Could the Ebola Virus Matrix Protein VP40 be a Drug Target?

Abstract

Filoviruses are filamentous lipid enveloped viruses and include Ebola (EBOV) and Marburg, which are morphologically identical but antigenically distinct. These viruses can be very deadly with outbreaks of EBOV having clinical fatality as high as 90%. In 2012 there were two separate Ebola outbreaks in the Democratic Republic of Congo and Uganda that resulted in 25 and 4 fatalities, respectively. The lack of preventive vaccines and FDA approved therapeutics has struck fear that the Ebola virus could become a pandemic threat. The Ebola genome encodes only seven genes, which mediate the entry, replication, and egress of the virus from the host cell. The EBOV matrix protein is VP40, which is found localized under the lipid envelope of the virus where it bridges the viral lipid envelope and nucleocapsid. VP40 is effectively a peripheral protein that mediates the plasma membrane binding and budding of the virus prior to egress. A number of studies have demonstrated specific deletions or mutations of VP40 to abrogate viral egress but to date pharmacological inhibition of VP40 has not been demonstrated. This editorial highlights VP40, which is the most abundantly expressed protein of the virus and discusses VP40 as a potential therapeutic target.

1. Introduction

Viral hemorrhagic fevers including that stemming from the Ebola virus (EBOV) pose a serious health threat in central and eastern Africa with fatality rates as high as 90%. EBOV is a filamentous lipid enveloped virus of the Filoviridae family and is one of most virulent pathogens that infect humans. With no current FDA approved vaccines or drugs for EBOV there is urgency towards developing treatments and preventive measures. Recent evidence suggests vaccines [1] or repositioning of FDA approved drugs [2] may be a viable means of preventing or treating infection, respectively. EBOV harbors a negative sense RNA genome encoding seven proteins including: nucleoprotein (NP), VP30, VP35 and L protein, which constitute the nucleocapsid (NC). The transmembrane glycoprotein (GP) is rooted in the lipid envelope of the virus and is responsible for entry of virions into the host cell. VP40 is the viral matrix protein, which regulates viral budding and NC recruitment as well as virus structure and stability. VP24 is a minor matrix protein that is also important for NC assembly and serves to antagonize interferon signaling by binding host cell karyopherin α proteins as well as the transcription factor STAT1 [3].

A number of strategies have been employed to combat or prevent EBOV infections. A small molecule inhibitor that binds the host cell Niemman-Pick C1 receptor was effective in disrupting its interaction with the EBOV GP to inhibit infection [4] while small molecular inhibitors of ERα-glucosidases reduced mortality of EBOV infections [5]. Rhesus macaques treated with anti-sense targeting specific for EBOV VP24 and VP35 protected them against EBOV challenge [6] while a siRNA cocktail for EBOV L polymerase, VP24, and VP35 protected against EBOV when given in seven post-exposure treatments [7]. However, there is some concern that EBOV has the ability to suppress siRNA through VP30, VP35, and VP40 [8] and may resist cellular RNAi treatments during replication. Antibodies have also been effective in neutralizing EBOV such as the MB-003 antibody cocktail, which was recently shown to significantly increase survival of nonhuman primates infected with EBOV [9]. In addition, eight glycoprotein-specific monoclonal antibodies to ZEBOV were generated and improved survival between 33–100% [10]. Small molecule inhibitors of cellular kinases have also shown promise in treating EBOV Infections. For instance, the c-Abl1 tyrosine kinase inhibitor nilotinib demonstrated efficacy in reducing EBOV infectivity presumably by inhibiting phosphorylation of the VP40 protein [11] while treatment of cells with the kinase inhibitors genistein and tyrophostin AG1478 inhibited EBOV infection [12]. Targeting of EBOV proteins or host machinery has demonstrated that we may be on the cusp of therapeutics to treat or prevent EBOV infections. In this editorial the potential of targeting VP40 in EBOV infections will be discussed.

2. VP40 Assembly and Budding

VP40 is a peripheral protein consisting of 326 amino acids and is the most abundantly expressed of the seven proteins of the virus. VP40 localizes to the inner leaflet of the plasma membrane of human cells where it guides formation of new viral particles although the molecular basis of its plasma membrane binding properties are not well understood. Expression of only VP40 in mammalian cells is enough to assemble and form virus like particles (VLPs) that are similar in size and shape and nearly indistinguishable from the authentic virus [13,14]. Therefore, understanding how VP40 regulates assembly of VLPs both in vitro and in live cells is critical for identifying therapeutic targets for inhibiting the replication and spread of the virus. The assembly of VLPs by Ebola VP40 also represents an attractive model for studying the assembly of the virus in a BSL-2 setting since the VLPs are noninfectious and high content screening of VLP formation can be performed using VP40 tagged with various reporters [15,16]. In addition to its role in assembly and budding VP40 has also been shown to regulate viral transcription, which may represent a unique structural target in the life cycle of the virus [17].

3. Structure and Function of VP40

The first crystal structure of VP40 revealed a structure with two distinct domains [18]. The N-terminal domain has been found to be critical to VP40 oligomerization [19] and a C-terminal domain is thought to be essential for membrane binding [20,21]. The N-terminal and C-terminal domains seem to be loosely connected as urea or RNA incubation can induce formation of a VP40 RNA-binding ring structure. Additionally, the N-terminal domain alone can drive formation of ring structures where each of the eight N-terminal domain subunits can bind a RNA trinucleotide [22]. The ring structure of VP40 was found in infected cells but not in purified EBOV [19,22] and its role has been ascribed to regulating viral transcription in infected cells [17]. The VP40 C-terminal domain has been shown to insert into the plasma membrane [23], where hydrophobic residues penetrate more than halfway through the monolayer of the plasma membrane inner leaflet [21]. These interactions are key to plasma membrane binding and viral egress as mutation of hydrophobic residues that reduce membrane penetration also reduce plasma membrane localization while abrogating VP40 oligomerization and VLP formation [21,23].

Recent and extensive structure-function analysis has revealed that VP40 can assemble into different structures that in turn regulate distinct functions in the EBOV life cycle [17]. Here, Saphire and colleagues found the predominant form of VP40 is a dimer (Figure 1) that structurally rearranges into a linear hexamer most likely in response to electrostatic interactions with the plasma membrane [17]. The linear hexamer, which forms a multilayered structure through domain displacement, is reminiscent of EBOV virion structures revealed by tomography. The VP40 dimeric interface (Figure 1) was shown to consist of a predominantly hydrophobic interface, mutation of which abolished VLP formation [17]. In order to facilitate budding and egress the VP40 dimer rearranges into linear hexamers that interact through a conserved C-terminal domain interface where mutations halted VLP egress but not plasma membrane localization of VP40. This new dimeric structure also revealed a large cationic patch in the C-terminal domain that likely interacts with the highly anionic interface of the cytoplasmic face of the plasma membrane. These newly solved VP40 structures may greatly enrich our understanding of how VP40 mediates membrane curvature changes in EBOV budding as VP40 filaments are structurally akin to BAR domains [24] that mediate membrane curvature changes in human cells.

Figure 1. Structure of the EBOV VP40 dimers.

The recent dimeric structure of VP40 (PDB ID: 4LDB [17]) is shown with the N-terminal domain in gray and the C-terminal domain in black. The VP40 N-terminal domain dimeric interface involves residues 52–65 and 108–117 both of which are part of alpha helices. These interactions have little H-bonding and are mostly hydrophobic in nature. Specifically residues involved are Ala⁵⁵, His⁶¹, Phe¹⁰⁸, Thr¹¹², Ala¹¹³, Met¹¹⁶, and Leu¹¹⁷ where Leu¹¹⁷ seems to be of key importance. The inset shows a close up of the dimeric interface with Leu¹¹⁷ shown in magenta. Mutation of Leu¹¹⁷ disrupts formation of VP40 dimers and abrogates viral budding.

4. Expert Opinion

How could VP40 be targeted by therapeutics? It’s clear that VP40 is required for assembly and budding of new viral particles where the VP40 structure bound on the plasma membrane interface likely guides EBOV particle morphology. VP40 has been shown to assemble into a filamentous structure, disruption of which halts viral egress [17]. Structural studies taken together with studies on the membrane binding properties of VP40 [21,23] reveal potentially druggable sites (Figure 2). Targeting the lipid binding sites of VP40 will be difficult but evidence from lipid-protein interaction studies suggest it may be possible. For instance, several computational studies have recently and effectively designed small molecule inhibitors of lipid-binding domains that demonstrated inhibition of membrane binding [25]. Because dimeric VP40 is used as a building block for C-terminal domain contacts in filamentous VP40, blocking N-terminal domain dimerization or C-terminal domain assembly would be effective at inhibiting EBOV assembly. All this being said my biggest concern with targeting VP40 specifically is that the virus will just synthesize more protein to overcome the pharmacological antagonism. Thus, cellular, biophysical and computational studies will be necessary to assess the feasibility of targeting the VP40 lipid-binding and/or oligomerization sites of with small molecules. However, if cellular levels of the VP40 octameric ring structure that regulates viral transcription are significantly lower than dimeric VP40, overproduction of VP40 octameric rings to circumvent pharmacological antagonism may not be as much of a concern. Furthermore, the RNA binding site may be druggable as a recent computational study generated ten lead compounds [26] that could warrant testing in vitro and in cells.

Figure 2. Potentially druggable sites of VP40.

The VP40 dimer is shown with some potential sites of inhibition to block viral budding. The VP40 dimeric interface is shown in magenta and the C-terminal domain hexameric interface mainly attributed to Met²⁴¹ and Ile³⁰⁷ is shown in green [17]. Disruption of the hexameric interface through mutagenesis reduces viral budding from the plasma membrane. The cationic patch exposed on the same interface of the dimer is shown in blue and mainly consists of Lys²²⁴, Lys²²⁵, Lys²⁷⁴, and Lys²⁷⁵ [17], which may interact through electrostatic interactions with the anionic inner leaflet of the plasma membrane. Mutation of these Lys residues greatly reduces viral budding. A hydrophobic loop shown in red has been shown to penetrate into the plasma membrane [21,23], a necessary step for viral budding. Still unknown is the orientation of the C-terminal domain at the plasma membrane interface with respect to the cationic patch and the hydrophobic loop.

Interfering with egress of the HIV-GAG protein, which regulates assembly and egress of HIV through PI(4,5)P₂ and PS binding at the plasma membrane [27] has seen success in in vitro, cellular, and animal studies. For instance, a peptide inhibitor of GAG oligomerization through capsid domain contacts [28], can alter HIV-GAG oligomerization a necessary step in HIV particle infectivity. Another strategy using the PI(4,5)P₂/PI4P antibody WR321 also neutralized HIV infections in cell culture [29]. These studies on HIV support the hypothesis that pharmacologically disrupting VP40 oligomerization particularly the dimeric interface or the C-terminal domain interface that mediates hexamer formation would disrupt budding and infectivity of EBOV. Once the lipid binding determinants of VP40 are more clearly elucidated lipid-antibody therapy for the plasma membrane lipids VP40 selectively binds would also in principle reduce budding and egress.

Could targeting the membrane bilayer itself halt VP40 mediated EBOV egress?

Recently, it was demonstrated that HIV-1 infection could be reduced by rigidifying liquid-ordered membrane domains in the cell [30]. This study used pharmacological inhibition of the enzyme dihydroceramide desaturase, which replaced sphingomyelin in the membrane with dihydrosphingomyelin. This membrane structural change made it more difficult for HIV insertion of the gp41 fusion peptide and reduced virus-cell membrane fusion. A similar principle could be applied to viral egress where bending of the plasma membrane to create a new viral particle is an important step in the life cycle and spread of infection. Because VP40 alone has been shown to induce membrane curvature changes to synthetic lipid vesicles that recapitulate the inner leaflet of the plasma membrane [21], altering the plasma membrane lipid composition of infected cells to increase their membrane rigidity may halt budding. Future studies geared towards the type of budding mechanism VP40 utilizes from the plasma membrane lipids and/or domains may provide important clues as to the selectivity of plasma membrane lipids required for VP40 assembly, membrane penetration, and egress. Because altering the plasma membrane structure or lipid composition to increase membrane rigidity would pharmacologically be a nonselective process among infected and non-infected cells it would likely cause cellular toxicity, limiting this option as a viable therapy. Nonetheless, studies aimed at understanding the viral protein mediated bending of the plasma membrane should lead to a better understanding of the underlying molecular mechanisms of viral budding.