Bauer A, Anderson JB, Derbyshire MK, DeWeese-Scott C, Gonzales NR, Gwadz M, Hao L, He S, Hurwitz DI, Jackson JD, Ke Z, Lanczycki CJ, Liebert CA, Liu C, Lu F, Lu S, Marchler GH, Mullokandov M, Song JS, Tasneem A, Thanki N, Yamashita RA, Zhang D, Zhang N, Bryant SH. CDD: a conserved domain database for interactive domain family analysis. Nucleic Acids Res 37: D205–D210, 2009.

25. Mege D, Lefevre F, Labonnardiere C. The porcine family of interferon-omega: cloning, structural analysis, and functional studies of five related genes. J Interferon Res 11: 341–350, 1991.

26. OIE Biological Standards Commission. Vesicular stomatitis. OIE Terrestrial Manual 2008, chapter 2.1.19. p.367–376. http://www.oie.int/Eng/Normes/Mmanual/A_summry.htm [2010].

27. Oritani K, Medina KL, Tomiyama Y, Ishikawa J, Okajima Y, Ogawa M, Yokota T, Aoyama K, Takahashi I, Kincade PW, Matsuzawa Y. Limitin: an interferon-like cytokine that preferentially influences B-lymphocyte precursors. Nat Med 6: 659–666, 2000.

28. Overend C, Mitchell R, He D, Rompato G, Grubman MJ, Garmendia AE. Recombinant swine beta interferon protects swine alveolar macrophages and MARC-145 cells from infection with porcine reproductive and respiratory syndrome virus. J Gen Virol 88: 925–931, 2007.

29. Patil AA, Cai Y, Sang Y, Blecha F, Zhang G. Cross-species analysis of the mammalian beta-defensin gene family: presence of syntenic gene clusters and preferential expression in the male reproductive tract. Physiol Genomics 23: 5–17, 2005.

30. Pestka S. The interferons: 50 years after their discovery, there is much more to learn. J Biol Chem 282: 20047–20051, 2007.

31. Reed LJ, Muench H. A simple method of estimating 50 per cent endpoints. Am J Hyg 27: 493–497, 1938.

32. Roisman LC, Jaitin DA, Baker DP, Schreiber G. Mutational analysis of the IFNAR1 binding site on IFNalpha2 reveals the architecture of a weak ligand-receptor binding-site. J Mol Biol 353: 271–281, 2005.

33. Ronni T, Matikainen S, Lehtonen A, Palvimo J, Dellis J, Van Eylen F, Goetschy JF, Horisberger M, Content J, Julkunen I. The proximal interferon-stimulated response elements are essential for interferon responsiveness: a promoter analysis of the antiviral MxA gene. J Interferon Cytokine Res 18: 773–781, 1998.

34. Rowland RR, Robinson B, Stefanick J, Kim TS, Guanghua L, Lawson SR, Benfield DA. Inhibition of porcine reproductive and respiratory syndrome virus by interferon-gamma and recovery of virus replication with 2-aminopurine. Arch Virol 146: 539–555, 2001.

35. Sadler AJ, Williams BR. Structure and function of the protein kinase R. Curr Top Microbiol Immunol 316: 253–292, 2007.

36. Sang Y, Patil AA, Zhang G, Ross CR, Blecha F. Bioinformatic and expression analysis of novel porcine beta-defensins. Mamm Genome 17: 332–339, 2006.

37. Sang Y, Yang J, Ross CR, Rowland RR, Blecha F. Molecular identification and functional expression of porcine Toll-like receptor (TLR) 3 and TLR7. Vet Immunol Immunopathol 125: 162–167, 2008.

38. Sang Y, Ross CR, Rowland RR, Blecha F. Toll-like receptor 3 activation decreases porcine arterivirus infection. Viral Immunol 21: 303–313, 2008.

39. Sang Y, Ruchala P, Lehrer RI, Ross CR, Rowland RR, Blecha F. Antimicrobial host defense peptides in an arteriviral infection: differential peptide expression and virus inactivation. Viral Immunol 22: 235–242, 2009.

40. Sen GC, Sarkar SN. The interferon-stimulated genes: targets of direct signaling by interferons, double-stranded RNA, and viruses. Curr Top Microbiol Immunol 316: 233–250, 2007.

41. Slutzki M, Jaitin DA, Yehezkel TB, Schreiber G. Variations in the unstructured C-terminal tail of interferons contribute to differential receptor binding and biological activity. J Mol Biol 360: 1019–1030, 2006.

42. Takaoka A, Yanai H. Interferon signalling network in innate defence. Cell Microbiol 8: 907–922, 2006.

43. Uzé G, Schreiber G, Piehler J, Pellegrini S. The receptor of the type I interferon family. Curr Top Microbiol Immunol 316: 71–95, 2007.

44. Zhao X, Cheng G, Yan W, Liu M, He Y, Zheng Z. Characterization and virus-induced expression profiles of the porcine interferon-omega multigene family. J Interferon Cytokine Res 29: 687–693, 2009.

45. Zuniga EI, Hahm B, Oldstone MB. Type I interferon during viral infections: multiple triggers for a multifunctional mediator. Curr Top Microbiol Immunol 316: 337–357, 2007.