286, 384–390 (2001) doi:10.1006/viro.2001.1012, available online at http://www.idealibrary.com on

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Vaccine Potential of Ebola VP24, VP30, VP35, and VP40

Julie A. Wilson, Mike Bray, Russell Bakken, and Mary Kate Hart1

Virology Division, United States Army Medical Research Institute of Infectious Diseases, 1425 Porter Street, Fort Detrick, Frederick, Maryland 21702-5011 Received February 23, 2001; returned to author for revision March 27, 2001; accepted May 16, 2001; published online July 5, 2001

Previous vaccine efforts with Ebola virus Zaire (EBOV-Z) emphasized the potential protective efficacies of immune responses to the surface and the nucleoprotein. To determine whether the VP24, VP30, VP35, and VP40 proteins are also capable of eliciting protective immune responses, these were expressed from alphavirus replicons and used to vaccinate BALB/c and C57BL/6 mice. Although all of the VP proteins were capable of inducing protective immune responses, no single VP protected both strains of mice tested. VP24, VP30, and VP40 induced protective immune responses in BALB/c mice, whereas C57BL/6 mice survived challenge only after vaccination with VP35. Passive transfer of immune sera to the VP proteins did not protect unvaccinated mice from lethal disease. The demonstration that the VP proteins are capable of eliciting protective immune responses to EBOV-Z indicates that they may be important components of a vaccine designed to protect humans from Ebola hemorrhagic fever. Key Words: Ebola virus; vaccine; VP24; VP30; VP35; VP40; VEE ; protection; immunity.

INTRODUCTION et al., 1996), presumably mediates into host cells by receptor binding and membrane fusion and is Ebola (EBOVs) are members of the family the only viral protein known to be on the surface of Filoviridae and are associated with outbreaks of highly virions and infected cells. The nonstructural glycoprotein lethal hemorrhagic fever in humans and nonhuman sGP, which is encoded by the unedited GP mRNA, is primates. Three different EBOV subtypes (Zaire, Su- synthesized in infected cell cultures (Volchkov et al., dan, and Ivory Coast) have been isolated from human infections, with the highest case fatality rates (greater 1995; Sanchez et al., 1996) and can be found in the sera than 80%) being associated with the Zaire subtype of patients infected with EBOV (Sanchez et al., 1999). (Johnson et al., 1977, Khan et al., 1999). The natural The NP, VP30, VP35, and L proteins associate with the reservoir of the EBOV is unknown and no vaccines or genomic RNA in a ribonucleoprotein complex (Elliott et effective therapeutic treatments are currently available al., 1985). Using an artificial replication system, Mu¨hl- for human use. Because of the severe pathogenicity of berger et al. (1999) demonstrated that the NP, VP35, and EBOV, live virus vaccines are not generally considered L proteins are essential for replication and encapsidation to be a feasible option for vaccination. Similarly, killed of the EBOV . Although VP30 was not essential vaccines have a risk associated both with production for replication, it appeared to be necessary for efficient and with incomplete inactivation of the virus. Based on transcription in this system. The NP, VP35, VP30, and L these concerns, EBOV vaccine efforts have shifted proteins have also recently been shown to be essential toward developing vaccines that contain only one or a for the recovery of infectious EBOV-Z from cloned cDNAs few EBOV proteins. (Volchkov et al., 2001). In addition to being an essential Seven structural proteins and one nonstructural pro- component of the replication complex, VP35 was also tein are encoded in the nonsegmented single-stranded recently implicated as an interferon antagonist (Basler et negative-sense RNA genome of EBOV (Sanchez et al., 2000). VP35 may therefore facilitate in al.,1993; Volchkov et al., 1995). The structural proteins infected cells by blocking the induction of antiviral im- consist of a type I transmembrane glycoprotein (GP), a mune responses normally induced by the production of nucleoprotein (NP), four virion structural proteins (VP24, interferon. VP30, VP35, and VP40), and an RNA-dependent RNA In contrast, the VP40 and VP24 proteins are mem- polymerase (L). The transmembrane GP, which is ex- brane-associated and are most likely located on the pressed after RNA editing (Volchkov et al., 1995; Sanchez inside of the membrane (Elliott et al., 1985). VP40 has been shown to associate with cell membranes, where it is believed to be involved in maturation of the virus by 1 To whom correspondence and reprint requests should be ad- inducing viral assembly at the plasma membrane of dressed. Fax: (301) 619-2290. E-mail: [email protected]. infected cells (Ruigrok et al., 2000; Dessen et al., 2000).

384 0042-6822 VACCINE POTENTIAL OF EBOLA VIRUS VP PROTEINS 385

The function of VP24 is not known but it may serve as a minor matrix protein, facilitating the interaction of VP40 and/or GP with the RNP complex, or function in the uncoating of the virion during infection (Peters et al., 1996). Most of the efforts to develop vaccines to EBOV have examined the protective capacity of the GP and the NP. Protection against EBOV-Z has been induced in rodent models by GP or NP expressed from Venezuelan equine encephalitis (VEE) virus replicons (Pushko et al., 1997a, 2000; Wilson and Hart, 2001), recombinant vi- ruses (Gilligan et al., 1997), or naked DNA constructs (Vanderzanden et al., 1998; Xu et al., 1998). However, replicons expressing the EBOV-Z GP, or a combination of FIG. 1. Immunoprecipitation of 35S-labeled proteins expressed in BHK EBOV-Z GP and NP, failed to protect cynomologus ma- cells with polyclonal rabbit anti-EBOV serum. BHK cells were trans- caques from a subcutaneous EBOV-Z challenge with a fected with replicon encoding either EBOV-Z VP24, VP30, VP35, dose of 103 plaque-forming units (PFU) (J. Smith and P. or VP40. The structural proteins from EBOV-Z (virion) grown in Vero E6 cells are shown. Proteins were resolved under reducing conditions on Jahrling, personal communication). Recently, a DNA an 11% SDS–polyacrylamide gel. prime-adenovirus boost strategy was used to protect cynomologus macaques from a 6 PFU intraperitoneal (ip) EBOV-Z challenge (Sullivan et al., 2000), suggesting that RESULTS it should be possible to develop a human-use vaccine for EBOV. Construction and characterization of replicon vectors Collectively, studies published to date suggest that expressing the EBOV VP proteins both antibody and cytotoxic T cell (CTL) responses To evaluate the ability of the EBOV-Z VP proteins to should be induced to confer optimal immunity to EBOV serve as protective , we expressed the EBOV-Z in humans. Using passive and adoptive transfer stud- VP24, VP30, VP35, and VP40 proteins from VEE virus ies, we demonstrated that monoclonal antibodies to replicon vectors. The advantages of this vaccine vector the GP (Wilson et al., 2000) and CTLs specific for the include a tropism for -presenting dendritic cells, NP (Wilson and Hart, 2001) can protect unvaccinated high levels of expression of heterologous proteins in the mice from a lethal challenge of EBOV. If CTLs have an cytoplasm of infected cells, and the induction of cellular, important role in the ability of humans to survive EBOV mucosal, and humoral immunity to heterologous anti- infection, then the optimal vaccine would be one that gens that are cloned in the place of the structural genes provides the most opportunities to induce CTLs. As of the VEE virus (Davis et al., 1996; Pushko et al., 1997b; only a limited number of CTL epitopes on a given viral Caley et al., 1997; MacDonald and Johnston, 2000). protein are capable of being presented by a particular To construct VEE virus replicons expressing the class I-MHC molecule, it may be necessary to include EBOV-Z VP genes, RT-PCR products containing the VP several EBOV proteins in a vaccine that is capable of genes of EBOV-Z (isolate Mayinga) were inserted down- stream of the 26S promoter in the VEE virus replicon inducing CTLs in human populations with diverse vector. SDS gel electrophoresis of proteins immunopre- MHC binding motifs. cipitated from cells transfected with the replicon con- To date, the ability of the EBOV VP proteins to induce structs confirmed that the expressed EBOV-Z proteins protection has been examined by expressing the pro- were of the predicted size (Fig. 1). Lower molecular teins from vaccinia virus and evaluating efficacy in a weight species were also observed in the lane contain- guinea pig model (Gilligan et al., 1997; Chepurnov et ing VP40. We did not determine whether these proteins al., 1997). There have been no reports of protective represented proteolytic degradation products or cellular immunity induced by the VP proteins. Because of the proteins that coimmunoprecipitated with VP40; however, potential importance of these proteins in the develop- similar findings were observed for VP40 expressed from ment of an optimal vaccine for EBOV, we evaluated the a recombinant vaccinia virus (Gilligan et al., 1997). DNA ability of the EBOV-Z VP24, VP30, VP35, and VP40 sequence analysis showed that the coding sequences of proteins to elicit protective immunity in two strains of the VP24, VP35, and VP40 genes cloned in the replicon mice. We selected mice that differ genetically, includ- vector were identical to those previously described (Gen- ing at the major histocompatibility locus, to evaluate Bank Accession Nos. L11365.1 and AF086833). The VP30 the efficacy of these proteins against different genetic protein expressed from the replicon vector was identical backgrounds. to that reported by Volchkov et al. (GenBank Accession 386 WILSON ET AL.

TABLE 1 Survival of Mice Vaccinated with VRPs Encoding EBOV-Z VP Proteins

ELISA GMTe Viremia, GMTf a b c d VRP Doses S/T MDD #Ab Positive (Log10) (Log10 PFU/mL)

A. BALB/c Mice EBOV VP24 3 18/20 8 Ϯ 1 19/20 2.4 Ϯ 0.2 5.2 Ϯ 0.3 2 19/20 6 18/20 2.4 Ϯ 0.2 4.8 Ϯ 1.0 EBOV VP30 3 17/20 7 Ϯ 0 20/20 2.6 Ϯ 0.3 6.2 Ϯ 0.7 2 11/20 7 Ϯ 1 18/20 2.3 Ϯ 0.3 6.5 Ϯ 0.8 EBOV VP35 3 5/19 7 Ϯ 1 17/19 3.6 Ϯ 0.5 6.9 Ϯ 0.9 2 4/20 7 Ϯ 2 16/20 2.9 Ϯ 0.6 6.5 Ϯ 0.9 EBOV VP40 3 14/20 8 Ϯ 1 20/20 4.3 Ϯ 0.4 4.6 Ϯ 1.5 2 17/20 7 Ϯ 1 20/20 4.1 Ϯ 0.4 5.6 Ϯ 0.4 Lassa N 3 0/20 7 Ϯ 1 0/10 Ͻ2.0 8.0 Ϯ 0.3 2 0/20 7 Ϯ 1 0/10 Ͻ2.0 8.4 Ϯ 0.4

B. C57BL/6 Mice EBOV VP24 3 0/20 7 Ϯ 0 11/20 2.5 Ϯ 0.3 8.6 Ϯ 0.2 EBOV VP30 3 2/20 8 Ϯ 1 18/20 2.7 Ϯ 0.5 7.7 Ϯ 1.3 EBOV VP35 3 14/20 8 Ϯ 1 20/20 3.0 Ϯ 0.6 4.5 Ϯ 1.0 EBOV VP40 3 1/20 7 Ϯ 1 19/20 3.4 Ϯ 0.5 7.8 Ϯ 0.5 Lassa N 3 1/20 7 Ϯ 1 0/10 Ͻ2.0 8.6 Ϯ 0.7

a Mice were injected two or three times at 1 month intervals with 2 ϫ 10 6 focus-forming units of VRPs encoding individual EBOV-Z VP proteins or the control Lassa N protein. b S/T, survivors/total challenged with 300 LD50 of mouse-adapted EBOV-Z. c MDD, mean day of death of mice that died after EBOV challenge. d Number of mice with detectable antibody (Ab) responses/number tested in ELISA at a serum dilution of 1:100. For Lassa N-vaccinated mice, all samples were negative from individual mice tested in one experiment (n ϭ 10) or pooled from separate experiments. e GMT, geometric mean titer calculated from endpoint titers of prechallenge sera in ELISA using irradiated, sucrose-purified EBOV-Z virions as antigen. For EBOV-vaccinated animals, mice without detectable antibodies at a serum dilution of 1:100 are not included. f GMT, geometric mean titer calculated from the sera of separate groups of mice sacrificed on day 4 (for BALB/c mice) or day 5 (for C57BL/6 mice) post-EBOV challenge. All mice were viremic when tested, n ϭ 5 for all groups except n ϭ 4 for Lassa N-vaccinated C57BL/6 mice and BALB/c mice that received two injections of EBOV-Z VP24 VRPs.

No. AF086833) with the exception that it contained 37 BALB/c mice from lethal EBOV challenge (Table 1A). The amino acids of irrelevant sequence in the place of its highest levels of protection (90–95%) were observed after C-terminal proline residue. However, the addition of vaccination with VRPs expressing EBOV-Z VP24 (Table these amino acids did not alter the reactivity of the 1A). In a similar experiment, two inoculations of VRPs expressed VP30 protein with anti-EBOV antisera which encoding the EBOV-Z VP24 protein also protected 5/5 ϫ 4 recognizes native EBOV-Z VP30 (Fig. 1). BALB/c mice from a 3 10 LD50 challenge dose and 5/5 ϫ 6 BALB/c mice from a 3 10 LD50 challenge dose. Three Evaluation of protective antigens in mice injections of VRPs encoding EBOV-Z VP30 induced pro- To evaluate the protective efficacy of the EBOV-Z VP tection from lethal disease in 85% of the BALB/c mice proteins, the replicon constructs expressing the VP pro- examined. However, when the vaccination schedule was teins from the Mayinga isolate of EBOV-Z were packaged decreased to two injections, only 55% of the mice immu- into virus-like replicon particles (VRPs) and injected into nized with VP30 survived challenge. Vaccination with two strains of mice that genetically differ, including at the VRPs encoding the VP40 protein protected 85 and 70% of MHC locus. Control mice were injected with VRPs ex- the BALB/c mice after either two or three injections, pressing the N protein of Lassa virus. BALB/c (H-2d)or respectively. In contrast, the VP35 protein was not effi- C57BL/6 (H-2b) mice were vaccinated two or three times cacious in the BALB/c mouse model, as only 20 and 26% at 1-month intervals with 2 ϫ 10 6 focus-forming units of of the mice were protected from lethal challenge after the VRPs and were challenged by ip inoculation of 300 either two or three doses, respectively. The mean day of

LD50 of mouse-adapted EBOV-Z (Bray et al., 1998) 1 death of the VP-vaccinated mice that succumbed to the month after the final booster injection. EBOV challenge was within 1 day of the control Lassa Vaccination with VRPs encoding the EBOV-Z VP24, N-vaccinated mice (Table 1A). When measured in a VP30, or VP40 protein protected the majority of the plaque assay, all of the BALB/c mice that were vacci- VACCINE POTENTIAL OF EBOLA VIRUS VP PROTEINS 387 nated with the EBOV-Z VP antigens had measurable TABLE 2 levels of infectious EBOV in the sera 4 days after chal- Protection from EBOV by Passive Transfer of Immune Sera lenge (Table 1A). However, vaccination reduced the level a b c d of viremia more than 2 log10 compared to controls. Donor sera S/T MDD ELISA GMT (Log10) In contrast to the efficacy observed in the BALB/c mice, C57BL/6 mice were protected from lethal EBOV A. BALB/c Mice challenge only after vaccination with the EBOV-Z VP35 EBOV GP 15/20 8 Ϯ 1 4.0 Ϯ 0.3 protein, with 70% of the mice protected after three inoc- EBOV VP24 0/20 6 Ϯ 1 2.6 Ϯ 0.0 Ϯ Ϯ ulations (Table 1B). When the viral titers were measured EBOV VP30 0/20 7 1 2.7 0.1 EBOV VP40 0/20 6 Ϯ 1 4.3 Ϯ 0.1 5 days after challenge, vaccination with VRPs encoding Lassa N 0/20 7 Ϯ 1 Ͻ2.0 the EBOV-Z VP35 protein reduced the by at least 4 log10 compared to control mice (Table 1B). In B. C57BL/6 Mice contrast, viral titers in the sera of C57BL/6 mice vacci- EBOV GP 17/20 7 Ϯ 2 3.8 Ϯ 0.1 nated with the EBOV-Z VP24, VP30, or VP40 VRPs were EBOV VP35 0/20 7 Ϯ 1 2.8 Ϯ 0.0 within 1 log of the control mice. Similar to what was Lassa N 0/20 7 Ϯ 1 Ͻ2.0 observed with the BALB/c mice, the mean day of death of a C57BL/6 mice that were not protected by VRPs encoding Donor sera were obtained 28 days after the third immunization with ϫ 6 the EBOV-Z VP proteins was within 1 day of the control 2 10 focus-forming units of VRPs encoding the indicated EBOV-Z protein or the Lassa N protein. One milliliter of pooled donor sera was mice injected with the Lassa N VRP (Table 1B). administered ip to unvaccinated, syngeneic mice 24 h before chal- lenge. b Antibodies to VP proteins S/T, survivors/total challenged with 300 LD50 of mouse-adapted EBOV-Z. Most of the mice had detectable EBOV-Z-specific se- c MDD, mean day of death of mice that died after EBOV challenge. rum antibodies after vaccination with VRPs encoding the d GMT, geometric mean titer calculated from endpoint titers of pooled EBOV-Z VP proteins (Table 1). For the vaccinated groups donor sera used in passive transfer experiments. ELISAs were per- that were protected from EBOV challenge, passive trans- formed using irradiated, sucrose-purified EBOV-Z virions as antigen. fer studies were performed to determine whether anti- bodies to the VP proteins are capable of conferring resistance to lethal EBOV disease. Sera were obtained any of the four VP proteins of EBOV-Z, but that protection from mice previously vaccinated with VRPs expressing was dependent on the mouse strain tested. either one of the EBOV-Z VP proteins, the negative- The observed VP-induced protection in mice differs control Lassa N or the positive-control EBOV-Z GP. One from previous studies (Gilligan et al., 1997; Chepurnov et milliliter of pooled immune sera was passively adminis- al., 1997) in which recombinant vaccinia viruses express- tered to syngeneic, unvaccinated mice 24 h before chal- ing either the EBOV-Z VP24, VP35, or VP40 proteins failed to protect guinea pigs from lethal challenge with guinea lenge with 300 LD50 of mouse-adapted EBOV-Z. When the pooled donor sera were tested for in vitro inhibition of pig-adapted EBOV-Z. It is not clear if this is due to the EBOV-Z plaque formation, a 1:20 dilution of antisera to different animal models examined or the different vector the GP or the VP proteins did not inhibit plaque formation and vaccination schedules used in the studies. However, by the mouse-adapted EBOV-Z by 80% compared to con- differences in protective efficacy were also observed trol wells. Despite the absence of in vitro neutralization previously when VRPs expressing EBOV-Z NP success- activity, most of the mice receiving the control immune fully protected BALB/c and C57BL/6 mice (Pushko et al., sera to GP survived lethal challenge (Table 2). In con- 2000; Wilson and Hart, 2001) but not strain 2 or strain 13 trast, antisera to the EBOV-Z VP24, VP30, VP35, and VP40 guinea pigs (Pushko et al., 1997a, 2000). Differences in proteins, which demonstrated ELISA titers similar to vac- protective efficacy of DNA vaccination with EBOV-Z NP cinated animals prior to EBOV challenge, failed to protect have also been observed in Hartley guinea pigs depend- the recipient mice from EBOV challenge or extend the ing on the vaccination schedule examined (Xu et al., time to death compared to control mice receiving Lassa 1998). We also observed differences in the ability of the N immune serum (Table 2). EBOV-Z VP proteins to induce protective responses in mice with different MHCs. We could not vaccinate with any one EBOV-Z VP protein to achieve protection in both DISCUSSION mouse strains, and only one of the four EBOV-Z VP EBOVs are extremely lethal pathogens for which no proteins was efficacious for C57BL/6 mice. As trans- vaccines or effective therapeutic measures are available. ferred antisera did not confer protection against lethal In this study, we examined the ability of the EBOV-Z VP challenge, it is possible that the failure of some mice and proteins to serve as protective antigens in a murine guinea pigs to survive challenge after vaccination with model of Ebola hemorrhagic fever. We found that high these proteins is due to a poor CTL response. The levels of protection could be induced by vaccination with mouse model used in this study has a very low LD50 388 WILSON ET AL.

(approximately one virion or 0.03 to 0.04 plaque-forming MATERIALS AND METHODS units) and viremias approaching 9 log10 are observed in infected mice (Table 1; Bray et al., 1998). We routinely Cells lines and viruses challenged the mice in this study with 300 LD50 of Baby hamster kidney (BHK, ATCC CCL 10), Vero 76 mouse-adapted EBOV-Z. It is possible that the higher (ATCC CRL 1587), and Vero E6 (ATCC CRL 1586) cell lines ϫ 3 challenge dose of guinea pig-adapted EBOV-Z (1 10 were maintained in minimal essential medium (MEM) ϫ 4 to 1 10 LD50) used in some guinea pig studies (Gilli- with Earle’s salts, 10% fetal bovine serum, and 50 ␮g/ml gan et al., 1997, Chepurnov et al., 1997, Pushko et al., of gentamicin sulfate. 1997a, 2000) may have overwhelmed the induced im- The mouse-adapted EBOV-Z was previously obtained mune responses. However, BALB/c mice vaccinated with by serial passage of the EBOV-Z (isolate Mayinga) in EBOV-Z VP24 were protected from lethal disease even mice (Bray et al., 1998). Frozen aliquots of a double- ϫ 5 when the challenge dose was increased up to 1 10 plaque-purified ninth-mouse-passage isolate were used ϫ 6 PFU (3 10 LD50). in all mouse challenge experiments and neutralization We could not demonstrate a role for antibodies in assays. This virus is uniformly lethal for immunocompe- mediating the VP-induced protection, which was not un- tent adult mice by ip inoculation, with an LD50 of approx- expected, as the VP proteins are not found on the surface imately one virion (0.03–0.04 Vero E6 plaque-forming of virions or infected cells and are presumed to be units) (Bray et al., 1998). The 1976 and 1995 isolates of inaccessible to the protective effects of antibodies. The EBOV-Z were provided by P. B. Jahrling. titers of the antibodies induced by vaccination with the EBOV-Z VP proteins were similar in both strains of mice Cloning of EBOV genes and construction of replicon tested (Table 1). Therefore, the inability of the VP proteins DNA to protect both strains of mice did not appear to be due to an inability of the humoral immune system to respond A plasmid encoding the VEE replicon vector contain- to the antigens. ing a unique ClaI site downstream from the 26S promoter The detection of viremia after challenge indicates that was described previously (Davis et al., 1996). The vaccination with the EBOV-Z VP proteins did not induce EBOV-Z GP and the Lassa N genes were previously sterile immunity. This finding was not surprising, given cloned into the VEE replicon vector as described the lack of protection observed in the antibody transfer (Pushko et al., 1997a,b). studies and the fact that other immune mechanisms, The EBOV-Z VP genes used in this study were cloned such as CTLs, would be involved only after viral infection by RT-PCR of RNA from EBOV-Z (strain Mayinga)-infected had occurred. However, vaccination with the EBOV-Z VP Vero E6 cells. First-strand synthesis was primed with proteins significantly reduced viral replication in the pro- oligo(dT) (Life Technologies). Second-strand synthesis tected groups and enabled the mice to survive the EBOV and subsequent amplification of viral cDNAs were per- challenge. It is likely that a strong CTL response to the formed with the -specific primers listed below. The VP proteins was capable of controlling viral replication primer sequences were derived from a previously pub- early in infection. It remains to be determined if the ability lished sequence for EBOV-Z (Sanchez et al., 1993) and of the VP proteins to protect against lethal EBOV chal- were designed to contain a ClaI restriction site for clon- lenge is due to the ability to mount an effective CTL ing the amplified VP genes into the ClaI site of the response, but our observation that the EBOV-Z VP pro- replicon vector. The VP24 forward primer was 5Ј-GG- teins differed in their ability to induce protective immune GATCGATCTCCAGACACCAAGCAAGACC-3Ј, the VP24 responses in MHC-disparate mice is compatible with reverse primer was 5Ј-GGGATCGATGAGTCAGCATATAT- this suggestion. GAGTTAGCTC-3Ј, the VP30 forward primer was 5Ј- The inability of some animals to mount protective re- CCCATCGATCAGATCTGCGAACCGGTAGAG-3Ј, the VP30 sponses to individual EBOV proteins highlights the need reverse primer was 5Ј-CCCATCGATGTACCCTCATCA- to thoroughly examine the immune mechanisms mediat- GACCATGAGC-3Ј, the VP35 forward primer was 5Ј-GG- ing protection as well as the ability of vaccine vectors to GATCGATAGAAAAGCTGGTCTAACAAGATGA-3Ј, the VP35 effectively induce those responses. It also indicates that reverse primer was 5Ј-CCCATCGATCTCACAAGTGTATCAT- an eventual EBOV vaccine will probably have to include TAATGTAACGT-3Ј, the VP40 forward primer was several EBOV proteins to induce protection in recipients 5Ј-CCCATCGATCCTACCTCGGCTGAGAGAGTG-3Ј, and with diverse MHC proteins. Our finding that protective the VP40 reverse primer was 5Ј-CCCATCGATATGTTAT- immunity to a lethal EBOV challenge can be induced in GCACTATCCCTGAGAAG-3Ј. mice by vaccination with the EBOV-Z VP24, VP30, VP35, PCR products containing individual EBOV-Z VP genes and VP40 proteins indicates that these proteins warrant were digested with ClaI and cloned into the ClaI site of further consideration as components of an eventual vac- the VEE replicon vector downstream from the VEE 26S cine candidate for EBOV. promoter. VACCINE POTENTIAL OF EBOLA VIRUS VP PROTEINS 389

Production of recombinant VEE virus replicons four-parameter curve fit (SOFTmax software, Molecular De- vices Corp.) of background-subtracted mean OD vs dilution, Replicon RNAs were packaged into particles as de- followed by extrapolation of the dilution at which the OD scribed (Pushko et al., 1997b). Briefly, capped replicon was 0.20. Mice that did not have detectable antibodies to RNAs were produced in vitro by T7 runoff transcription of EBOV are noted but were not included in the geometric NotI-digested plasmid templates using the RiboMAX T7 mean titer calculations. RNA polymerase kit (Promega). BHK cells were cotrans- Plaque assays were performed as described (Moe et fected with the replicon RNAs and two RNAs expressing al., 1981) with confluent Vero E6 cells. To evaluate the the VEE virus structural proteins. The cell culture super- presence of EBOV-neutralizing antibodies, serial dilu- natants were harvested approximately 30 h after trans- tions of mouse sera were mixed with 100 PFU of mouse- fection and the replicon particles were concentrated and adapted EBOV-Z at 37°C for 1 h and used to infect partially purified by centrifugation through a 20% sucrose confluent Vero E6 cells. None of the samples reduced cushion. Packaged VRPs were suspended in phosphate- the number of plaques by 80% compared to control wells. buffered saline (PBS) and titers were determined as immunofluorescent foci after infection of Vero cells as described (Pushko et al., 1997b) using either EBOV-im- Vaccination and challenge of mice mune rabbit serum or mouse monoclonal antibodies to Specific pathogen-free 6- to 8-week-old female VP24 (Z-AC01-BG11-01), VP35 (M-HC01-AF11), or VP40 BALB/c or C57BL/6 mice (NCI, Frederick, MD) were (M-HD06-AD10). housed in cages equipped with microisolators and pro- vided food and water ad libitum. Mice were acclimated Metabolic labeling and radioimmunoprecipitation of for 1 week before vaccination. Research was conducted EBOV proteins in compliance with the Animal Welfare Act and other Vero E6 cells (75 cm2 flasks) were infected with the federal statutes and regulations relating to animals and 1995 strain of EBOV-Z at a multiplicity of infection of 1–3 experiments involving animals and adheres to principles PFU/cell. After 24 h of infection, the growth medium was stated in the Guide for the Care and Use of Laboratory removed and cells were starved for 30 min in medium Animals, National Research Council, 1996. The facility lacking methionine and cysteine. To label viral proteins, where this research was conducted is fully accredited by cells were incubated for 24 h in MEM containing 2% the Association for Assessment and Accreditation of heat-inactivated fetal bovine serum, 0.1 mCi/ml 35S-la- Laboratory Animal Care International. Animal experi- beled methionine, and 0.1 mCi/ml 35S-labeled cysteine. ments were performed a minimum of two times. The medium was harvested and centrifuged to remove Groups of 10 BALB/c or C57BL/6 mice per experiment cell debris (15 min at 1500 g). Labeled EBOV-Z virions were subcutaneously injected at the base of the neck ϫ 6 were pelleted through a 20% sucrose cushion (3 h at with 2 10 focus-forming units of VRPs encoding the 36,000 rpm in a SW41 rotor) and suspended in Zwitter- EBOV-Z genes, or with a control replicon encoding the gent lysis buffer (10 mM Tris–HCl, pH 8.0, 0.5 M NaCl, 1 Lassa N gene. Booster immunizations were adminis- mM EDTA, 4% Zwittergent 3-14, and protease inhibitors). tered at 1-month intervals. Survival data are recorded as To label proteins expressed by VReps, BHK cells were the combined results of two separate experiments. Anes- electroporated with VRep RNAs expressing each of the thetized mice were bled from the retro-orbital sinus 1 EBOV-Z VP proteins. After 16 h of incubation, cells were week before challenge to examine antibody titers to the starved for 30 min in medium lacking methionine and cys- viral proteins. One month after the final booster injection, teine and then incubated for4hinMEMmedium contain- mice were transferred to a BSL-4 containment area and ing 0.1 mCi/ml 35S-labeled methionine and cysteine. Cell challenged by ip inoculation of 300 LD50 (10 PFU) of monolayers were lysed in Zwittergent lysis buffer and the mouse-adapted EBOV-Z. The mice were observed daily EBOV-Z VP proteins were analyzed by immunoprecipitation for 28 days, which is approximately four times longer as described (Hevey et al., 1997) using EBOV-immune rab- than the mean day of death of control mice, and morbid- bit serum. Immunoprecipitated proteins were resolved by ity and mortality were recorded. Surviving mice were electrophoresis on an 11% SDS–polyacrylamide gel and back-challenged with 300 LD50 of mouse-adapted were visualized by autoradiography. EBOV-Z and monitored for an additional 28 days. All mice that survived the initial challenge virus also survived the subsequent back-challenge. Serological assays To determine serum viral titers after challenge, four or Enzyme-linked immunosorbent assays (ELISAs) were five mice from vaccinated and control groups were anes- performed essentially as described (Hevey et al., 1997) on thetized and exsanguinated on day 4 (BALB/c mice) or plates coated with irradiated, sucrose-purified EBOV-Z 1995 day 5 (C57BL/6 mice) after challenge. The viral titers in virions. Absorbance values 0.2 above control samples were individual sera were determined by plaque assay. For considered positive. Endpoint titers were calculated using a passive transfer studies, donor sera were obtained from 390 WILSON ET AL. anesthetized mice 28 days after the third inoculation with geting in Venezuelan equine encephalitis virus pathogenesis. J. Virol. VRPs. One milliliter of pooled donor sera was adminis- 74, 914–922. tered ip to naive, syngeneic mice 24 h before ip chal- Moe, J. B., Lambert, D. B., and Lupton, H. W. (1981). Plaque assay for Ebola virus. J. Clin. Microbiol. 13, 791–793. lenge with 300 LD50 of mouse-adapted EBOV-Z. Anesthe- Mu¨hlberger, E., Weik, M., Volchkov, V. E., Klenk, H.-D., and Becker, S. tized recipient mice were bled from the retro-orbital si- (1999). Comparison of the transcription and replication strategies of nus 5 h prior to challenge to determine serum antibody Marburg virus and Ebola virus by using artificial replication systems. titers to the EBOV-Z proteins. J. Virol. 73, 2333–2342. Peters, C. J., Sanchez, A., Rollin, P. E., Ksiazek, T. G., and Murphy, F. A. (1996). Filoviridae: Marburg and Ebola viruses, In “Fields Virology, 3rd ACKNOWLEDGMENTS ed.” (B. N. Fields, D. M. Knipe, and P. M. Howley, Eds.), pp. 1161–1176. Lippincott-Raven, Philadelphia. We thank P. Pushko and J. Smith for providing the EBOV-Z GP and Pushko, P., Bray, M., Ludwig, G. V., Parker, M., Schmaljohn, A., Sanchez, Lassa N replicon constructs used in these studies and M. Hevey for A., Jahrling, P. B., and Smith, J. F. (2000). Recombinant RNA replicons providing the RNA from EBOV-Z (Mayinga isolate)-infected Vero E6 derived from attenuated Venezuelan equine encephalitis virus pro- cells used for cloning the EBOV-Z VP genes. We also thank E. Shew- tect guinea pigs and mice from Ebola hemorrhagic fever virus. bridge and M. Gibson for care of the animals used in this study. J.A.W. Vaccine 19, 142–53. was supported by a National Research Council fellowship. The views Pushko, P., Parker, M., Geisbert, J., Negley, D., Schmaljohn, A., Jahrling, of the authors do not purport to reflect the positions of the Army or the P., Sanchez, A., and Smith, J. F. (1997a). Venezuelan equine enceph- Department of Defense. alitis virus replicon vector: Immunogenicity studies with Ebola NP and GP genes in guinea pigs. In “Vaccines 97,” pp. 253–258. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York. REFERENCES Pushko, P., Parker, M., Ludwig, G. V., Davis, N. L., Johnston, R. E., and Smith, J. F. (1997b). Replicon-helper systems from attenuated Vene- Basler, C. F., Wang, X., Mu¨hlberger, E., Volchkov, V., Paragas, J., Klenk, zuelan equine encephalitis virus: Expression of heterologous genes H.-D., Garcı´a-Sastre, A., and Palese, P. (2000). The Ebola virus VP35 in vitro and immunization against heterologous pathogens in vivo. protein functions as a type I IFN antagonist. Proc. Natl. Acad. Sci. Virology 239, 389–401. USA 97, 12289–12294. Ruigrok, R. W. H., Schoehn, G., Dessen, A., Forest, E., Volchkov, V., Bray, M., Davis, K., Geisbert, T., Schmaljohn, C., and Huggins, J. (1998). Dolnik, O., Klenk, H.-D., and Weissenhorn, W. (2000). Structural char- A mouse model for evaluation of prophylaxis and therapy of Ebola acterization and membrane binding properties of the matrix protein hemorrhagic fever. J Infect. Dis. 178, 651–661. VP40 of Ebola virus. J. Mol. Biol. 300, 103–112. Caley, I. J., Betts, M. R., Irlbeck, D. M., Davis, N. L., Swanstrom, R., Sanchez, A., Kiley, M. P., Holloway, B. P., and Auperin, D. D. (1993). Frelinger, J. A., and Johnston, R. E. (1997). Humoral, mucosal, and Sequence analysis of the Ebola virus genome: Organization, genetic cellular immunity in response to a human immunodeficiency virus elements, and comparison with the genome of Marburg virus. Virus type I immunogen expressed by a Venezuelan equine encephalitis Res. 29, 215–240. virus vaccine vector. J. Virol. 71, 3031–3038. Sanchez, A., Ksiazek, T. G., Rollin, P. E., Miranda, M. E. G., Trappier, S. G., Chepurnov, A. A., Ternovoi, V. A., Dadaeva, A. A., Dmitriev, I. P., Sizikova, Khan, A. S., Peters, C. J., and Nichol, S. T. (1999). Detection and molec- L. P., Volchkov, V. E., Kudoyarova, N. M., Rudzevich, T. N., and ular characterization of Ebola viruses causing disease in human and Netesov, S. V. (1997). Immunobiological properties of vp24 protein of nonhuman primates. J. Infect. Dis. 179(Suppl. 1), S164–S169. Ebola virus expressed by recombinant vaccinia virus. Vopr. Virusol. 3, Sanchez, A., Trappier, S. G., Mahy, B. W., Peters, C. J., and Nichol, S. T. 115–120. (1996). The virion of Ebola viruses are encoded in two Davis, N. L., Brown, K. W., and Johnston, R. E. (1996). A viral vaccine reading frames and are expressed through transcriptional editing. vector that expresses foreign genes in lymph nodes and protects Proc. Natl. Acad. Sci. USA 93, 3602–3607. against mucosal challenge. J. Virol. 70, 3781–3787. Sullivan, N. J., Sanchez, A., Rollin, P. E., Yang, Z. Y., and Nabel, G. J. Dessen, A., Volchkov, V., Dolnik, O., Klenk, H.-D., and Weissenhorn, W. (2000). Development of a preventative vaccine for Ebola virus infec- (2000). Crystal structure of the matrix protein VP40 from Ebola virus. tion in primates. Nature 408, 605–609. EMBO J. 19, 4228–4236. Vanderzanden, L, Bray, M., Fuller, D., Roberts, T., Custer, D., Spik, K., Elliott, L. H., Kiley, M. P., and McCormick, J. B. (1985). Descriptive Jahrling, P., Huggins, J., Schmaljohn, A., and Schmaljohn, C. (1998). analysis of Ebola virus proteins. Virology 147, 169–176. DNA vaccines expressing either the GP or NP genes of Ebola virus Gilligan, K. J., Geisbert, J. B., Jahrling, P. B., and Anderson, K. A. (1997). protect mice from lethal challenge. Virology 246, 134–144. Assessment of protective immunity conferred by recombinant vac- Volchkov, V. E., Volchkova, V. A., Mu¨hlberger, E., Kolesnikova, L. V., Weik, cinia viruses to guinea pigs challenged with Ebola virus. In “Vaccines M., Dolnik, O., and Klenk, H.-D. (2001). Recovery of infectious Ebola 97,” pp. 87–92. Cold Spring Harbor Laboratory Press, Cold Spring virus from complementary DNA: RNA editing of the GP gene and viral Harbor, New York. cytotoxicity. Science 291, 1965–1969. Hevey, M., Negley, D., Geisbert, J., Jahrling, P., and Schmaljohn, A. Volchkov, V. E., Becker, S., Volchkova, V. A., Ternovoj, V. A., Kotov, A. N., (1997). Antigenicity and vaccine potential of Marburg virus glycopro- Netesov, S. V., and Klenk, H.-D. (1995). GP mRNA of Ebola virus is tein expressed by baculovirus recombinants. Virology 239, 206–216. edited by the Ebola virus polymerase and by T7 vaccinia virus Johnson, K. M., Lange, J. V., Webb, P. A., and Murphy, F. A. (1977). polymerases. Virology 214, 421–430. Isolation and partial characterization of a new virus causing acute Wilson, J. A., and Hart, M. K. (2001). Protection from Ebola virus medi- haemorrhagic fever in Zaire. Lancet 12, 569–571. ated by cytotoxic T lymphocytes to the viral nucleoprotein. J. Virol. 75, Khan, A. S., Tshiokoo, F. K., Heymann, D. L., Le Guenno, B., Nabeth, P., 2660–2664. Kerstiens, B., Fleerackers, Y., Kilmarx, P. H., Rodier, G. R., Nkuku, O., Wilson, J. A., Hevey, M., Bakken, R., Guest, S., Bray, M., Schmaljohn, Rollin, P. E., Sanchez, A., Zaki, S. R., Swanepoel, R., Tomori, O., Nichol, A. L., and Hart, M. K. (2000). Epitopes involved in antibody-mediated S. T., Peters, C. J., Muyembe-Tamfum, J. J., and Ksiazek, T. G. (1999). protection from Ebola virus. Science 287, 1664–1666. The reemergence of Ebola hemorrhagic fever, Democratic Republic Xu, L., Sanchez, A., Zhi-Yong, Y., Zaki, S. R., Nabel, E. G., Nichol, S. T., of the Congo, 1995. J. Infect. Dis. 179(Suppl. 1), S76–S86. and Nabel, G. J. (1998). Immunization for Ebola virus infection. Nat. MacDonald, G. H., and Johnston, R. (2000). Role of dendritic cell tar- Med. 4, 37–42.