Hypoxia-inducible factor 2 triggers the production of highly infectious native-like hepatitis C virus particles

Cells and treatments

Huh7.5 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 4.5 g/L D-glucose, 4 mM L-glutamine (Invitrogen), and supplemented with 100 U/mL penicillin, 100 µg/mL streptomycin (Invitrogen) and 10% fetal calf serum (Thermo Fisher Scientific). Standard de-differentiated Huh7.5 cells were cultured at atmospheric oxygen tension (Std—normoxia—21% O₂ [vol/vol], 5% CO₂ [vol/vol], 74% N₂ [vol/vol]). These cells were used to seed 12-well plates at a density of 4 × 10⁵ cells/well 24 h before use. Differentiated Huh7.5 cells cultured in hypoxia (Hypo-Diff) were seeded in 12-well plates coated with rat type I collagen (Sigma-Aldrich), at a density of 8 × 10⁴ cells/well. At 40% confluence, the cell culture medium was replaced with complete DMEM supplemented with 1% DMSO (vol/vol). Cells were cultured for 15 days, during which time the medium was replaced with DMEM supplemented with 1% DMSO every two days. Nine days after DMSO exposure, Huh7.5 cells were transferred into a hypoxia workstation (HypoxyLab—Oxford Optronix, Abingdon, UK) and cultured for another six days at low oxygen tension (1% [vol/vol] O_2, 5% [vol/vol] CO₂, 94% [vol/vol] N₂).

RNA Interference

Four days post transfer into the hypoxia workstation, Hypo-Diff Huh7.5 cells were transfected with 10 nM of small interfering RNAs (siRNAs) targeting HIF-1α (siHIF-1α—Assay ID: s6541; Ambion), HIF-2α (siHIF-2α Assay ID: s4699; Ambion) or a non-targeting control siRNA (All Stars negative control siRNA, Qiagen) using Lipofectamine RNAiMax (Invitrogen). The medium was replaced by complete DMEM supplemented with 1% DMSO 6 h post-transfection according to the manufacturer’s instructions.

Plasmids and transfection

HA-HIF2alpha-P405 A/P531 A-pcDNA3 and HA-HIF1alpha P402 A/P564 A-pcDNA3 (named thereafter HIF-1α mut and HIF-2α mut), a gift from William Kaelin (Addgene plasmid # 18,956; http://n2t.net/addgene:18956; RRID:Addgene_18956 and Addgene plasmid # 18955; http://n2t.net/addgene:18955; RRID:Addgene_18955, respectively), are CMV promoter-driven pcDNA3 plasmids expressing stabilized HIF-1α and HIF-2α [18]. These plasmids included 2 mutated amino acids, which were then no longer hydroxylated by the PHD. The mutant proteins are O₂ insensitive and thus resistant to proteasome-specific degradation under normoxic conditions. Standard de-differentiated Huh7.5 cells cultivated in normoxia were used to seed a 12-well plate at a density of 1 × 10⁵ cells/well. Transfections were performed with 1.5 µg of HIF-1α mut, HIF-2α mut or pcDNA3 plasmids using JetPEI (Life Technologies) as transfectant, according to the manufacturer’s instructions.

HCVcc production and infection

For HCVcc production, Huh7.5 cells were transfected with JFH1 RNA by electroporation (single pulse of 900 µF and 270 V), as previously described [5, 19–21]. Supernatants were collected 48 and 72 h post-transfection, passed through a filter with 0.45 µm pores, and stored at −80 °C. JFH1-transfected Huh7.5 cell supernatants were harvested after three post-transfection passages, to obtain high HCVcc titers. HCVcc were titrated as previously described [4].

Except otherwise stated, HCVcc infection experiments were performed for 6 h, at a multiplicity of infection (MOI) of 0.1, 72 h post transfer into the hypoxia workstation for Hypo-Diff Huh7.5 cells and 24 h or 72 h post-seeding for Std Huh7.5 cells. The cells were then washed and cultured for an additional 48 h or 72 h in the presence or absence of 1% DMSO, in normoxia or hypoxia, according to the cell culture conditions.

Antibodies

The following antibodies (Ab) were purchased: mouse monoclonal anti-HCV core protein (C7–50, MA1–080, Thermo Fisher Scientific), mouse monoclonal anti-HIF-1α (clone 54, BD Biosciences), rabbit anti-HIF-2α (Ab199, Abcam), rabbit anti-β-actin (ab8227, Abcam, 1:2000), Alexa Fluor™ 594 donkey anti-rabbit IgG (H + L) (Invitrogen), Alexa Fluor™ 594 donkey anti-mouse IgG (H + L) (Invitrogen).

RT-qPCR

Total RNA was extracted with the RNeasy Mini Kit (Qiagen) and viral RNA with the QIAamp Viral RNA Mini Kit (Qiagen), according to the manufacturer’s instructions. Complementary DNA was synthesized with the ProtoScript II First-Strand cDNA Synthesis Kit (New England Biolabs), according to the manufacturer’s instructions. Quantitative PCR was performed with LightCycler 480 SYBR Green Master Mix (Roche Diagnostics). Amplification signals were analyzed with LightCycler 480 software. Gene and viral expression assay information are available in Table 1 (Supplemental Material).

SDS-PAGE and immunoblots

The cells were lysed in RIPA buffer (20 mM Tris, pH 7.5, 20 mM EDTA, 150 mM NaCl, 1% NP40, and 1% sodium deoxycholate) supplemented with protease inhibitor cocktail tablets (Thermo Fisher Scientific). The resulting lysate was centrifuged, and the protein concentration of the supernatant was assessed using the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Reducing buffer was added to samples before incubating at 95 °C for 5 min. Forty µg of proteins were separated on 12% polyacrylamide gels and transferred to PVDF membranes (Amersham). Membranes were blocked in PBST, 5% BSA, and proteins detected using specific primary anti-HIF-1α (clone 54; 1:250), anti-HIF-2α (Ab199; 1:500), anti-β-actin (ab8227, Abcam, 1:2000) and corresponding HRP-secondary antibodies. Protein bands were revealed by enhanced chemiluminescence on an Imagequant LAS500 apparatus (GE Healthcare).

Indirect immunofluorescence and lipid droplet staining

Cells were grown on 12-mm glass coverslips and were fixed and permeabilized by incubation with 4% paraformaldehyde in PBS for 10 min at room temperature. After washing three times in PBS, cells were permeabilized and non-specific sites were saturated by incubation for 30 min at room temperature in 0.3% BSA – 0.4% Triton X100 in PBS. Cells were then labeled with an anti-HIF-1α (clone 54; 1:100) or an anti-HIF-2α (Ab199; 1:200) antibody, or a mouse monoclonal anti-HCV core protein (C7-50; 1:250) in PBS-BSA 0.2% by incubation for 1 h at room temperature. The cells were washed three times in PBS and incubated for 1 h at room temperature with an Alexa Fluor™-conjugated secondary antibody (1:2000) in the same buffer. Lipid droplets were stained by incubating cells with Bodipy 493/503 (Fisher Scientific) at a dilution of 1:10,000 in PBS for 5 min at room temperature. Cells were washed three times in PBS and were then mounted in a mixture of equal volumes of Fluoromount-G and DAPI Fluoromount-G (SouthernBiotech). Confocal microscopy was performed with an SP8 confocal microscope (Leica).

The size and number of lipid droplets per cell were analyzed with Image J software (NIH, Bethesda, MD, USA) after being defined and selected, for five representative fields in each set of conditions. Histograms of lipid droplet sizes were generated from data normalized by the number of cells in each image.

Infectivity of the collected supernatants

Fifty µL of fresh supernatant from JFH1-infected Std and HIF-1α or HIF-2α-silenced Hypo-Diff Huh7.5 cells were transferred for 6 h onto 15,000 JFH1-naive Huh7.5 cells, cultured in Std cell culture conditions in 96-well plates. The cells were thoroughly washed with PBS and incubated in complete DMEM for 72 h. Cells were fixed and permeabilized by incubation with cold 50% acetone—50% ethanol (vol/vol) for 10 min at room temperature. Indirect immunofluorescence analyses were then performed, as described above. Focus-forming units (FFU) were counted (Olympus IX 51, Tokyo, Japan), and the results are expressed as FFU/mL/10⁵ HCV genome copies of supernatant from JFH1-infected cells.

Buoyant density on iodixanol gradient ultracentrifugation

Fresh supernatants from JFH1-infected Std and HIF-1α or HIF-2α-silenced Hypo-Diff Huh7.5 cells were overlaid on iodixanol gradients generated from equal volumes (2 mL) of 4, 13, 22, 31 and 40% (weight/weight) Opti-Prep density gradient medium (Merk) solutions in gradient buffer (0.25 M sucrose, 0.5 M Tris pH = 8.0, 0.1 M EDTA pH = 8.8, 2 M MgCl₂, 2 M MgSO₄). Equilibrium was achieved by ultracentrifugation for 18 h at 250,000 × g in an SW41 rotor at 4 °C in a Beckman Coulter Optima XPN80 ultracentrifuge. Twelve fractions (1 mL each) were collected from the top. The viral genome was quantified in these fractions and the infectivity of the viral particles in each fraction was analyzed as described above. The density of each fraction was determined by measuring its mass.

Statistical analysis

Statistical analyses were performed with Prism software (GraphPad Software Inc., San Diego, California, USA). The results are expressed as the mean ± SEM of at least three sample replicates unless otherwise stated. The statistical significance of differences was assessed in two-tailed Student's t-tests or Welch t-tests, depending on the normality of the data and the homoscedasticity of the variance, as determined by Shapiro–Wilk tests and F tests, respectively. All p values below 0.05 were considered significant.

The silencing approach of HIF-1α and HIF-2α is efficient and specific in hepatocyte-like cells exposed to sustained hypoxia

Our previous work demonstrated the high efficiency to combine both Huh7.5 cells partial differentiation by exposure to DMSO and to culture cells under sustained hypoxia (Hypo-Diff) to enable the production of highly lipidated and infectious HCV-LVPs, compared to standard Huh7.5 cell-culture conditions without DMSO and in normoxia (Std) [12]. Here we have reproduced these experimental conditions to decipher the precise role of HIF-1α and HIF-2α in these viral characteristics using RNA interference (Fig. 1A). Both specificity and efficacy of siRNAs to achieve HIF-1α and HIF-2α down-regulation were controlled by complementary approaches including HIFα mRNA and protein detection (i.e. western blot and immunofluorescence) after adapted transfections four days post hypoxia (Fig. 1A, B and C). Interestingly, HIF-1α and HIF-2α were not detected in Std cell nuclei on immunofluorescence analyses, whereas both transcription factors were found to have an extensive nuclear distribution in Hypo-Diff cells, a sub cellular localization that was lost after specific transfection of siHIF-1α and siHIF-2α (Fig. 1B and C). The functional impact of this interference was confirmed since the quantities of HIF-1-specifically transcribed mRNA of glucose transporter 1 (GLUT1) and phosphoglycerate kinase 1 (PGK1) and of the HIF-2-targeted superoxide dismutase-2 (SOD-2) were dramatically and specifically down-regulated of about 55, 77 and 55% after transfection with siHIF-1α and siHIF-2α in hypoxic conditions, respectively (Fig. 1D and E). The silencing strategy thus shown to be effective to specifically and independently down-regulate HIF-1 and HIF-2 transcription factors.

The size and number of cytoplasmic lipid droplets are controlled by HIF-2α in Hypo-Diff Huh7.5 cells

Intracellular neutral lipids are key elements required for HCV LVP assembly, morphogenesis and egress from infected cells [22, 23]. In DMSO-differentiated Huh7.5 cells exposed to a physiological hypoxia, compared to Std cell culture conditions, we previously showed an increase of both the size and number of cytoplasmic lipid droplets (LDs), the organelles responsible for intracellular neutral lipid storage [12]. Using the specific siRNA-mediated knockdown of HIF-1α and HIF-2α we therefore determined their importance in the morphological characteristics of intracellular LDs. Neutral lipids were stained with Bodipy 493/503, images were captured and the number of LDs per cell was determined (Fig. 2A and Fig. 2B). Analyses confirmed that Hypo-Diff Huh7.5 cells contained significantly more LDs than Std-cultured cells (median 14.7 ± 2.7 LDs/cell and 6.7 ± 0.5 LDs/cell, respectively, P < 0.01) (Fig. 2B). Remarkably, while these amounts were unchanged in HIF-1α knockdown Hypo-Diff Huh7.5, the number of LDs per cell significantly decreased 1.7-fold in siRNA HIF-2α-transfected cells compared to control Hypo-Diff Huh7.5. The values obtained for siRNA HIF-2α-transfected cells were comparable to and not statistically different from Std cell culture conditions (median 8.3 ± 1.8 LDs/cell and 6.7 ± 0.5 LDs/cell, respectively, ns). A mean-volume analysis confirmed that LDs contained in Hypo-Diff Huh7.5 cells were significantly larger than in Std cells (3.55 ± 0.48 µm² and 1.71 ± 0.31 µm², respectively, P < 0.001) (Fig. 2C). Noteworthy, the down-modulation of HIF-1α in Hypo-Diff Huh7.5 cells further increased, moderately although significantly, these sizes (4.4 ± 0.56 µm² and 3.55 ± 0.48 µm², respectively, P < 0.05) further arguing against a role of HIF-1α and suggesting the responsibility of HIF-2α in the morphogenesis of large LDs. HIF-2α-depleted Hypo-Diff cells values largely confirmed this hypothesis since LDs size dramatically and significantly dropped as compared to both HIF-1α siRNA transfected cells and Hypo-Diff counterparts (2.14 ± 0.31 µm², 4.4 ± 0.56 µm² and 3.55 ± 0.48 µm², respectively, P < 0.001) to reach levels similar to Std cell culture conditions (1.71 ± 0.31 µm²).

These results indicate that physiological hypoxia and predominantly HIF-2α participate in the generation of neutral lipid storage in the form of LDs in partially differentiated Huh7.5 cells cultured in physiological hypoxia. Given the importance of LDs in the intracellular HCV life cycle, the importance of HIF-2α was therefore explored.

The highly infectious viral particles produced by HCV-infected Hypo-Diff Huh7.5 cells is dependent on HIF-2α

The requirement of LDs to serve as an assembly platform for HCV is well documented [22–24]. We thus explored the susceptibility of HIF-2α (and HIF-1α) down-regulated Hypo-Diff Huh7.5 cells to support HCV infection and the ability of these cells to produce infectious LVPs relative to Std and wild-type Hypo-Diff cell culture conditions. To this end, Std and Hypo-Diff cells were infected with the cell culture-adapted HCV strain JFH1 at a multiplicity of infection (MOI) of 0.1 a day before HIFs siRNA transfection (Fig. 3A). Two days post transfection, neutral lipid-rich LDs together with the HCV core protein were detected by confocal fluorescence microscopy using Bodipy 493/503 and the mouse monoclonal anti-HCV core protein C7-50, respectively (Fig. 3B). The expression of the HCV core protein was comparable between the standard cell culture condition and both untransfected and siHIF-1α-transfected Hypo-Diff cells. However, siHIF-2α-transfected cells showed a significant downregulation, with approximately 25% reduction compared to the other cell culture conditions (Fig. 3B, left of the dotted line and Fig. 3C). Under all cell culture conditions, JFH1-infected cells exhibited an exclusive localization of core proteins at the surface of LDs, that formed ring-like structures with a perinuclear localization (Fig. 3B, right of the dotted line). Particularly, no different core localization was observed in HIF-1α and HIF-2α down-regulated Hypo-Diff Huh7.5 cultures, indicating that the HCV core-LDs association is independent of both transcription factors, associated target genes and the level of neutral lipid storage.

Two days post-siRNA transfections and three days post-HCV infection, intracellular and extracellular HCV genomes were quantified by RT-qPCR. Due to their long-term exposure to DMSO, Hypo-Diff Huh7.5 cells are almost quiescent, unlike Std cell cultures (data not shown). Thus, to exclude bias in viral load analyses, cells were counted at the time of RNA extraction and the results were standardized per 10⁵ cultured cells. Infections of Hypo-Diff Huh7.5 cells were nearly as efficient as in Std cells since a weak but not significant decrease in the intra- and extracellular HCV genome amounts was measured (Fig. 3D and Fig. 3E) (intracellular: 3.9 × 10⁷ ± 9.7 × 10⁶ HCV RNA copies/10⁵ cells and 8.4 × 10⁷ ± 4.8 × 10⁶ HCV RNA copies/10⁵ cells, respectively, ns; extracellular: 4.4 × 10⁵ ± 6.5 × 10⁴ HCV RNA copies/10⁵ cells and 1.5 × 10⁶ ± 1.9 × 10⁵ HCV RNA copies/10⁵ cells, respectively, ns). These values were close and not statistically different in HIF-1α down-regulated Hypo-Diff and in scramble-siRNA transfected Huh7.5 cells. Intra- and extracellular HCV genome amounts were in contrast dramatically decreased in HIF-2α down-regulated Hypo-Diff cells compared to Std and HIF-1α transfected Huh7.5 cells (intracellular: 5.95 × 10⁵ ± 1.4 × 10⁵ HCV RNA copies/10⁵ cells, 8.4 × 10⁷ ± 4.8 × 10⁶ HCV RNA copies/10⁵ cells, P < 0.001 and 2.8 × 10⁷ ± 9.8 × 10⁶ HCV RNA copies/10⁵ cells, respectively, P < 0.05; extracellular: 2.1 × 10⁴ ± 5 × 10³ HCV RNA copies/10⁵ cells, 1.5 × 10⁶ ± 1.9 × 10⁵ HCV RNA copies/10⁵ cells, P < 0.05 and 3.0 × 10⁵ ± 5.3 × 10⁴ HCV RNA copies/10⁵ cells, respectively, P < 0.05). These data, along with those showing lower expression of the core protein in HIF-2α-depleted Hypo-Diff cells (Fig. 3B, left of the dotted line and Fig. 3C), strongly suggest that HIF-2α is a key factor in HCV replication under physiological conditions of oxygenation. HCVcc-containing extracellular medium from JFH1-infected Std and the different Hypo-Diff cultures was incubated with naïve Huh7.5 cells to complete progeny viral infectivity. Since extracellular amounts of HCV genome copies of these different cultures were not equivalent (Fig. 3E), naïve cells were infected with standardized 10⁵ genome copies for each infection, allowing to measure specific infectious viral efficacy. Three days post-infection, the HCV core protein was stained to calculate FFU, and viral infectivity was determined (Fig. 3F). With an equal number of viral genome used for infections performed with the different supernatants, our results show that progeny viruses produced by HCV infected-Hypo-Diff Huh7.5 cells were eightfold more infectious than those produced by JFH1-infected Std Huh7.5 (9456 ± 824 vs. 1159 ± 166 FFU/mL/10⁵ genome copies, respectively, P < 0.01), confirming our previous findings [12]. Transfection of HCV-producing Hypo-Diff cells with scramble or HIF-1α siRNA did not significantly change these major differences. In contrast, progeny viruses produced by HIF-2α down-regulated Hypo-Diff cells lost these high levels of specific infectivity. In that case, values regressed to those generated for viruses produced in Std conditions (1986 ± 1383 vs. 1159 ± 166 FFU/mL/10⁵ genome copies, respectively, ns). These results thus indicate that HIF-2α has a major role for HCV replication, the production and specific infectivity of progeny viruses generated by hepatocyte-like cells exposed to physiological hypoxia.

To confirm the role of HIF-2α independently from other changes associated with hypoxia, oxygen-insensitive HIF-1α and HIF-2α mutated variant proteins (HIF-1α mut and HIF-2α mut, respectively) were used (Fig. 4). These proteins contain a double mutation (P402 A/P564 A in HIF-1α and P405 A/P531 A in HIF-2α) described as responsible for HIF-1α and HIF-2α stabilization in normoxia [19–21]. Huh7.5 cells in normoxia (Std cells) were thus transfected with these constructions and infected a day later with JFH1 at MOI 1, before harvesting both cells and supernatants (Fig. 4A). Immunoblot analyses displayed high levels of HIF-1α mut (Fig. 4B) and HIF-2α mut (Fig. 4C) protein expressions compared to untransfected and pcDNA3-transfected cells and demonstrated both the role of mutated amino acids in the stabilization of HIF-1/2α in Huh7.5 cells and the efficiency of the transfection approach to stably express these mutated transcription factors in normoxia. We therefore assessed their role in the morphological characteristics of intracellular LDs. Neutral lipids were stained with Bodipy 493/503, images were acquired, and the number and size of LDs per cell were quantified (Fig. 4D, E, and F). The analysis clearly showed that while the stabilized expression of HIF-1α in normoxia did not alter either LD characteristic, HIF-2α mut-transfected cells exhibited significantly more and larger LDs compared to the standard cell culture conditions (median 20.5 ± 2.1 LDs/cell vs. 9.7 ± 1.1 LDs/cell, respectively, P < 0.001; 2.18 ± 0.35 µm² vs. 1.1 ± 0.25 µm², respectively, P < 0.001). These findings confirm that HIF-2α contributes to the formation of neutral lipid storage in Huh7.5 cells and reveal that its expression is both necessary and sufficient for this process. JFH1-infection of HIF-1α mut and HIF-2α mut expressing Std cells had no effect on intra- and extracellular HCV genome amounts when quantified by RT-qPCR and compared to untransfected cells (Fig. 4G and H). These data indicate that HIF-1α mut and HIF-2α mut do not perturb viral replication and LVPs egress in the condition of (non-physiological) atmospheric oxygen pressure. In contrast, progeny viruses produced by HIF-2α mut-expressing Std cells displayed a significantly twofold higher infectivity compared to the other conditions (15,733 ± 1527 FFU/mL for HIF-2α mut vs. 6933 ± 611 FFU/mL for untransfected Std, 6733 ± 577 FFU/mL for pcDNA3 and 7600 ± 529 FFU/mL for HIF-1α mut, P < 0.001) (Fig. 4I), indicative for qualitative instead of quantitative LVPs modifications under the control of HIF-2α.

Overall, these results clearly demonstrate that HIF-2α is both necessary and sufficient to produce highly infectious HCV progeny viruses, alongside an increased number and size of intracellular lipid droplets. In our physiological culture model, we demonstrated that HCV infectivity is dependent on morphological characteristics such as low LVPs density [12]. We thus investigated whether the loss of infectivity of viruses produced by HIF-2α-depleted cells was associated with a defect of lipidation in the Hypo-Diff cell culture model.

HIF-2α in Hypo-Diff Huh7.5 cells controls the production of highly lipidated HCV-LVPs and their related broad infectivity

Compared to Std cell culture conditions, HCV-infected Hypo-Diff cells produce highly lipidated and broadly infectious LVPs [12]. As the depletion of HIF-2α in Hypo-Diff cells reduces HCV infectivity, the question then raised as to whether HIF-2α controls the lipidation of HCV LVPs and their associated infectivity. We first determined the relative amounts of HCV RNA present in fractions of various densities isolated from the supernatants of HCV-infected HIF-1α and HIF-2α down-regulated Hypo-Diff cells after iodixanol density gradient centrifugation as summarized in Fig. 5A. As we previously reported [12], we confirmed here that the HCV genome was significantly more abundant for Hypo-Diff cells than for Std cells in low-density fractions (d < 1.03 g/mL) (i.e., highly lipidated LVPs) (42.8 ± 5.6% and 16.8 ± 4%, respectively, P < 0.0001) (Fig. 5B). The depletion of HIF-1α did not significantly change this difference compared to Std cells (P < 0.001). In contrast, the downregulation of HIF-2α in Hypo-Diff cells abrogated the production of highly lipidated HCV LVPs, the density of which was identical to those of Std cell culture conditions (17.9 ± 1.7% and 16.8 ± 4%, respectively, ns). Interestingly, for the density fractions mainly produced by cells (1.03 g/mL < d < 1.08 g/mL), these results were reversed. While Std cells generated a larger proportion of HCV LVPs of lower density (70 ± 8.9%), these proportions dropped for wild-type and HIF-1α down-regulated Hypo-Diff cells (47 ± 2.8% and 56.4 ± 2%, P < 0.0001, P < 0.01, respectively), while they were nearly unchanged for HIF-2α down-regulated Hypo-Diff cells (65.2 ± 1.9%, ns). Thus, overall, HIF-2α-depleted Hypo-Diff cells produce HCV LVPs with a density close to Std cells and significantly higher (i.e., lower lipidated LVPs) than those produced by wild-type and HIF-1α down-regulated Hypo-Diff cells.

We then determined the infectivity of these fractions, using naïve Huh7.5 cells to calculate FFU/mL and % of viral infectivity (Fig. 5C). Our results confirmed a clear relationship between HCV RNA-associated low buoyant density fractions and the associated infectivity. The proportion of infectious progeny viruses produced by Hypo-Diff, siRNA scramble-transfected and HIF-1α-depleted Hypo-Diff cells associated with the lowest buoyant density fractions (d < 1.03 g/mL) were significantly greater than those produced by Std cells (52.3 ± 3.4%, 48.9 ± 3.5%, 36.9% ± 2.6 and 19.1 ± 4.8%, P < 0.001 and P < 0.01, respectively). These proportions were inverted within higher densities (1.03 < d < 1.08 g/mL) (i.e., lower lipidated LVPs), confirming that Hypo-Diff cells produce significantly more lipidated and infectious HCV LVPs than Std conditions and suggesting that HIF-1α is not a major actor of this feature. In contrast, HCV LVPs produced by HIF-2α-depleted Hypo-Diff cells presented an infectivity identical to that of Std conditions both in lowest buoyant density fractions (d < 1.03 g/mL) (21.5 ± 2.2% vs 19.1 ± 4.8%, ns, respectively) and in higher densities (1.03 < d < 1.08 g/mL) (75.8 ± 2.1% vs 79.4 ± 5%, ns, respectively).

Overall, these results provide a demonstration that HIF-2α controls both the production and infectivity of very low mean buoyant density (i.e. highly lipidated) HCV LVPs from cells that combine two key physiological characteristics observed during natural intrahepatic HCV infection (i.e., cellular differentiation and prolonged hypoxia).

Ethical approval

Not applicable.

Consent for publication

Not applicable.

Consent to participate

Not applicable.

Competing interest

The authors have no relevant financial or non-financial interests to disclose.