Role of Toll-Like Receptors in Activation of Porcine Alveolar Macrophages by Porcine Reproductive and Respiratory Syndrome Virus

Abstract

Control of virus replication initially depends on rapid activation of the innate immune response. Toll-like receptor (TLR) ligands are potent inducers of innate immunity against viral infections. Porcine reproductive and respiratory syndrome virus (PRRSV), a positive-sense RNA virus, initiates infection in porcine alveolar macrophages (PAMs), elicits weak immune responses, and establishes a persistent infection. To understand the role of single-stranded RNA and double-stranded RNA (dsRNA) intermediates in eliciting host immunity, we sought to determine if TLRs, particularly those that respond to viral molecular patterns, are involved in PRRSV infection. Activation of TLR3 in PAMs with dsRNA increased gene expression for alpha interferon and suppressed PRRSV infectivity. In contrast, TLR4 activation by the treatment of PAMs with lipopolysaccharide did not influence PRRSV infectivity.

Porcine reproductive and respiratory syndrome virus (PRRSV), the causative agent of porcine reproductive and respiratory syndrome in swine, is a positive-stranded RNA virus of the Arteriviridae family in the order Nidovirales. PRRSV causes highly significant losses to the swine industry worldwide (31) as a result of both reproductive failure (late-term abortions and stillbirths) in pregnant sows and respiratory disease (pneumonia) in pigs of all ages (37). Infection with PRRSV also predisposes pigs to infection by bacterial pathogens as well as other viral pathogens (5), and it is a key etiological agent of the porcine respiratory disease complex. PRRSV has a tropism for cells of a phagocytic lineage, especially porcine alveolar macrophages (PAMs), a cell type that plays an essential role in microbial defense of the pig lung. The major obstacles to controlling PRRSV are the genetic variability of the virus (and, therefore, antigenic variability) (1) and the dysregulation of the host response by the virus (9, 19). Despite extensive research, currently available PRRSV vaccines are not protective against all field strains, resulting in less-than-perfect control programs. Unraveling the mechanisms of PRRSV-host cell interactions will be a critical step in developing predictable control and prevention strategies.

The type I interferons (IFN) alpha interferon (IFN-α) and beta interferon (IFN-β) are potent antiviral cytokines that represent an important part of innate immunity (32). In humans and mice, it has been demonstrated that IFN-α/β also affects adaptive immunity, for example, by the maturation of dendritic cells and promotion of a Th1-like immune response (4, 6). Production of IFN-α can be induced by many viruses and bacteria, including components of these microorganisms such as glycoproteins, double-stranded RNA (dsRNA), and DNA containing unmethylated CpG dinucleotides. Toll-like receptor (TLR) ligands have been investigated for their ability to initiate innate immune responses as an early line of defense against viral replication. TLRs are pattern recognition receptors directed against key pathogen-associated molecules which are evolutionarily conserved. When a TLR ligand binds to the receptor, an intracellular signal transduction cascade is triggered, altering the pattern of gene expression in the cell. Many adjuvants are believed to be mimics of TLR ligands, so TLRs turn out to be important for immune responses to vaccines as well as natural disease. A subset of TLRs, TLR3, TLR7/8, and TLR9, is involved in antiviral responses by triggering the production of antiviral cytokines such as type I IFN. Interestingly, contact between a virus and a TLR-expressing cell is often sufficient to induce type I IFN production without the need for infection, allowing uninfected cells to participate in the antiviral responses (24). This feature can be exploited to study virus-recognizing TLRs by mimicking virus-TLR interactions using synthetic ligands such as polyriboinosine-polyribocytidylic acid [poly(I·C)] (TLR3 ligand), single-stranded RNA40 (TLR7/8 ligand), or oligodeoxyribonucleotides containing unmethylated CpG dinucleotides (CpG-ODN) (TLR9 ligands) (7). TLR3 recognizes dsRNA, a molecular pattern associated with viral infection. Poly(I·C), a synthetic analog of dsRNA, is the ligand of choice for TLR3. dsRNA is known to induce the activation of NF-κB and the production of IFN-β through distinct mechanisms that are MyD88 dependent or MyD88 independent (36). TLR4 is the receptor for lipopolysaccharide (LPS) from gram-negative bacteria with its toxic moiety, lipid A (29). This recognition involves the binding of LPS with LPS-binding protein and subsequently with CD14, which physically associates with a complex including TLR4 and MD2 (30). Formation of the TLR4-centered LPS receptor complex induces the production of proinflammatory cytokines through the MyD88 pathway. LPS signaling also involves a MyD88-independent cascade that mediates the expression of IFN-inducible genes. TLR7 and TLR8 are also involved in responses to viral infections and recognize GU-rich, short single-stranded RNA as well as small synthetic molecules such as imidazoquinolines and nucleoside analogues. The water-soluble derivative of the imidazoquinoline compound R848, CL097, is a TLR7 and TLR8 ligand that induces the activation of NF-κB. TLR9 recognizes specific unmethylated CpG-ODN sequences that distinguish microbial DNA from mammalian DNA (3) and can induce B-cell proliferation, NK cell activation, and the production of cytokines, including type I IFN (17).

The critical role that TLRs can play in host immune responses against viruses is suggested by how the hepatitis C virus (HCV) is capable of establishing a persistent infection. HCV, through a virally encoded protease, induces specific proteolysis of a Toll-interleukin-1 (IL-1) receptor domain-containing adaptor to inhibit poly(I·C)-activated signaling, which potentially promotes persistent infection (20). In view of the fact that PRRSV also causes persistent infections with shedding, which is problematic for the effective control of the spread of the disease, we sought to better understand the ability of TLR ligands to trigger IFN responses and the role of selected TLR ligands with PAMs and how they might be affected by PRRSV infection. The goal of the present study was to determine the importance of certain properties of TLR ligands for their IFN-α-inducing capacity in PAMs. For that purpose, a panel of TLR ligands that had previously shown IFN-α-inducing capacity in other species was selected and analyzed for IFN-α induction in PAMs in vitro. The approach described may contribute to research on PRRSV pathogenesis and the development of new adjuvants for better and more-effective vaccines.

MATERIALS AND METHODS

Cells.

Each experiment was carried out with PAMs obtained from at least three conventionally reared 6- to 8-week-old pigs, originating from at least two different litters, from a PRRSV-free farm. Pigs were housed in isolation rooms at the National Animal Disease Center (Ames, IA) according to Institutional Animal Care and Use Committee guidelines. For the collection of lung lavage fluid, pigs were euthanized with an overdose of pentobarbital given intravenously. Primary PAMs were isolated from lung lavage fluid, cultured, and infected, as previously described (12). Briefly, lungs were lavaged with 100 ml of minimum essential medium (MEM; JRH Biosciences, Lenexa, KS), and ∼50 ml was recovered. Collected lavage fluid was centrifuged at 400 × g for 10 min, and cells were washed once with MEM and resuspended in supplemented medium (Dulbecco's modified Eagle's medium with 5% fetal bovine serum [FBS]; Gibco-Invitrogen, Carlsbad, CA) and 1% antibiotic/antimycotic solution (Gibco-Invitrogen). PAMs were tested by PCR for porcine circovirus type 2 and Mycoplasma spp. (27, 33) and found to be free of both. Cell counts were performed with a hemacytometer, and cytospin preparations were made by using centrifugation at 650 rpm (45 × g) for 4 min in a cytocentrifuge (Thermo Fisher Scientific Inc., Waltham, MA) and stained with a hematoxylin and eosin stain (Hema 3 stain; Thermo Fisher Scientific Inc.) for differential cell counting. Typical yields per lavage were 10⁸ to 10⁹ PAMs, with >95% viability. PAMs were diluted to 10⁶/ml of MEM and added into either 6-well plates or 25-cm² flasks with 2 or 5 ml of cells, respectively. PAMs were cultured overnight at 37°C with 5% CO₂ in Dulbecco's MEM with 5% FBS (Gibco-Invitrogen) and 1% antibiotic/antimycotic solution (Gibco-Invitrogen). Nonadherent cells were removed by gentle aspiration of the culture supernatant and replaced with fresh media (with or without the various treatments described below).

The Madin-Darby bovine kidney cells (ATCC CCL-22) were cultured in MEM containing Earle's salts, 2 mM l-glutamine, 100 U/ml penicillin G, 100 U/ml streptomycin sulfate (Gibco BRL, Gaithersburg, MD), and 10% FBS (Atlanta Biologicals, Norcross, GA).

Viruses.

The virus strain used for this experiment was PRRSV VR-2332. Virus stocks were propagated on MARC-145 cells, a clone of the African monkey kidney cell line MA-104 that is considered highly permissive to PRRSV (16), cultured, and maintained in culture medium composed of MEM supplemented with 10% FBS and 50 mg/liter of gentamicin (Gibco-Invitrogen).

Reagents used for IFN-α induction.

Synthetic dsRNA, or poly(I·C) (InvivoGen, San Diego, CA), was used to ligate TLR3 at a concentration of 25 μg/ml culture medium. To ligate TLR4, LPS from Escherichia coli O55:B5 (Sigma-Aldrich, St. Louis, MO) was used at a concentration of 100 ng/ml. The TLR7/8 ligand CL097 (InvivoGen, San Diego, CA) was used at a concentration of 5 μg/ml, and the TLR9 ligand CpG-ODN (M362; InvivoGen) was used at a concentration of 2.5 μg/ml. Exogenous human purified recombinant IFN-β1a protein (hrIFN-β; R&D Systems, Inc.) was used at 20 U/ml.

PAMs were infected with PRRSV at a multiplicity of infection (MOI) of 10 with or without TLR ligands or hrIFN-β treatments listed above at selected time points. Noninfected and unstimulated cells were included at each time point as controls.

Relative quantification of IFN-α mRNA.

At each indicated time point, PAMs were removed by scraping and centrifuged at 400 × g, and the supernatant was collected and stored at −80°C. Upon being thawed, the cells were lysed with RLT buffer, the first step for RNA isolation (RNeasy mini-isolation kit; Qiagen, Valencia, CA). RNA was isolated and DNase treated (RNase-free DNase; Qiagen). SYBR green-based real-time reverse transcription-PCR (RT-PCR) was conducted for IFN-α mRNA targets according to the manufacturer's recommendations (SuperScript III Platinum SYBR green one-step qRT-PCR kit; Invitrogen, Carlsbad, CA). Briefly, SuperScript III RT/Platinum Taq mix, SYBR green reaction mix, primers, and template were mixed in a 25-μl reaction mixture and cycled as follows: 50°C for 3 min; 95°C for 5 min; 95°C for 15 s, followed by 60°C for 1 min (45 cycles); and 40°C for 1 min. A final dissociation analysis was performed to identify the presence of primer dimers and analyze the specificity of the reaction. Fluorescence was measured following each cycle and displayed graphically by Applied Biosystems 7500 system sequence detection software version 1.3.1 (Applied Biosystems, Foster City, CA). All samples were run in triplicate. Levels of mRNA were calculated using the threshold cycle (2^-ΔΔC_T) method, which expresses mRNA in treated cells relative to that in nonstimulated cells after normalizing to 18S rRNA (QuantumRNA universal 18S internal standard; Ambion, Austin, TX) (21). Primers were designed using Primer Express, purchased from Integrated DNA Technologies, and used at a final concentration of 10 μM. Primer pair efficiency was confirmed using the method described by Livak and Schmittgen (21). PCR products were <100 bp in size, and the primers were IFN-α forward, 5′-GGCTCTGGTGCATGAGATGC-3′, and IFN-α reverse, 5′CAGCCAGGATGGAGTCCTCC-3′, which allow for the detection of the targeted porcine IFN-α gene (IFN-α1, IFN-α2, IFN-α3, IFN-α5, IFN-α8, IFN-α11, IFN-α12, and IFN-α14) expression (11). In the present study, the data are presented as the change in target gene expression in stimulated PAMs that are normalized to the internal control gene (18S rRNA) and relative to the mock control expressed by the 2^−ΔΔC_T method (2, 21). Increased mRNA expression was defined as a change of ≥2.0-fold, “normal” expression was a change ranging from 0.5001- to 1.9999-fold, and decreased mRNA expression was a change of ≤0.5-fold.

IL-1β ELISA.

IL-1β protein was measured using the porcine IL-1β/IL-1F2 DuoSet enzyme-linked immunosorbent assay (ELISA) development kit from R&D Systems, Inc., Minneapolis, MN, by following the manufacturer's protocol.

IFN-α ELISA.

IFN-α protein was measured with a porcine IFN-α-specific ELISA by using F17 monoclonal antibody (MAb) and K9 MAb (R&D Systems, Inc.) as previously described (13). MAb K9 was conjugated with horseradish peroxidase using a peroxidase labeling kit (Roche Molecular Biochemical, Indianapolis, IN). Immulon 2 flat-bottomed 96-well plates (Fisher Scientific, Houston, TX) were coated with F17 at a concentration of 3 μg/plate in coating buffer (100 mM carbonate buffer, pH 9.6; Sigma Inc., St. Louis, MO) overnight at 4°C. After being blocked with 1% nonfat dried milk and 0.05% Tween 20 in phosphate-buffered saline (PBS) for 1 h at 37°C, the plates were washed five times with 0.05% Tween 20 in PBS. Samples (50 μl) were added into each well containing 50 μl of 1% nonfat dried milk and 0.05% Tween 20 in PBS and incubated for 2 h at 37°C. Following five washes, 100 μl of peroxidase-conjugated K9 was added to each well. After 1 h of incubation at 37°C and after five washes, 100 μl of substrate solution tetramethylbenzidine (KPL, Inc., Gaithersburg, MD) was added to each well. After 30 min, the reaction was stopped with tetramethylbenzidine stop solution (KPL, Inc., Gaithersburg, MD), and the optical density at 450 nm was measured by an ELISA plate reader. Quantified recombinant porcine IFN-α (R&D Systems, Inc., Minneapolis, MN) was used as a standard, and IFN-α concentrations were calculated based upon a standard curve. One unit/milliliter of recombinant porcine IFN-α is equivalent to 26 pg/ml.

Virus quantification.

Supernatant collected at various times postinfection was retained for extracellular virus quantification. To obtain viral RNA, the QIAamp viral RNA minikit (Qiagen) was used as described in the kit instructions. A commercially available real-time, single-tube RT-PCR assay for the detection of U.S. PRRSV was obtained from Tetracore, Inc. (Rockville, MD), and used to detect PRRSV RNA as described previously (35). PRRSV RNA was transcribed in a 96-well format using a 25-μl reaction volume consisting of Tetracore U.S. PRRSV mastermix (20 μl mastermix, 1 μl enzyme mix 3) and 4 μl of extracted RNA. Reaction plates were loaded into an Applied Biosystems 7500 sequence detection system (Foster City, CA). The thermal cycler program for the U.S. PRRSV assay consisted of 60°C for 20 min, 95°C for 15 s and 40 cycles at 95°C for 15 s, and 60°C for 1 min. Known amounts of the serially diluted RNA transcript obtained in vitro (10^-1 through 10⁸ copies/μl) were used to generate a standard curve.

Statistical analyses.

Treatment means for each in vitro condition studied were compared using a two-way analysis of variance. Significance was established at a P value of <0.05.

RESULTS AND DISCUSSION

Induction of IFN-α response by TLR ligands in noninfected PAMs.

A strong type I IFN response to viral infection is critical for the production of downstream antiviral mediators. It has been demonstrated that PRRSV does not induce a strong IFN-α response (8, 34), yet the virus is sensitive to the effects of type I IFN and can suppress type I IFN responses induced by another virus (10, 18, 28). TLRs recognize microbial products and initiate transcription of type I IFN. In order to understand whether PRRSV infection in vitro alters the capacity of TLRs on PAMs to respond to their ligands, we investigated IFN-α gene transcription and protein expression, following stimulation of PAMs with TLR3/4/7/8/9 ligands. In addition, hrIFN-β was used to determine whether exogenous IFN altered type I IFN transcription. Figure 1A illustrates the effects of TLR ligand stimulation on IFN-α transcript levels in PAMs. No significant increase in transcription of IFN-α was detected following treatment with TLR4, TLR7/8, TLR9, or IFN-β. In contrast, transcription of IFN-α was increased (P < 0.05) at 8 and 12 h posttreatment with the TLR3 ligand poly(I·C). IL-1β protein expression was measured for the impact of the TLR ligands on proinflammatory cytokine production. Slight increases in the concentration of IL-1β protein expression in the PAM supernatant following treatment with TLR3, TLR4, and TLR7/8 over mock-treated cells were detected by ELISA (Fig. 1B). No significant increase in the concentration of IFN-α protein in the PAM supernatant following treatment with TLR4, TLR7/8, TLR9, or IFN-β was detected by ELISA (Table 1). The concentration of IFN-α protein in the PAM supernatant increased to a mean of 13.8 ± 6.5 pg/ml and 73.4 ± 27.5 pg/ml at 8 and 12 h posttreatment, respectively, with TLR3 ligand poly(I·C). A dose titration of TLR3 agonist, 10 to 100 μg/ml poly(I·C), showed IFN-α protein expression to increase from 20 ng/ml to 41 ng/ml, whereas the IFN-α protein expression in response to TLR4 agonist, 100 ng/ml to 1 μg/ml LPS, was below the range of sensitivity for the assay (Table 1).

FIG. 1.

IFN-α transcription and IL-1β protein expression by TLR ligands in PAMs. PAMs from three pigs were treated with 25 μg/ml poly(I·C) (TLR3), 100 ng/ml LPS (TLR4), 5 μg/ml CL097 (TLR7/8), 2.5 μg/ml CpG-ODN M362 (TLR9), or 20 U/ml hrIFN-β (IFN-β) or mock treated with media. (A) Relative quantitative RT-PCR was used to quantitate the mRNA levels for IFN-α transcript abundance. The relative level of mRNA was normalized to that of mock-infected cells for each time point. (B) The concentration of IL-1β protein in the supernatant was quantified by ELISA. Each experiment was performed in triplicate. Data are presented as means ± standard errors of the means (n = 5).

TABLE 1.

Effect of dose titrations for TLR3 and TLR4 agonists on IFN-α protein in PAMs

Dose titration for TLR agonist	IFN-α concn (pg/ml)
Poly(I·C) (μg/ml)
10	20,165 ± 528
25	32,550 ± 1,967
50	37,306 ± 178
100	41,037 ± 1,179
LPS (ng/ml)
100	<16.25
250	<16.25
500	<16.25
1,000	<16.25

IFN-α transcript induction in PAMs by TLR3 and TLR4 stimulation compared to PRRSV infection.

dsRNA is known to accumulate within infected cells, and it has the required physical and biological properties needed to induce antiviral responses and pathological inflammatory processes (14). Thus, it appeared important to examine whether PRRSV-induced activation of IFN-α in PAMs shares any characteristics with activation stimulated by a TLR3 agonist such as poly(I·C). Poly(I·C) triggered induction of IFN-α transcription in a time-dependent manner (P < 0.05; Fig. 1 and 2). IFN-α transcript levels induced by poly(I·C) (TLR3) peaked at 8 h poststimulation and were diminished by 24 h (Fig. 2A). IFN-α protein levels induced by poly(I·C) (TLR3) peaked at 16 h poststimulation (Fig. 2B). Surprisingly, LPS (TLR4) treatment failed to induce a significant increase of IFN-α transcript levels, and therefore, IFN-α protein responses to TLR4 stimulation were not measured. Infection of PAMs with PRRSV VR-2332 at an MOI of 10 resulted in a significant increase in IFN-α transcription at 16 h postinfection that diminished by 24 h. However, no significant increase in IFN-α protein expression was observed (Fig. 2B). Our results agree with previous studies, showing minimal, if any, IFN-α induction by PRRSV infection in vivo (26). Induction of IFN-α transcription by poly(I·C) was observed at 4 h poststimulation, whereas with PRRSV VR-2332 at an MOI of 10, IFN-α transcription was not significantly increased until 16 h postinfection. This delay may be consistent with the time required to generate dsRNA within the infected cell during the replication of the virus. Previous studies have shown that production of inflammatory mediators in response to viral infection can occur in the presence or absence of the multiplication of the pathogen (15). Loving et al. (22) found that IFN-α transcript in PAMs induced by poly(I·C) was expressed similar to our findings here at a >35- and >4-fold increase relative to that of mock-infected cells at 12 and 24 h, respectively. In addition, PAMs contained biologically active type I IFN protein, but Mx1 transcription in PAMs was not upregulated in response to poly(I·C), suggesting that PAMs regulate the type I IFN response not by limiting IFN-α/β production but by activating a phosphorylated STAT1 inhibitory response after poly(I·C) stimulation.

FIG. 2.

Comparison of the induction of IFN-α in PAMs by TLR3 and TLR4 ligands to PRRSV. PAMs from 13 pigs were treated with 25 μg/ml poly(I·C) (TLR3), 100 ng/ml LPS (TLR4), or PRRSV VR-2332 at an MOI of 10 or mock treated with media. (A) Relative quantitative RT-PCR was used to quantitate the mRNA levels for IFN-α transcript abundance. The relative level of mRNA was normalized to that of mock-infected cells for each time point. (B) The concentration of IFN-α protein in the supernatant was quantified by ELISA. Each experiment was performed in triplicate, and data are presented as means ± standard errors of the means (n = 13).

PRRSV infection alters IFN-α response to TLR3 ligand.

To determine whether a lack of induction or, alternatively, specific interference with induction of IFN-α genes was occurring, PAMs were exposed to both TLR3 and TLR4 ligands and PRRSV. The experiments were performed in two ways: first, by exposing cells to poly(I·C) or LPS and then infecting them with PRRSV 2 h later or, second, by infecting cells with PRRSV and then exposing them to poly(I·C) or LPS 6 h later. At 6 h (Fig. 3A) after TLR3 ligand treatment, a change of 121 (±77)-fold (n = 12) mean IFN-α transcript abundances was measured in PAMs infected posttreatment. PRRSV infection posttreatment with TLR4 ligand had a slight effect on transcription of IFN-α, but the magnitude of increase did not compare with those with the TLR3 ligand treatment (Fig. 3A). Infection with PRRSV suppressed induction of IFN-α transcription by TLR3 ligand: at 4 h posttreatment (Fig. 3B), mean IFN-α transcript abundance was measured as a change of 46 (±12)-fold (n = 12) in PAMs infected prior to treatment compared to a change of 121 (±77)-fold (n = 12) in PAMs infected posttreatment (Fig. 3A). In contrast, PRRSV infection did not significantly interfere with induction by TLR4 ligand (P > 0.05), which was measured as a change of only 6 (±2)-fold (n = 12) when infected pretreatment (Fig. 3A) compared to 14 (±10)-fold (n = 12) in PAMs infected posttreatment with TLR4 ligand (Fig. 3B). Infection with PRRSV suppressed IFN-α protein expression by TLR3 ligand before and after treatment. TLR4 ligand stimulation did not induce IFN-α protein expression (Fig. 3C).

FIG. 3.

PRRSV infection alters IFN-α response to TLR3 ligand. (A) PAMs from 12 pigs were treated with 25 μg/ml poly(I·C) (TLR3) and 100 ng/ml LPS (TLR4) for 2 h prior to infection with PRRSV VR-2332 at an MOI of 10 or mock treated with media. (B) PAMs from 12 pigs were infected with PRRSV VR-2332 at an MOI of 10 or mock treated with media 6 h prior to treatment with 25 μg/ml poly(I·C) (TLR3) or 100 ng/ml LPS (TLR4). Relative quantitative RT-PCR was used to quantitate the mRNA levels for IFN-α transcript abundance. The relative level of mRNA was normalized to that of mock-infected cells for each time point. (C) The concentration of IFN-α protein in the supernatant was quantified by ELISA. Each experiment was performed in triplicate, and data are presented as means ± standard errors of the means (n = 12). h p.i., hours postinfection.

Increased transcription of IFN-α in response to TLR3 and TLR4 ligands suppresses PRRSV infectivity.

PAMs were exposed to poly(I·C) or LPS for 2 h and then infected with PRRSV VR-2332. RNA and supernatant were collected at 6 h and 24 h posttreatment. A change of less than onefold in PRRSV titer between 6 and 24 h measured by the detection of PRRSV ORF7 transcript was observed for treatment with TLR3 or TLR4 ligands (data not shown). Replication of PRRSV was not markedly increased in the short time interval that was observed.

In this study, we conclude that infection of PAMs with PRRSV significantly reduced the ability of PAMs to respond to TLR3 ligation after as little as 2 h postinfection. Downregulation of IFN-α production facilitates PRRSV replication, since the elevation of type I IFN by in vivo stimulation or exogenous administration substantially reduces viral growth and enhances humoral immune responses (10, 23, 25, 28). Given that interference with TLR signaling is thought to play a role in establishing persistent infection by the HCV (20), future studies are needed to identify the mechanism by which PRRSV reduces TLR3 signaling in PAMs and whether this bears a relationship to persistent infection in the host.