An accessory to the 'Trinity': SR-As are essential pathogen sensors of extracellular dsRNA, mediating entry and leading to subsequent type I IFN responses

Abstract

Extracellular RNA is becoming increasingly recognized as a signaling molecule. Virally derived double stranded (ds)RNA released into the extracellular space during virus induced cell lysis acts as a powerful inducer of classical type I interferon (IFN) responses; however, the receptor that mediates this response has not been identified. Class A scavenger receptors (SR-As) are likely candidates due to their cell surface expression and ability to bind nucleic acids. In this study, we investigated a possible role for SR-As in mediating type I IFN responses induced by extracellular dsRNA in fibroblasts, a predominant producer of IFNbeta. Fibroblasts were found to express functional SR-As, even SR-A species thought to be macrophage specific. SR-A specific competitive ligands significantly blocked extracellular dsRNA binding, entry and subsequent interferon stimulated gene (ISG) induction. Candidate SR-As were systematically investigated using RNAi and the most dramatic inhibition in responses was observed when all candidate SR-As were knocked down in unison. Partial inhibition of dsRNA induced antiviral responses was observed in vivo in SR-AI/II(-/-) mice compared with WT controls. The role of SR-As in mediating extracellular dsRNA entry and subsequent induced antiviral responses was observed in both murine and human fibroblasts. SR-As appear to function as 'carriers', facilitating dsRNA entry and delivery to the established dsRNA sensing receptors, specifically TLR3, RIGI and MDA-5. Identifying SR-As as gatekeepers of the cell, mediating innate antiviral responses, represents a novel function for this receptor family and provides insight into how cells recognize danger signals associated with lytic virus infections. Furthermore, the implications of a cell surface receptor capable of recognizing extracellular RNA may exceed beyond viral immunity to mediating other important innate immune functions.

Figure 1. Primary murine fibroblasts express functional SR-As.

(A) SR-A transcripts were detected using RT-PCR and primers specific to SR-AI (lane I), SR-AII (II), MARCO (M), SCARA3 (3), SCARA4 (4), and SCARA5 (5) in murine embryonic fibroblasts (MEFs) derived from C57Bl/6 or balb-c mice, the macrophage-like cell line RAW 264.7, and primary lung fibroblasts and splenocytes derived from C57Bl/6 mice. GAPDH was used as an internal control for all cell types (G). (B) Cell associated AcLDL was observed by live-cell fluorescence microscopy in MEFs derived from C57Bl/6 mice treated with Alexafluor 488 labeled AcLDL for 1h (2.5 µg/mL), in the presence or absence of SR-A competitive ligands (fucoidin, DxSO₄), their corresponding non-competitive counterparts (fetuin, ChSO₄) or nucleic acids (poly IC, poly dA:dT) all at 100 µg/mL. Magnification 200X. (C) AcLDL entry was measured using a fluorescence plate reader assay. Cells were treated with Alexafluor 488 labeled AcLDL (2.5µg/ mL) for 1h in the presence of poly IC (pIC), fucoidin (fuc), fetuin (fet), DxSO₄, ChSO₄ or poly dA:dT (100 µg/mL). Cells treated with AcLDL alone were set at 100% fluorescence. These data include three independent experiments ± SEM. Statistical analysis was performed by a one-way ANOVA with Tukey post test (* p<0.05, *** p<0.001).

Figure 2. SR-As mediate viral dsRNA binding, entry and resulting ISG induction in MEFs using a mechanism dependent on clathrin-mediated endocytosis.

(A) Cellular binding of viral dsRNA was observed by fluorescence microscopy in MEFs derived from C57Bl/6 mice treated with either 200 bp or 1000 bp Alexafluor 488 labeled viral dsRNA (v200 or v1000 respectively, both at 1 µg/mL) for 1h, in the presence or absence of fucoidin or fetuin (both at 100 µg/mL). Cells were fixed and nuclei counterstained with Hoechst 33258. Magnification 400X. (B) Fluorescently labeled dsRNA entry was quantified using cells treated with Alexafluor 488 labeled v200 or v1000 (both at 1 µg/mL) for 1h in the presence of fucoidin or fetuin (both at 100 µg/mL). Cells treated with dsRNA alone (dsRNA) were set at 100% fluorescence. These data include three independent experiments ± SEM. Statistical analysis was performed by a one-way ANOVA with Tukey post test (*** p<0.001) (C) MEFs were treated with v1000 (1 µg/mL) for 2h in the presence of endocytosis pathway inhibitors, chlorpromazine (CP), cytochalasin D (cyto D) and bafilomycin A1 (Baf A). IP10 transcript levels were measured using real time PCR and reported as a fold change difference from mock treated cells (ctrl), whose fold change was set to 1. These data are the average of three independent experiments ± SEM. Statistical analysis was performed by one-way ANOVA with a Dunnett's post test using v1000 as the control comparison, * p<0.05 (D) ISG15, IRF-7 and ISG56 transcript levels were measured by real time PCR in C57Bl/6 MEFs treated with viral dsRNA (v1000, 0.5 µg/mL) for 4 h in the presence or absence of dextran sulfate (DxSO₄), fucoidin (Fuc.) or fetuin (Fet.). These data are the average of three independent experiments ± SEM. Statistical analysis was performed by one-way ANOVA and a Dunnett's post test, comparing all treatments to v1000 alone, ** p<0.01.

Figure 3. dsRNA binding is dependent on all candidate SR-A family members.

DsRNA binding, entry and subsequent induction of ISGs was measured in balb-c derived MEFs treated with media alone (OM), Dharmafect alone (DF), SR-AI/II specific siRNA (I/II) or a combination of SR-AI/II, SCARA3, SCARA4 and SCARA5 siRNA oligomers (-I/II,3,4,5) for 24h. (A) Binding of dsRNA was measured in MEFs treated with SR-A siRNA for 24h followed by Alexafluor 488 labeled v1000 (1 µg/mL) for 1h. Nuclei were counterstained with Hoechst 33258. Magnification 400X. (B) Entry of Alexafluor 488 labeled v1000 was measured by fluorescence plate reader, in balb-c MEFs treated with siRNA for 24h followed by 1h treatment with fluorescently labeled v1000 (1 µg/mL). Results are reported as a % of DF alone (set to 100%). These data include three independent experiments and are reported as an average ± SEM. Statistical analysis was performed by a one-way ANOVA with a Dunnett's post test, with DF being the control comparison (** p<0.01). (C) Levels of ISG56 and IP10 transcript expression were measured by real time PCR in balb-c MEFs treated with SR-A siRNA for 24h followed by stimulation with 1 µg/mL v1000 for 6 h. Results are reported as a fold change difference from DF treated cells (DF), whose fold change was set to 1. These data include three independent experiments and are reported as an average ± SEM. Statistical analysis was performed by a one-way ANOVA with a Dunnett's post test, with DF being the control comparison (* p<0.05). (D) SR-A transcript levels were measured by real time PCR in balb-c MEFs treated with SR-A siRNA for 24h. Results are reported as a fold change difference from DF treated cells (DF), whose fold change was set to 1. These data include three independent experiments and are reported as an average ± SEM. Statistical analysis was performed by a one-way ANOVA with a tukey's post hoc test (* p<0.05, ** p<0.01).

Figure 4. SR-As are involved in dsRNA induced antiviral responses in vivo.

(A) SR-A expression was measured at the transcript level in both wild type (WT) and SR-AI/II^-/- mouse whole lungs using RT-PCR and primers specific to SR-AI (lane I), SR-AII (II), MARCO (M), SCARA3 (3), SCARA4 (4), SCARA5 (5) and GAPDH (G). (B) Poly IC treated mice were sacrificed 12h post-treatment and type I IFN bioactivity in the BALF was determined by antiviral assay. Data is presented as % VSV replication at the indicated BALF dilutions. The absence of VSV replication indicates the presence of type I IFN in BALF. Statistical analysis was performed using a one-way ANOVA with a tukey's post hoc test (* p<0.05). (C) IP10, ISG56 and ISG15 transcript levels in the lung of poly IC treated WT and SR-AI/II^-/- mice were measured 12h post treatment. Results are reported as a fold change comparison with PBS treated WT or SR-AI/II^-/- mice. Statistical analysis was performed using a t test; *** p<0.001, * p<0.05. Data show one representative experiment and are reported as means ± SEM, n = 4.

Figure 5. DsRNA induced antiviral responses are mediated by TLR3 and the RLRs.

MEFs derived from (A) wild type C57Bl/6 (WT), TLR3^-/-, TRIF^-/- and (B) WT, RIGI^-/-, MDA5^-/- and IPS-1^-/- mice were treated with serially diluted nM concentrations of poly IC for 6 h. (C) To compare length effects, WT, RIGI^-/- and MDA5^-/- MEFs were treated with serially diluted nM concentrations of in vitro transcribed dsRNA (v200). Cells were then challenged with VSV-GFP (MOI = 0.1) and GFP fluorescence intensity was measured at 24h pi. Data is reported for each dsRNA concentration tested as a % of VSV replication in mock treated cells. These data include three independent experiments and are reported as an average ± SEM.

Figure 6. Human fibroblasts express SR-As that mediate dsRNA binding via the collagenous domain.

(A) SR-A transcripts were detected in primary human fibroblasts (Hel) using RT-PCR and primers specific to SR-AI (lane I), SR-AII (II), MARCO (M), SCARA3 variant 1(3v1), SCARA3 variant 2 (3v2), SCARA4 (4), and SCARA5 (5). SR-A transcripts detected from immortalized human embryonic kidney cells (293) and primary human peripheral blood mononuclear cells (hPBMCs) were included for comparison purposes. (B) SR-A expression at the protein level was measured in human fibroblasts by western blot analysis using a whole cell extract (WCE). A single band ∼70kDa in size was detected. (C) Alexafluor 488 labelled v1000 (1 µg/mL) binding was observed in HEL cells by fluorescence microscopy with or without pre-treatment with the SR-A competitive ligand fucoidin or fetuin, its non-competitive counterpart (both at 100 µg/mL). Also, v1000 binding was monitored in the presence of the anti-human SRA antibody (αSRA) and a normal goat serum (ngs) control (both diluted at 1∶50).

Figure 7. Proposed model of SR-A-mediated antiviral activity.

Lytic viral infections lead to dsRNA in the extracellular environment. (1) Surface expressed SR-As bind extracellular dsRNA and facilitate entry via clathrin-mediated endocytosis. (2) Endosomal TLR3 binds dsRNA and induces an antiviral response via TRIF. (3) dsRNA escapes the endosome and is detected in the cytoplasm by the RLRs RIG-I and MDA5, which induce an antiviral response via IPS-1. (4) While the carrier function of SR-As is likely the dominant mode of action, the direct contribution of SR-As to antiviral immunity by binding to TRIF and IPS-1 or by inducing independent, non-classical antiviral responses cannot be ruled out.