Structural basis of interaction between the hepatitis C virus p7 channel and its blocker hexamethylene amiloride
Hepatitis C virus (HCV) infection, which causes hepatitis C and can chronically lead to serious and life-threatening diseases including liver cirrhosis and hepatocellular carcinoma (Lauer and Walker, 2001), is a rising global health problem. More than 170 million people are infected by HCV worldwide and 3–4 million people are infected each year. No effective vaccines are available to prevent HCV infection. Moreover, HCV is a fast mutating RNA virus with seven distinct genotypes and many subtypes within each genotype. The high degree of genetic diversity can lead to further viral resistance to the current therapies within individual patients (Li et al., 2012). Hence, there remains a strong desire in the medical community to explore new therapeutic opportunities.
p7, the only viroporin encoded by the HCV genome, is a 63-residue protein that oligomerizes in membrane to form ion channels with cation selectivity (Moradpour and Penin, 2013). The p7 channel has been shown to facilitate efficient assembly and release of infectious virions, thus has been sought after as a potential anti-HCV drug target. Several inhibitors have been found to inhibit p7 channel activity with varying efficacies, including the amantadine and rimantadine (Griffin et al., 2003) that also block the influenza M2 channel (Pielak et al., 2011), hexamethylene amiloride (HMA) (Premkumar et al., 2004), and the long-alkyl-chain iminosugar derivatives (Pavlovic et al., 2005). Another novel small molecule known as BIT225, which has partial structural resemblance to HMA, also inhibits p7 activity and has been pursued as an anti-HCV drug candidate in clinical trials (Luscombe et al., 2010).
For many years, the inhibition mechanism of the known p7 inhibitors remained elusive in the absence of a high-resolution p7 structure. Previous solution NMR studies revealed that the adamantane compounds bind to six equivalent hydrophobic pockets (due to the 6-fold symmetry of the p7 hexamer) near the kink of the central pore-forming helices, consisting of elements from different helical segments and from different subunits (OuYang et al., 2013). A more recent study showed that insertion of rimantadine into these pockets could function as “molecular wedge” that attenuates conformational breathing needed for cation flow through the N-terminal constriction of the p7 channel (Dev et al., 2015). In addition to the studies on the p7 hexamer, a combined NMR and modeling study modeled the binding of rimantadine and several new compounds in the context of a p7 heptamer complex (Foster et al., 2014). Compared to rimantadine, HMA has a completely different chemical structure and was reported to inhibit p7 channel in different genotypes with stronger inhibition at similar drug concentration (Griffin et al., 2008). Earlier NMR titrations showed that HMA caused larger chemical shift perturbations in the loop region, while amantadine and NN-DNJ have a larger effect on the chemical shifts of the terminal regions (Cook et al., 2013). Therefore, HMA may act on a different site of p7 than rimantadine, and it is of substantial interest to identify this binding site.
We then performed more direct experiments to accurately pinpoint the HMA binding site by measuring protein-drug NOEs. Detection of HMA-specific NOEs was made possible by using a sample that contains (15N, 2H)-labeled p7 (5a) and deuterated or “NMR invisible” DPC. The 3D 15N-edited NOESY-TROSY-HSQC spectrum of this sample showed strong HMA NOE crosspeaks corresponding to the hexamethylene for Leu24, Trp48, Leu52, and Val53 and weaker hexamethylene NOE crosspeaks for Thr27 and Lys33 (Fig. 1D). Our earlier NMR investigation of rimantadine binding to p7 (5a) observed NOE crosspeaks between the adamantane protons and the amide protons of Val26, Leu56, and Arg57 (OuYang et al., 2013). The differences in intermolecular NOE patterns between HMA and rimantadine are consistent with the differences in NMR titration data between the two inhibitors shown in Fig. 1A. Altogether, the NMR data indicate that HMA binds to a different site compared to the one that adamantane compounds bind to. We could not detect any NOEs to the NH2 groups of HMA due to fast solvent exchange, and thus the orientation of the bound HMA could not be determined using the NOE data.
Our NMR and MD results show that HMA and rimantadine bind to different but nearby peripheral pockets of the p7 channel. Although the NMR data was collected for the p7 (5a), the high sequence identity between 5a and 1a and the fact that the p7 (5a) residues involved in HMA binding are mostly conserved in p7 (1a) provided us with strong confidence that the HMA binding site reported in this study is relevant to the identified HMA inhibition and likely applicable to other p7 variants. The binding of HMA and rimantadine both involve anchoring of their hydrophobic groups to a hydrophobic pocket adjacent to the channel cavity. In both cases, the binding site is not inside the channel cavity and thus there are six equivalent binding sites due to the six-fold symmetry of the p7 channel. Overall, HMA binding can be characterized by insertion of the relatively flat HMA molecule into a crevice near the wider mouth of the channel cavity where the hydrophobic hexamethylene interacts with the hydrophobic pocket deep inside the crevice and the polar amiloride fills the solvent accessible portion of the crevice (Fig. 2D). We note that previous studies of drug binding of p7 using a heptameric hairpin structure model yielded different peripheral binding sites for inhibitors. The differences between the two models might relate to the genotypes and sample environment. However, some key residues are conserved in both models, for example, Trp48 and Leu52 (Foster et al., 2014).
Although different binding sites, NMR chemical shift perturbation data in Fig. 1A–C indicate that the two inhibitors induce very similar long-range conformational changes near the narrow opening of the cavity. For both inhibitors, the binding site is far away from the narrow end of the cavity. In fact, compared to the rimantadine binding site, the HMA site is even farther away. We thus hypothesize that the mechanism of inhibition by HMA is also allosteric, either by restricting movement of helices needed for ion flow or by inducing structural rearrangements that block ion conduction, or both.
It is interesting to note that the HMA and rimantadine binding sites both contain several hydrophobic residues and the two residues they have in common are Leu52 and Val53. Mutations of L(50–55)A that decreased hydrophobicity have been shown to result in the loss of rimantadine sensitivity, and other observed mutation with decreased hydrophobicity also confers drug resistance in p7 (Foster et al., 2011). Since the known inhibitors of p7 channel, although structurally very different, contain a hydrophobic component, we believe one of the key determinants of p7 allosteric inhibitors is hydrophobic interaction. In the case of HMA, besides the hydrophobic interactions of hexamethylene with Leu24, Leu52 and Val53, the amiloride of HMA forms hydrogen bond with Lys33 and interacts with Leu36 and Leu45 by Van der Waals force (Fig. 2E). The extra polar interactions between HMA and p7 may account for the stronger inhibition. Further functional validation of HMA binding would be highly desirable to validate these interactions to provide a future direction for structure-guided inhibitor design.
Finally, obtaining structural information of weak inhibitors binding to ion channels is notoriously difficult. The combined use of NMR to identify the proximal ligand binding site and MD simulation to refine the channel-ligand complex as demonstrated here for the HCV p7 channel represents an effective solution to harvest useful information from many known weak inhibitors of ion channels.
We thank Bin Wu, Hongjuan Xue, and Songzi Jiang for technical assistance and insightful discussions. This work was supported by grants from CAS-Shanghai Science Research Center, Grant No: CAS-SSRC-YJ-2015-02 (to B. O.), the National Natural Science Foundation of China (Grant No. 31270905 to Z. J. L), the strategic Priority Research Program of the Chinese Academy of Sciences (Grant No. XDB08030301), NIH Grant AI112960 (to J. J. C.) and sponsored by the Shanghai Pujiang Program (to B. O.).
B. O. and J. J. C. conceived of the study; L. Zhao, L. D. and L. Zhou prepared samples; S. W. performed MD simulation and homology model study; L. Zhao and L. D. performed NMR titrations; L. Zhao, J. D. and Z. L. collected NOE data; L. Zhao and B. O. analysed the NMR data; J. J. C. and B. O. wrote the paper and all authors contributed to editing of the manuscript.
Linlin Zhao, Shuqing Wang, Lingyu Du, Jyoti Dev, Liujuan Zhou, Zhijun Liu, James J. Chou, and Bo OuYang declare that they have no conflict of interest. This article does not contain any studies with human subjects performed by any of the authors.
- DassaultSystèmesBIOVIA (2015) Discovery studio modeling environment. Dassault Systèmes, San DiegoGoogle Scholar
- Foster TL, Thompson GS, Kalverda AP, Kankanala J, Bentham M, Wetherill LF, Thompson J, Barker AM, Clarke D, Noerenberg M et al (2014) Structure-guided design affirms inhibitors of hepatitis C virus p7 as a viable class of antivirals targeting virion release. Hepatology 59:408–422CrossRefPubMedPubMedCentralGoogle Scholar
- Luscombe CA, Huang Z, Murray MG, Miller M, Wilkinson J, Ewart GD (2010) A novel Hepatitis C virus p7 ion channel inhibitor, BIT225, inhibits bovine viral diarrhea virus in vitro and shows synergism with recombinant interferon-alpha-2b and nucleoside analogues. Antiviral Res 86:144–153CrossRefPubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.