Class A Scavenger Receptor–Mediated Double-Stranded RNA Internalization Is Independent of Innate Antiviral Signaling and Does Not Require Phosphatidylinositol 3-Kinase Activity

Double-stranded RNA, a replication by-product of all viruses (1), acts as a pathogen-associated molecular pattern (PAMP) and has long been implicated as a potent stimulus for both innate and adaptive antiviral immune responses against viral infection (2). The innate immune system recognizes PAMPs by the germline-encoded pattern recognition receptors (PRRs) present either on the cell surface or within distinct intracellular compartments, resulting in the induction of a type I IFN response (3). PRRs that recognize dsRNA include endosomal TLR3, cytoplasmic retinoic acid–inducible gene-I (RIG-I)–like receptors (RLRs) RIG-I, MDA-5, and LGP2, and the nucleotide oligomerization domain-like receptor Nalp3 (4). The inherent stability of duplex RNA structures within the host cell and when released into the extracellular milieu after virus-induced cell lysis facilitates induction of antiviral responses in neighboring, uninfected cells (5). Similar to the effects of viral dsRNA, mislocalized circulating endogenous dsRNA acts as a danger-associated molecular pattern and a potent stimulator of autoimmune diseases and conditions like preeclampsia (6–8).

During a viral infection or after the uptake of extracellular dsRNA by the cell, PRR-mediated dsRNA binding results in the recruitment of intracellular adaptor proteins. When dsRNA is endosomal, TLR3 is recruited from the endoplasmic reticulum to the endosome where it binds dsRNA and triggers intracellular signaling pathways through a Toll/IL-1R domain-containing adaptor inducing IFN-β (TRIF)–dependent mechanism (9). RIG-I and melanoma differentiation-associated gene 5 recognize dsRNA in the cytoplasm and signal through IFN-β promoter stimulator 1 (IPS-1), an adaptor molecule associated with the mitochondria (10) that is also known as mitochondrial antiviral signaling protein, virus-induced signaling adaptor, and CARD adaptor inducing IFN-β. These pathways lead to the activation of transcription factors, including IFN regulatory factor (IRF) 3 and IRF7, the induction of type I IFNs and IFN-stimulated genes (ISGs), and the establishment of an antiviral state (11). Of interest, alternative pathways of dsRNA-mediated IFN or ISG induction, independent of the classic PRRs and/or IRFs, also exist (12–14). Regardless of the intracellular dsRNA signaling pathways, the mechanism by which extracellular dsRNA funnels into these pathways remained elusive until we identified class A scavenger receptors (SR-As) as essential mediators of dsRNA internalization (15). Our finding was consistent with a previous report that SR-A–mediated uptake of extracellular dsRNA leads to inflammatory cytokine induction (16).

Although these findings explain how cells internalize extracellular dsRNA, it remains unknown whether SR-As contribute to intracellular signaling beyond ligand internalization, particularly because a partial antiviral response was observed in cells depleted for either TRIF or IPS-1 (15). Because there is precedence for the cytoplasmic tail of SR-As to associate with several cellular proteins (17, 18), it is possible that SR-As contribute to signaling. However, the use of deletion or site-directed mutagenesis to express SR-As lacking the cytoplasmic tail is problematic because multiple regions within the cytoplasmic tail are critical for SR-A surface localization and internalization (19, 20). In addition to SR-A cytoplasmic domains, studies have implicated PI3K activation in SR-A–mediated cell adhesion and surface localization (21, 22). PI3K exhibits both protein kinase and lipid kinase activities (23, 24) and phosphorylates membrane lipids that act as second messengers that regulate metabolism, proliferation, and survival (25, 26). Although other studies have implicated PI3K signaling in virus and polyinosinic:polycytidylic acid (poly IC) uptake and dsRNA-induced antiviral responses (21, 27–30), the role of PI3K in SR-A–mediated uptake of extracellular dsRNA remains obscure. A full understanding of dsRNA trafficking is critical to develop novel strategies to either use dsRNA as an adjuvant or to block the immune sequelae of extracellular dsRNA associated with infection and autoimmune disorders.

In this study, we examined whether SR-As simply serve as carrier molecules to deliver dsRNA to the TLR3 and RLR pathways or whether they can mediate dsRNA-induced antiviral responses independent of these pathways. We have also examined the requirement of PI3K activity for dsRNA uptake and subsequent antiviral signaling. Our results suggest that SR-As function to deliver extracellular dsRNA to intracellular sensors, but do not mediate signaling independent of TLR3 and the RLRs. Moreover, blocking PI3K activation did not affect SR-A–mediated dsRNA uptake but decreased downstream antiviral responses.

Murine embryonic fibroblasts (MEFs) were derived from wild type (WT) C57BL/6, TLR3/IPS-1^−/−, and TLR3/IPS-1/MyD88^−/− mice and were maintained in α-MEM supplemented with 10% FBS, 100 U ml⁻¹ penicillin, 100 μg ml⁻¹ streptomycin, and 2 mM l-glutamine. The A549 human lung carcinoma cell line (American Type Culture Collection) was maintained in α-MEM supplemented with 10% FBS. Vero cells (ATCC) were maintained in DMEM supplemented with 5% FBS. All cells were incubated at 37°C in a humidified 5% CO₂ incubator.

Long poly IC, 2–5 kb, was purchased from GE Healthcare (Buckinghamshire, U.K.) and short poly IC, 0.5–1.6 kb was purchased from Sigma. Doxycycline hyclate and oligomers were purchased from Sigma-Aldrich. Puromycin was purchased from Gibco (Life Technologies). The PI3K inhibitor wortmannin (Novex; Life Technologies) was reconstituted in DMSO. Human recombinant platelet-derived growth factor (PDGF-BB; Life Technologies) was reconstituted in PBS containing 0.1% BSA at a concentration of 50 ng/μl. Real-time PCR TaqMan probes for human ISG56, ISG15, and GAPDH were purchased from Applied Biosystems (Streetsville, ON, Canada).

dsRNA treatments were performed in serum-free OptiMEM media (Life Technologies) for specified time periods, with the first hour occurring in the presence of 50 μg/ml diethylaminoethyl (DEAE)-dextran (Pharmacia). DEAE-dextran is a cationic polymer that binds negatively charged nucleic acids and enables a closer association between the negatively charged cell membrane and the nucleic acid of interest (31). The use of DEAE-dextran does not bypass the requirement for SR-A as depletion of SR-As blocks the binding and entry of dsRNA even in the presence of DEAE-dextran (15). In all experiments, DEAE-dextran alone was used as a control to ensure that the polymer was not influencing subsequent cellular responses. After treatment, cells were washed, replaced with media, and harvested at the indicated times.

For whole cell extracts, cells were washed twice with ice-cold PBS and scraped into radioimmunoprecipitation assay buffer (10 mM Tris-HCl, pH 7.2, 150 mM NaCl, 0.1% SDS, 1.0% Triton X-100, 1% deoxycholate, 5 mM EDTA, 1 mM PMSF, 1 mM sodium orthovanadate, 2 mM DTT, and 1× protease inhibitor mixture [Sigma]). Lysates were incubated on ice for 10 min, passed through a 22-gauge needle, and centrifuged at 18,000 × g for 10 min at 4°C. Extracts were quantified using Bradford assay (Bio-Rad Laboratories).

Twenty-five micrograms of protein extracts was fractionated on 10% denaturing polyacrylamide gels, transferred onto nitrocellulose membranes (Millipore), and blocked in 5% skim milk or 5% BSA for anti–β-actin. Membranes were probed with 1:500 of the following primary Abs: anti-NS3 [hepatitis C virus (HCV)] (AdipoGen), anti-TRIF, and anti–mitochondrial antiviral signaling protein (Cell Signaling Technology). For the β-actin (Cell Signaling Technology) primary Ab, a 1:10,000 dilution was used. Secondary Abs conjugated to HRP were used and the signal was visualized using an ECL system (ECLplus kit; Millipore).

RNA was harvested using TRIzol reagent (Invitrogen). A total of 2.5 μg RNA was DNase-treated (DNA-free kit; Ambion) and one fifth of each sample was reverse transcribed with 0.2 ng random hexamer primer and 50 U Superscript II (Invitrogen) in a total reaction volume of 20 μl. Real-time quantitative PCR was performed in triplicate, in a total reaction volume of 25 μl, using Universal PCR Master Mix and gene-specific oligomers (Applied Biosystems). PCR was run in the ABI PRISM 7900HT Sequence Detection System using the Sequence Detector Software version 2.2 (Applied Biosystems). Data were analyzed using the ΔΔ cycle threshold method. Specifically, gene expression was normalized to the housekeeping gene (GAPDH) and expressed as fold change over the control group. Previous work validated that treatments do not affect GAPDH transcript levels (data not shown).

A cDNA fragment comprising nt 3418–5475 (aa 1027–1711) of the HCV H strain (genotype 1a) corresponding to the NS3-4A gene was amplified by PCR from pUHDNS3-4A (provided by Dr. Moradpour Darius) using the sense primer 5′-GGGGACAAGTTTGTACAAAAAAGCAGGCTCACCATGGCGCCCATCAC-3′ (the att site is underlined; the initiation codon is double underlined) and antisense primer 5′ GGGGACCACTTTGTACAAGAAAGCTGGGTTTTAGCACTCTTCCATCTCATCG-3′ (the att site is underlined; the ochre stop codon is double underlined). The amplification product was cloned into the piggyBac (PB) vector (pB-TET) using pDONR221 as an intermediate to yield the expression construct PB-TET-NS3-4A, as described in the Gateway recombination protocol by Life Technologies (32). The PB-TAG and PB-CAG-rtTA vectors were provided by Dr. Jonathan Draper (McMaster University). The transposase expression vector pCyL43 PBase was obtained from Sanger (http://www.sanger.ac.uk/technology/clonerequests).

A549 cells were grown to 80% confluence in a six-well dish and cotransfected with vector constructs PB-TET-NS3-4A (response plasmid, which bears the gene of interest, NS3-4A), PB-CAG-rtTA (plasmid that allows expression of a regulatory protein, rtTA), and pCyL43 (plasmid that encodes for transposase) in the ratio of 10:5:2 using Lipofectamine 2000 reagent (Invitrogen) according to the manufacturer’s protocol. In the presence of doxycycline, a tetracycline derivative, regulatory protein binds to tetracycline-response element and activates transcription of the NS3-4A complex. Transfected cells were selected using puromycin (Life Technologies), which was added to the medium 24 h posttransfection at a concentration of 1 μg/ml. Puromycin-resistant clones that appeared 10 d posttransfection were isolated and further screened for induction of NS3-4A expression. Clones were grown to 60% confluence and doxycycline was added at a concentration of 1 μg/ml. Forty-eight hours later, NS3-4A expression was measured in cell lysates using Western blot analysis.

To measure dsRNA-induced antiviral responses, we seeded cells in 12-well dishes and treated them with serial dilutions of poly IC for 6 h (MEFs) or overnight (A549). Supernatants were removed and cells were challenged with vesicular stomatitis virus (VSV)-GFP (MOI of 0.1 PFU/cell) in serum-free medium for 1 h. Viral inoculate was then removed and replaced with DMEM containing 1% methylcellulose. GFP fluorescence intensity was measured 24 h later on a Typhoon Trio (GE Healthcare) and quantified using Image Quant TL software. To measure production of type I IFN, we transferred supernatants from treated cells to Vero cells for 12 h, followed by challenge with VSV-GFP as indicated earlier.

A549 cells were grown to confluence and serum-starved for 12–16 h. Cells were then pretreated with wortmannin for 1 h at 37°C. To monitor PI3K activity, we assessed Akt phosphorylation by Western blot analysis with a phospho-specific anti-Akt (T³⁰⁸) rabbit mAb (Cell Signaling Technology). To ensure maintenance of inhibitor activity, we treated control samples with PDGF (50 ng/ml) 30 min before protein harvest.

A549 cells were seeded onto coverslips to reach 50% confluency after 24 h. After treatment, cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 3% goat serum, 3% BSA, and 0.02% Tween 20. Cells were incubated with 1:400 primary Ab [NF-κB (p65); Santa Cruz Biotechnology] followed by 1:500 anti-rabbit Alexa Fluor 488 secondary Ab (Invitrogen). All Ab dilutions were performed in 3% goat serum. Nuclei were stained with Hoechst stain (1:10,000 dilution) before mounting onto slides. All images were captured using a Leica DM-IRE2 inverted microscope and analyzed using Openlab software (Improvision).

dsRNA was labeled with Alexa Fluor 488 using the Ulysis nucleic acid labeling kit (Invitrogen). Excess labeling reagent was removed using Micro Biospin P-30 columns (Bio-Rad, Hercules, CA). A549 cells were seeded and grown to confluence in 96-well plates and serum-starved for 12–16 h. The next day, cells were pretreated with wortmannin for 1 h at 37°C before addition of fluorescently labeled dsRNA. After 1 h, unbound dsRNA was removed by washing cells with PBS. Total fluorescence was measured using the fluorescence plate reader (SpectraMax i3) before removal of unbound dsRNA. To quench the extracellular fluorescence signal, we subsequently added 0.025% trypan blue, thereby measuring the intracellular fluorescence. Results were reported as a percentage of total fluorescence relative to control cells.

Although the prototypic induction of IFN and ISGs involves TLR or RLR activation of IRF3, IFN and ISG induction independent of these cellular factors can occur (12–14, 33, 34). We previously observed that the antiviral response to extracellular dsRNA was partially inhibited when evaluated in either TRIF^−/− MEFs or IPS-1^−/− MEFs. Although it has been suggested that SR-As have signaling capacity (17, 18), it is unknown whether SR-As contribute to the partial antiviral response observed when either TRIF or IPS-1 is depleted. Hence, to investigate whether SR-As can mediate a dsRNA-induced response independent of TLR3 and RLRs, we performed depletion of both key downstream adaptors. The HCV serine protease NS3-4A inactivates both TLR3- and RIG-I–mediated signaling by cleaving TRIF and IPS-1 adaptor proteins without affecting additional signaling factors (35, 36).

To achieve stable, inducible NS3-4A expression, we used the Tet-On gene expression system based on the Escherichia coli tetracycline-resistance operon (37) where gene expression is under the control of a tetracycline response element. Human lung adenocarcinoma A549 cells were chosen because they are competent for IFN signaling and respond to extracellular dsRNA due to the expression of the SR-As SCARA3 and SCARA5 (data not shown). Two stable cell lines were created to evaluate the influence of NS3-4A expression on the degradation of TRIF and IPS-1 adaptor proteins: rtTA, expressing the regulatory protein; and NS3-4A, expressing rtTA and NS3-4A. rtTA control and NS3-4A cells were grown for 24 h in the presence or absence of doxycycline. Full-length NS3-4A protein (∼70 kDa) was detected in the presence of doxycycline in the NS3-4A cell line, but not the control rtTA cell line (Fig. 1A). In the absence of doxycycline, no NS3-4A was detected, confirming tight regulation of this inducible expression system.

Because we observed tight regulation of the NS3-4A protein in doxycycline-induced cells, we next examined the expression of TRIF and IPS-1 adaptor proteins to determine the cleavage efficiency of the NS3-4A protein complex. In the absence of doxycycline, TRIF was detected in both cell lines as a predominant 98-kDa band by Western blot analysis. However, doxycycline-induced NS3-4A expression resulted in the efficient cleavage of TRIF (Fig. 1B). The IPS-1 adaptor protein exists in three isoforms: cleaved (51–54 kDa), endogenous (57 kDa), and aggregated (75 kDa). Expression of NS3-4A resulted in cleavage of all three isoforms (Fig. 1C). These results clearly demonstrate that induction of the NS3-4A protein complex leads to cleavage and depletion of both TRIF and IPS-1 adaptor proteins.

To determine whether SR-As can mediate antiviral signaling in response to extracellular dsRNA independent of the classical TLR3 and RLR pathways, an antiviral assay was performed in both NS3-4A and rtTA cell lines in the presence and absence of doxycycline. Cells treated with increasing amounts of dsRNA were challenged with VSV-GFP and the resulting fluorescence, representing viral replication, was quantified 24 h postinfection. Control rtTA cells induce a complete antiviral response at poly IC concentrations of ≥10 nM, both in the presence and absence of doxycycline (Fig. 2A). In the absence of doxycycline, NS3-4A cells respond to poly IC in a concentration-dependent manner, with a full antiviral response observed with 100 nM (Fig. 2B). However, NS3-4A cells fail to induce an antiviral response regardless of poly IC concentration in the presence of doxycycline.

TLR3 and RLRs recognize foreign nucleic acid within endosomal and cytoplasmic compartments, respectively, initiating a signaling cascade that involves the induction of type 1 IFN through IRF3 and NF-κB. To confirm our antiviral data at the molecular level, we examined the activation of NF-κB by monitoring its translocation from the cytoplasm to the nucleus by immunofluorescence microscopy. In untreated (Mock) NS3-4A cells, NF-κB was detected uniformly throughout the cytoplasm in the presence and absence of doxycycline. Upon long or short poly IC treatment, NF-κB nuclear localization was detected in NS3-4A cells in the absence of doxycycline. However, in the presence of doxycycline, localization of NF-κB was inhibited (Fig. 3), confirming the lack of signaling by dsRNA upon cleavage of TRIF and IPS-1.

Further, we examined whether the impaired antiviral response in doxycycline-induced NS3-4A cells corresponds to a decrease in ISG induction using real-time RT-PCR. Transcript levels of ISG56 were measured in control rtTA and NS3-4A cells mock-treated or treated with dsRNA (100 nM) in the presence or absence of doxycycline. Furthermore, cellular supernatants were transferred to Vero cells, a monkey kidney epithelial cell line that can respond to but not make IFN (38, 39). Treated Vero cells were challenged with VSV-GFP, and the resulting fluorescence, representing viral replication, was quantified 24 h postinfection. Control rtTA cells showed a significant induction of ISG56 transcripts (Fig. 4A), as well as production of IFN (Fig. 4C), both in the presence or absence of doxycycline in response to poly IC relative to mock treatment. In contrast, NS3-4A cells failed to induce ISG56 transcripts (Fig. 4B) or produce IFN (Fig. 4D) in response to poly IC in the presence of doxycycline relative to the doxycycline-negative cells.

Unlike A549 cells that express only SCARA3 and SCARA5, all six SR-A transcripts are readily detected in C57BL/6 MEFs (15). Hence the capability of SR-As to mediate downstream antiviral signaling in response to dsRNA was also investigated in MEFs. Consistent with results from NS3-4A–expressed A549 cells, the antiviral response to extracellular dsRNA was completely inhibited in both TLR3/IPS-1^−/− (Fig. 5A) and TLR3/MyD88/IPS-1^−/− (Fig. 5B) MEFs relative to WT MEFs. Together, these data suggest that in human and mouse cells, although SR-As are essential for the uptake of extracellular dsRNA (15), they cannot mediate antiviral responses independent of TLR3/TRIF or IPS-1.

Based on previous observations from our laboratory and other studies, SR-A–mediated internalization of extracellular dsRNA occurs via endocytosis, and depletion of SR-As blocks the binding and internalization of extracellular dsRNA (15, 40). Corroborating our previous findings, we found that internalization of fluorescently labeled poly IC begins within 10 min of addition to culture medium (data not shown). Because PI3K has been implicated in regulating endocytosis, cell adhesion, and intracellular membrane trafficking in macrophages (41, 42), we quantified extracellular dsRNA entry in the absence of PI3K activity. Previous studies found that when used at concentrations of ≤100 nM, wortmannin irreversibly inhibits PI3K activity without affecting other lipid or protein kinases (43, 44). In preliminary experiments, cells were pretreated with wortmannin at increasing concentrations followed by stimulation with PDGF, a known inducer of PI3K activity; 50 nM wortmannin was found sufficient to block phosphorylation of Akt after PDGF stimulation (Fig. 6A). Thus, subsequent studies used 50 nM wortmannin to limit off-target effects. When entry of fluorescently labeled poly IC was quantified 1 h posttreatment, 16.2 and 7.9% of total fluorescence was cell associated and internalized, respectively (Fig. 6B). When wortmannin-treated cells were quantified for poly IC uptake, 16.7 and 8.0% of total fluorescence was cell associated and internalized, respectively (Fig. 6B), indicating that SR-A–mediated poly IC uptake does not require PI3K signaling. Similar results were observed with the reversible PI3K pathway inhibitor LY294002 (data not shown).

Once extracellular dsRNA is internalized, both endosomal TLR3 and cytosolic RLRs contribute to IFN-β and ISG induction (15). Studies have implicated PI3K activity in TLR3- and RIG-I–dependent activation of IRF3 in dsRNA-induced antiviral responses (29, 30). To assess the requirement of PI3K activity downstream of SR-As, we transfected control or wortmannin-treated cells with poly IC using Lipofectamine 2000 and determined ISG induction after 8 h. Because of its short half-life, wortmannin was replenished every hour for the duration of the experiment to ensure complete inhibition of PI3K. Consistent with the other reports (29, 30), a reduction in ISG56 and ISG15 transcript levels was observed in wortmannin-treated cells compared with control-treated cells (Fig. 6C). Together, these results suggest that PI3K activity is not required for SR-A–mediated internalization of extracellular dsRNA but is required for full antiviral signaling downstream of SR-As.

Extracellular dsRNA acts as a PAMP or danger-associated molecular pattern, depending on its origin, modulating innate immune responses in many cell types. It is now well appreciated that SR-As are essential components of extracellular dsRNA-induced cellular responses via their ligand recognition and internalization activities. However, the downstream intracellular signaling capacity of SR-As remains largely elusive. In this study, we demonstrate that SR-As bind, internalize, and deliver dsRNA to intracellular sensors, but do not modulate antiviral responses independent of TLRs and RLRs. Further, many studies have implicated PI3K activation in SR-A–mediated cell adhesion, and recently it was shown that PI3K inhibition resulted in blockage of poly IC uptake in macrophages (30). Our findings suggest that PI3K does not influence SR-A–mediated dsRNA uptake in fibroblast and epithelial cells but contributes to downstream signal transduction.

Whereas SR-As were initially characterized on phagocytic cells such as macrophages, we previously found that all cell types examined express at least one SR-A family member (15), which is not surprising because the innate immune system consists of many cell types in addition to macrophages. Indeed, nonphagocytic cells such as fibroblasts and epithelial cells rapidly produce type I IFN and ISGs in response to extracellular dsRNA. We previously noted that the antiviral response was partially inhibited in either TRIF^−/− or IPS-1^−/− MEFs, suggesting that both the TLR3 and the RLR pathways contribute to extracellular dsRNA-induced antiviral responses (15). Because SR-As use clathrin-mediated endocytosis, delivery of extracellular dsRNA to endosomal TLR3 is easy to conceptualize. It is unclear, however, how endosomal dsRNA stimulates the cytosolic RLR pathways. Moreover, we observed efficient ISG induction in the absence of IPS-1 or the downstream transcription factor IRF3 in response to long extracellular dsRNA molecules (12). These findings suggest that pathways distinct from those containing TLRs, RLRs, and IRF3 exist in cells. Because viruses have evolved immune-evasion strategies against key signaling proteins, it is not surprising that cells have evolved alternative pathways. Because there is precedence for association of the SR-A cytoplasmic tail with cellular proteins (17, 18), it is possible that SR-As may contribute to dsRNA signaling independent of the prototypic pathways.

However, antiviral signaling in response to extracellular dsRNA was completely abrogated when both TRIF and IPS-1 adaptor proteins were cleaved in human epithelial cells using the HCV NS3-4A protease, or removed through generation of knockout animals in mouse fibroblasts. Corroborating the antiviral data, cells deficient for TRIF and IPS-1 fail to induce the nuclear accumulation of NF-κB in response to both long and short poly IC. Further, the transcript levels of ISGs were also completely inhibited in these cells after stimulation with extracellular dsRNA. The simplest interpretation of our data is that SR-As are involved in ligand internalization but have no signaling capacity. Indeed, in the absence of TLR3 or RLR signaling cascades, we fail to detect NF-κB activation, ISG and IFN induction, or the ensuing antiviral response. However, we cannot rule out the possibility that the cytoplasmic tail of SR-As may contribute to TLR3 or RLR signaling. The requirement of specific domains within the cytoplasmic tail for proper cellular localization and ligand internalization (20) precludes the use of tail-deficient mutants.

Moreover, although lipid-based transfection of dsRNA bypasses the requirement of SR-As for ligand internalization, it is not possible to directly compare ISG induction after deposition of dsRNA into the medium (SR-A–mediated entry) versus dsRNA transfection (SR-A–independent entry), because the efficiency of internalization and subsequent ligand trafficking are likely very different between these two approaches. Consistent with this notion, although we fail to observe cellular toxicity with increased concentrations of dsRNA added to culture medium, we rapidly induce cell death with increased amounts of transfected dsRNA, consistent with our findings that membrane perturbation with lipid-based particles is sufficient to trigger cell signaling events (45).

Although a recent study implicated PI3K activation in SR-A–mediated cell adhesion, but not ligand internalization in macrophages (21), a different group showed that ligand internalization via Mac-1 (a surface integrin receptor) required PI3K signaling in macrophages (30). In nonphagocytic cells, it is unknown whether PI3K activity is required for the binding and uptake of extracellular dsRNA. In A549 cells, PI3K activity was not required for SR-A–mediated dsRNA binding or internalization. Whereas most SR-A studies in macrophages focus on macrophage receptor for collagenous structure, A549 cells express SCARA 3 and SCARA 5, which are not as well studied. Along with previous findings, our study suggests that SR-As internalize ligands in a PI3K-independent manner, regardless of cell type or family member. In addition to ligand entry, PI3K activity has been implicated in dsRNA-mediated signaling through IRF3 (27, 29, 46). Indeed, wortmannin partially inhibited antiviral signaling upon transfection of poly IC into cells, a process that bypasses SR-A–mediated entry. This observation suggests that reduced antiviral signaling mediated by PI3K inhibitors occurs downstream of dsRNA entry. Although the PI3K pathway does not play a significant role in SR-A–mediated dsRNA uptake, it remains to be determined whether other cofactors are required for extracellular dsRNA entry.

In summary, this study reveals that SR-A–mediated dsRNA internalization is independent from downstream signaling and that in the absence of TLR3 and RLR signaling, SR-As do not contribute to antiviral responses. Moreover, we found that PI3K signaling contributes to ISG induction downstream of dsRNA internalization. Because derivatives of poly IC are being developed as antivirals and adjuvants for clinical use despite the involvement of circulating dsRNA in autoimmunity (47, 48), it is crucial to understand how these molecules are recognized by the immune system. A clear understanding of how host cells recognize and respond to extracellular dsRNA is paramount for the development of effective therapeutics.

We thank Drs. John Draper and Moradpour Darius for reagents.

The authors have no financial conflicts of interest.