Hepatitis C Virus Alters Macrophage Cholesterol Metabolism Through Interaction with Scavenger Receptors

Cells, virus, reagents

THP-1 human monocytic cell line was purchased from ATCC and maintained in RPMI-1640 Media (Gibco, cat. no. 11875093) supplemented with 0.05 mM 2-mercaptoethanol (Sigma, cat. no. M7522), 10% fetal bovine serum (FBS) (HyClone/Cytiva, cat. no. SH30071.03), 100 U/mL penicillin/streptomycin (Corning, cat. no. 30001CI), and 2 mM L-glutamine (Gibco, cat. no. 25030081). Cells were differentiated with 100 nM phorbol 12-myristate 13-acetate (PMA; LC Labs) for 72 h. Macrophages were rested 24 h without PMA before use in experiments. Unless otherwise noted, THP-1 macrophage-like cells were differentiated and rested as described above for all experiments, then exposed to 0.1 multiplicity of infection (MOI) HCV or mock supernatant for 48 h.

Human hepatoma cell line Huh7.5.1 was maintained in Dulbecco's modified Eagle's medium (Corning, cat. no. 10-017-CV) with 10% FBS (HyClone/Cytiva, cat. no. SH30071.03), 100 U/mL penicillin/streptomycin (Corning, cat. no. 30001CI), 2 mM L-glutamine (Gibco, cat. no. 25030081), and 1% nonessential amino acids (Gibco, cat. no. 11140050).

Primary human MDMs were generated from peripheral blood mononuclear cells (PBMCs). Deidentified donor blood was obtained from a commercial source (Virginia Blood Services). PBMCs were isolated using Ficoll in SepMate separation tubes (StemCell) followed by enrichment of monocytes using the EasySep Human Monocyte Enrichment Kit without CD16 depletion (StemCell). Monocytes were further enriched by plastic adherence for 2–4 h, followed by removal of nonadherent cells and culture in X-VIVO 15 (Invitrogen) supplemented with M-CSF (PeproTech) for a total of 14 days with half-media changes every 3–4 days to obtain mature resting macrophages.

All experimental protocols were approved by the Institutional Review Board of the University of Virginia (IRB no. 8233). All Methods were carried out in accordance with relevant guidelines and regulations required by the National Institutes of Health (NIH) and the University of Virginia.

JFH-1 HCV was kindly provided by Takaji Wakita (69). Virus stocks were prepared by infecting Huh7.5.1 cells. Virus was collected following 6–8 days of infection and tittered by Focus-Forming assay on Huh7.5.1 cells exposed to virus for 3 days. Cells were fixed and stained with anti-Core antigen (Thermo Fisher) followed by APC-labeled anti-mouse IgG (Life Technologies) (48). Virus stocks were cleared of debris by centrifugation, aliquoted, and stored at −80°C until use. Mock supernatants were prepared from uninfected cells maintained under identical conditions.

siRNA reagents, including control and scavenger receptor class B type 1 (SR-B1) siRNAs (sc-44752), transfection reagent, and media were from Santa Cruz.

Real-time reverse transcriptase quantitative polymerase chain reaction

Total RNA was extracted using RNA-Bee as per the manufacturer's instructions (Tel-Test) and converted to first-strand cDNA with the High-Capacity cDNA Reverse Transcription Kit (Life Technologies). Real-time reverse transcriptase quantitative polymerase chain reaction (RT-qPCR) was run on StepOnePlus System with SybrGreen (Life Technologies). Primers were HCV inNeg F1 ACTGTCTTCACGCAGAAAGCGCC and HCV inNeg R1 CAAGCGCCCTATCAGGCAGTACC (modified for Genotype 2A) for HCV (JFH-1) 5′ untranslated region (5′-UTR) and qNS3-2aF CTGCCACCCTGGGGTTTGGG and qNS3-2aR GCAGCCCCCATCGGCGAGAA for HCV nonstructural protein 3 transcript region (NS3) (19,52).

Target gene expression analysis with SybrGreen chemistry was normalized to housekeeping gene hypoxanthine phosphoribosyltransferase 1 (HPRT) with primers HPRT-F CATTATGCTGAGGATTTGGAAAGG and HPRT-R AGACGTTCAGTCCTGTCCATAA. TaqMan gene expression assays for ABCA1, CD81, Claudin-1 (CLDN1), dendritic cell-specific intercellular adhesion molecule-3-grabbing nonintegrin (DC-SIGN) (CD209), HPRT, LXRα (NR1H3), SR-B1 (SCARB1), and SREBP-1c (SREBF1) were obtained from (Thermo Fisher).

Western blot

Cell lysates were prepared in lysis buffer (HEPES pH:7.4, NaCl, EDTA, DTT, Triton X-100). Proteins were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) with TGX precast gels (Bio-Rad) and were transferred to polyvinylidene difluoride membranes (Bio-Rad) using a semidry transfer cell (Bio-Rad). Membranes were blocked with milk in TBS-T, followed by incubation in primary antibodies and incubation with secondary antibodies for development. Blots were developed using enhanced chemiluminescent (ECL) detection.

Antibodies used for western blotting were: Hep C cAg (C7-50; Santa Cruz), actin (Santa Cruz), HCV E2 (Santa Cruz), HCV NS3 (Abcam), SR-B1 (GeneTex), DC-SIGN (Cell Signaling), CD81 (Systems Biosciences), CLDN1 (ProSci) LXRα (R&D Systems), LXRβ (R&D Systems), ABCA1 (Novus), Vinculin (Cell Signaling), and SREBP-1 (Santa Cruz).

Flow cytometry

For analysis of surface expression, THP-1 macrophage-like cells were differentiated as above. Adherent macrophages were washed and incubated in Cell Dissociation Buffer, enzyme free (Thermo Fisher/Gibco) for 20 min on ice with gentle disruption with a vacuum pipette every 5 min. Cells were collected, washed, and resuspended in TruStain FcX Fc Receptor Blocking Solution (BioLegend) in fluorescence activated cell sorting (FACS) buffer for 15 min. Fc Blocking reagent was removed and cells were incubated in primary antibodies for 20 min. Cells were washed twice in FACS buffer and analyzed on a CytoFlex Flow Cytometer (Beckman Coulter). Antibodies used were PerCP CD45 (Clone 2D1; Tonbo), PE CD36L1 (SCARB1, SR-B1, Clone m1B9; BioLegend), FITC CD81 (TAPA-1, Clone 5A6; BioLegend), and CD209 (DC-SIGN)-APC-Vio770 (Clone DCN47.5; Miltenyi).

Oil Red O staining

Cells in tissue culture plates were stained in 1% Oil Red O (Sigma) in 60% isopropanol for 1 h. Staining was differentiated in three changes of 60% isopropanol followed by a running water wash. Slides were counterstained in Hematoxylin, washed, and blued in running water and imaged in water before extraction. Neutral Lipids (Oil Red O) were extracted using 100% isopropanol. A separate aliquot was also used for determination of cell protein. Total cell protein was similar between mock- (1.031 mg/mL) and HCV-exposed (1.027 mg/mL) cell groups and results are presented non-normalized as Raw Absorbance values.

Cholesterol content analysis

Free and esterified cholesterol mass analysis was conducted according to published methods (38). Media were removed, and cells were washed twice. Lipids were extracted in hexane/isopropanol (3:2) for 1 h with mixing. Lipid extracts were collected, and cells were reextracted with hexane/isopropanol. Pooled extracts were dried under nitrogen and lipid films resuspended in reaction buffer (0.1 M potassium phosphate, pH 7.4, 50 mM NaCl, 5 mM cholic acid, and 0.1% Triton X-100) for 5 h with shaking.

Aliquots were assayed for free cholesterol and cholesterol ester using Amplex Red cholesterol assay (Molecular Probes) in the presence or absence of cholesterol esterase, respectively. Free cholesterol was calculated as difference of total cholesterol and esterified cholesterol for matched samples. Cells were dissolved in NaOH/SDS, and an aliquot used for protein determination by BCA protein assay. Cholesterol mass values were normalized to cellular protein.

Cholesterol staining

Cells were grown in a chamber slide (Lab-Tek II, Nunc). Media were removed and cells were washed twice before fixing and staining with the Cholesterol Cell-Based Detection Assay Kit (Cayman) according to the manufacturer's recommendations. Cholesterol (Filipin III)-stained cells were imaged with an LSM710 laser-scanning confocal microscope (Zeiss). Nuclei were counterstained with DRAQ5 (BioStatus).

Blocking antibody experiments

For blockade of E2 envelope interaction, HCV cell-free supernatants were incubated with anti-E2 (1:30) for 1 h before exposure to cells and antibody maintained throughout exposure. Cells were treated with blocking antibodies to target HCV host receptors for 1 h before exposure to HCV and maintained throughout exposure. Blocking antibodies used included: anti- E2 (GeneTex, cat. no. GTX103353), CD81 (Santa Cruz sc-23962, 1:50), SR-BI (GeneTex, EP1556Y GTX61567, 1:50), and DC-SIGN (R&DMAB161, 10 μg/mL).

Statistical analyses

Statistical analyses were performed using GraphPad Prism software. Samples for qPCR and metabolic assessments were measured in triplicate (three biological replicates, n = 3, each sample assayed in technical duplicate) and results presented as mean + standard error. Experiments presented graphically were performed a minimum of two times with biological triplicate replicates. Measurements of biological triplicates from a single representative experiment are presented. Individual measures for all graphs in main Figures are included as Supplementary Fig. S9. t-Tests at 95% Confidence Interval (CI) were used to assess differences in the means between two grouped variables. One-way analysis of variance (ANOVA) with post hoc Tukey's multiple comparison analyses at 95% CI was used to assess differences in means between multiple grouped variables. *^,# p < 0.05, **^,## p < 0.01, ***^,### p < 0.001.

Software use

All graphs were prepared using statistical software Prism (GraphPad). Photographic films were scanned with an Epson professional desktop scanner and Epson Scan software. Flow Cytometry data were collected on a CytoFlex flow cytometer (Beckman Coulter) and processed with FlowJo Software (FlowJo). Light microscopic images were collected using instrument-associated software (Leica Microsystems).

Spectrophotometric Absorbance readings were collected using a SpectraMax plate reader and Softmax software. Confocal microscopy images were processed with instrument-specific software (Zen, Zeiss Microscopy) and further processed for publication using ImageJ (NIH). Real-time quantitative PCR data were collected and analyzed with instrument-specific software (ABI) and statistical comparisons used in analysis of RT-qPCR were conducted with Prism software (GraphPad).

HCV RNA accumulates and persists in macrophages without viral protein expression

Induction of innate immune mediators, including CCL5, interferons, IL-1β, IL-6, IL-8, IL-10 (TGF-β), and tumor necrosis factor alpha in macrophages upon exposure to HCV has been reported along with a gradual decline of viral RNA between 24 and 72 h (36,49,57 –59,61,73). To characterize viral RNA dynamics immediately upon virus exposure, we determined HCV RNA kinetics in macrophages differentiated from THP-1 promonocytic cells within 24 h of HCV exposure.

THP-1 macrophage-like cells were exposed to clarified supernatants from HCV-infected Huh7.5.1 hepatoma cells and viral RNA and protein were analyzed in macrophage lysates. Viral genomic RNA was detectable by reverse transcriptase quantitative PCR (RT-qPCR) of the 5′-UTR and NS3 in macrophages within 1 h of HCV (0.1 MOI–MOI) exposure and gradually decreased throughout 6 h (Fig. 1A).

OPEN IN VIEWER

FIG. 1.

(A) THP-1 macrophages were exposed to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for the indicated times. Viral RNA was quantified by RT-qPCR using primers for the 5′-UTR or the NS3. Fold change in specific viral RNA relative to HPRT is presented by setting mock-infected supernatant as 1. (B) Expression of core protein was assessed from cell lysates and supernatants of THP-1 macrophages exposed to 1 MOI HCV for the indicated times. Actin was assessed as a loading control for cell lysates and ApoE for supernatants. (C) THP-1 macrophages exposed to 0.1 MOI HCV for 1 h (left) or 4 h (right) were treated with trypsin for 5 min followed by extensive washing. Viral RNA was quantified by RT-qPCR using either 5′-UTR or NS3 primers. Fold change in viral RNA Relative to HPRT is presented as a percentage of the non-trypsin-treated sample set as 100%. (D) THP-1 macrophages were exposed to 0.1 MOI HCV for 4 h in the presence or absence of chloroquine (50 μM), BafA1 (2.5 μM), or vehicle (DMSO). Viral RNA was quantified by RT-qPCR using 5′-UTR primers. Fold Change in specific viral RNA relative to HPRT is presented as a percentage of the DMSO-treated sample set as 100%. Experiments presented graphically were performed a minimum of 2 times with biological triplicate replicates. Measurements of biological triplicates from a single representative experiment are presented. One-way ANOVA at 95% CI was used to assess differences in means between multiple grouped variables (A). t-Tests at 95% CI were used to assess differences in the means between two grouped variables (C–D). *^,# p < 0.05, **^,## p < 0.01. ANOVA, analysis of variance; ApoE, apolipoprotein E; BafA1, bafilomycin A1; CI, confidence interval; DMSO, dimethyl sulfoxide; HPRT, hypoxanthine phosphoribosyltransferase 1; MOI, multiplicity of infection; NS3, nonstructural protein 3 region; RT-qPCR, real-time reverse transcriptase quantitative polymerase chain reaction; 5′-UTR, 5′ untranslated region.

When we extended our studies to examine the presence of viral genomic RNA during a 3-day exposure to 0.1 MOI HCV in THP-1 macrophage-like cells, viral RNA above background was detectable throughout time course, gradually declined throughout 24 h, and stabilized at a low level by 72 h (Supplementary Fig. S1A). In contrast, viral protein (NS3 or core) was not detectable at any time with this dose of HCV (Supplementary Fig. S1B). Increasing viral inocula (1 MOI) allowed detection of low levels of viral protein in macrophages, potentially derived from inocula, with a transient increase in viral protein at 6 h postexposure (Fig. 1B) (36). These results suggest that virus exposure in macrophages leads to accumulation and maintenance of viral genomic RNA in these cells that persists at least for 72 h.

To compare the level of HCV RNA detected in macrophages upon HCV exposure with HCV infection in hepatocytes, we examined viral genomic RNA and protein during HCV infection in hepatoma cells. In Huh7.5.1 cells exposed to HCV, viral genome was detectable within 24 h, increased rapidly between 24 and 72 h, and peaked between 96 and 120 h postinfection before a precipitous drop with marked cytopathic effect at 144 h (Supplementary Fig. S1C).

Viral life cycle is intact in this system as negative-strand RNA replication intermediates were detectable from 72 h until peaking at 120 h postinfection (Supplementary Fig. S1D). Viral polyprotein (high mw reactivity to NS3 antibody) was detectable within 24 h and polyprotein processing (multiple bands of lower mw reactivity to NS3 antibody) was evident by 48 h, peaking with the presence of viral genomes between 72 and 120 h postinfection (Supplementary Fig. S1E).

Structural (Core), nonstructural (NS3), and Envelope (E2) proteins were all detectable demonstrating intact and complete nature of the viral replication cycle within Huh7.5.1 cells (Supplementary Fig. S1E). This contrasts with the viral replication cycle in THP-1 macrophage-like cells, which is either abortive or significantly delayed under the conditions of these experiments.

Based on common features of the expression of surface receptor for HCV entry in macrophages and hepatocytes, it is possible that macrophages may become productively infected. However, viral genomic RNA did not increase significantly throughout 3 days of exposure indicating THP-1 macrophage-like cells do not become productively infected.

Envelope-independent internalization of viral RNA in macrophages and degradation in endo/lysosomes

Viral RNAs can nonspecifically attach to the surface of macrophages. Thus, the presence of viral RNA in cell homogenates does not directly address whether these RNAs are entering cells. In addition, a recent study demonstrated that HCV can attach to the surface of both primary monocytes and macrophages but only macrophages internalize the virus (73). To measure viral binding and uptake, macrophages (Fig. 1C) or monocytes (Supplementary Fig. S2A) were incubated with HCV followed by treatment with trypsin to remove viral particles that remained attached to the cell surface as in Zhang et al. (73). Bound or internalized virus was quantified by qPCR for viral genomic sequences in the absence or presence of trypsin treatment, respectively.

The quantity of virus bound (-trypsin) was normalized to 1 and uptake (+ trypsin) presented as a percentage of total attached HCV. Based on our studies indicating detection of highest viral RNA at 1 h postexposure, we examined trypsin sensitivity at both 1 and 4 h as in Zhang et al. (73). At 1 h postexposure ∼80% of viral RNA in macrophages was sensitive to trypsinization, indicating that virus is surface associated, reflecting binding (Fig. 1C). By 4 h postexposure ∼70% of viral RNA in macrophages was insensitive to trypsinization, indicating virus uptake and internalization (Fig. 1C) (73).

To test the possibility that a difference in virus uptake between monocytes and macrophages (73) might reflect upregulation of a factor required for virus recognition upon differentiation, we analyzed virus uptake in monocytes exposed to virus. At 4 h postexposure monocytes internalized considerably less virus than macrophages as a percentage of the amount of trypsin-insensitive RNA bound to cells as well as the ratio of trypsin-insensitive viral RNA to housekeeping gene HPRT (Supplementary Fig. S2A). Blockade of cell surface entry factor CD81 is incapable of preventing HCV protein or RNA uptake in macrophages but can prevent infection and uptake in other cell types (17,35,36,49,64). Studies from other groups have been inconsistent on the requirement for viral E2 protein in HCV uptake and cytokine induction.

Thus, we examined whether preincubating virus with an E2-blocking antibody previously demonstrated to block HCV infection in hepatocytes (confirmed in Supplementary Fig. S2B) could prevent virus binding or uptake in macrophages (12). Preincubating virus with E2-blocking antibody failed to prevent HCV binding or uptake in macrophages (Supplementary Fig. S2C). These results suggest that unlike results in DCs, receptor-mediated entry of HCV is not responsible for uptake in macrophages. It also implicates an alternative process distinct from the envelope-mediated process that leads to productive infection in hepatocytes (35).

Envelope–receptor interactions lead to fusion to the cellular membrane and release of viral genomes into the cytoplasm where they are degraded by cellular factors or used as templates for viral replication (28). We next examined whether inhibition of endosomal acidification with chloroquine or lysosomal acidification with the v-ATPase inhibitor Bafilomycin A1 (BafA1) could prevent loss of viral RNAs during 4 h exposure in macrophages (7,12,24,36,46,49,53,68,71,73).

Viral RNA in chloroquine- or BafA1-treated cells at 4 h postexposure was increased ∼1.5–2-fold relative to vehicle-treated macrophages (Fig. 1D and Supplementary Fig. S2D). Thus, chloroquine or BafA1 treatment, disrupting degradation in the endo/lysosomal network, increases recoverable viral RNA at 4 h, implicating these disposal pathways in the overall reduction of viral RNA between 1 and 24 h (Fig. 1A and Supplementary Fig. S1A). Combined, these results suggest that internalization of HCV RNA in macrophages does not require viral E2 interaction and RNA is subsequently degraded by acidification in the endo/lysosomal network.

Macrophages express receptors for HCV

As macrophages specifically internalize HCV independent of viral E2 entry factor, differentiation may upregulate surface factor(s) that enable cells to recognize HCV. Expression of SCARB1 (SR-B1), CD209 (DC-SIGN), CD81, and CLDN1 have been reported to mediate interaction between HCV particles and hepatocytes (20,37). We analyzed the expression of these receptors in THP-1 macrophage-like cells, monocytes, and Huh7.5.1 hepatocytes. Notably, SR-B1/SCARB1 was expressed in all three cell types, DC-SIGN/CD209 and CD81 were expressed at higher levels in myeloid cells, and CLDN1/CLDN1 expression was restricted to hepatocytes (Fig. 2A, B). We hypothesized that surface expression differences of SR-B1, DC-SIGN, and CD81 in macrophages relative to monocytes may explain the increase in virus uptake in these cells.

OPEN IN VIEWER

FIG. 2.

(A) Gene expression of SR-B1 (SCARB1), DC-SIGN (CD209), CD81, and CLDN1 in undifferentiated THP-1 monocytes, PMA-differentiated macrophages and Huh7.5.1 hepatocytes were analyzed by qRT-PCR. Fold change in GOI relative to HPRT is presented relative to undifferentiated monocyte (or hepatocyte for CLDN1) value set as 1. (B) Protein levels of SR-B1, DC-SIGN, CD81, and CLDN1 in undifferentiated THP-1 monocytes, PMA differentiated macrophages and Huh7.5.1 hepatocytes were determined by western blot. (C) Cell surface expression of SR-B1, DC-SIGN, and CD81 in THP-1 monocytes and PMA differentiated macrophages detected by flow cytometry. Gray shaded histograms represent isotype control antibody staining. (D) Time course of cell surface expression of DC-SIGN expression by flow cytometry on THP-1 monocytes through 72 h macrophage differentiation with PMA, followed by 24 h without PMA to produce “resting” macrophages. Macrophages in the “resting” state are used throughout the studies presented herein. Gray shaded histograms represent isotype control antibody staining. Experiments presented graphically were performed a minimum of 2 times with biological triplicate replicates. Measurements of biological triplicates from a single representative experiment are presented. t-Tests at 95% CI were used to assess differences in the means between two grouped variables (A). *p < 0.05. CLDN1, claudin-1; DC-SIGN, dendritic cell-specific intercellular adhesion molecule-3-grabbing nonintegrin; GOI, gene-of-interest; PMA, phorbol 12-myristate 13-acetate; SR-B1, scavenger receptor class B type 1.

Despite lower protein expression compared with hepatocytes, cell surface expression of SR-B1 was detectable in THP-1 monocytes and macrophage-like cells by Flow Cytometry (Fig. 2C). CD81 was also detectable on the surface of both cell types (Fig. 2C). Surprisingly, THP-1 macrophage-like cells displayed minimal surface expression of DC-SIGN similar to that observed in monocytes (Fig. 2C). DC-SIGN surface expression showed an initial upregulation for THP-1 macrophage-like cells during PMA treatment but was rapidly lost upon removal of PMA to achieve a “resting” macrophage phenotype (Fig. 2D). These results suggest that macrophages express receptors involved in recognition of HCV.

Macrophage exposure to HCV alters lipid metabolism causing cholesterol accumulation

HCV particles, replicons, and proteins induce accumulation of multiple lipids, including fatty acids and cholesterol in hepatocytes (1,10,32,39,44,62,63). We hypothesized that virus exposure in macrophages may also elicit metabolic disturbance. To test this possibility, we extracted and quantified Oil Red O as a measure of total neutral lipids in macrophages exposed to virus. Oil Red O extraction and quantification as a measure of neutral lipids indicated increased lipids in macrophages exposed to HCV for 48 h (Fig. 3A). Microscopic observation demonstrated that HCV-exposed macrophages were larger with increased cytoplasm and prominent lipid inclusions (Supplementary Fig. S3). We determined no difference in triglyceride levels (Supplementary Fig. S5A), however, total cholesterol demonstrated a 20% increase between mock- and HCV-exposed macrophages (Fig. 3A).

OPEN IN VIEWER

FIG. 3.

(A) THP-1 macrophages were exposed to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for 48 h and then stained for neutral lipid content with Oil Red O (left). Arrows in insets delineate minimal staining in mock- and increased aggregate perinuclear staining in HCV-exposed macrophages, respectively. Neutral lipids were extracted and quantified by absorbance (top, middle) and cholesterol was extracted and quantified by Amplex Red cholesterol assay (bottom, middle). Free and esterified cholesterols were individually quantified by Amplex Red cholesterol assay (right). Results from a representative experiment, of two performed, with biological triplicate replicates is presented. t-Tests at 95% CI were used to assess differences in the means between two grouped variables (A). *p < 0.05. (B) THP-1 macrophages were exposed to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for 48 h and then stained for cholesterol content with Filipin; nuclei were counterstained with DRAQ5. Arrows in insets delineate membrane staining in mock and aggregate perinuclear staining in HCV-exposed macrophages, respectively. Full-field images in (B) include a scale bar = 50 μm. Insets are zoomed images at the same magnification with scale bars = 5 μM. All confocal images were collected at 1,024 × 1,024 resolution, 12-bit dimension with a plan-apochromat 20 × lens on a Zeiss LSM710. Laser-scanning confocal microscope.

We next determined cholesterol species accumulated in macrophages exposed to HCV by analyzing free versus esterified cholesterols. Cholesterol esters accumulated in macrophages exposed to HCV such that increase in total cholesterol included 40% increase in cholesterol esters coupled with 15% increase in free cholesterol (Fig. 3A).

We performed filipin staining in macrophages to determine if virus exposure induced discreet intracellular cholesterol accumulations in these cells. Confocal microscopy demonstrated diffuse intracellular accumulations of cholesterol in a perinuclear region in HCV-exposed macrophages, in contrast to weaker staining localized to membranes, in mock-exposed macrophages (Fig. 3B and Supplementary Fig. S4, arrows in insets).

Fluorescence traces through mock-exposed cells demonstrated two polar peaks of Filipin staining corresponding to cell membrane with a nuclear peak in the center of cells. In contrast, HCV-exposed cells demonstrated a dramatic skewing of Filipin staining preferentially toward one side of cells indicating perinuclear accumulation (Supplementary Fig. S4). Increased cholesterol biogenesis, reduced cholesterol efflux, or increased lipid uptake can increase cellular cholesterol.

To determine whether cholesterol accumulation in macrophages occurred from internal or exogenous sources, we assayed uptake of DiI-labeled LDL and C12-NBD-labeled cholesterol. No significant increase in the capacity of virus-exposed cells to take up lipoproteins or cholesterol was detected when these features were assayed with commercially available kits utilizing fluorescently labeled lipoprotein and cholesterol species (Supplementary Fig. S5B). These data indicate that HCV exposure causes macrophages to accumulate cholesterol in diffuse pools derived from an intracellular process. These data prompted us to examine pathways of cholesterol flux in macrophages to determine if HCV-induced cholesterol accumulation could activate these pathways.

HCV exposure in macrophages increases proteins in the LXRs—dependent metabolic pathway

LXRs regulate macrophage cholesterol efflux and are involved in immune responses (65). LXR activation can induce or repress gene expression. Considering HCV-induced accumulation of cholesterol in macrophages (Fig. 3), we wanted to examine effects on LXR protein expression and activation. HCV exposure had no effect on LXRβ, but resulted in a slight increase in LXRα in macrophages (Supplementary Fig. S6A). Next, we examined the effect of HCV exposure on LXR targets in macrophages. HCV induced upregulation of LXR target ABCA1 beginning at 24 h and increasing throughout 48 h postexposure in THP-1 macrophage-like cells (Supplementary Fig. S6B). Cell culture virus preparations contain factors released from producer cells. Thus, we validated our findings using sucrose-gradient purified HCV isolates from Genotypes 1A, 1B, and 3A.

Within 24 h of exposure to primary isolates, THP-1 macrophage-like cells demonstrated a roughly two- to three-fold increase in ABCA1 protein expression compared with mock-exposed cells indicating that LXR activation is a specific effect of virus exposure (Fig. 4A and Supplementary Fig. S6C). THP-1 macrophage-like cells exposed to HCV also demonstrated increased SREBP-1c (Fig. 4B). We examined LXR activation in primary MDMs exposed to HCV. Forty-eight hours exposure to HCV caused upregulation of multiple LXR target proteins in MDMs, ABCA1, SREBP-1c, and LXRα itself (Fig. 4C).

OPEN IN VIEWER

FIG. 4.

(A) THP-1 macrophages were exposed to 0.1 MOI sucrose-gradient purified HCV-containing supernatant of the indicated genotype or mock-infected Huh7.5.1 supernatants for 24 h and LXR target ABCA1 protein expression was analyzed by western blot. (B) THP-1 macrophages were exposed to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for 24 h (as in Supplementary Fig. S6A) and SREBP-1c protein levels were analyzed by western blot. Panels represent two samples per condition, with panels separated to demonstrate that mock- and HCV-exposed samples were run in nonadjacent lanes with all samples analyzed from the same gel. (C) Primary human MDMs were exposed to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for 48 h and LXRα and LXR targets SREBP-1c and ABCA1 protein expression were analyzed by western blot. MDMs, monocyte-derived macrophages.

To confirm the widely reported role of LXR activation on expression of these metabolic targets in THP-1 macrophage-like cells, cells were treated with LXR agonists GW3965 (GW) or TO-901317 (TO) and protein expression was examined. As expected, LXR agonists increased expression of proteins involved in cholesterol efflux (ABCA1) and lipogenesis (SREBP-1c) (Supplementary Fig. S6D), confirming that LXR activation regulates these proteins involved in cholesterol and fatty acid metabolism paralleling the observed effects of HCV in macrophages. These results confirm that LXR regulates metabolic signaling in macrophages and suggest that HCV activates expression of proteins in this pathway in macrophage models and primary macrophages.

Macrophages recognize HCV through SR-B1 resulting in LXR activation

Blockade of viral envelope protein E2-CD81 interaction did not inhibit viral RNA uptake in macrophages (Supplementary Fig. S2C), although CD81 is expressed in this system (Fig. 2C). Resting macrophages also displayed minimal surface DC-SIGN (Fig. 2C). Therefore, we examined the role of SR-B1 in HCV uptake in macrophages. HCV produced in hepatocytes through the apolipoprotein pathway is most infectious as lipoviral particles containing lipoprotein constituents (6,11,13,18,21,25). In the liver, SR-B1 functions to uptake high-density lipoproteins from the circulation (2,60). SR-B1 also plays a role in HCV infection in hepatocytes. We hypothesized that macrophages sense HCV similar to a lipoprotein through SR-B1.

Therefore, we sought to determine if targeting SR-B1 could reduce or delay accumulation of viral RNA observed upon exposure of macrophages to HCV. Macrophages and hepatocytes express SR-B1, although the role of this receptor in recognition of HCV in immune cells has not been described. Inhibitors of SR-B1, termed blockers of lipid transport (BLT) uncovered the role of SR-B1 in HCV uptake in hepatocytes. Blocker of lipid transport 4 (BLT-4) reduces HDL enhancement of pseudo-particle infectivity and blocks soluble E2 binding to SR-B1-expressing cells. Among CD81, DC-SIGN, and SR-B1, SR-B1 has been described to have a role in lipid binding at the cell surface of macrophages and hepatocytes.

We examined whether BLT-4 could reduce or delay HCV RNA binding upon exposure in macrophages. As we observed the greatest amount of viral RNA binding to macrophages at 1 h postexposure (Fig. 1A, C) and blockade of lipid binding by SR-B1 would be expected to effect viral binding, we examined the effect of BLT-4 treatment at this time point. For comparison, we utilized an inhibitor of actin polymerization, cytochalasin D (CytoD), which has been demonstrated to inhibit HCV genome uptake in macrophages (36,49).

One hour pretreatment with BLT-4 (Sigma, cat. no. SML0512) reduced bound total viral RNA in association with macrophages to ∼75% of levels in vehicle-treated cells similar to the effect of CytoD (Fig. 5A and Supplementary Fig. S7A). We also tested the effect of blocking antibodies to DC-SIGN and CD81 on HCV RNA uptake in macrophages. In agreement with minimal surface expression, we observed no effect of DC-SIGN blockade on virus uptake (Supplementary Fig. S7B). Blocking CD81 also had no effect on virus uptake, consistent with data showing that blocking E2 did not affect this process (Supplementary Figs. S2C and S7B).

OPEN IN VIEWER

FIG. 5.

(A) THP-1 macrophages were pretreated with BLT-4 (50 μM) for 1 h before and throughout exposure to 0.1 MOI HCV for 1 h. Viral RNA was quantified by RT-qPCR with either 5′-UTR or NS3 primer. Fold change in viral RNA relative to HPRT is presented as a percentage of the non-compound-treated sample set as 100%. (B) THP-1 macrophages were pretreated for 1 h with blocking antibody to SR-B1 (1:50) or isotype control antibody before exposure to 0.1 MOI HCV or mock-infected Huh7.5.1 supernatants for 48 h in the continued presence of antibody and ABCA1 protein expression was analyzed by western blot. Experiments presented graphically were performed a minimum of two times with biological triplicate replicates. Measurements of biological triplicates from a single representative experiment are presented. t-Tests at 95% CI were used to assess differences in the means between two grouped variables (A). **p < 0.01. BLT-4, blocker of lipid transport 4.

These results suggest that macrophages recognize HCV through scavenger receptors involved in lipoprotein uptake. Data with CytoD confirm the requirement for phagocytosis in HCV uptake in macrophages (36,49). Taken together, this suggests a mechanism whereby lipid-associated virus binds to cells and is internalized through a mechanism involving SR-B1 and phagocytosis.

LXR activation by HCV in macrophages has not been previously described. This finding was unexpected as HCV exposure increases cholesterol in macrophages and LXR activation should reduce cellular cholesterol, suggesting virus may act directly on LXR. To demonstrate if the effect of HCV exposure on LXR activation was specific to virus, we assessed the ability of antibody blockade of SR-B1 to prevent HCV-induced ABCA1 protein induction alongside reduction in viral RNA accumulation by BLT-4. We used anti-SR-B1 antibody demonstrated to prevent HCV infection of Huh7.5 hepatocytes (12).

In the presence of nonblocking isotype, HCV exposure induced the expected increase in ABCA1 protein compared with mock supernatants (Fig. 5B). The addition of SR-B1 blocking antibody from 1 h before infection throughout 48 h of virus exposure prevented the HCV-induced increase in ABCA1. SR-B1 blocking antibody reduced ABCA1 levels independently of virus, perhaps indicating that this antibody interferes with cellular recognition of apolipoproteins, which could independently reduce ABCA1 expression.

We also assessed the involvement of SR-B1 in recognition of virus and induction of LXR activation in macrophages by utilizing SR-B1 knockdown THP-1 macrophage-like cells. Notably, knockdown of SR-B1 similarly reduced ABCA1 independent of HCV (Supplementary Fig. S8). These data suggest that blocking SR-B1 can generally reduce ABCA1 protein and also prevent the HCV-specific induction of ABCA1.