Inhibition of hepatitis C virus p7 membrane channels in a liposome-based assay system

Abstract

Chemotherapy for patients chronically infected with hepatitis C virus (HCV) is ineffective in over 50% of cases, generating a high demand for new drug targets. The p7 protein of HCV displays membrane channel activity in vitro and is essential for replication in vivo though its precise role in the virus life cycle is unknown. p7 channel activity can be specifically inhibited by several classes of compounds, making this protein an attractive candidate for drug development, though techniques used to date in characterising this protein are unsuited to compound library screening. Here we describe an assay for the channel forming ability of p7 based on the release of a fluorescent indicator from liposomes. We show that recombinant p7 from genotype 1b HCV causes a dose-dependent release of dye when mixed with liposomes and that this property is enhanced at acidic pH. We demonstrate that this activity is due to the formation of a size-selective pore rather than non-specific disruption of liposomes and that activity can be blocked by amantadine and several other compounds, validating it as a measure of p7 channel function. This system provides the first convenient in vitro assay for exploiting p7 as a therapeutic target.

Fig. 1. Association of purified FLAG-p7 with liposomes. Ficoll density gradient separation was used to demonstrate association of purified FLAG-p7 with liposomes.

(A) Ten microgram of purified FLAG-p7 was analysed by SDS-PAGE and Coomassie staining (left panel) or immunoblot with p7-specific antibody, 1055 (right panel). In addition to monomeric FLAG-p7 (black arrow), higher molecular weight forms were evident (lighter arrows) representing oligomers of p7 formed in the presence of SDS micelles. (B) Five microgram of purified FLAG-p7 was mixed with PA:PC liposomes, adjusted to 20% Ficoll and layered at the bottom of a discontinuous density gradient then subjected to ultra-centrifugation. Eight equal fractions (lanes 1, bottom-8, top) were taken from the gradient and 20 μl aliquots analysed by immunoblot (top panel) using an anti-FLAG monoclonal antibody alongside 10% of the input (I). Again, as well as monomeric FLAG-p7 (black arrow) higher molecular weight forms were detected with increased protein concentration (lighter arrow). In addition, a sample was removed from fractions for fluorimetric detection of rhodamine-labelled liposomes to confirm migration to the 10% Ficoll/aqueous interface (bottom panel, red trace). Control reactions where no liposomes were added resulted in FLAG-p7 remaining at the bottom of the gradient (middle panel and bottom panel, black trace). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the article.)

Fig. 2. Dose-dependent release of CF from liposomes by addition of FLAG-p7. FLAG-p7/liposome mixes were monitored in real time for CF release by fluorimetry at λ_ex 485 nm/λ_em 520 nm.

(A) Dose-dependent release of CF caused by increasing amounts of FLAG-p7 shown in real time (left hand panel) and by initial rate calculated from the linear part of the curve (right hand panel). (B) Comparison of FLAG-p7-mediated CF release with that caused by addition of parental GST-FLAG-p7 fusion protein to the same concentration. (C) Assessment of reproducibility for FLAG-p7-mediated CF release. Initial rates for CF release were calculated for six batches of purified FLAG-p7 protein, each mixed with freshly prepared liposomes in triplicate. L: CF release from liposomes alone; S: CF release from liposomes where same volume of methanol was added as a solvent control; ΔFU s⁻¹: change in arbitrary fluorescence units/second. Error bars represent standard deviation of the mean.

Fig. 3

FLAG-p7 displays size constraints for transport of fluorescent dextrans (FD). The ability of FLAG-p7 and mellitin to effect release of increasing molecular weight fluorescent dextrans from liposomes was assessed using fluorimetry at λ_ex 485 nm/λ_em 520 nm on liposome-free supernatants (normalised for each dextran to Triton X-100 lysed liposomes as 100% reading) after incubation at 37 °C for 1 h. Top panel: 50 nM FLAG-p7 with 4 kDa dextran (FD-4), 10 kDa dextran (FD-10), 70 kDa dextran (FD-70); bottom panel shows results for 10 nM mellitin. Negative controls for liposomes alone (taken as 0%) and where methanol was added as a solvent control are also shown for each FD. Error bars represent standard deviation of the mean.

Fig. 4. Enhanced FLAG-p7-mediated membrane permeability at reduced pH. To assess the effect of pH on FLAG-p7-induced CF release, protein and liposomes were incubated in buffers of reduced pH at 37 °C for 1 h. Liposome-free supernatants were then re-adjusted to pH 7.4 by the addition of quenching amounts of 1 M Tris–HCl pH 8.0 to restore CF fluorescence.

(A) Restoration of CF fluorescence in solution following treatment at acidic pH as indicated by addition of neutral buffer. (B) FLAG-p7 (50 nM) mediated CF release from liposomes at acidic pH following restoration of CF fluorescence with Triton X-100 lysed liposomes as positive and methanol solvent plus liposome only negative controls. Error bars represent standard deviation of the mean.

Fig. 5

Inhibition of FLAG-p7-mediated CF release by amantadine. To validate the liposome assay as a means of screening p7 inhibitors, amantadine was titrated into reactions containing 50 nM FLAG and CF release monitored in real time by fluorimetry at λ_ex 485 nm/λ_em 520 nm. Initial rates of CF release were calculated from the linear portions of resultant curves (right panel) and used to calculate an IC₅₀ for amantadine for this concentration of FLAG-p7 as ~2 μM. Traces summarised in key at bottom. ΔFU s⁻¹: change in arbitrary fluorescence units/second. Error bars represent standard deviation of the mean.

Fig. 6. Effects of candidate p7 inhibitors on FLAG-p7-mediated CF release. The ability of 50 nM FLAG-p7 to cause release of CF in the presence of 1 μM candidate inhibitory compounds (red trace) was assessed in real time by fluorimetry at λ_ex 485 nm/λ_em 520 nm. Traces summarised in key at bottom left.

(A) Rimantadine. (B) HMA (NB: see Section 2). (C) GSK-1. (D) GSK-2. (E) GSK-3. (F) Plot of initial rates for FLAG-p7-mediated CF release in presence/absence of potential inhibitory compounds. ΔFU s⁻¹: change in arbitrary fluorescence units/second. Error bars represent standard deviation of the mean.