Shedding of Ebola Virus Surface Glycoprotein Is a Mechanism of Self-regulation of Cellular Cytotoxicity and Has a Direct Effect on Virus Infectivity

Transcription

1 SUPPLEMENT ARTICLE Shedding of Ebola Virus Surface Glycoprotein Is a Mechanism of Self-regulation of Cellular Cytotoxicity and Has a Direct Effect on Virus Infectivity Olga Dolnik, 1,2 Valentina A. Volchkova, 1 Beatriz Escudero-Perez, 1 Philip Lawrence, 1 Hans-Dieter Klenk, 2 and Viktor E. Volchkov 1 1 Molecular Basis of Viral Pathogenicity, CIRI, INSERM U CNRS UMR5308, Université de Lyon, Université Claude Bernard Lyon 1, Ecole Normale Supérieure de Lyon, France; and 2 Institut für Virologie, Philipps Universität Marburg, Germany The surface glycoprotein (GP) is responsible for Ebola virus (EBOV) attachment and membrane fusion during virus entry. Surface expression of highly glycosylated GP causes marked cytotoxicity via masking of a wide range of cellular surface molecules, including integrins. Considerable amounts of surface GP are shed from virusinfected cells in a soluble truncated form by tumor necrosis factor α converting enzyme. In this study, the role of GP shedding was investigated using a reverse genetics approach by comparing recombinant viruses possessing amino acid substitutions at the GP shedding site. Virus with an L635V substitution showed a substantial decrease in shedding, whereas a D637V substitution resulted in a striking increase in the release of shed GP. Variations in shedding efficacy correlated with observed differences in the amounts of shed GP in the medium, GP present in virus-infected cells, and GP present on virions. An increase in shedding appeared to be associated with a reduction in viral cytotoxicity, and, vice versa, the virus that shed less was more cytotoxic. An increase in shedding also resulted in a reduction in viral infectivity, whereas a decrease in shedding efficacy enhanced viral growth characteristics in vitro. Differences in shedding efficacy and, as a result, differences in the amount of mature GP available for incorporation into budding virions did not equate to differences in overall release of viral particles. Likewise, data suggest that the resulting differences in the amount of mature GP on the cell surface led to variations in the GP content of released particles and, as a consequence, in infectivity. In conclusion, fine-tuning of the levels of EBOV GP expressed at the surface of virus-infected cells via GP shedding plays an important role in EBOV replication by orchestrating the balance between optimal virion GP content and cytotoxicity caused by GP. Keywords. Ebola; Ebolavirus; glycoprotein GP; shedding; shed GP; infectivity; cytotoxicity. The genera Ebolavirus, Marburgvirus, and Cuevavirus constitute the family Filoviridae, a group of negativestrand, nonsegmented RNA viruses responsible for severe hemorrhagic fever in human and nonhuman primates [1, 2]. Ebola virus (EBOV) belongs to the species Correspondence: Viktor E. Volchkov, PhD, Laboratory Molecular Basis of Viral Pathogenicity, International Centre for Research in Infectiology, INSERM U1111, UCBL1, 21 avenue Tony Garnier, Lyon cedex 07, France (viktor.volchkov@ inserm.fr). The Journal of Infectious Diseases 2015;212:S322 8 The Author Published by Oxford University Press on behalf of the Infectious Diseases Society of America. All rights reserved. For Permissions, please journals.permissions@oup.com. DOI: /infdis/jiv268 Zaire ebolavirus and is most pathogenic for humans with a mortality rate reaching up to 90%. The surface glycoprotein (GP) is responsible for attachment and membrane fusion and thus is indispensable for virus replication [3 6]. EBOV GP is a highly N- and O-glycosylated type 1 glycoprotein forming trimeric viral spikes composed of disulfide-linked subunits GP 1 and GP 2 generated by proteolytic cleavage of the GP precursor by the cellular protease furin [7].SurfaceGPisencoded bythegpgeneand is expressed as the result of transcriptional RNA editing, when an extra, nontemplate adenosine is inserted by the viral polymerase at a specific editing site during transcription [8, 9]. GP messenger RNAs (mrnas) synthesized without editing are used to express a major GP S322 JID 2015:212 (Suppl 2) Dolnik et al

2 gene-specific, nonstructural product, secreted glycoprotein sgp [8, 9]. Recently, we have demonstrated that, in addition to fulllength GP mrnas, a shorter RNA is also synthesized in virusinfected cells, representing >40% of the total amount of GP mrna [10]. This RNA is a truncated version of the full-length GP mrna that is polyadenylated at the editing site and thus represents a nonstop mrna. Surface GP of EBOV has been shown to interfere with functions of cell surface proteins, eventually causing strong cytotoxic effects [11 14]. Recently, it has been demonstrated that the cytotoxic properties of GP are associated with its overexpression and explained by its ability to cause steric hindrance and masking of cellular surface molecules, thus preventing their normal functions [13, 15]. A recombinant mutant of EBOV containing mutations at the GP gene editing site that significantly increase the synthesis of surface GP was shown to be extremely cytotoxic, owing to overloading and exhausting of the cellular endoplasmic reticulum with increased amounts of the highly glycosylated protein [16]. This no-editing mutant of EBOV has lower growth kinetics in vitro and is less proficient at cell-to-cell spreading. Significant amounts of surface GP are shed from virus-infected cells via proteolytic cleavage mediated by cellular tumor necrosis factor α converting enzyme (TACE) [17]. Cleavage results in removal of the transmembrane anchor from the GP 2 subunit and release of a soluble trimeric complex consisting of GP 1 and truncated GP 2. Shed GP preserves its antigenic properties and is recognized by anti-gp antibodies, including virusneutralizing antibody KZ52 [18]. Shed GP was detected in the blood of infected guinea pigs and monkeys [17, 19] andcan play a decoy function for anti-gp antibodies, including virus neutralizing antibodies, and presumably can also interfere during treatment with anti-gp antibodies. Recently, it has been shown that shed GP triggers immune activation and increased vascular permeability and is thus an important pathogenicity factor [20]. In the present study, we generated and analyzed recombinant EBOVs displaying 2 alternative patterns of GP shedding, ranging from a considerable increase in shedding efficiency to the almost complete loss of shedding. We demonstrate that shed GP plays an important role in EBOV replication by orchestrating viral cytotoxicity, optimal virion GP content, and viral infectivity. MATERIALS AND METHODS Cells Vero E6 cells were maintained in Dulbecco s modified Eagle s medium (DMEM) supplemented with 10% fetal calf serum (FCS). BSR T7/5 cells used for rescue of recombinant viruses were kindly provided by K. K. Conzelmann and were grown in Glasgow medium supplemented with 10% newborn calf serum. Generation and Rescue of Recombinant Viruses Mutations at the GP gene TACE cleavage site were introduced into the intermediate plasmid pkss25 by site-directed mutagenesis, using the primer pairs 5 -GTTGATAAAACCCTTCCGCT CCAGGGGGACAATGACAATTG and 5 -CAATTGTCATTGT CCCCCTGGAGCGGAAGGGTTTTATCAAC for 637D/V exchange and 5 -CATGATTTTGTTGATAAAACCGTACCGG ACCAGGGGGACAATG and 5 -CATTGTCCCCCTGGTCC GGTACGGTTTTATCAACAAAATCATG for 635L/V exchange. Full-length EBOV antigenomic complementary DNAs were then assembled essentially as described elsewhere [16]. To ensure that these recombinant viruses replicate in guinea pigs, 3 substitutions inthevp24geneknowntoenhanceebovreplicationandvirulence in these animals were also introduced [21]. The recombinant viruses: EBOV-LS (low shedding, recebov-gp/l635v-vp24/ 8mc), EBOV-HS (high shedding, recebov-gp/d637v- VP24/8mc), and EBOV-WT (wild type, recebov-gp/wt- VP24/8mc) were rescued using a previously established reverse genetics system for EBOV [16] and amplified in parallel in Vero E6 cells to ensure the same passage history. The presence of the introduced mutations was verified by direct sequencing of reverse transcription polymerase chain reaction fragments generated from viral RNA isolated from the culture supernatants of virus-infected Vero E6 cells. Several consecutive passages of the viruses in Vero E6 cells showed that the mutations introduced were preserved over at least 5 passages. Western Blot Analysis The expression of viral proteins was analyzed by Western blotting and visualized by means of the enhanced chemiluminescence technique, using SuperSignal WestDura enhanced substrate. Virus Infectivity Titration Titration of cell culture supernatants was performed by end point dilution in 96-well plates. Vero E6 cells were inoculated with serial 10-fold dilutions of samples and incubated for 1 hour at 37 C, and fresh medium containing 5% FCS was added to the cultures. After culture for 7 8 days, CPE were evaluated under a light microscope, and the 50% tissue culture infective dose (TCID 50 ) was calculated according to the Karber formula [22], with virus titers expressed as infectious units (IU) per milliliter. Growth Kinetics Growth kinetics for recombinant EBOV stocks were determined for single- and multicycle infections of Vero E6 cell cultures. Viruses were inoculated on subconfluent monolayers of Vero E6 cells at multiplicities of infection (MOIs) of 1 and 0.01, respectively. Supernatants were collected at 24-hour intervals until day 6 after infection, and virus titers were assessed by the TCID 50 titration method. Immunostaining Fresh monolayers of Vero E6 cells were infected at an MOI of 0.1 with EBOV-WT, EBOV-HS, or EBOV-LS in 96-well plates and incubated for 1 hour at 37 C. Inocula were removed after Shedding of Ebola Virus GP JID 2015:212 (Suppl 2) S323

3 1 hour of infection, and cells were overlaid with DMEM containing 1.5% carboxymethyl cellulose and 2.5% FCS. Cells were fixed with 4% paraformaldehyde 4 days after infection, permeabilized with 0.1% Triton X-100, and incubated with monoclonal mouse anti-np antibody (dilution, 1:200) and then with peroxidase-conjugated goat anti-mouse antibody (dilution, 1:1000). Foci of virus replication were visualized after incubation with TrueBlue peroxidase substrate. RESULTS AND DISCUSSION Generation of Recombinant EBOVs Possessing Mutations in the GP TACE Cleavage Site We have previously demonstrated that substitutions within the TACE cleavage site of EBOV GP could dramatically alter the efficacy of shedding in a transient expression system [17]. In that study, it was shown that an L635V exchange results in decreased shedding, whereas a D637V mutation leads to increased shedding (Figure 1A). In this study, using a reverse genetics approach [16], we investigated the effect of these substitutions on the replication of recombinant EBOV mutants: EBOV-LS, EBOV-HS, and EBOV-WT. Of note, EBOV-LS showed an increased tendency to convert into a variant containing 8 uridine residues at the GP gene editing site during passaging in Vero E6 cells, compared with EBOV-HS and EBOV-WT [23]. To exclude a possible variation between the viruses with respect to the state of the editing site, all of the following experiments were performed with low-passage viral stocks. Recombinant EBOV Mutants Display Sharply Different Patterns of GP Shedding To compare the viruses described above with respect to the kinetics and efficiency of GP shedding, Vero E6 cells were infected at either a low (0.01) or high (1) MOI with recombinant wildtype or mutant EBOV stocks. The samples of cells and culture supernatants were collected at different intervals and analyzed by Western blot, using anti-gp 2, anti-vp24, or anti-np antibodies (Figure 1B). Use of anti-gp 2 antibodies allowed easy discrimination between virion- and cell-associated mature GP and shed GP containing a truncated GP 2 subunit that thus migrates faster than the TACE-uncleaved GP on polyacrylamide gels. As expected, infection by EBOV-LS did not induce any significant Figure 1. Rescue of recombinant viruses with tumor necrosis factor α converting enzyme (TACE) cleavage site mutations. A, Schematic representation of EBOV glycoprotein (GP) indicating the native amino acid sequence at the TACE cleavage site and mutations introduced in recombinant mutant viruses: D637V for high-shedding (HS) virus, and L635V for low-shedding (LS) virus. The fusion domain (FD) and the transmembrane domain (TM) of GP 2 are indicated, as well as the furin cleavage site. B, Recombinant EBOV mutants display sharply different patterns of GP shedding. Western blot analysis of samples of cells and culture supernatants of Vero E6 cells infected with different viruses over time after infection, using GP 2 -, NP-, and VP24-specific antibodies as indicated. Positions of full-length and truncated GP 2, as well as VP24 and NP, are indicated. Abbreviations: MOI, multiplicity of infection; WT, wild type. S324 JID 2015:212 (Suppl 2) Dolnik et al

4 release of shed GP. Infection with EBOV-WT resulted in an accumulation of shed GP in the culture medium, as demonstrated by increasing amounts of truncated GP 2 over time (Figure 1B). In comparison to EBOV-WT, infection with EBOV-HS showed a dramatic increase in GP shedding (Figure 1B). Accordingly, a low level of full-length GP 2 was observed in EBOV-HS-infected cells (Figure 1B), whereas increased amounts of GP 2 were detected in EBOV-LS infected cells. An increase in shedding led to a decrease in the release of virion GP into the culture medium, whereas a decrease in shedding had the opposite effect, as more full-length GP 2 could be seen in the medium of EBOV- LS infected cells, compared with EBOV-WT infected cells. At a high MOI, early after infection (days 3 4), lower levels of full-length GP 2 were observed in the culture medium for EBOV-HS in comparison with the two other viruses, whereas late after infection similar quantities of uncleaved, virion GP 2 could be seen with all 3 viruses. Remarkably, at both MOIs, EBOV-HS showed lower levels of VP24 in the culture medium early after infection but higher VP24 quantities later after infection, compared with EBOV-WT and EBOV-LS. Alterations in GP Shedding Efficacy Affect Both CPE Development and Virus Growth Both EBOV mutants showed distinct differences in terms of development of virus-induced cytopathic effects (CPE). EBOV- HS caused the least and EBOV-LS the most pronounced CPE (Figure 2A). In addition, EBOV-LS appeared to cause the appearance of CPE much earlier than the 2 other viruses. These results are in line with our previous finding concerning the cytotoxic properties of overexpressed EBOV in transient expression systems [11]. In the case of EBOV-LS, an increased amount of GP expressed on the cell surface and unable to be cleaved by TACE would certainly contribute to cytotoxicity. Similarly, reduced cytotoxicity in the case of EBOV-HS could most likely be explained by a lower concentration of GP at the surface of infected cells, owing to increased shedding. To understand how shedding may influence virus growth, culture supernatants of Vero E6 cells infected at low or high MOIs were collected at different intervals after infection, and virus titers were assayed using the TCID 50 method (Figure 2B). EBOV-WT and EBOV-LS showed rather similar growth kinetics. Notably, at a high MOI, viral titers reached their maximum on day 3 after infection, before starting to decrease (Figure 1B). In these cases, the decrease in virus titers is probably explained by the death of infected cells and the resulting acidification of culture medium that likely partially neutralizes the virus. EBOV-HS showed lower titers than the other 2 viruses over the first 4 days of infection. However, this virus continued to release infectious virions into the medium, and all 3 viruses reached similar titers on day 6 after infection. Overall, data obtained thus far would suggest that the lower CPE allows EBOV- HS to replicate longer, resulting ultimately in release of a rather high level of viral proteins into the medium, regardless of the apparently lower presence of mature GP in cells (Figure 2B). When infection was performed at a low MOI, EBOV-HS uniformly showed lower titers (approximately 1 log), compared with the 2 other viruses, which would appear to be consistent with the lower levels of virion GP observed in the culture medium (Figure 1B). Interestingly, at the end of the experiment, EBOV-HS showed a slightly higher level of viral proteins released into the medium. EBOV-LS reached the highest titers, which again would appear consistent with the higher level of virion GP detected in the medium. Shedding Affects Virus Infectivity Through GP Content in Virions To investigate whether alterations in shedding efficacy have an effect on GP content in the recombinant viruses generated, samples of culture medium containing equal amounts of structural protein VP24 were assayed for the presence of virion-associated full-length GP 2. Two-fold dilutions of culture medium containing viral particles and shed GP were prepared and analyzed using Western blot with anti-gp 2 antibodies (Figure 3A). The samples of EBOV-HS were found to contain approximately 4-fold less virion GP 2 and about 4 times more shed GP, compared with EBOV-WT. No significant difference was found between EBOV-LS and EBOV-WT with respect to GP 2 content,exceptthatsamplesfromebov-lsdidnotcontain truncated GP 2. Next we compared the infectivity of recombinant viruses. Two-fold dilutions of culture medium containing equal amounts of IU, as determined by TCID 50 titration of viral particles, were prepared and analyzed using Western blot with anti-vp24 antibodies (Figure 3B). For an equal quantity of IU, about 4-fold more VP24 was detected with EBOV-HS in comparison to EBOV-WT, meaning that around 4-fold more viral particles were needed with the EBOV-HS virus to achieve the same titer as EBOV-WT or EBOV-LS and suggesting a direct link between EBOV infectivity and virion GP content. While the difference in the infectivity of the viruses was not readily expected, it correlated well with the data presented above, revealing that GP shedding efficiency somewhat determines the quantity of mature GP that is present in virus-infected cells and available for incorporation into newly budding viral particles. Remarkably, increased shedding of mutated GP from the surface of the cells does not visibly affect budding efficiency. A low level of GP on the cell surface apparently results in budding of virions with lower GP content and would help to explain the difference in infectivity seen for the 2 recombinant viruses. The difference between the viruses in terms of their infectivity correlates well with the viruses growth kinetics data and may explain the slower growth of EBOV-HS. Remarkably, the data clearly indicate that GP incorporation into virions has an upper limit. Higher amounts of mature GP present in cells infected by EBOV-LS do not lead to higher levels of GP in virions, Shedding of Ebola Virus GP JID 2015:212 (Suppl 2) S325

5 Figure 2. Alterations in glycoprotein (GP) shedding efficacy affect both the development of cytopathic effects (CPE) and virus growth. A, CPE on Vero E6 cells infected with wild-type EBOV (EBOV-WT), EBOV-LS (low shedding), and EBOV-HS (high shedding) at multiplicities of infection (MOIs) of 1 and 0.01 on day 3 after infection. Pictures were taken using a light microscope. B, Virus growth kinetics of EBOV-WT and shedding mutants at MOIs of 1 and Vero E6 cells were infected with EBOV-WT, EBOV-LS, and EBOV-HS at the indicated MOIs, and samples of culture supernatant were collected over the 6-day period after infection. Supernatants were titrated by the 50% tissue culture infective dose method. The titers were calculated according to the Karber formula as infectious units (IU) per milliliter and are displayed as log 10 IU/mL. The data are representative of 3 independent experiments. compared with EBOV-WT infected cells. The fact that these 2 viruses reveal very similar infectivity strongly supports our conclusions. A recent publication used a pseudotyping system instead of EBOV infection [24], however somewhat contradict our findings. The authors demonstrated that overexpression of EBOV GP could result in overloading of particles with surface GP that causes their lower infectivity, compared with particles produced in cells expressing lower levels of GP. One possible explanation of this contradiction could be that, during EBOV infection, virions are produced in a more organized and self-regulated fashion, whereas during pseudotyping, GP incorporation appears to be less controlled, causing inadequate GP integration into particles. The reduced infectivity of EBOV-HS was further analyzed using an immunoplaque assay that allowed us to evaluate the efficiency of EBOV-HS spread in cell culture. Vero E6 cells were infected at a low MOI with EBOV-HS, EBOV-WT, and EBOV-LS, and after incubation for 4 days under methylcellulose to allow only cell-to-cell spread of the virus, cells were stained for the presence of the NP protein. The observed plaques were about 30% smaller for EBOV-HS, compared with those for EBOV-WT and EBOV-LS, supporting the notion that EBOV-HS is less infectious than the 2 other viruses (Figure 3C). Interestingly, the smaller plaque size of EBOV-HS could also be associated with the effect of shed GP on neighboring cells. Shed GP mimics surface GP in sequence and structure and differs only by the absence S326 JID 2015:212 (Suppl 2) Dolnik et al

6 Figure 3. Shedding affects virus infectivity through glycoprotein GP content in virions. A, Western blot analysis of 2-fold dilutions of culture supernatants of different viruses normalized for the presence of viral protein 24 (VP24), using anti-vp24 specific antibodies (upper panel) and anti-gp 2 specific antibodies (lower panel). Positions of proteins are indicated. B, Western blot analysis of 2-fold dilutions of culture supernatants of different viruses normalized for virus titers using the 50% tissue culture infective dose method with anti-vp24 antibodies. C, Immunoplaque assay. Vero-E6 cells were infected at a multiplicity of infection of 0.1 with wild-type EBOV (EBOV-WT) and EBOV-HS (high shedding). Pictures of the plaques were taken using a light microscope with 10 objective (left panel), and the mean diameters were determined (histogram). of the domain cleaved by TACE, comprising 11 membrane-proximal amino acids, the transmembrane domain, and a short cytoplasmic tail. It is very plausible that shed GP is able to interact with the same surface molecules that are engaged in interaction with EBOV and thus that shed GP could potentially interfere with virus attachment to cells. The study of whether shed GP can modulate virus interaction with cells is ongoing. It is now well established that EBOV GP is cytotoxic for cells, and this is particularly evident when it is overexpressed in transient expression systems [11, 12, 15, 16]. However, during EBOV infection, the cytotoxic effects in cell culture became visible only from day 3 after infection onward, and while cell rounding and partial cell detachment can also be observed, these cells continue to survive and to release viral particles. Thus, it appears that, although GP is intrinsically cytotoxic for the infected cell, EBOV has developed mechanisms that precisely allow it to downregulate expression of GP. We have shown earlier that the high expression rate of GP seen with a recombinant EBOV deficient for the editing site results in high cytotoxicity that significantly reduced virus plaque size, suggesting impaired virus spread [16]. It was postulated that transcriptional RNA editing, which is required for EBOV GP expression, is at the same time involved in the control of GP-induced cytotoxicity in that it controls and reduces the expression of mrna encoding surface GP. In accordance with the concept presented above, here we demonstrate that GP shedding also contributes to the reduction of GP content in cells and, importantly, represents an additional control mechanism for GP-induced cytotoxicity at the protein level. Overall, it appears that GP shedding provides the means to achieve the necessary balance of GP concentration at the cell surface to guarantee sufficient levels for the formation of optimally infectious particles. The intensity of GP shedding is determined by the cleavability of the amino acid sequence at the TACE cleavage site. It is interesting to note that the EBOV TACE cleavage site appears to be suboptimal for complete cleavage, compared with natural, cellular substrate targets of TACE. As demonstrated in our study, this property allows the necessary balance of GP expression on the cell surface and, as a result, the required GP content in budding virions for optimal infectivity. Moreover, it has been shown that TACE itself is activated in EBOV-infected cells [19], Shedding of Ebola Virus GP JID 2015:212 (Suppl 2) S327

7 and this may provide a means to overcome the suboptimal nature of the cleavage site in the conditions of continuously increasing GP expression seen during EBOV replication. An important consequence of shedding is the release of high amounts of soluble GP into the blood of an infected individual or experimentally infected animal. Our previous work suggested that shed GP can act as decoy by interacting with GP-specific antibodies that otherwise would interact with virus particles or infected cells [17]. In a recent study, we also demonstrated that shed GP plays an important role in both activation of noninfected immune cells for systemic release of proinflammatory and antiinflammatory cytokines and an increased permeability of endothelial cells [20], thus contributing to the septic-like shock syndrome observed during EBOV infection. It would be of importance in the future to use the generated recombinant viruses for analysis of EBOV-induced pathogenesis, using an appropriate animal model. Of note, viruses generated in this study contained mutations in the gene encoding VP24, which allows EBOV replication in guinea pigs, and thus they are important tools for further investigation. In conclusion, our data provide evidence that fine-tuning the levels of EBOV GP expressed at the surface of virus-infected cells via GP shedding plays an important role in EBOV replication by orchestrating both optimal virion GP content and viral cytotoxicity. In consideration of our finding that GP shedding affects EBOV infectivity and spread, it is justified to investigate putative inhibitors that would have the potential for therapeutic use in the treatment of EBOV-induced hemorrhagic fever. Notes Acknowledgments. All experiments involving live Ebola virus were performed at the INSERM biosafety level 4 laboratory Jean Merieux in Lyon, France. Financial support. This work was supported by the INSERM, the European Union FP7 project ANTIGONE (278976), the Agence Nationale de la Recherche (ANR-07-MIME ), the Fondation pour la Recherche Médicale (DMI ), and the Alexander von Humbold Foundation (Feodor Lynen Fellowship to O. D.). Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Kuhn JH, Bao Y, Bavari S, et al. Virus nomenclature below the species level: a standardized nomenclature for natural variants of viruses assigned to the family Filoviridae. Arch Virol 2013; 158: Feldmann H, Sanchez A, Geisbert TW. Filoviridae: Marburg and Ebola viruses. In: Knipe DM, Howley PM, eds. Fields virology. Philadelphia: Lippincott Williams and Wilkins, 2013: Chan SY, Speck RF, Ma MC, Goldsmith MA. Distinct mechanisms of entry by envelope glycoproteins of Marburg and Ebola (Zaire) viruses. J Virol 2000; 74: Chandran K, Sullivan NJ, Felbor U, Whelan SP, Cunningham JM. Endosomal proteolysis of the Ebola virus glycoprotein is necessary for infection. Science 2005; 308: Ito H, Watanabe S, Sanchez A, Whitt MA, Kawaoka Y. Mutational analysis of the putative fusion domain of Ebola virus glycoprotein. J Virol 1999; 73: Yang Z, Delgado R, Xu L, et al. Distinct cellular interactions of secreted and transmembrane Ebola virus glycoproteins. Science 1998; 279: Volchkov VE, Feldmann H, Volchkova VA, Klenk HD. Processing of the Ebola virus glycoprotein by the proprotein convertase furin. Proc Natl Acad Sci U S A 1998; 95: Volchkov VE, Becker S, Volchkova VA, et al. GP mrna of Ebola virus is edited by the Ebola virus polymerase and by T7 and vaccinia virus polymerases. Virology 1995; 214: Sanchez A, Trappier SG, Mahy BW, Peters CJ, Nichol ST. The virion glycoproteins of Ebola viruses are encoded in two reading frames and are expressed through transcriptional editing. Proc Natl Acad Sci U S A 1996; 93: Volchkova VA, Vorac J, Repiquet-Paire L, Lawrence P, Volchkov VE. Ebola virus GP gene polyadenylation versus RNA editing. J Infect Dis 2015; 212(suppl 2):S Alazard-Dany N, Volchkova V, Reynard O, et al. Ebola virus glycoprotein GP is not cytotoxic when expressed constitutively at a moderate level. J Gen Virol 2006; 87: Chan SY, Ma MC, Goldsmith MA. Differential induction of cellular detachment by envelope glycoproteins of Marburg and Ebola (Zaire) viruses. J Gen Virol 2000; 81: Francica JR, Matukonis MK, Bates P. Requirements for cell rounding and surface protein down-regulation by Ebola virus glycoprotein. Virology 2009; 383: Simmons G, Wool-Lewis RJ, Baribaud F, Netter RC, Bates P. Ebola virus glycoproteins induce global surface protein down-modulation and loss of cell adherence. J Virol 2002; 76: Reynard O, Borowiak M, Volchkova VA, Delpeut S, Mateo M, Volchkov VE. Ebolavirus glycoprotein GP masks both its own epitopes and the presence of cellular surface proteins. J Virol 2009; 83: Volchkov VE, Volchkova VA, Muhlberger E, et al. Recovery of infectious Ebola virus from complementary DNA: RNA editing of the GP gene and viral cytotoxicity. Science 2001; 291: Dolnik O, Volchkova V, Garten W, et al. Ectodomain shedding of the glycoprotein GP of Ebola virus. Embo J 2004; 23: Maruyama T, Rodriguez LL, Jahrling PB, et al. Ebola virus can be effectively neutralized by antibody produced in natural human infection. J Virol 1999; 73: Rubins KH, Hensley LE, Wahl-Jensen V, et al. The temporal program of peripheral blood gene expression in the response of nonhuman primates to Ebola hemorrhagic fever. Genome Biol 2007; 8:R Escudero-Perez B, Volchkova VA, Dolnik O, Lawrence P, Volchkov VE. Shed GP of Ebola virus triggers immune activation and increased vascular permeability. PLoS Pathog 2014; 10:e Mateo M, Carbonnelle C, Reynard O, et al. VP24 is a molecular determinant of Ebola virus virulence in guinea pigs. J Infect Dis 2011; 204(suppl 3):S HierholzerJC,KillingtonRA.Virusisolationandquantitation.In: Mahy BWJ, Kangro HO, eds. Virology methods manual. London: Academic Press Limited, 1996: Volchkova VA, Dolnik O, Martinez MJ, Reynard O, Volchkov VE. Genomic RNA editing and its impact on Ebola virus adaptation during serial passages in cell culture and infection of guinea pigs. J Infect Dis 2011; 204(suppl 3):S Mohan GS, Ye L, Li W, et al. Less is more: Ebola virus surface glycoprotein expression levels regulate virus production and infectivity. J Virol 2015; 89: S328 JID 2015:212 (Suppl 2) Dolnik et al