RSV A2 strain was obtained from ATCC (Manassas, MA). EZ vaccine strain was from Serum Institute of India (Pune, India). Vero and HEp-2 cells were maintained in Dulbecco’s modified Eagles’s medium (Invitrogen, Carlsbad, CA) supplemented with 2% fetal bovine serum (FBS), 2mM L-glutamine, and 1% penicillin/streptomycin at 37°C with 5% CO₂.

Antigenomic cDNA was synthesized by DNA2.0 (Menlo Park, CA) based on the sequence of live attenuated measles virus Edmonston-Zagreb strain (NCBI DNA accession number F266290). The T7 RNA polymerase promoter sequence was added to the 5’ end of the antigenomic cDNA and the hepatitis delta virus ribozyme and a T7 RNA polymerase terminator sequences were added to the 3’ end of the antigenomic cDNA. Sequence at nucleotides 5502 (Sca I) and 9343 (Mlu I) was modified to facilitate cDNA cloning. The three amino acid residues at positions 3682, 3862 and 10555 (3682_K→R, 3862_S→N, 10555_K→R) that were different from the EZ vaccine strains were corrected by site-directed mutagenesis. The MV antigenomic cDNA was cloned into pUC18 vector and denoted as pEZ (Fig. 1). Supporting plasmids encoding measles N, P and L genes were cloned into pCITE plasmids under the control of T7 promoter as previously described and were designed pN, pP and pL respectively [30Schneider H, Spielhofer P, Kaelin K, et al. Rescue of measles virus using a replication-deficient vaccinia-T7 vector J Virol Methods 1997; 64(1 ): 57-64.].

The ectodomain of RSV F and EBV gp350 ectodomains were inserted into the non-coding region 3' proximal of the N gene or in the junction between the P and M genes of pEZ (Fig. 1). Each inserted gene contained measles virus transcriptional unit consisting of the sequences of N gene stop (GE-N) and P gene start (GS-P). To insert the RSV F gene into the promoter proximal location, a pUC-SnaBI/SacII (pUC-S/S) subclone that contained the T7 promoter and MV sequence of 1 to 2044 with a silent mutation at nt 821 to knockout EcoRV enzyme site was generated. A gene cassette containing the ectodomain of RSV-F (aa 1-524) followed by GE-N and GS-P sequence were inserted into EcoRV site at nt 99. The genomic DNA was in multiples of 6 nucleotides (the rule of six). The SnaB I-Sac II fragment was swiped back to pEZ antigenomic cDNA (Fig. 1A) and the resulting cDNA was denoted as pEZ-RSVsF1.

To insert the RSV F gene and EBV gp350 gene into pEZ between the P and M genes, a pUC-SacII/NarI (pUC-S/N) subclone that contained the MV sequence of 2045 nt to 4924 nt was constructed. A gene cassette containing gene end (GE) of N, gene start (GS) of P and ectodomain of RSV-F (aa 1-524), full length (aa 1 -861) or truncated (aa 1-470) ectodomain of gp350 at SpeI site (nt 3372). The Sac II to Nar I fragment was swiped back to pEZ antigenomic cDNA (Fig. 1A) and the resulting cDNA was denoted as pEZ-RSVsF3, pEZ-EBVgp350 and pEZ-EBVgp350tr respectively.

Rescue of the recombinant viruses from MV antigenomic cDNA was performed as previously described [16Zuniga A, Wang Z, Liniger M, et al. Attenuated measles virus as a vaccine vector Vaccine 2007; 25(16 ): 2974-83., 30Schneider H, Spielhofer P, Kaelin K, et al. Rescue of measles virus using a replication-deficient vaccinia-T7 vector J Virol Methods 1997; 64(1 ): 57-64., 31Radecke F, Spielhofer P, Schneider H, et al. Rescue of measles viruses from cloned DNA EMBO J 1995; 14(23 ): 5773-84.]. Briefly, HEp-2 cells were infected with MVA-T7 at a multiplicity of infectivity (moi) of 5 for an hour prior to transfection. A DNA mixture of pEZ (0.8 μg), pN (0.4 μg), pP (0.4 μg) and pL (0.2 μg) were co-transfected into MVA-T7 infected HEp-2 cells using Lipofectamine 2000 (Invitrogen, Carlsbad, CA). The culture medium was replaced with DMEM containing 2% FBS after overnight incubation at 37^oC. The transfected cells were incubated for an additional 3 to 4 days and the culture supernatant was used to infect fresh Vero cells to amplify the rescued virus. The rescued viruses were then plaque purified twice in Vero cells. The viruses rescued from pEZ, pEZ-RSVsF1, pEZ-RSVsF3, pEZ-EBVgp350 and pEZ-EBVgp350tr were designated as rEZ, sF1, sF3, gp350 and gp350tr, respectively.

Replication kinetics of rEZ, gp350, gp350tr, sF1 or sF3 viruses was compared in Vero cells. Vero cells in six-well plates were infected with each virus at a MOI of 0.01 in triplicates. After a 1 hour adsorption at room temperature, the cells were washed three times with phosphate-buffered saline (PBS) and incubated with 2 mL of complete DMEM at 35^oC with 5% CO₂. Aliquots of culture supernatant were collected daily from each well for 7 days and stored at -80^oC until virus titration. Viral titers were determined by plaque assay in Vero cells as previously described [32Mann GF, Allison LM, Copeland JA, Agostini CF, Zuckerman AJ. A simplified plaque assay system for measles virus J Biol Stand 1980; 8(3 ): 219-5.]. Replication kinetic was performed twice in triplicate and each supernatant sample was titered in duplicate.

Vero cells were infected at a MOI of 3.0 with rEZ, gp350, gp350tr, sF1 or sF3 for 18 hours at 37°C. Culture supernatant and cells were harvested separately. The cells were scraped and pelleted for 10 sec at 6,000 rpm in microfuge tubes and lysed in lysis buffer (50mM Tris-HCl, 150 mM NaCl, 1% Triton X-100, 0.5% [vol/vol] protease inhibitor cocktail, pH 8.0) (Sigma, St. Louis, MO) for 10 min on ice. The resulting cell lysates were cleared by centrifugation at 13,000 rpm for 5 min.

Proteins from culture supernatant or cell lysates were separated by electrophoresis and transferred onto a PVDF membrane. Viral proteins were detected with anti-measles nucleoprotein (clone 2D7, Santa Cruz Biotechnology), polyclonal rabbit EBV gp350 [33Jackman WT, Mann KA, Hoffmann HJ, Spaete RR. Expression of Epstein-Barr virus gp350 as a single chain glycoprotein for an EBV subunit vaccine Vaccine 1999; 17(7-8 ): 660-8.] or monoclonal antibody against RSV F proteins (palivizumab, MedImmune, Gaithersburg, MD) antibodies. Intensities of the bands were quantified by ImageQuant software (GE Healthcare, Piscataway, NJ). Relative intensity was determined by comparing pixel intensity of the band relative to the rEZ control.

Specific-pathogen-free 6- to 8-week-old cotton rats were purchased from Harlan (Indianapolis, IN). The animal study protocol was approved by MedImmune’s Institutional Animal Care and Use Committee. Animals were anesthetized with isoflurane by inhalation and vaccinated intramuscularly (i.m.) with 10⁵ PFU of rEZ, gp350tr, gp350, sF1 or sF3 in a 200 μl inoculum volume. The control groups were inoculated intramuscularly (i.m) with PBS, 10⁵ PFU of UV-inactivated sF3, or intransally (i.n) with 10⁶ PFU of RSV A2 strain. The animals were boosted with the same dose of viruses 4 weeks later.

The animals were observed for clinical signs daily and bled at 14-day intervals to measure antibody responses. Twenty-eight days after the second immunization, the cotton rats from measles virus vectored RSV vaccine groups were challenged with 10⁶ PFU of RSV A2 strain i.n. The lungs were harvested on day 4 post-challenge and viral titers were determined by plaque assay.

Adult rhesus macaques aged 3-9 year-old were used for the immunogenicity study at Valley Biosystems (West Sacramento, CA). All experimental procedures were approved by the Institutional Animal Use and Care Committee at MedImmune and Valley Biosystems. Measles vaccine immunization history of the rhesus macaques was confirmed by neutralization assay prior to the study. A total of 16 rhesus macaques (8 measles virus seropositive, 8 naïve) were used in the study. Half of the animals (4 measles virus sero-positive, 4 naïve) received two doses of gp350 intramuscularly using 10⁵ PFU/inoculation and the other half received two doses of sF3 at the same dose intramuscularly. The inoculations were performed at 4 week intervals. Blood was drawn from each animal every 14 days to assess immune responses. Animals were anesthetized with ketamine intramuscularly at each procedure. Clinical symptoms were monitored by veterinarians for the duration of the study.

EBV gp350- or RSV F-specific enzyme-linked immunosorbent assay (ELISA) were performed to determine the levels of gp350 or F protein-specific antibodies. Briefly, 100 ng of purified recombinant soluble EBV gp350 or RSV F protein expressed from mammalian cells was adsorbed onto Immulon 2B plates overnight in PBS at 4°C. The plate was blocked with 5% BSA in PBS. Serial dilutions of animal sera in PBS/1%BSA were added to the plate and incubated for an hour at room temperature. After washing, HRP-conjugated anti-cotton rat IgG (Immunology Consultants Laboratory, Newberg, OR) or anti-rhesus macaques IgG (Rockland Immunochemicals, Boyertown, PA) were added and incubated for 1 hour. Finally, the plate was developed with tetramethylbenzidine peroxidase substrate (KPL, Rockford, IL) and stopped with HCl. Absorbances of the samples were read at 450 nm and the titer for each sample was calculated by determining the dilution at which the absorbance reached 2 times over the reading of no serum control. The lower limits of detection were 3.3 log₂ and 6.6 log₂ for EBV gp350 ELISA and RSV F ELISA, respectively.

Measles virus specific neutralizing antibodies were determined as described previously [34Christe KL, McChesney MB, Spinner A, et al. Comparative efficacy of a canine distemper-measles and a standard measles vaccine for immunization of rhesus macaques (Macaca mulatta) Comp Med 2002; 52(5 ): 467-72.] in 96-well plates using Vero cells. Neutralizing antibody titer is defined as the highest dilution that inhibits 50% CPE. RSV specific neutralizing antibody was determined by the plaque reduction neutralization assay as described previously [35Tang RS, Schickli JH, MacPhail M, et al. Effects of human metapneumovirus and respiratory syncytial virus antigen insertion in two 3' proximal genome positions of bovine/human parainfluenza virus type 3 on virus replication and immunogenicity J Virol 2003; 77(20 ): 10819-28.]. The lower limit of detection was 4.3 log₂ and 3.0 log₂ for MV microneutralization and RSV plaque reduction assays, respectively. EBV specific neutralizing antibody titer was determined by the method described by Sashihara et al.,[36Sashihara J, Burbelo PD, Savoldo B, Pierson TC, Cohen JI. Human antibody titers to Epstein-Barr Virus (EBV) gp350 correlate with neutralization of infectivity better than antibody titers to EBV gp42 using a rapid flow cytometry-based EBV neutralization assay Virology 2009; 391(2 ): 249-56.].

ELISpot assays were performed with spleens from vaccinated and control cotton rats 14 days after the second dose and peripheral blood mononucleocyte cells (PBMC) of rhesus monkeys 28 days after the second dose. Splenocytes were isolated as previously described [37Mok H, Lee S, Utley TJ, et al. Venezuelan equine encephalitis virus replicon particles encoding respiratory syncytial virus surface glycoproteins induce protective mucosal responses in mice and cotton rats J Virol 2007; 81(24 ): 13710-22.]. Splenocytes were counted using Vi-Cell (Beckman Coulter, Brea, CA). Gamma interferon (IFN-γ) and Interleukin 4-secreting T cells were quantified in an enzyme-linked immunospot (ELISpot) assay using R&D DuoSet ELISA System for cotton rat IFN-γ and IL-4 antibodies (R&D Systems, Minneapolis, MN) as previously described using 10 μg/mL of EBV gp350 or RSV F proteins for stimulation [38Zhan X, Slobod KS, Krishnamurthy S, et al. Sendai virus recombinant vaccine expressing hPIV-3 HN or F elicits protective immunity and combines with a second recombinant to prevent hPIV-1 hPIV-3 and RSV infections Vaccine 2008; 26(27-28 ): 3480-8.]. Spots were scanned and counted by ImmunoSpot spot Analyzer (Cellular Technology, Cleveland, OH). The numbers of IFN-γ or IL-4 secreting cells were expressed as the number cells per 10⁶ cells.

GraphPad Prism software was used to analyze the data (GraphPad Software Inc., San Diego, CA). All data were expressed as the mean and standard error. The data were also analyzed by Mann-Whitney rank sum test to compare data distribution between any two experimental groups. Titer comparison between viruses in replication kinetic experiment was performed using the Kruskal-Wallis test, assuming non-parametric distribution.

The reverse genetics system for measles Edmonston-Zagreb (EZ) vaccine virus was established by transfecting

MV antigenomic plasmid pEZ together with pP, pN and pL into MVA-T7 infected HEp-2 cells to recover infectious EZ virus. To generate rEZ expressing RSV F or EBV gp350, the F ectodomain of RSV and EBV gp350 were cloned into the MV antigenomic cDNA (Fig. 1A). The F was cloned at the two positions of the viral genome: one proximal to the N gene (1^st position) and the other between the P and M genes (3^rd position). EBV gp350 ectodomain and its N-terminal half of the gp350 (amino acid 1-470, gp350tr) that encoded the receptor binding site to CD21 [39Tanner J, Whang Y, Sample J, Sears A, Kieff E. Soluble gp350/220 and deletion mutant glycoproteins block Epstein-Barr virus adsorption to lymphocytes J Virol 1988; 62(12 ): 4452-64.] were also cloned into the EZ genome at the 1^st and 3^rd positions. Recombinant EZ with RSV F, EBV gp350 and EBV gp350tr inserted at the 1^st or 3^rd position were generated. However, viruses with gp350 and gp350tr inserted at the 1^st position replicated poorly in vitro and were not evaluated further. The rescued recombinant viruses expressing RSV F or EBV gp350 or gp350tr were designed as sF1 (F at the 1^st position), sF3 (F at the 3^rd position), gp350 and gp350tr (both at the 3^rd position), respectively.

To confirm that these recombinant measles viruses expressed the desired antigens, the expression of the inserts in infected Vero cells was examined by Western blotting analysis. In addition to MV N protein that was detected in infected cell lysate (Fig. 1B), RSV F and EBV gp350 were detected in the culture supernatant of the infected cells (Fig. 1B), indicating the production of secreted form of RSV F and EBVgp350 proteins from virus infected cells. The observation of lower MV N expression in Vero cells infected with sF1 compared to other viruses may be attributed to the insertion of RSV F proximal to the MV N gene, which moved the MV N gene to the 2^nd position of the genome and reduced its mRNA transcript level. The relative expression of RSV F to MV N, however, was shown to be higher for sF1 compared to sF3. This is consistent with the observation that genes proximal to 3’ end of the genome are transcribed more abundantly than those downstream and more RSV F was produced from sF1 than sF3 per virion.

Viral replication efficiency of these recombinant viruses was assessed by growth kinetics in Vero cells. Both recombinant EZ (rEZ) and EZ vaccine viruses grew to a peak titer of 7.9 log₁₀ PFU/mL on day 5 (Fig. 3A). rEZ with either RSVF (Fig. 2A) or EBVgp350 (Fig. 2B) inserts had comparable growth kinetics but slightly lower (between 0.1-0.7 log₁₀ lower) than rEZ. The peak titers between the viruses were not statistically different. sF3 had a faster growth kinetics and reached peak titer on day 3 instead of day 5. Therefore, insertion of RSV F or EBV gp350 into the genome of EZ did not significantly reduce viral growth in vitro.

Cotton rats had previously been demonstrated to be semi-permissive to measles virus infection [40Wyde PR, Ambrose MW, Voss TG, Meyer HL, Gilbert BE. Measles virus replication in lungs of hispid cotton rats after intranasal inoculation Proc Soc Exp Biol Med 1992; 201(1 ): 80-7.] and are the best available small animal model for MV infection. In addition, cotton rats are susceptible to RSV infection and have been used in the evaluation of RSV vaccines [41Jin H, Zhou H, Cheng X, Tang R, Munoz M, Nguyen N. Recombinant respiratory syncytial viruses with deletions in the NS1 NS2 SH and M2-2 genes are attenuated in vitro and in vivo. Virology 2000; 273(1 ): 210-8., 42Murphy BR, Sotnikov AV, Lawrence LA, Banks SM, Prince GA. Enhanced pulmonary histopathology is observed in cotton rats immunized with formalin-inactivated respiratory syncytial virus (RSV) or purified F glycoprotein and challenged with RSV 3-6 months after immunization Vaccine 1990; 8(5 ): 497-502.] and anti-RSV antibody investigated for prophylactic application [43Wu H, Pfarr DS, Johnson S, et al. Development of motavizumab an ultra-potent antibody for the prevention of respiratory syncytial virus infection in the upper and lower respiratory tract J Mol Biol 2007; 368(3 ): 652-.]. The ability of rEZ vectored vaccines to induce humoral responses to measles virus and RSV F or EBV gp350 protein were assessed in cotton rats. Cotton rats were immunized

with two doses of rEZ viruses intramuscularly at 4 weeks interval. Sera were collected 4 weeks post dose 1 and 2 weeks post dose 2. Measles virus neutralizing antibodies (MV Ab) were induced rapidly after one dose administration, reaching titers of 6.3-8.7 log₂. No significant differences in MV Ab titers were observed among the animals receiving sF1, sF3, gp350 or gp350tr compared to rEZ. The 2^nd dose increased MV Ab titers by 2-4 folds (Table 1). The control animals vaccinated with UV-inactivated sF3 or RSV A2 virus did not have detectable MV Ab titer (data not shown).

RSV-specific antibody responses were measured by RSV F-specific ELISA and microneutralization assay. sF1 or sF3 induced RSV F-specific IgG antibody in the vaccinated cotton rats (about 15.5-15.9 log₂ post dose 1 and 22.1-23.7 log₂ post dose 2). The RSV F-specific IgG antibody titers were similar between the sF1 and sF3 vaccinated groups. RSV neutralizing antibody titers were also measured against RSV A2 strain. After the first dose, sF1 and sF3 induced RSV neutralizing Ab and titers were boosted to 7.5-log₂ and 8.9-log₂ after the second dose (Table 1). There was no statistical difference in antibody response between sF1 and sF3 vaccinated animals and those intranasally infected with 2 doses of RSV A2 strain (9.4 log₂ ± 1.5, data not shown). Animals vaccinated with UV-inactivated sF3 did not have detectable RSV antibody (data not shown), indicating that virus immunogenicity was dependent on live virus infection.

EBV-specific antibody response was measured by EBV gp350-specific ELISA assay. EBV specific ELISA Ab response was detected after one dose administrations of gp350 or gp350tr; after the 2^nd dose the titers reached 7.3 and 4.3 log₂ for gp350 and gp350tr vaccinated animals, respectively (Table 1). Gp350 induced a significantly higher EBV gp350-specific IgG titer than gp350tr. EBV specific neutralizing antibody could not be detected by the EBV neutralization assay.

Cell mediated immune response is important for viral clearance and may play an important role in vaccine mediated protective immune response [44Munoz JL, McCarthy CA, Clark ME, Hall CB. Respiratory syncytial virus infection in C57BL/6 mice: clearance of virus from the lungs with virus-specific cytotoxic T cells J Virol 1991; 65(8 ): 4494-7.]. RSV F-specific and EBV gp350-specific T cell responses in splenocytes were examined by IFN-γ and IL-4 ELISpot on day 7 post the 2^nd dose. The frequencies of RSV F-specific IFN-γ secreting cells were statistically higher in sF3 vaccinated group (average 130 spots/10⁶ splenocytes) compared to sF1 vaccinated (average 9 spots/10⁶ splenocytes) or RSV A2-infected animals (average 45 spots/10⁶ splenocytes) (Fig. 3A). The RSV F-specific IL-4 secreting cells were low, averaged about 20 spots/10⁶ splenocytes for both sF1 and sF3 groups. Surprisingly, no RSV-specific IL-4 secreting cells detected in RSV A2-infected animals (Fig. 3C). Thus, sF3 and RSV induced a Th1-biased response based on the higher IFN-γ to IL-4 secreting cell ratio.

The frequencies of EBV gp350-specific IFN-γ and IL-4 secreting cells for gp350 (average 23 spots and 15 spots/10⁶ splenocytes for IFN-γ and IL-4 respectively) and gp350tr vaccinated groups (average 49 spots and 41 spots/10⁶ splenocytes for IFN-γ and IL-4 respectively) were not statistically different, but were statistically higher than the PBS control group for the gp350tr group (Fig. 3B, D).

Animal model is not available for assessing EBV vaccine induced protective immune response. Cotton rat is the best available small animal model to evaluate RSV vaccine induced protection against wt RSV challenge infection. To assess if the immune responses induced from sF1 and sF3 could provide protection against RSV A2 challenge, vaccinated cotton rats were challenged with RSV A2 intranasally and virus titers in the lungs were measured. Similar to RSV infected animals, sF1 or sF3 offered complete protection against RSV A2 challenge with viral titers in the lungs that was at least 4.5 log₁₀ lower than the control animals. In contrast, cotton rats vaccinated with 10⁵ PFU of UV-inactivated sF3 had viral titer similar to the PBS control group, indicating that viral replication is necessary for induction of protective immune response (Fig. 4).

Rhesus macaques are fully permissive to measles virus infection and have been used for evaluation of measles vaccine and measles virus pathogenesis studies [45de Vries RD, Lemon K, Ludlow M, et al. In vivo tropism of attenuated and pathogenic measles virus expressing green fluorescent protein in macaques J Virol 84(9 ): 4714-24.-47Zhu YD, Heath J, Collins J, et al. Experimental measles.II. Infection and immunity in the rhesus macaque Virology 1997; 233(1 ): 85-92.]. The ability of rEZ vectored RSV F and EBV gp350 to induce measles virus specific or F or gp350 specific immune responses was examined in this more permissive animal model. Since sF3 demonstrated better immunogenicity than sF1 in cotton rats, only sF3 was evaluated in measles virus seropositive (MV+) rhesus macaques that had prior measles vaccination and measles virus seronegative (MV-) rhesus macaques. No abnormal clinical signs were observed and serum chemistries were within the normal range in all monkeys throughout the study (data not shown). sF3 or gp350 vaccinated monkeys induced a similar level of MV neutralizing antibody in both MV- and MV+ monkeys. A second dose of vaccination with sF3 or gp350 increased MV neutralizing Ab titers by about 4-fold in all groups (Table 2). The Ab titer differences between the groups were not statistically significant.

Antibody responses to the inserted genes, however, were relatively low. sF3 inoculated animals had RSV F-specific IgG titer higher than the pre-vaccination level (titers were under the detection limit). F specific IgG Ab titers were 9.3 log₂ and 8.8 log₂ in MV+ and MV- groups after dose 1, respectively. After a second dose, three out of 4 animals in MV+ group had more than 4-fold increase in RSV F-specific IgG Ab titer but none of the animals in MV- group had 4-fold increase in Ab response (Fig. 5A). After two doses of vaccination, RSV neutralizing Ab titers in all groups were very low in both MV+ and MV- groups. Seven out of 8 animals had a detectable level of RSV neutralizing Ab. The RSV Nt Ab titers were not statistically different between the two groups (5.0 log₂ for MV+ versus 4.5 log₂ for MV- groups) (Table 2). Ab responses to EBV gp350 in the gp350 vaccinated monkeys were not detected at all time points. Similar to antibody responses, IFN-γ secreting cells in PBMC to measles viral antigens were significantly higher (~10-fold) than the responses to either RSV F or EBV gp350 (Fig. 5B) while PBMC, when stimulated with measles viral antigens, RSV F or EBV gp350, did not secrete any IL-4. Thus, compared to the cotton rat studies, rEZ vectored viruses were poorly immunogenic in the primate model.