Summary
The metabolic oxidative degradation of cellular lipids severely restricts replication of hepatitis C virus (HCV), a leading cause of chronic liver disease, but little is known about the factors regulating this process in infected cells. Here we show that HCV is restricted by an iron-dependent mechanism resembling the one triggering ferroptosis, an iron-dependent form of non-apoptotic cell death, and mediated by the non-canonical desaturation of oleate to Mead acid and other highly unsaturated fatty acids by fatty acid desaturase 2 (FADS2). Genetic depletion and ectopic expression experiments show FADS2 is a key determinant of cellular sensitivity to ferroptosis. Inhibiting FADS2 markedly enhances HCV replication, whereas the ferroptosis-inducing compound erastin alters conformation of the HCV replicase and sensitizes it to direct-acting antiviral agents targeting the viral protease. Our results identify FADS2 as a rate-limiting factor in ferroptosis, and suggest the possibility of pharmacologically manipulating the ferroptosis pathway to attenuate viral replication.
Graphical Abstract
eTOC Blurb
Yamane et al. demonstrate that cellular metabolic processes that regulate ferroptosis noncytolytically restrict replication of hepatitis C virus in hepatocytes. The authors show that the ferroptosis-like signals are regulated primarily by canonical and non-canonical fatty acid desaturation catalyzed by FADS2 and iron-dependent oxidation of highly-unsaturated fatty acids.
Introduction
The oxidative degradation of polyunsaturated fatty acids (PUFAs) occurs as a consequence of cellular metabolism. The process of lipid peroxidation (LPO) and the products that result from it influence pleiotropic cellular functions, modulating immunity, tumor suppression, aging and cell death (Ayala et al., 2014). The regulatory mechanisms controlling LPO have attracted considerable attention with the recognition that cancer cells resistant to chemotherapy may be susceptible to ferroptosis, a non-apoptotic form of cell death triggered by an iron-dependent production of LPO (Hangauer et al., 2017; Viswanathan et al., 2017). Both LPO and the cellular susceptibility to ferroptosis are regulated by the independent activities of glutathione peroxidase 4 (GPX4), that neutralizes lipid peroxides, and ferroptosis suppressor protein 1 (FSP1), an oxidoreductase that reduces coenzyme Q10 (Bersuker et al., 2019; Doll et al., 2019; Yang et al., 2014). The cellular susceptibility to ferroptosis is also influenced by the fatty acid content of membrane phospholipids (Kagan et al., 2017; Magtanong et al., 2019), but the fatty acid metabolic pathways that regulate LPO have not been explored in depth.
LPO is a crucial determinant of cellular permissiveness for replication of hepatitis C virus (HCV), a highly pathogenic RNA virus that is a leading cause of chronic liver disease (Huang et al., 2007; Yamane et al., 2014). HCV infection can be treated effectively with oral direct-acting antivirals (DAAs) targeting several components of its replicase. However, over 70 million individuals are chronically infected with HCV worldwide, and as many as ~5% may have drug-resistant infections (Spearman et al., 2019). This makes it important to understand viral and cellular factors that regulate the activity of the HCV replicase, a membrane-bound, multiprotein complex with both protease and polymerase activities that mediates amplification of the viral genome, and its susceptibility to inhibition by DAAs. A positive-strand RNA virus, HCV, is unique in that its replicase activity is fine-tuned by LPO occurring within the replicase membranes, which are derived from the endoplasmic reticulum (ER) (Yamane et al., 2014). Membrane-proximal regions of the NS3/4A protease and NS5B RNA-dependent RNA polymerase act as “LPO sensors” that shut down replicase activity through conformational changes occurring in the replicase upon exposure to LPO, a unique feature of HCV that we have demonstrated both in human hepatoma cells and primary human hepatocytes (Yamane et al., 2014). The LPO-sensitive phenotype of HCV, a conserved feature of clinical isolates of diverse HCV genotype (Saeed et al., 2015; Yamane et al., 2014), reduces the replicative activity of the virus and promotes its persistence in damaged liver tissue. Loss of the ability to downregulate replication upon exposure to LPO accounts for the robust cell culture replication phenotype of laboratory-adapted strains of HCV, such as JFH1, TNcc, and H77D (Yamane et al., 2014).
Despite the crucial role of LPO in HCV infection, the cellular processes that promote endogenous LPO and restrict HCV replication remain poorly characterized. Although LPO can result from both enzymatic and non-enzymatic reactions, we show here that HCV replication is predominantly regulated by a non-enzymatic pathway involving the Fenton reaction. Chelating iron with deferoxamine (DFO) abolishes LPO and promotes HCV replication. This iron-dependent production of LPO is reminiscent of ferroptosis (Dixon et al., 2012), but does not lead to cell death. Noncytolytic concentrations of a ferroptosis-inducing compound, erastin, disrupt HCV replicase function by altering the conformation of the membrane-anchored viral protease, enhancing its susceptibility to DAAs, and facilitating viral clearance. In contrast, ferroptosis inhibitors lower the binding affinity of these DAAs and reduce their activity against the virus. We also show that FADS2, which catalyzes the first, rate-limiting step in highly-unsaturated fatty acid (HUFA; ≥20 carbon fatty acids with ≥3 double bonds) biosynthesis, functions as a master-regulator of HCV replication by promoting LPO, and is a key factor in sensitizing cells to ferroptosis.
Results
The 5-lipoxygenase inhibitor BWA4C restricts HCV replication via promotion of LPO.
We used chemical compounds that inhibit lipoxygenase (LOX) or cyclooxygenase (COX) to determine the contribution of enzymatic LPO pathways to the restriction of HCV strains with disparate sensitivities to LPO in human hepatoma (Huh-7.5) cells (Figure 1A). An initial screen showed that LOX inhibitors possessing antioxidant properties (PD146176, AA-861 and Zileuton) stimulated replication of an LPO-sensitive virus, H77S.3 (Figure 1B). By contrast, BWA4C, an iron-binding LOX inhibitor that lacks antioxidant activity, potently suppressed its replication. These results suggest that the ability of LOX inhibitors to stimulate replication of an LPO-sensitive virus is dependent upon antioxidant properties, rather than enzymatic inhibition of LOXs. Consistent with this, replication of HJ3–5, a cell culture-adapted LPO-resistant HCV variant (Yamane et al., 2014), was neither enhanced nor suppressed by these compounds (Figure 1B). The absence of detectable transcripts from LOX genes (ALOX5, ALOX12, ALOX12B, ALOX15, ALOX15B, and ALOX3E) in these cells provides further support for the LOX-independent activity of these inhibitors (Figure S1A).
Figure 1. The iron ligand-type 5-LOX inhibitor BWA4C restricts HCV via promotion of LPO.
(A) Diagrams of HCV RNA genomes expressing Gaussia luciferase (GLuc) reporter with differential sensitivity to LPO. Arrowheads indicate cell culture–adaptive mutations. (B) Dose-dependent effects of cyclooxygenase (COX) and lipoxygenase (LOX) inhibitors on GLuc expression from Huh-7.5 cells transfected with either H77S.3/GLuc or HJ3–5/GLuc RNAs at 72 h. Data are presented as the mean from two independent experiments. (C) Validation of dose-dependent effects of BWA4C on GLuc expression from Huh-7.5 cells transfected with the LPO-sensitive (H77S.3, N.2) and resistant (HJ3–5, H77D) HCV/GLuc RNAs. *p < 0.05, **p < 0.01 versus control (two-way ANOVA with Dunnett’s multiple comparisons test). (D) Suppression of infectious virus production in BWA4C-treated Huh-7.5 cells transected with H77S.3 RNA. Data are the mean ±SD from three independent experiments. **p < 0.01 versus control (two-sided Student’s t-test) (E) Dose-dependent increase of malondialdehyde abundance and cell viability in Huh-7.5 cells treated with BWA4C. (F) Growth kinetics of H77S.3/GLuc and N.2/GLuc RNAs in the presence of BWA4C (1 μM) or BWA4C plus vitamin E (VE, 1 μM) in Huh-7.5 cells. *p < 0.05, **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). Unless otherwise indicated, data are presented as the mean ±SD from three biological replicates.
It was noteworthy that BWA4C, an iron ligand-type 5-LOX inhibitor, specifically inhibited replication of LPO-sensitive HCVs (HCV-N.2 and H77S.3), but not the LPO-resistant HJ3–5 or H77D, which contains LPO-resistant substitutions derived from the TNcc strain in the H77S.3 genetic background (Li et al., 2012; Yamane et al., 2014) (Figure 1C). BWA4C induced a ten-fold reduction in the yield of infectious virus released by H77S.3 RNA–transfected cells (Figure 1D). The impact of BWA4C on viral genome amplification was replicated in experiments with subgenomic replicons that express only the replicase proteins of LPO-sensitive H77S versus resistant JFH1 strains (Figure S1B). BWA4C promoted LPO in a dose-dependent manner, as measured by increases in cell-associated malondialdehyde, without affecting cell viability (Figure 1E). Importantly, the antiviral activity of BWA4C against LPO-sensitive HCVs was completely reversed when combined with a lipophilic antioxidant, vitamin E (VE, α-tocopherol) (Figure 1F), whereas replication of LPO-resistant HCVs was not affected by either treatment (Figure S1C). Efforts to map BWA4C-responsive substitutions within H77S.3 versus the related LPO-resistant H77D replicase suggested the A1672S substitution, located within the transmembrane region of NS4A and previously linked to LPO resistance (Yamane et al., 2014), as a key determinant of BWA4C sensitivity (Figure S1D). Taken collectively, these data provide strong evidence that BWA4C suppresses HCV replication by promoting LPO, rather than inhibiting 5-LOX.
Iron is required for the antiviral activity of BWA4C and for constitutive LPO
BWA4C is unique among other LOX inhibitors in that it is an iron-ligand type acetohydroxamate inhibitor lacking antioxidant property. Thus, we hypothesized that BWA4C-induced LPO might require iron as a cofactor to promote non-enzymatic oxidation. As anticipated, deferoxamine (DFO), an iron chelating agent, abolished the antiviral activity of BWA4C against LPO-sensitive HCVs, H77S.3 and N.2, while having little effect on the LPO-resistant HCVs, HJ3–5 and H77D (Figures 2A, 2B, and S2A). Consistent with a role for labile ferrous iron (Fe2+) in catalyzing the generation of hydroxyl and peroxyl radicals via the Fenton reaction to initiate LPO (Figure 2C) (Collin, 2019), DFO treatment suppressed basal as well as BWA4C-induced LPO (Figure 2D). Importantly, the iron sequestration resulting from treatment with DFO alone was sufficient to enhance replication of LPO-sensitive HCVs (Figures 2E, 2F, and 2G). This enhancement was abolished when DFO treatment was combined with VE, supporting the notion that DFO acts through suppressing LPO (Figure 2H). In contrast, DFO alone, as well DFO in combination with VE, had no effect on the LPO-resistant HJ3–5 virus (Figures 2E, S2B, and S2C). These results highlight the importance of iron as a cofactor for BWA4C-induced LPO in downregulating HCV replication, and also as a crucial regulator of basal LPO that constitutively suppresses HCV replication.
Figure 2. Iron mediates BWA4C-induced suppression of HCV and promotion of constitutive LPO.
(A) Effect of BWA4C (1 μM) or BWA4C plus 10 μM deferoxamine (DFO) on GLuc activities of the indicated HCVs at 72 h. **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). (B) Growth kinetics of H77S.3/GLuc and N.2/GLuc RNAs in the presence of BWA4C (1 μM) or BWA4C plus DFO (10 μM) in Huh-7.5 cells. **p < 0.01 versus vehicle control (two-way ANOVA with Dunnett’s multiple comparisons test). (C) Schematic of non-enzymatic LPO processes involving the Fenton reaction and termination by lipophilic antioxidants. (D) Malondialdehyde abundance of Huh-7.5 cells treated with DFO (10 μM) or DFO plus BWA4C (3 μM) for 48 h. **p < 0.01 (one-way ANOVA with Dunnett’s multiple comparisons test). (E) Dose-dependent effects of DFO on GLuc activities secreted from Huh-7.5 cells transfected with the indicated HCV RNAs at 72 h. **p < 0.01 versus control (two-way ANOVA with Dunnett’s multiple comparisons test). (F) Growth kinetics of GLuc activities secreted from Huh-7.5 cells transfected with H77S.3/GLuc or N.2/GLuc RNAs and treated with DFO (10 μM). **p < 0.01 versus DMSO control (two-way ANOVA with Dunnett’s multiple comparisons test). (G) Effects of DFO on viral RNA abundance in H77S.3 RNA-transfected Huh-7.5 cells. **p < 0.01 (two-sided Student’s t-test). (H) Dose-dependent effects of DFO on replication of H77S.3/GLuc and N.2/GLuc RNAs in the absence and presence of VE (1 μM) at 72 h in Huh-7.5 cells. **p < 0.01 versus no DFO (two-way ANOVA with Dunnett’s multiple comparisons test). Data are presented as the mean ± SD from three biological replicates.
Fatty acid desaturases promote LPO via HUFA synthesis, thereby restricting HCV replication.
Carbon-carbon double bonds in PUFAs are attacked by free radicals to initiate LPO. Thus, we questioned how cellular fatty acid desaturases, that catalyze the formation of such double bonds, affect the iron-dependent production of LPO and susceptibility to HCV infection. Mammalian cells lack an enzyme for de novo synthesis of n-3 and n-6 PUFAs, thus most PUFAs are generally considered to be derived from exogenous sources (i.e. fetal bovine serum (FBS) in the context of cell culture). Essential fatty acids, linoleic acid (LA, 18:2n-6) and α-linolenic acid (ALA, 18:3n-3), are subject to a series of desaturation events mediated by FADS1 and FADS2, and elongation of their acyl tails by ELOVL2/5, to produce HUFAs such as arachidonic acid (ARA, 20:4n-6) or docosahexaenoic acid (DHA, 22:6n-3) (Figure 3A). Desaturation of saturated fatty acids (SFAs) to the monounsaturated fatty acids (MUFAs), oleic acid and palmitoleic acid, is mediated by stearoyl-CoA desaturase (SCD) (Figure 3A). We confirmed that the antiviral activity of PUFAs against HCV correlates significantly with the number of carbon-carbon double bonds catalyzed by FADS1/2 (LA < dihomo-γ-linolenic acid (DGLA) < ARA) (Figure 3B), whereas exogenously added oleic acid, which suppresses LPO, enhances HCV replication (Magtanong et al., 2019; Yamane et al., 2014).
Figure 3. Fatty acid desaturases promote LPO and restrict HCV replication.
(A) Schematic overview of fatty acid metabolic pathways catalyzing synthesis of polyunsaturated fatty acids (PUFAs, left) or mono-unsaturated fatty acids (right). (B) Dose-dependent effects of PUFAs with different levels of desaturation on GLuc expression from H77S.3/GLuc-transfected Huh-7.5 cells at 72 h (right). *p < 0.05, **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). (C) Fatty acid profiles of fetal bovine serum (FBS) and lipoprotein-deficient serum (LPDS). Area of circles is proportional to the concentration of total fatty acids (μg ml−1) in each serum. (D) Effects of siRNAs directed to indicated fatty acid desaturases on GLuc expression from Huh-7.5 cells transfected with H77S.3/GLuc RNA and cultured in the media supplemented with normal FBS or LPDS for 72 h. *p < 0.05, **p < 0.01 versus Control siRNA (two-way ANOVA with Dunnett’s multiple comparisons test). (E) Effects of siRNAs directed to indicated fatty acid desaturases on the viral RNA abundance in Huh-7.5 cells transfected with H77S.3/GLuc RNA at 72 h. **p < 0.01 versus control (one-way ANOVA with Dunnett’s multiple comparisons test). (F) Malondialdehyde abundance in Huh-7.5 cells transfected with indicated siRNAs for 72 h or treated with a SCD inhibitor, CAY10566 (10 μM) for 48 h. *p < 0.05, **p < 0.01 versus control (one-way ANOVA with Dunnett’s multiple comparisons test). (G) Dose-dependent effects of CAY10566 on GLuc expression in Huh-7.5 cells transfected with indicated HCV/GLuc RNAs at 72 h. *p < 0.05, **p < 0.01 versus DMSO control (two-way ANOVA with Dunnett’s multiple comparisons test). (H) Effects of silencing of indicated fatty acid desaturases with LPDS medium in the absence and presence of 1 μM VE. *p < 0.05, **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). (I) Effects of indicated siRNAs on replication of wild type H77c/GLuc (genotype 1a) and HCV-N/GLuc (genotype 1b) RNAs in Huh-7.5/SEC14L2 cells. Data are pooled from three biological replicates from two independent experiments (line at mean). *p < 0.05, **p < 0.01 versus control (one-way ANOVA with Dunnett’s multiple comparisons test). (J) Growth kinetics of HCV-N/GLuc RNA in Huh-7.5/SEC14L2 cells transfected with indicated siRNAs (DAA, 30 μM sofosbuvir). *p < 0.05, **p < 0.01 versus control siRNA (two-way ANOVA with Dunnett’s multiple comparisons test). Unless otherwise indicated, data are presented as the mean ±SD from three biological replicates.
We depleted these desaturases via specific siRNAs (Figures S3A and S3B) and determined how this effects viral replication in cells cultured in medium supplemented with either normal FBS or lipoprotein-depleted FBS (LPDS, in which 87% of PUFA are removed) in order to minimize the effect of exogenous fatty acids (Figure 3C). Despite an overall negative impact on cell proliferation (Figure S3C), depletion of FADS1, FADS2, and SCD enhanced H77S.3 replication (Figures 3D and 3E). Silencing FADS2, which catalyzes the first and rate-limiting step in the synthesis of HUFAs, had the strongest effect. By contrast, most of these siRNAs, except for a few siRNAs targeting FADS1 (#1 and #2), suppressed replication of the LPO-resistant HCVs, HJ3–5 and H77D, consistent with previous reports showing a proviral role for the JFH1-based virus (Hofmann et al., 2018; Lyn et al., 2014; Nguyen et al., 2014) (Figure S3D). Results were similar in cells cultured in media supplemented with normal FBS or LPDS containing minimal exogenous fatty acids (Figures 3D and S3D), suggesting that impaired intracellular desaturation of fatty acids accounted for the increases in H77S.3 replication.
While SCD is not known to catalyze the synthesis of PUFAs under normal conditions, we found that siRNA depletion of SCD significantly reduced the cellular LPO level, albeit with lower efficiency than either FADS1 or FADS2 depletion (Figure 3F). Likewise, pharmacologic inhibition of SCD with CAY10566 reduced the LPO level (Figure 3F) and recapitulated the effects of SCD siRNA on the LPO-sensitive (H77S.3 and N.2) versus resistant (HJ3–5) viruses (Figure 3G). Eliminating LPO via VE supplementation significantly attenuated the enhanced replication of H77S.3 in cells depleted of these desaturases, supporting an LPO-dependent mechanism of action (Figure 3H). More importantly, depleting these desaturases similarly enhanced the replication of fully-wild type HCVs (genotype 1a H77c and genotype 1b HCV-N) in Huh-7.5 cells stably expressing SEC14L2, a protein that facilitates efficient replication of clinical HCV isolates (Saeed et al., 2015) (Figures 3I and 3J). These latter data provide strong support for the relevance of these desaturases to clinical HCV isolates, and their potential role in HCV infections in vivo.
SCD-dependent synthesis of 5,8,11-eicosatrienoic acid, 20:3n-9 (Mead acid), regulates LPO and HCV replication.
It was surprising to find that H77S.3 replication was enhanced in SCD-depleted cells, given that SCD has been reported to be a pro-viral factor (Hofmann et al., 2018; Lyn et al., 2014; Nguyen et al., 2014), and has not been recognized to play a canonical role in HUFA synthesis. Fatty acid profiling of Huh-7.5 cells transfected with HCV RNA revealed abundant Mead acid (5,8,11-eicosatrienoic acid, 20:3n-9), an atypical long-chain HUFA synthesized from SCD-catalyzed oleic acid followed by FADS1/2-dependent desaturation (Figures 4A and 4B) (Ichi et al., 2014). Mead acid was the third most abundant HUFA, after ARA and DHA, in these cells (Figure 4A). This was unexpected, given that Mead acid is typically absent under normal conditions; its synthesis is activated by a deficiency of essential fatty acids (Figure 4B). We confirmed a similar abundance of Mead acid (~1% of total fatty acids) in primary human hepatocyte cultures (Figure 4A).
Figure 4. Non-canonical fatty acid desaturation pathway catalyzing Mead acid synthesis regulates LPO and HCV replication in vitro.
(A) Relative fatty acid abundances in Huh-7.5 cells transfected with indicated HCV RNAs or replication-incompetent GND control RNA for 48 h versus primary human hepatocytes (PHH). (B) Schematic overview of fatty acid metabolic pathways catalyzing synthesis of Mead acid. (C) Fatty acid abundances in Huh-7.5 cells transfected with indicated siRNAs and cultured with medium supplemented with LPDS for 3 days. (D) Levels of indicated glycerophospholipids and phosphatidylinositol phosphates (PIPs) determined by LC-MS/MS in Huh-7.5 cells. (E) Effects of Mead acid (50 μM) on GLuc expression from Huh-7.5 cells transfected with H77S.3/GLuc and HJ3–5/GLuc RNAs with and without 1 μM vitamin E (VE) for 72 h. *p < 0.05, **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). Data are presented as the mean ±SD from three biological replicates.
Mead acid was among the HUFAs most reduced by siRNA depletion of FADS1, FADS2 or SCD (Figure 4C), with FADS2 depletion causing a 90% reduction in its abundance, in contrast to ~30% reductions in ARA or DHA. This suggests that the intracellular abundance of multiple HUFAs, including Mead acid and ARA, accounts for the efficient reduction in the malondialdehyde levels following FADS2-depletion in Huh-7.5 cells (Figure 3F). The distribution of Mead acid (C20:3) among the glycerophospholipids was similar to that of ARA (C20:4), with particular enrichment in phosphatidylethanolamine (PE) and phosphatidylinositol (PI) (Figure 4D). Interestingly, the enrichment of C20:3 and ARA was found in PI-phosphate species including PI(4)P, a crucial molecule required for formation of the double-membrane vesicles of the HCV replication organelle. While C20:3 can represent both Mead acid and DGLA, fatty acid profiling identified a predominance of Mead acid in Huh-7.5 cells (Figure 4A). Similar to other HUFAs, exogenous Mead acid efficiently suppressed H77S.3 replication, but not the LPO-resistant HJ3–5, in a VE-reversible manner (Figure 4E). These results establish Mead acid as an important HUFA synthesized by SCD and FADS1/2 and a key molecule that suppresses HCV replication in Huh-7.5 cells.
Fatty acid profiling of humanized liver tissue from HCV-infected chimeric mice revealed an increased abundance of HUFAs associated with infection (Figure S4). In contrast to hepatoma cells, LA accounted for the majority of PUFAs, similar to primary hepatocytes (Figure 4A), whereas Mead acid comprised only 0.2% (HCV-infected) versus 0.1% (noninfected) of total fatty acids (Figure S4). Nonetheless, the elevated abundance of FADS1/2-catalyzed HUFAs is indicative of specific activation of FADS-dependent pathways in the HCV-infected liver, consistent with increased oxidative stress and LPO products observed in HCV-infected livers (Farinati et al., 1995; Paradis et al., 1997).
FADS2 regulates ferroptosis and HCV replication
Our data show that HCV infection is restricted by iron-dependent promotion of LPO (Figures 2D, 2E, 2F, 2G, and 2H), a process reminiscent of ferroptosis (Dixon et al., 2012). We thus questioned whether FADS2-regulated HUFA synthesis can affect the cellular susceptibility to ferroptosis. Although FADS2 has been described as anti-ferroptotic (Jiang et al., 2017), we unexpectedly found that siRNA depletion of FADS2, like treatment with DFO or the ferroptosis inhibitor ferrostatin-1 (Fer1) (Dixon et al., 2012), protected cells from ferroptosis induced by the cystine/glutamate transporter (xCT) inhibitor erastin and the GPX4 inhibitor RSL3 in Huh-7.5 cells (Figures 5A, 5B, and S5A). FADS2 depletion similarly protected immortalized primary hepatocytes (PH5CH8) and lung carcinoma cells (A549) from erastin-induced cell death (Figure S5A). In contrast to FADS2, siRNA silencing of SCD or FADS1 failed to protect Huh-7.5 cells from ferroptosis (Figure S5B). Protection elicited by FADS2 depletion was as effective as siRNA depletion of ACSL4, an essential regulator of ferroptosis that catalyzes synthesis of long chain PUFA-CoA (Doll et al., 2017; Kagan et al., 2017) (Figure S5B). We confirmed the effect of FADS2 silencing using 3 independent siRNAs as well as CRISPR/Cas9-genome editing (Figures 5C and S5C), and demonstrated that susceptibility to erastin-induced cell death could be restored by exogenous supplementation with ARA (Figure 5C). Conversely, we found that human embryonic kidney cells (293FT), which are resistant to ferroptosis and express only low levels of FADS2, could be rendered susceptible to erastin-induced cell death by lentiviral expression of Flag-tagged FADS2 (Figure 5D and 5E). Importantly, this susceptibility was efficiently blocked by DFO or Fer1 (Figure S5D). Whereas FADS2 expression significantly increased downstream production of HUFAs, including Mead acid, ARA and DGLA in 293FT cells (Figure 5F), exogenous HUFA supplementation was sufficient to confer susceptibility to erastin-induced ferroptosis (Figures 5G and S5E). In contrast, LA failed to induce ferroptosis in the absence of FADS2 expression (Figure 5G). Thus, FADS2 expression is a crucial determinant of ferroptosis in 293FT cells. Consistent with this, publicly available data from the Cancer Therapeutics Response Portal (CTRP) (Rees et al., 2016) indicate that FADS2 expression correlates with sensitivity to erastin-induced cell death in large intestine cancer cell lines (Figure S5F). Taken together, these results indicate that FADS2 expression determines ferroptosis sensitivity in multiple cell types.
Figure 5. FADS2-dependent synthesis of HUFAs regulates ferroptosis and restricts HCV replication.
(A) Cytotoxicity of erastin and RSL3 in Huh-7.5 cells transfected with FADS2 siRNA or control assessed 24 h after treatment using WST-8 reagent. **p < 0.01 versus DMSO control (one-way ANOVA with Dunnett’s multiple comparisons test). (B) Microscopic images of erastin-treated cells as in A (original magnification, 40 ×). Scale bar, 40 μm. (C) Cell viability of FADS2-depleted Huh-7.5 cells or control after treatment with 10 μM erastin or erastin plus 20 μM arachidonic acid (ARA) for 24 h. **p < 0.01 (two-way ANOVA with Dunnett’s multiple comparisons test). (D) Immunoblots showing lentiviral expression of Flag-tagged FADS2 protein in 293FT cells. (E) Dose-dependent cytotoxicity of erastin in 293FT cells transduced with FADS2-Flag or empty vector assessed 48 h after treatment. Microscopic images of erastin-treated 293FT cells is shown on right (original magnification, 40×). Scale bar, 80 μm. **p < 0.01 versus vector control (two-way ANOVA with Dunnett’s multiple comparisons test). (F) Fatty acid abundances in 293FT cells stably expressing FADS2-Flag versus empty vector control. *p < 0.05, **p < 0.01 (two-way ANOVA with Sidak’s multiple comparisons test). (G) Cell viability of 293FT cells treated with DMSO or 10 μM erastin plus either ethanol (Vehicle), 50 μM linoleic acid (LA), dihomo-γ-linolenic acid (DGLA), arachidonic acid (ARA), or Mead acid (MA). **p < 0.01 versus DMSO control (one-way ANOVA with Dunnett’s multiple comparisons test). (H) Effects of indicated PUFAs on H77S.3/GLuc replication at 72 h in Huh-7.5 cells transfected with FADS2 or control siRNA. **p < 0.01 versus vehicle control (two-way ANOVA with Dunnett’s multiple comparisons test). (I) Dose-dependent effects of erastin (left) and ferrostatin-1 (Fer1, right) on GLuc activities secreted from Huh-7.5 cells transfected with the indicated HCV RNAs at 72 h. **p < 0.01 versus DMSO control (two-way ANOVA with Dunnett’s multiple comparisons test). (J) Effects of erastin (1 μM) and Fer1 on GLuc activities secreted from Huh-7.5 cells transfected with FADS2 siRNA or control at 72 h. *p < 0.05, **p < 0.01 (one-way ANOVA with Dunnett’s multiple comparisons test (erastin) or two-way ANOVA with Sidak’s multiple comparisons test (Fer1)). Data are presented as the mean ±SD from three biological replicates.
Given the essential role of FADS2 in the execution of ferroptosis, we hypothesized that FADS2 might restrict HCV replication in a manner similar to its regulation of ferroptosis. Like ferroptosis, we found that exogenous LA requires FADS2 to suppress HCV replication, in contrast to DGLA, ARA or Mead acid, which significantly inhibit it in the absence of FADS2 expression (Figure 5H). Moreover, subcytotoxic concentrations of erastin (~1 μM) specifically inhibited replication of the LPO-sensitive H77S.3, but not resistant HJ3–5 (Figure 5I, left panel), whereas inhibiting ferroptosis with Fer1 specifically enhanced H77S.3 (Figure 5I, right panel). Importantly, the contrary effects of erastin and Fer1 on H77S.3 replication were completely ablated in FADS2-depleted cells (Figure 5J). Collectively, these data establish FADS2 as an important pro-ferroptotic factor that also restricts HCV replication.
Compounds that target the ferroptosis pathway regulate HCV replicase conformation and alter the sensitivity to the DAAs.
LPO regulates HCV replicase activity by modulating the conformation of the replicase components, NS3/4A protease and NS5B polymerase, which are both tethered on the ER membrane (Yamane et al., 2014). Thus, we questioned whether chemically manipulating the ferroptosis pathway could influence HCV replicase conformation, and alter the affinity of DAAs for their ligands. Treating H77S.3- or N.2-infected cells with the ferroptosis inhibitors, DFO or Fer1, significantly increased the 50% effective concentration (EC50) of glecaprevir, an NS3/4A protease inhibitor, and sofosbuvir, a potent NS5B polymerase inhibitor, thereby attenuating their antiviral effects (Figures 6A, 6B, 6C, and S6A). Conversely, a subcytotoxic concentration of erastin lowered the EC50 of glecaprevir, facilitating clearance of H77S.3 and N.2, while having no effect on sofosbuvir (Figures 6A, 6B, 6C, 6D, and S6A). The ferroptosis inhibitors also increased the EC50 of glecaprevir in cells depleted of FADS2, significantly attenuating its antiviral effect against both H77S.3 and N.2 (Figures 6E, 6F, and 6G). This effect was also specific to glecaprevir, and not observed with sofosbuvir (Figures 6E and 6G). Changes in the EC50 of the DAAs were not due to drug-drug interactions, or altered drug uptake/metabolism, as there were no changes in the EC50 against the LPO-resistant HJ3–5 (Figures S6B and S6C). In an effort to validate these findings in vivo, we infected mice with chimeric human livers with LPO-sensitive HCV, and after viremia was well established, assessed the response to the NS3/4A inhibitor glecaprevir with or without combination treatment with the ferroptosis inducer, imidazole-ketone-erastin (IKE) (Zhang et al., 2019). HCV viremia levels were monitored for 7 days following the start of treatment. Despite the small numbers of animals in each group (n = 3), the IKE-treated group demonstrated a strong trend toward lower levels of viremia on 4 day post-treatment with the p-value close to statistical significance (p = 0.064), without noticeable changes in serum albumin levels or body weight (Figures 6H and 6I). Increases in the level of viremia in both treatment groups at day 7 were indicative of the emergence of glecaprevir resistance. Taken collectively, these results show that ferroptosis signals are critically linked to HCV replicase function.
Figure 6. FADS2-regulated ferroptosis pathway affects the EC50 of direct-acting antiviral agents.
(A) Dose-dependent inhibition of H77S.3 (upper panels) and N.2 (lower panels) replication by the NS3/4A inhibitor glecaprevir (left) and the NS5B inhibitor sofosbuvir (right) in the presence of 1 μM erastin, 1 μM ferrostatin-1 (Fer1), 10 μM deferoxamine (DFO) or DMSO vehicle. (B) Inhibition of H77S.3 (upper panels) and N.2 (lower panels) replication by glecaprevir (10 nM, left panels) and sofosbuvir (300 nM, right panels) in the presence of erastin, Fer1 (each 1 μM) or DFO (10 μM), assessed by quantifying GLuc secreted at 72 h after drug addition. DAA, direct-acting antiviral. (C) EC50 values of direct-acting antivirals against H77S.3, N.2 and HJ3–5 viruses in the presence of erastin, Fer1 (each 1 μM), or DFO (10 μM). Assays were carried out as in A. Red bars represent limits of the 95% confidence intervals of EC50 values calculated from Hill plots. (D) Kinetics of H77S.3 (top) and N.2 (bottom) replication. Assays were carried out as in B. (E) Effect of siRNAs targeting FADS2 or non-targeting control siRNA on inhibition of H77S.3 replication by glecaprevir (left) and sofosbuvir (right). (F) Inhibition of H77S.3 (top) and N.2 (bottom) replication by 10 nM glecaprevir in Huh-7.5 cells transfected with FADS2 or non-targeting control siRNA. (G) EC50 values of direct-acting antivirals against H77S.3 in cells transfected with FADS2 or non-targeting control siRNA. Assays were carried out as in E. Red bars represent limits of the 95% confidence intervals of EC50 values calculated from Hill plots. (H) Effect of imidazole-ketone-erastin (IKE, n = 3) or vehicle control (n = 3) on glecaprevir-induced suppression of viral replication in HCV-infected chimeric mice with humanized livers (two-sided Student’s t-test). (I) Serum albumin levels and body weight changes of the chimeric mice shown in H. Unless otherwise indicated, data are presented as the mean ±SD from three technical replicates and represent two (A-D,F) or three (E) independent experiments.
Discussion
The replication of HCV, like all hepatotropic viruses, can be restricted by cell-intrinsic innate immune responses in hepatocytes. However, HCV is unique among these viruses in terms of its proclivity for persistent infection, and the inhibition of its replication cycle by metabolic factors that enhance LPO (Yamane et al., 2014). We demonstrate here that LPO-mediated restriction of HCV replication is driven primarily by the iron-mediated peroxidation of FADS2-dependent HUFAs. Moreover, we show that FADS2-catalyzed desaturation of fatty acids, involving non-canonical desaturation of oleate to produce Mead acid in addition to the canonical synthesis of arachidonate, is a rate-limiting step controlling cellular susceptibility to ferroptosis. The complete elimination of erastin-induced suppression of HCV replication, coupled with inhibition of ferroptotic cell death, following FADS2 depletion supports a model in which FADS2-regulated ferroptosis restricts HCV.
Membrane proximal domains of the NS3/4A and NS5B proteins, essential components of the HCV replicase complex, act as LPO sensors that undergo conformational alterations upon exposure to LPO (Yamane et al., 2014). Results from experiments with ferroptosis inhibitors (DFO or Fer1) and an inducer (erastin) suggest that NS3/4A is a more versatile sensor than NS5B, and capable of sensing both ‘constitutive’ and ‘induced’ LPO. It is likely that an LPO sensor domain exists within the transmembrane domain of NS4A (Yamane et al., 2014) that functions through a direct interaction with oxidation-sensitive lipids, such as PE, the primary target for oxidation in the ER compartment (Doll et al., 2017; Kagan et al., 2017), or possibly PI(4)P, a key component of the replicase membrane that is highly enriched in HUFA acyl tails (Figure 4D). Further studies of peroxidation-sensitive membrane lipids in liver tissue are likely to expand our understanding of the molecular switch that regulates the viral replicase and its affinity for DAAs.
A preliminary experiment carried out in a human liver chimeric mouse model provides additional support for the role of ferroptosis in LPO-mediated restriction of HCV replication. These results suggest that co-treatment of infected mice with IKE in combination with glecaprevir improves antiviral suppression by the DAA in vivo (Figure 6H). Experiments with larger numbers of animals, more intensive monitoring of viremia and efforts to optimize the timing, dosage and type of ferroptosis agonist used to enhance the antiviral response, will be required to determine the extent to which ferroptosis agonists may boost the response to extremely potent DAAs like glecaprevir in vivo. We speculate that manipulating antioxidant parameters related to glutathione-independent mechanisms, such as FSP1-regulated coenzyme Q10 metabolism and dietary intake of lipophilic antioxidants, might be required to maximize the response to ferroptosis agonists in vivo.
Despite the inhibition of HCV replication, the impact of hepatic ferroptosis on viral pathogenesis is uncertain. Hepatic iron deposition, which results from reduced synthesis of hepcidin and increased serum ferritin in chronic hepatitis C, can cause oxidative stress and may be exacerbated by alcohol consumption (Milic et al., 2016). While increased iron accumulation and LPO could be beneficial in reducing HCV load, hepatic ferroptosis could promote liver damage and proinflammatory responses, as recently demonstrated in non-alcoholic steatohepatitis (Tsurusaki et al., 2019).
Although it has a predominant role in catalyzing HUFA synthesis, FADS2 paradoxically has been described only as a suppressor of ferroptosis (Jiang et al., 2017). The importance of this HUFA biosynthetic pathway has been highlighted recently by the discovery that loss of FADS1 is a factor driving ferroptosis resistance in gastric cancers (Lee et al., 2020). In contrast, we show that FADS1 depletion leads to an accumulation of DGLA (20:3n-6) (Figure 4C) and thus fails to confer resistance to erastin-induced death in hepatoma cells. Our data, derived from both depletion and overexpression experiments, demonstrate that FADS2 plays a pivotal role in both basal and induced LPO restricting HCV replication, as well as ferroptotic cell death in multiple cell lines, including A549, PH5CH8, and 293FT. Furthermore, we show that LA, an essential PUFA, is by itself functionally inert, and must be desaturated by FADS2 to elicit pro-ferroptotic effects and suppress HCV.
We unexpectedly identified Mead acid as the third most abundant HUFA in multiple cell types, including primary human hepatocytes (Figure 4A), and the most responsive to the expression status of FADS2 among other HUFAs in multiple cell types (Figures 4C and 5F). These results suggest that the non-canonical metabolism of n-9 HUFAs by FADS2 predominates in cell culture, including both established cell lines and primary hepatocytes. This could reflect a general deficiency of essential fatty acids in cultured cells, in which low levels of essential fatty acids are supplied from FBS (Else, 2020), compared to liver tissue in situ where LA comprises the majority of PUFAs (Figure S4). It would be of interest to determine whether cancer cells or certain immortalized cell lines exploit the desaturation of MUFA to benefit their proliferation in environments with low essential fatty acids. Given the role of FADS2 in increasing cancer plasticity (Vriens et al., 2019), additional investigations to determine the spatiotemporal control of FADS2 expression and its metabolites in vivo should be a priority. Such studies would expand our understanding of the mechanisms underlying ferroptosis sensitivity in cancers, and the unique ability of HCV to downregulate its replication and persist in its host.
Significance
The oxidative degradation of polyunsaturated fatty acids (PUFAs), designated lipid peroxidation (LPO), is recognized as a crucial cellular process that regulates ferroptosis, an iron-dependent non-apoptotic cell death, and that downregulates hepatitis C virus (HCV) replication. Endogenous LPO alters the conformation of the viral RNA replicase tethered to endoplasmic reticulum-derived membranes, impairing its ability to replicate the HCV genome. Yet, how LPO is regulated to induce ferroptosis or restricts HCV replication remains incompletely characterized. In this manuscript, we define metabolic processes that promote LPO and show that LPO restriction of HCV infection is iron-dependent and closely parallels ferroptosis. Whereas ferroptosis has not been recognized to have an antiviral effect on any virus, we show that factors promoting ferroptosis disrupt the HCV replicase, altering its conformation and rendering it more susceptible to inhibition by direct-acting antivirals targeting several components of its replicase. Through siRNA-based depletion of fatty acid desaturases involved in PUFA metabolism, we demonstrate that FADS2, by catalyzing the canonical and non-canonical desaturation of fatty acids, functions as a master-regulator of HCV replication in infected cells and is a key factor in sensitizing cells to ferroptosis. Importantly, fatty acid profiling reveals the unexpected predominance of n-9 fatty acid metabolism that synthesizes Mead acid in the promotion of endogenous LPO and ferroptosis sensitivity in multiple cell types. Our data demonstrate the functional importance of the FADS2-regulated PUFA biosynthesis pathway in ferroptosis, and suggest the possibility that this pathway might be pharmacologically manipulated to attenuate viral replication.
STAR Methods
RESOURCE AVAILABILITY
Lead Contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Dr. Daisuke Yamane (yamane-ds@igakuken.or.jp).
Material Availability
All unique reagents generated in this study are available from the Lead Contact without restriction.
Data and Code Availability
All data reported in this paper will be shared by the lead contact upon request. This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.
EXPERIMENTAL MODEL AND SUBJECT DETAILS
Cells
Huh-7.5 human hepatoma cells, Human embryonic kidney 293FT cells, PH5CH8 immortalized human hepatocytes, and A549 lung carcinoma cells were mycoplasma-free and cultured in pyruvate-free Dulbecco’s modified Eagle’s medium (DMEM), High Glucose supplemented with 10% fetal bovine serum (FBS), 1×GlutaMAX-I and 1×MEM Non-Essential Amino Acids Solution (Thermo Fisher Scientific).
Human liver chimeric uPA-SCID mice
Chimeric mice were purchased from PhoenixBio Co. (Hiroshima, Japan). All animal work was performed by PhoenixBio, in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the local Animal Ethics Committee of PhoenixBio. The chimeric mice were generated by transplanting human primary hepatocytes into severe combined immunodeficient mice (3-week-old male) carrying the urokinase plasminogen activator transgene under control of an albumin promoter (Alb-uPA). Liver tissue samples isolated from animals inoculated with HCV-RMT (genotype 1a, GenBank accession number AB520610) and HCVgenotype4a/KM (genotype 4a, AB795432) were described previously (Katsume et al., 2013).
To examine the effects of the ferroptosis inducing compound imidazole-ketone-erastin (IKE) on HCV replication in vivo, 4-month old male chimeric mice were inoculated with serum containing 1 × 105 genome copies of genotype 1b HCV strain, and serum HCV RNA reached >3 × 106 copies per ml in all infected animals at 8 week post inoculation. Mice were administered orally a viral protease inhibitor (glecaprevir, 2.85 mg/kg, once daily) in combination with intraperitoneal injection of IKE (23 mg/kg dissolved in 5% DMSO/95% Hanks’ Balanced Salt Solution at pH 4.0, once daily) or vehicle control, starting at 8 week post inoculation for 7 consecutive days. Mouse serum samples were obtained for HCV RNA or human albumin determination. HCV RNA was extracted and quantified as described previously (Katsume et al., 2013).
METHOD DETAILS
Reagents and antibodies
AA-861, BWA4C, α-tocopherol, linoleic acid, arachidonic acid, dimethyl sulfoxide were from Sigma. CAY10566, Mead acid, dihomo-γ-linolenic acid, erastin, ferrostatin-1, and (1S,3R)-RSL3 were from Cayman Chemical. SC560, SC236, MK886, Zileuton, BAY-X1005, PD146176, and deferoxamine were from Tocris Bioscience. PSI-7977 (Sofosbuvir) and Glecaprevir were from Chemscene. Cell viability was determined using Cell Counting Kit-8 (DOJINDO, Japan). Lipoprotein-deprived serum (LPDS) was prepared by incubating the heat-inactivated FBS with fumed silica (Sigma, S5130) overnight, followed by removal of the silica by centrifugation at 2,000g for 20 min and filtration using a 0.22 μm filter device.
Primary antibodies to SCD (1:500 dilution, #2438) was from Cell Signaling Technology; FADS1 (1:1,000 dilution, #27533) was from Cayman Chemical; FADS2 (1:1,000 dilution, A10270) were from ABclonal; FADS2 (1:1,000 dilution, 28034–1-AP) was from Proteintech; HCV NS3 (1:500, ab13830) was from Abcam; GAPDH was from Wako (1:4,000 dilution, 016–25523). IRDye 680 or 800 secondary antibodies including #926–32211, #926–32212, #926–32214, #926–68020 and #926–68073 (1:20,000) were from LI-COR.
Viruses
pH77S.3, pH77S.3/GLuc, pHCV-N.2/GLuc, pJFH1-QL/GLuc, pHJ3–5/GLuc, pH77D/GLuc, pH77c/GLuc, and pHCV-N/GLuc were described previously (Yamane et al., 2014).
HCV infectivity assays.
Huh-7.5 cells were seeded at 5×104 cells per well into 48-well plates 24 h before inoculation with 100 μl of culture medium. Cells were fed with media containing 100 mM HEPES and 1 μM VE 24 h later to facilitate visualization of core protein expression, fixed with methanol/acetone (1:1) at −20°C for 10 min 72 h post-inoculation, and stained for intracellular core antigen with a mouse monoclonal antibody (Clone 31–2, 5 μg per m1). Clusters of infected cells identified by staining for core antigen were considered to constitute a single infectious focus, and the data expressed as focus-forming units (FFU) per ml.
Viral RNA transcription and transfection
In vitro transcription of HAV or HCV RNA was carried out using T7 RiboMAX™ Express Large Scale RNA Production System (Promega) using 2 μg of SmaI (HAV)- or XbaI (HCV)-linearized DNA template in a 20 μl reaction volume and incubated at 37°C for 1 h. Transfection of viral RNA was performed by electroporating 5 μg of in vitro transcribed viral RNA into 5×106 Huh-7.5 cells with a Gene Pulser Xcell Total System (250V, 950 μF and 50 Ω). Alternatively, Huh-7.5 cells plated on a 96-well plate were transfected with 50 ng viral RNA using TransIT®-mRNA Transfection Kit (Mirus) at a ratio of 1:2:2 (RNA:mRNA Boost Reagent:TransIT-mRNA Reagent) as per manufacturer’s instruction (Yamane et al., 2014).
Lentiviral plasmids, production, and transduction
The lentiviral transfer plasmids encoding FADS2 gene fused with C-terminal Flag sequence was created by PCR amplifying the host genes using cDNA derived from Huh-7.5 cell total RNA as template and primers flanked by XbaI and NheI restriction sites as described (Yamane et al., 2019). To generate FADS1/2 knockout Huh-7.5 cells, sgRNAs listed in Table S1 were cloned into BsmBI-digested lentiCRISPRv2 plasmid (Shalem et al., 2014). Lentivirus production was carried out by co-transfection of individual lentiviral transfer plasmids and lentiviral packaging mix (Sigma) into 293T cells. Lentiviral transduction was performed by supplementation of 8 μg ml−1 polybrene, followed by antibiotic selection with 6 μg ml−1 puromycin.
RNA extraction and quantitative RT-PCR
Total RNA extraction was performed with the RNeasy mini Kit (Qiagen). Quantification of HCV genome RNA and FADS1, FADS2, GPX4, AIFM2, ACSL4, ACTB and ALOX genes was carried out by a One-step quantitative RT-PCR analysis with the Luna Universal One-Step RT-qPCR Kit (NEB) using the primer pairs listed in Table S1.
Immunoblots
Cell lysates were prepared in lysis buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 20 mM sodium fluoride, 1 mM Na3VO4) supplemented with Complete protease inhibitor mixture (Roche). Western blotting was performed with standard methods. Odyssey Infrared Imaging System (Li-COR Biosciences, Lincoln) was used for visualization.
RNA interference
siRNAs listed in Table S1 were obtained from Dharmacon or ThermoFisher Scientific. Cells were transfected with 20 nM siRNA using Lipofectamine RNAiMAX Transfection Reagent (Thermo Fisher Scientific) according to the manufacturer’s protocol.
Luciferase assay
Secreted Gaussia luciferase (GLuc) activity was measured in 20 μl aliquots of the supernatant fluids using the Renilla Luciferase Assay System (Promega) according to the manufacturer’s protocol. The luminescent signal was measured on a Mithras LB940 Multimode Microplate Reader (Berthold).
Lipid peroxidation assay
Malondialdehyde, a product of lipid peroxidation, was quantified by the thiobarbituric acid reactive substances (TBARS) Assay Kit (Cayman Chemical) as described (Yamane et al., 2014). In brief, cells scraped into PBS containing complete protease inhibitor cocktail (Roche) were homogenized by sonication on ice using BioRuptor (Diagenode). The amount of malondialdehyde in 100 μl of cell homogenates was analyzed by a fluorescent method as described by the manufacturer. Lipid peroxidation levels were expressed as the amount of malondialdehyde normalized to the amount of total protein.
Fatty acid analysis
Total cellular lipids, including both free and esterified fatty acids, were extracted by the Bligh-Dyer method. Isolated lipids were methylated with 2.5% sulfuric acid in methanol, and the resulting fatty acid methyl esters extracted in hexane were quantified by gas chromatography–mass spectrometry (GC-MS) analysis with a GC-MS QP2010 (Shimadzu) as described (Ichi et al., 2014). Liver tissue samples from chimeric mice were described previously (Katsume et al., 2013). Unless otherwise indicated, data are presented as relative percentages. For genetic depletion and ectopic expression experiments, data are presented as absolute concentrations to better reflect biological functions of the desaturases.
Phospholipid preparation
Methods for comprehensive phospholipids (PLs) analysis were described previously (Yamane et al., 2019). Briefly, total PLs were extracted from the culture cells with the Bligh-Dyer method. An aliquot of the lower/organic phase was evaporated to dryness under N2, and the residue was dissolved in methanol for LC/MS/MS measurements of PC and PE. To analyze PA, PS, PI and PIPs, another aliquot of the same lipid extract was added with an equal volume of methanol before being loaded onto a DEAE cellulose column (Santa Cruz Biotechnology) pre-equilibrated with chloroform. After successive washes with chloroform/methanol (1:1, v/v), the acidic PLs were eluted with chloroform/methanol/HCl/water (12:12:1:1, v/v), followed by evaporation to dryness to give a residue, which was resolved in methanol. The resultant fraction was subjected to a methylation reaction with TMS-diazomethane before LC/MS/MS analysis.
Mass spectrometric analyses
LC-electrospray ionization-MS/MS analysis was performed with an UltiMate 3000 LC system (Thermo-Fisher Scientific) equipped with HTC PAL autosampler (CTC Analytics). A 10 μL aliquot of the lipid samples was injected and the lipids were separated on Waters X-Bridge C18 column (3.5 μm, 150 mm × 1.0 mm i.d.) at room temperature (25°C) using a gradient solvent system as follows: mobile phase A (isopropanol/methanol/water (5/1/4 v/v/v) supplemented with 5 mM ammonium formate and 0.05% ammonium hydroxide)/mobile phase B (isopropanol supplemented with 5 mM ammonium formate and 0.05% ammonium hydroxide) ratios of 60%/40% (0 min), 40%/60% (0–1 min), 20%/80% (1–9 min), 5%/95% (9–11 min), 5%/95% (11–30 min), 95%/5% (30–31 min), 95%/5% (31–35 min) and 60%/40% (35–45 min). Flow rate was 25 μL/min. PLs species were measured by the selected reaction monitoring (SRM) in positive ion mode with a triple-stage quadrupole mass spectrometer (TSQ Vantage AM, Thermo-Fisher Scientific). The characteristic fragments of individual PLs were detected by the product ion scan (MS/MS mode). Chromatographic peak areas were used for comparative quantitation of each molecular species (e.g. 38:6, 40:6) in a given class of the phospholipids (e.g. PA, PC).
QUANTIFICATION AND STATISTICAL ANALYSIS
Unless noted otherwise, all between-group comparisons were carried out by ANOVA or two-sided Student’s t test using Prism 6.0 software (GraphPad Software, Inc.). p < 0.05 was considered statistically significant. Statistical details of experiments and statistical tests used are described in the figure legends, figures and Results section. p-values can be found in the figure legends and figures.
Supplementary Material
KEY RESOURCES TABLE.
| REAGENT or RESOURCE | SOURCE | IDENTIFIER |
|---|---|---|
| Antibodies | ||
| HCV NS3 (Mouse monoclonal Ab) | Abcam | Cat#ab13830 |
| SCD1 (Rabbit polyclonal Ab) | Cell Signaling Technology | Cat#2438; Lot#:2 |
| FADS1 (Mouse monoclonal Ab) | Cayman Chemical | Cat#27533; Lot#0587925-1 |
| FADS2 (Rabbit polyclonal Ab) | ABclonal | Cat#A10270; Lot#3508656001 |
| FADS2 (Rabbit polyclonal Ab) | Proteintech | Cat#28034-1-AP; Lot#59518 |
| GAPDH (Mouse monoclonal Ab) | Wako | Cat#014-25524; Lot#SAL0162 |
| Actin (Rabbit polyclonal Ab) | Sigma Aldrich | Cat#A2066; Lot#106M4770V |
| Bacterial and virus strains | ||
| Hepatitis C virus (HCV) H77S.3/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) H77S.3 | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) N.2/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) HJ3-5/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) H77D/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) H77c/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) HCV-N/GLuc | Yamane et al., 2014 | N/A |
| Hepatitis C virus (HCV) HCV-RMT GENBANK: AB520610 | Katsume et al., 2013 | N/A |
| Hepatitis C virus (HCV) genotype4a-KM GENBANK: AB795432 | Katsume et al., 2013 | N/A |
| Chemicals, peptides, and recombinant proteins | ||
| CAY10566 | Cayman Chemical | Cat#10012562 |
| Mead acid | Cayman Chemical | Cat#90190 |
| Dihomo-γ-linolenic acid | Cayman Chemical | Cat#90230 |
| Erastin | Cayman Chemical | Cat#17754 |
| Ferrostatin-1 | Cayman Chemical | Cat#17729 |
| (1S,3R)-RSL3 | Cayman Chemical | Cat#19288 |
| Glecaprevir | Chemscene | Cat#CS-8098 |
| PSI-7977 (sofosbuvir) | Chemscene | Cat#CS-0554 |
| Imidazole-ketone-erastin | Selleck Chemicals | Cat#S8877 |
| AA-861 | Sigma Aldrich | Cat#A3711 |
| BWA4C | Sigma Aldrich | Cat#B7559 |
| (±)-α-Tocopherol | Sigma Aldrich | Cat#T3251 |
| Linoleic acid | Sigma Aldrich | Cat#L1376 |
| Arachidonic acid | Sigma Aldrich | Cat#A3611 |
| Silica, fumed powder | Sigma Aldrich | Cat#S5130 |
| Dimethyl sulfoxide | Sigma Aldrich | Cat#D2650 |
| SC560 | Tocris Bioscience | Cat#1550 |
| SC236 | Tocris Bioscience | Cat#3919 |
| MK886 | Tocris Bioscience | Cat#1311 |
| Zileuton | Tocris Bioscience | Cat#3308 |
| BAY-X1005 | Tocris Bioscience | Cat#3541 |
| PD146176 | Tocris Bioscience | Cat#2850 |
| Deferoxamine | Tocris Bioscience | Cat#138-14-7 |
| SYTOX Green | Thermo Fisher Scientific | Cat#S7020 |
| Critical commercial assays | ||
| TBARS assay kit | Cayman Chemical | Cat#: 10009055 |
| Experimental models: Cell lines | ||
| Huh-7.5 | Apath, LLC | N/A |
| 293FT | Thermo Fisher Scientific | Cat#R70007 |
| PH5CH8 | Dr Nobuyuki Kato (Okayama University) | N/A |
| A549 | JCRB | Cat#JCRB0076 |
| Experimental models: Organisms/strains | ||
| uPA/SCID mice with humanized livers | PhoenixBio Katsume et al., 2013 | N/A |
| Primary human hepatocytes (PXB-cells) | PhoenixBio | N/A |
| Oligonucleotides | ||
| See Table S1 | ||
| Software and algorithms | ||
| GraphPad Prism 6.0 | GraphPad Software | https://www.graphpad.com/dl/96314/10B92408/ |
| Image Studio 3.1.4 | LI-COR | N/A |
Highlights.
Iron reacts with FADS2-catalyzed HUFAs to promote endogenous lipid peroxidation
HCV is restricted by lipid peroxidation driven primarily by the Fenton reaction
FADS2 expression level determines the cellular sensitivity to ferroptosis
Promoting ferroptosis sensitizes the HCV replicase to a viral protease inhibitor
Acknowledgements
The authors thank C. M. Rice, T. Wakita, and N. Kato for reagents, R. Maezawa for technical assistance, and Y. Tokunaga, T. Munakata and K. Yamaji for technical advice. This work was supported in part by JSPS KAKENHI (JP17H05070 and JP21H02746 to DY and JP17K00848 and JP20K02383 to II), AMED (JP19fk0210062 to DY), and NIH grants R01-AI095690 (SML).
Footnotes
Declaration of Interests
The authors declare no competing interests.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- Ayala A, Munoz MF, and Arguelles S (2014). Lipid peroxidation: production, metabolism, and signaling mechanisms of malondialdehyde and 4-hydroxy-2-nonenal. Oxid Med Cell Longev 2014, 360438. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bersuker K, Hendricks JM, Li Z, Magtanong L, Ford B, Tang PH, Roberts MA, Tong B, Maimone TJ, Zoncu R, et al. (2019). The CoQ oxidoreductase FSP1 acts parallel to GPX4 to inhibit ferroptosis. Nature 575, 688–692. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Collin F (2019). Chemical Basis of Reactive Oxygen Species Reactivity and Involvement in Neurodegenerative Diseases. Int J Mol Sci 20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, Patel DN, Bauer AJ, Cantley AM, Yang WS, et al. (2012). Ferroptosis: an iron-dependent form of nonapoptotic cell death. Cell 149, 1060–1072. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Doll S, Freitas FP, Shah R, Aldrovandi M, da Silva MC, Ingold I, Grocin AG, Xavier da Silva TN, Panzilius E, Scheel CH, et al. (2019). FSP1 is a glutathione-independent ferroptosis suppressor. Nature 575, 693–698. [DOI] [PubMed] [Google Scholar]
- Doll S, Proneth B, Tyurina YY, Panzilius E, Kobayashi S, Ingold I, Irmler M, Beckers J, Aichler M, Walch A, et al. (2017). ACSL4 dictates ferroptosis sensitivity by shaping cellular lipid composition. Nat Chem Biol 13, 91–98. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Else PL (2020). The highly unnatural fatty acid profile of cells in culture. Prog Lipid Res 77, 101017. [DOI] [PubMed] [Google Scholar]
- Farinati F, Cardin R, De Maria N, Della Libera G, Marafin C, Lecis E, Burra P, Floreani A, Cecchetto A, and Naccarato R (1995). Iron storage, lipid peroxidation and glutathione turnover in chronic anti-HCV positive hepatitis. J Hepatol 22, 449–456. [DOI] [PubMed] [Google Scholar]
- Hangauer MJ, Viswanathan VS, Ryan MJ, Bole D, Eaton JK, Matov A, Galeas J, Dhruv HD, Berens ME, Schreiber SL, et al. (2017). Drug-tolerant persister cancer cells are vulnerable to GPX4 inhibition. Nature 551, 247–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hofmann S, Krajewski M, Scherer C, Scholz V, Mordhorst V, Truschow P, Schobel A, Reimer R, Schwudke D, and Herker E (2018). Complex lipid metabolic remodeling is required for efficient hepatitis C virus replication. Biochim Biophys Acta Mol Cell Biol Lipids 1863, 1041–1056. [DOI] [PubMed] [Google Scholar]
- Huang H, Chen Y, and Ye J (2007). Inhibition of hepatitis C virus replication by peroxidation of arachidonate and restoration by vitamin E. Proc Natl Acad Sci U S A 104, 18666–18670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ichi I, Kono N, Arita Y, Haga S, Arisawa K, Yamano M, Nagase M, Fujiwara Y, and Arai H (2014). Identification of genes and pathways involved in the synthesis of Mead acid (20:3n-9), an indicator of essential fatty acid deficiency. Biochim Biophys Acta 1841, 204–213. [DOI] [PubMed] [Google Scholar]
- Jiang Y, Mao C, Yang R, Yan B, Shi Y, Liu X, Lai W, Liu Y, Wang X, Xiao D, et al. (2017). EGLN1/c-Myc Induced Lymphoid-Specific Helicase Inhibits Ferroptosis through Lipid Metabolic Gene Expression Changes. Theranostics 7, 3293–3305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kagan VE, Mao G, Qu F, Angeli JP, Doll S, Croix CS, Dar HH, Liu B, Tyurin VA, Ritov VB, et al. (2017). Oxidized arachidonic and adrenic PEs navigate cells to ferroptosis. Nat Chem Biol 13, 81–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Katsume A, Tokunaga Y, Hirata Y, Munakata T, Saito M, Hayashi H, Okamoto K, Ohmori Y, Kusanagi I, Fujiwara S, et al. (2013). A serine palmitoyltransferase inhibitor blocks hepatitis C virus replication in human hepatocytes. Gastroenterology 145, 865–873. [DOI] [PubMed] [Google Scholar]
- Lee JY, Nam M, Son HY, Hyun K, Jang SY, Kim JW, Kim MW, Jung Y, Jang E, Yoon SJ, et al. (2020). Polyunsaturated fatty acid biosynthesis pathway determines ferroptosis sensitivity in gastric cancer. Proc Natl Acad Sci U S A 117, 32433–32442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Li YP, Ramirez S, Jensen SB, Purcell RH, Gottwein JM, and Bukh J (2012). Highly efficient full-length hepatitis C virus genotype 1 (strain TN) infectious culture system. Proc Natl Acad Sci U S A 109, 19757–19762. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lyn RK, Singaravelu R, Kargman S, O’Hara S, Chan H, Oballa R, Huang Z, Jones DM, Ridsdale A, Russell RS, et al. (2014). Stearoyl-CoA desaturase inhibition blocks formation of hepatitis C virus-induced specialized membranes. Sci Rep 4, 4549. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Magtanong L, Ko PJ, To M, Cao JY, Forcina GC, Tarangelo A, Ward CC, Cho K, Patti GJ, Nomura DK, et al. (2019). Exogenous Monounsaturated Fatty Acids Promote a Ferroptosis-Resistant Cell State. Cell Chem Biol 26, 420–432 e429. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Milic S, Mikolasevic I, Orlic L, Devcic E, Starcevic-Cizmarevic N, Stimac D, Kapovic M, and Ristic S (2016). The Role of Iron and Iron Overload in Chronic Liver Disease. Med Sci Monit 22, 2144–2151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Nguyen LN, Lim YS, Pham LV, Shin HY, Kim YS, and Hwang SB (2014). Stearoyl coenzyme A desaturase 1 is associated with hepatitis C virus replication complex and regulates viral replication. J Virol 88, 12311–12325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Paradis V, Mathurin P, Kollinger M, Imbert-Bismut F, Charlotte F, Piton A, Opolon P, Holstege A, Poynard T, and Bedossa P (1997). In situ detection of lipid peroxidation in chronic hepatitis C: correlation with pathological features. J Clin Pathol 50, 401–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rees MG, Seashore-Ludlow B, Cheah JH, Adams DJ, Price EV, Gill S, Javaid S, Coletti ME, Jones VL, Bodycombe NE, et al. (2016). Correlating chemical sensitivity and basal gene expression reveals mechanism of action. Nat Chem Biol 12, 109–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Saeed M, Andreo U, Chung HY, Espiritu C, Branch AD, Silva JM, and Rice CM (2015). SEC14L2 enables pan-genotype HCV replication in cell culture. Nature 524, 471–475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shalem O, Sanjana NE, Hartenian E, Shi X, Scott DA, Mikkelson T, Heckl D, Ebert BL, Root DE, Doench JG, et al. (2014). Genome-scale CRISPR-Cas9 knockout screening in human cells. Science 343, 84–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spearman CW, Dusheiko GM, Hellard M, and Sonderup M (2019). Hepatitis C. Lancet 394, 1451–1466. [DOI] [PubMed] [Google Scholar]
- Tsurusaki S, Tsuchiya Y, Koumura T, Nakasone M, Sakamoto T, Matsuoka M, Imai H, Yuet-Yin Kok C, Okochi H, Nakano H, et al. (2019). Hepatic ferroptosis plays an important role as the trigger for initiating inflammation in nonalcoholic steatohepatitis. Cell Death Dis 10, 449. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Viswanathan VS, Ryan MJ, Dhruv HD, Gill S, Eichhoff OM, Seashore-Ludlow B, Kaffenberger SD, Eaton JK, Shimada K, Aguirre AJ, et al. (2017). Dependency of a therapy-resistant state of cancer cells on a lipid peroxidase pathway. Nature 547, 453–457. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vriens K, Christen S, Parik S, Broekaert D, Yoshinaga K, Talebi A, Dehairs J, Escalona-Noguero C, Schmieder R, Cornfield T, et al. (2019). Evidence for an alternative fatty acid desaturation pathway increasing cancer plasticity. Nature 566, 403–406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yamane D, Feng H, Rivera-Serrano EE, Selitsky SR, Hirai-Yuki A, Das A, McKnight KL, Misumi I, Hensley L, Lovell W, et al. (2019). Basal expression of interferon regulatory factor 1 drives intrinsic hepatocyte resistance to multiple RNA viruses. Nat Microbiol 4, 1096–1104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yamane D, McGivern DR, Wauthier E, Yi M, Madden VJ, Welsch C, Antes I, Wen Y, Chugh PE, McGee CE, et al. (2014). Regulation of the hepatitis C virus RNA replicase by endogenous lipid peroxidation. Nat Med 20, 927–935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yang WS, SriRamaratnam R, Welsch ME, Shimada K, Skouta R, Viswanathan VS, Cheah JH, Clemons PA, Shamji AF, Clish CB, et al. (2014). Regulation of ferroptotic cancer cell death by GPX4. Cell 156, 317–331. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Zhang Y, Tan H, Daniels JD, Zandkarimi F, Liu H, Brown LM, Uchida K, O’Connor OA, and Stockwell BR (2019). Imidazole Ketone Erastin Induces Ferroptosis and Slows Tumor Growth in a Mouse Lymphoma Model. Cell Chem Biol 26, 623–633 e629. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data reported in this paper will be shared by the lead contact upon request. This paper does not report original code. Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.