17,β‐estradiol inhibits hepatitis C virus mainly by interference with the release phase of its life cycle

Abstract Background & Aims Oestrogen and oestrogen‐mediated signalling protect from hepatitis C virus through incompletely understood mechanisms. We aimed to ascertain which phase(s) of hepatitis C virus life cycle is/are affected by oestrogens. Methods Huh7 cells infected with the JFH1 virus (genotype 2a) were exposed to dehydroepiandrosterone, testosterone, progesterone and 17β‐estradiol (tested with/without its receptor antagonist fulvestrant). Dose–response curves were established to calculate half maximal inhibitory concentration values. To dissect how 17β‐estradiol interferes with phases of hepatitis C virus life cycle, its effects were measured on the hepatitis C virus pseudo‐particle system (viral entry), the subgenomic replicon N17/JFH1 and the replicon cell line Huh7‐J17 (viral replication). Finally, in a dual‐step infection model, infectious supernatants, collected from infected cells exposed to hormones, were used to infect naïve cells. Results Progesterone and testosterone showed no inhibitory effect on hepatitis C virus; dehydroepiandrosterone was only mildly inhibitory. In contrast, 17β‐estradiol inhibited infection by 64%‐67% (IC 50 values 140‐160 nmol/L). Fulvestrant reverted the inhibition by 17β‐estradiol in a dose‐dependent manner. 17β‐estradiol exerted only a slight inhibition (<20%) on hepatitis C virus pseudo‐particles, and had no effect on cells either transiently or stably (Huh7‐J17 cells) expressing the N17/JFH1 replicon. In the dual‐step infection model, a significant half maximal inhibitory concentration decline occurred between primary (134 nmol/L) and secondary (100 nmol/L) infections (P=.02), with extracellular hepatitis C virus RNA and infectivity being reduced to a higher degree in comparison to its intracellular counterpart. Conclusions 17β‐estradiol inhibits hepatitis C virus acting through its intracellular receptors, mainly interfering with late phases (assembly/release) of the hepatitis C virus life cycle.


| INTRODUCTION
In contrast with the authentic revolution we are facing in the field of antiviral therapy of hepatitis C, several long-standing issues concerning the natural history of HCV infection remain incompletely understood. Specifically, it has always been puzzling why gender affects so deeply the course of hepatitis C. It has been consistently shown along the years that the ability to spontaneously clear HCV infection is greater in women than in men. [1][2][3][4][5] Moreover, during early chronic infection (1-year postinfection), HCV RNA levels are higher in men than in women, 6 while cirrhotic progression rarely occurs in pre-menopausal women. 7 Based on these observations, it has been proposed that oestrogen and oestrogen-mediated signalling may play a role as protective factors, reducing the disease progression or increasing the chance of clearing the virus. 8 The mechanisms by which this protection occurs are unknown; the prevailing hypothesis advocates that sex hormones bind to specific receptors expressed in immune cells, thereby influencing adaptive and, most importantly, innate immune responses. 9,10 However, oestrogens are not key players in the major mechanism of innate antiviral response, the JAK/STAT pathway, 11 and in other settings, the major impact of sex steroids is on the target tissue, not on immune modulation. 12 Thus, it is not inconceivable that oestrogens may determine a suboptimal environment for viral replication by completely different mechanisms. This possibility is nicely exemplified by a recent study demonstrating that oestrogens favour the cleavage of the tight junction protein occludin, one of the proteins HCV uses to gain access to the hepatocyte. 13 Better understanding of the mechanisms by which factors modulating the natural history of HCV infection act remains a compelling issue in the DAA era, since new anti-HCV drugs are so expensive that many countries restrict their use to patients with significant fibrosis progression. With the present in vitro study, we aimed to better define the antiviral properties of sex hormones on HCV infection, and in particular to ascertain which phase(s) of HCV life cycle is/are affected by oestrogens.

| Plasmidsandinvitrotranscription
The plasmid pJFH1, containing the full-length genomic cDNA sequence of the HCV genotype 2a strain and the N17/JFH1 17,18 plasmid were linearized using XbaI enzyme (New England Biolabs) and then treated with Mung Bean Nuclease (NEB) prior purification. Linearized plasmids were used as a template to generate in vitro transcribed RNA using MEGAscript T7 (Life Technologies, Milan, Italy). The 10 μg of RNA were electroporated into Huh7 cells as previously described. 19

| AntiviralevaluationonHCVcc
Huh7 cells were infected with JFH1 virus at a multiplicity of infection (MOI) of 0.1. The infection was performed as follow using three different models to derive hormone dose-response scales, as shown in Figure 1. In model #1, Huh7 cells were infected in the absence of hormones: after 3 hours the inoculum was replaced with fresh complete medium containing various hormone concentrations. In model #2, 1 hour before infection the cells were pre-treated with each hormone. Subsequently the cells were infected in the presence of the compounds and incubated at 37°C. After the 3 hours the viral inoculum was removed, cells washed and re-fed with complete medium and incubated at 37°C for 72 hours. In model #3, Huh7 cells were pre-treated overnight with hormones, then washed and infected in the absence of drugs.
Viral inhibition was evaluated counting the focus forming units (FFU) after immune staining on fixed cells; results were normalized to DMSO-treated cells (drug vehicle control).

| Combinationtreatment
Huh7 cells were exposed to 17,β-estradiol at the concentration of 200 nmol/L, plus different concentrations of DHEA-S or testosterone as reported in Table 1; in addition, the infections were performed according to the models #2 and #3.

Key points
• Among sex hormones, only 17,β-estradiol exhibits antiviral properties against HCV in two different models of infection.
• The oestrogen-related antiviral effect is abrogated in a dose-dependent manner by the oestrogen-receptor modulator, fulvestrant, confirming that the antiviral action of 17,β-estradiol is indirect.
• Oestrogen stimulation partially reduces viral entry, but has no effect on viral RNA replication.

| Dualstepinfection
Briefly, Huh7 cells were infected and treated with different concentrations of 17,β-estradiol following the models #2 and #3 described above. After 3 days, cells were fixed and stained to obtain data from primary infection. The supernatants, collected from infected cells, were used to perform a secondary infection on naïve Huh7, without any treatment. After other 3 days, the antiviral effect on secondary infection was visualized as described below.

| Immunoperoxidasestaining
Infected cells were fixed in 4% paraformaldehyde. After permeabilization with Triton X-100 0.5%, treatment with H 2 O 2 0.3% and saturation with BSA 3%, cells were visualized with AEC Staining Kit (Sigma Aldrich) using as primary antibody a serum of HCV positive patient, and as secondary antibody a rabbit anti-human IgG labelled with peroxidase (Dako).

| RNAinhibition
Huh7 cells were treated with 17,β-estradiol at the concentration of 200 nmol/L and infected following models #2 and #3 as described above. After 72 hours intracellular RNA was purified using TRIreagent (Sigma-Aldrich), while RNA from the released viral particles was extracted using PureLink Viral RNA kit (Life Technologies).
Reverse transcription was performed with random primers using the same volume (10 μL) for RNAs extracted from supernatant and 1 μg for intracellular RNAs. Viral quantification of RNA was determined by RT-qPCR using the Fast Sybr Master Mix (Life Technologies) with specific primers for the HCV 5′ NTR: 5′-TCCCGGGAGAGCCATAGTG-3′ (sense) 5′-TCCAAGAAAGGACCCAGTC-3′ (antisense) as described previously. 20 The copy number of viral RNA was obtained by absolute quantification using a linear regression on serial dilutions of the pJFH1 plasmid at known concentrations.

| Infectivityinhibitionandspecificinfectivity
Huh7 cells were treated with 17,β-estradiol as described above for RNA inhibition. After 72 hours supernatants were collected and titrated in a 96-well plate following five-fold dilutions or quantitated by RT-qPCR as described above. Infected cells were washed, lysed and titrated as previously reported. 21,22 Specific infectivity was calculated as the ratio between the viral titre, expressed as FFU/mL and the number of RNA copies.
F I G U R E 1 Schematic representation of infection models. In model #1 cells were infected and then treated with E2 for 72 h. In model #2 cells were pretreated for 1 h and then exposed with E2 during the first 3 h of infection. In model #3 cells were pretreated overnight and then infected in absence of E2

| HCVpseudoparticles
HCV pseudoparticles (HCVpp) were generated as previously reported. 23 To evaluate viral inhibition, Huh7 cells were pre-treated for 1 hour with 17,β-estradiol or BAF at the concentration of 10 nmol/L and then infected for 3 hours with HCVpp at a MOI of 0.1 according to model #2 described above. After 72 hours cells lysed and luciferase readings taken. Results were normalized to DMSO-treated cells.

| Statisticalanalysis
Statistical analysis of data was performed using the software package Stata Rel. 13

| Antiviralpropertiesofsexhormonesin relationshipwiththetimingofinfectionandthe exposuretohormones
The sex hormones 17,β-estradiol, progesterone, testosterone and DHEA-S were tested in the three different models of infection, as described in Methods, using dose-response scales covering the physiological ranges listed in Table 1 and Table S1. All the hormones were non-cytotoxic when tested by the MTT assay (data not shown).
In Interestingly, neither testosterone nor DHEA-S proved to be able to boost 17,β-estradiol antiviral effects (Figure 4).

| Roleofintracellularoestrogensignallingin determininganantiviralstate
To assess whether the antiviral effect of the 17,β-estradiol was directly exerted on the virus rather than being dependent on intracellular signalling pathways, we investigated the role of the oestrogen intracellular receptor (ERαβ) in the antiviral response using a selective αβ-oestrogen receptor degrader, Fulvestrant. 24 To this purpose, Huh7 cells were treated with 200 nmol/L 17,β-estradiol in combination with different concentrations of Fulvestrant. The experiment was performed according to the models #2 and #3 described above. As reported in Figure 3B, the inhibition induced by 17,β-estradiol was completely abolished by Fulvestrant, in a dose-dependent manner for both models tested, whereas the treatment with Fulvestrant alone had no effect.

| Dissectinghowoestrogensinterferewith phasesofHCVlifecycle
We initially tested 17,β-estradiol effect on HCV pseudoparticle system (HCVpp), which allows evaluation of the viral entry. As shown in Figure 5A Figure 5C. Based on these results, we performed a dual-step infection, in which infectious supernatants, collected from infected cells exposed to hormones following models #2 and #3, are used to infect naïve cells.
The results obtained in the model #2 showed a significant difference between primary and secondary infection across groups, with IC 50 values of 134 and 100 nmol/L respectively ( Figure 6A). In contrast, in model #3 there was no significant variation between primary and secondary infections. The IC 50 were 120 and 112 nmol/L respectively ( Figure 6B). To validate our findings, we evaluated the inhibitory effect at extracellular and intracellular level, analysing the HCV RNA amount.
As illustrated in Figure 7A,   infectivity of 40% and 50% respectively. Noteworthy, these infectivity data correlate with those of viral RNA levels shown in Figure 7A.
Finally, we calculated the specific infectivity, expressed as the ratio between viral titre and RNA copies; no differences were detected between control and E2 specific infectivity ( Figure S1).

| DISCUSSION
In the present paper, we show that 17,β-estradiol is able to inhibit HCV life cycle in vitro, whereas no effects were observed from exposure to DHEA-S, testosterone or progesterone. 17,β-estradiol acts through its intracellular receptor signalling pathway, triggering an antiviral state able to interfere with HCV assembly and/or release and partially with viral entry, but not on viral RNA replication. Hence, the effect on HCV is specific of oestrogens, and is medi- Interestingly, we observed, as already reported by Ulitzky et al., 13 that exposure to oestrogens leads to a lower number of foci, characterized also by a reduced size ( Figure S2). On these basis, we hypothesized that 17,β-estradiol can interfere with assembly/release of HCV infection. 17,β-estradiol does block HCV infection in a late phase, as shown by measuring viral inhibition in exposed cells and, in parallel, viral infectivity of released particles.
We observed a more profound inhibition in released particles and extracellular RNA than in intracellular RNA and infectivity, a finding confirmed by RT-qPCR and by infectious unit titration assay.
Overall, these data suggest that oestrogen can impair a late phase of HCV life cycle, mainly interfering with the release of new particles, as shown by the significant difference between intracellular and extracellular reduction (Figure 7). This observation finds further evidence showing a moderate inhibition in the intracellular infectivity. To support this hypothesis, we found no effect on specific infectivity ( Figure S1), confirming that the de novo viral particles are assembled properly. It is tempting to speculate that interference with the release phase might involve the SREBP pathway, which has been shown to be activated by oestrogen. 31 Interestingly, SREBP is able in vitro to modulate MTP, 32 which is crucial for the correct release of infectious particles, 21 playing a role in the regulation of the presence of ApoE on the nascent particles. 33 In addition, the protein cholesterol-25-hydroxylase, known to alter the cholesterol content of membranes, 34 it has been recently reported to affect the formation of the membranous web. 35 Moreover, the antiviral properties could be dependent on the protective effect on lipid peroxidation on liver 36,37 resulting in an endogenous regulation of HCV life cycle, as recently reported. 38 These hypotheses need to be tested in appropriately designed experiments; moreover, it is necessary to verify that the antiviral effects of 17,β-estradiol are similar in all viral genotypes. Meanwhile, the safety (and possibly the benefits) of oestrogen therapy in women with chronic viral hepatitis, discouraged in the distant past because of the theoretical risk of provoking or enhancing cholestasis, finds further support in our findings.
In conclusion, our study shows that 17,β-estradiol impairs HCV life cycle acting through its intracellular receptors and determining a massive reduction in released particles through interference with early (viral entry) and, above all, late (assembly/release) phases of the HCV life cycle.