and Modified Vaccinia Virus Ankara-Human - Journal of Virology

JOURNAL OF VIROLOGY, July 2007, p. 7022–7033 0022-538X/07/$08.00⫹0 doi:10.1128/JVI.02654-06 Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vol. 81, No. 13

Direct Comparison of Antigen Production and Induction of Apoptosis by Canarypox Virus- and Modified Vaccinia Virus Ankara-Human Immunodeficiency Virus Vaccine Vectors䌤 Xiugen Zhang,1 Farah Cassis-Ghavami,2 Mike Eller,3 Jeff Currier,3 Bonnie M. Slike,3 Xuemin Chen,1 James Tartaglia,4 Mary Marovich,3 and Paul Spearman1* Departments of Pediatrics and Microbiology and Immunology, Emory University School of Medicine, Atlanta, Georgia1; Department of Pediatrics, Vanderbilt University, Nashville, Tennessee2; U.S. Military HIV Research Program, Rockville, Maryland3; and Sanofi Pasteur, Toronto, Ontario, Canada4 Received 30 November 2006/Accepted 29 March 2007

Recombinant poxvirus vectors are undergoing intensive evaluation as vaccine candidates for a variety of infectious pathogens. Avipoxviruses, such as canarypox virus, are replication deficient in mammalian cells by virtue of a poorly understood species-specific restriction. Highly attenuated vaccinia virus strains such as modified vaccinia virus Ankara (MVA) are similarly unable to complete replication in most mammalian cells but have an abortive-late phenotype, in that the block to replication occurs post-virus-specific DNA replication. In this study, an identical expression cassette for human immunodeficiency virus gag, pro, and env coding sequences was placed in canarypox virus and MVA vector backbones in order to directly compare vector-borne expression and to analyze differences in vector-host cell interactions. Antigen production by recombinant MVA was shown to be greater than that from recombinant canarypox virus in the mammalian cell lines and in the primary human cells tested. This observation was primarily due to a longer duration of antigen production in recombinant MVA-infected cells. Apoptosis induction was found to be more profound with the empty canarypox virus vector than with MVA. Remarkably, however, the inclusion of a gag/pro/env expression cassette altered the kinetics of apoptosis induction in recombinant MVA-infected cells to levels equal to those found in canarypox virus-infected cells. Antigen production by MVA was noted to be greater in human dendritic cells and resulted in enhanced T-cell stimulation in an in vitro antigen presentation assay. These results reveal differences in poxvirus vector-host cell interactions that should be relevant to their use as immunization vehicles.

based vectors under evaluation as HIV candidate vaccines. Because of potential safety concerns with wild-type vaccinia virus strains, highly attenuated vaccinia virus strains are actively being pursued as HIV vaccine candidates. NYVAC is a highly attenuated vaccinia virus strain that was derived by the precise deletion of 18 open reading frames from the Copenhagen vaccine strain of vaccinia virus (47). The attenuation and immunogenicity of NYVAC have been demonstrated in several animal models (26, 47), and the block to replication appears to be less complete than that for canarypox virus in mammalian cells (47). Modified vaccinia virus Ankara (MVA) was generated by passaging the Ankara strain of vaccinia virus more than 500 times on chicken embryo fibroblasts (CEFs) (36). MVA replicates permissively in CEF and Syrian hamster BHK cells but does not complete its replication cycle in other mammalian cell lines or in human primary cells (9). In contrast to the avipoxvirus vectors, both NYVAC and MVA demonstrate an abortive-late phenotype in most human cell lines (45, 47). Thus, although both avipoxvirus-based vectors and highly attenuated vaccinia virus vectors are replication deficient in human cells, they are blocked at distinct stages of replication. Canarypox virus vectors have been tested extensively as HIV vaccine candidates in human trials (5, 7, 13, 16, 17, 23, 34). The safety profiles of these vectors have in general been excellent (15), and HIV-specific immune responses have been elicited (7). The magnitudes and persistence of these responses have

The search for a safe and effective human immunodeficiency virus (HIV) vaccine has stimulated the development of recombinant live vectors as vehicles for the induction of specific cellular immune responses. Attenuated poxvirus vectors have a number of desirable features as HIV vaccine candidates, including promising safety profiles and the ability to incorporate substantial genetic material for the expression of foreign gene products (35, 38, 46, 48). The leading poxvirus vectors are derived from two genera within the Poxviridae family. Canarypox virus and Fowlpox virus vectors belong to the Avipoxvirus genus. Avipoxviruses are naturally restricted to replication in avian species. In human cell lines, avipoxviruses fail to replicate their DNA genomes (47). The precise mechanisms underlying this host cell restriction have not been elucidated. Despite a complete block to replication, avipoxvirus recombinants express appropriately engineered foreign gene products in human cells and can elicit specific cellular and humoral immune responses to inserted genes in animals and in human volunteers (22, 49). Vectors derived from Vaccinia virus, a member of the Orthopoxvirus genus, constitute another leading group of poxvirus* Corresponding author. Mailing address: Departments of Pediatrics and Microbiology and Immunology, Emory University School of Medicine, Atlanta, GA 30322. Phone: (404) 727-5642. Fax: (404) 7278249. E-mail: paul.spearman@emory.edu. 䌤 Published ahead of print on 4 April 2007. 7022

VOL. 81, 2007

COMPARISON OF POXVIRUS-BASED HIV VACCINE VECTORS

been modest, however, leading the NIH National Institute of Allergy and Infectious Diseases-sponsored HIV Vaccine Clinical Trials Network to forego plans for phase III testing of a canarypox virus-based HIV vaccine regimen (44). Nevertheless, a phase III trial is presently under way in Thailand to test the efficacy of a canarypox virus-based HIV vaccine candidate employed as a priming agent in combination with a bivalent recombinant subunit gp120 protein booster (39). Vectors based on the avipoxvirus fowlpoxvirus are also under evaluation as HIV vaccines for human use (42). MVA vectors have generated very promising results in simian models of HIV infection (1, 4, 25), and on this basis, they are also currently undergoing evaluation in human trials. The promise of this approach is highlighted by the ability of a DNA prime/MVA boost regimen to protect macaques from simian-human immunodeficiency virus 89.6P challenge (1), and this regimen is now the model for a human trial sponsored by the National Institutes of Health-sponsored HIV Vaccine Trials Network. MVA-based vaccination regimens are also safe, as reported in a phase I human clinical trial (10). Despite intense interest in poxvirus vectors as candidate vaccines for HIV and other infectious diseases, few direct comparisons of poxvirus-based vaccine candidates have been reported to date. A recent report compared NYVAC and MVA vectors in cell culture systems. Significant differences were discovered in vector-generated cytopathic effect, growth characteristics, phosphorylation of the translational initiation factor eIF2-␣, stage of replication block, and the degree of apoptosis induced by these vaccinia virus-derived vectors (37). These findings can now form the basis for a detailed understanding of the vector-specific factors that may affect the immunogenicity of vectors employed in vaccine trials. The finding of a number of significant differences between the closely related MVA and NYVAC vectors makes it even more likely that there will be distinct differences in antigen production and host interactions seen between the more distantly related orthopoxvirus vectors and avipoxvirus vectors. Direct comparisons of vector-elicited antigen expression and host cell effects generated by avipoxvirus vectors and attenuated vaccinia virus vectors are warranted in order to choose the optimal vector to elicit a desired response and to design improved future generations of vaccines. In this study, we performed a head-to-head comparison of a canarypox virus vector and an MVA vector bearing an identical HIV gene expression cassette. We found significant differences in the timing and magnitudes of antigen expression exhibited by canarypox virus and MVA vectors. While apoptosis was lower, as expected, with the naked MVA vector, the inclusion of a gag/pro/env expression cassette had a major influence on apoptosis mediated by either recombinant vector. Recombinant MVA generated more antigen in most human cell lines and in primary cells, including human dendritic cells (DC), than did the canarypox virus recombinant and resulted in enhanced cellular responses in an in vitro antigen presentation system. MATERIALS AND METHODS Construction of recombinant viruses. The construction of the recombinant canarypox virus vector vCP205 has been described previously (19). This vector expresses the HIV type 1 (HIV-1) LAI gag and pro open reading frames under the control of the vaccinia virus I3L promoter. The expression cassette includes

7023

a modified HIV-1 MN env gene corresponding to the gp120 coding region linked to the gp41 transmembrane anchor sequence from HIV-1 LAI (28 amino acids) and under the control of the vaccinia virus H6 promoter. The gag/pro and env expression cassettes were placed between canarypox virus flanking regions in shuttle plasmid pHIV32, and homologous recombination was used to generate the recombinant canarypox virus vCP205. The recombinant MVA virus MVA205 was generated by inserting the identical expression cassette used for creating vCP205 into the deletion 3 locus of the MVA genome. To do this, the entire expression cassette was removed from the vCP205 transfer vector pHIV32 by restriction digestion with BamHI and XbaI and was then inserted by blunt-end cloning into the MVA transfer vector pMC03 (generously provided by Linda Wyatt and Bernard Moss, NIH), between SbfI and AscI sites. The resulting transfer construct lacked the synthetic vaccinia virus early/late promoter of pMC03 but retained the selection marker beta-glucuronidase (8). After six rounds of plaque purification and selection using staining with X-Gluc (5-bromo4-chloro-3-indolyl-␤-D-glucuronic acid), a master stock of MVA205 was generated. One experiment in this paper describes the use of vCP1452, a vector that includes the vaccinia virus E3L and K3L genes, whose construction was described in detail previously (20). MVA-GFP was generously provided by Mark Feinberg (11). vCP-GFP is a recombinant canarypox virus vector, previously termed vCP1540 and developed by Sanofi Pasteur, expressing the green fluorescent protein (GFP) gene under the control of the H6 promoter. MVA/T7 is a recombinant vector expressing T7 polymerase (50) and was provided by Linda Wyatt and Bernard Moss. Virus propagation and plaque titration. CEFs were prepared from 10- to 11-day-old chicken embryos and were maintained in minimum essential medium (MEM; Gibco, Gaithersburg, MD) with 10% fetal calf serum (FCS), 100 U/ml penicillin, and 100 g/ml streptomycin. To generate MVA205 and vCP205 stocks, 150-cm2 flasks were seeded from the master stocks at a multiplicity of infection (MOI) of 0.1. The infected cells were incubated at 37°C for 3 to 5 days until maximum cytopathic effect was evident. Working virus stocks were generated by freeze-thawing and sonication of the infected CEF cell pellets three times, followed by removal of cell debris by centrifugation at 1,000 ⫻ g for 10 min at 4°C. Stocks were plaque titrated by preparing serial 10-fold dilutions of stock virus in MEM with 2% FCS. Diluted virus was placed on CEFs in 60-mm2 dishes for 1 h and rocked three or four times. The inoculum was then removed and the dishes overlaid with 2% MEM containing 1% agarose. vCP205 plaques were counted after being stained with 0.1% neutral red, and the plaque titers of stocks were recorded; MVA205 titers were tested by X-Gluc staining. The resulting virus stocks were stored at ⫺80°C until use. Gag and Env expression by MVA205 and vCP205 stocks was confirmed by Western blotting using pooled sera from HIV-positive individuals for detection, and the presence of intact gene inserts was confirmed by sequencing the viral DNA extracted from cells infected with the master stock. Quantitation of antigen expression. Gag-Env expression was measured as previously described (19, 20). Cells were plated on the night prior to infection, and infections were performed when the cells were noted to be 80% confluent on the dish. An accurate count of the number of cells present per dish was achieved by counting the cells from one 10-cm2 dish at the time of infection (about 1.2 ⫻ 107 CEF cells or 0.8 ⫻ 107 HeLa cells at 80% confluence). Cells were infected with MVA205 and vCP205 at an MOI of 10 in 2 ml of MEM with 2% FCS. After 1 hour of infection with intermittent gentle agitation, the medium was removed, and 10 ml of fresh 2% MEM was added to each dish. Medium and/or cells were harvested at the time points outlined in Results for detection of p24 antigen or for other assays. Samples for protein analysis were prepared as follows. Supernatants were filtered through a 0.45-␮m filter and pelleted by centrifugation through a 5-ml 20% sucrose cushion in a Beckman SW28 rotor at 100,000 ⫻ g for 3 h. The pellets were then resuspended in phosphate-buffered saline (PBS). Analysis of cellular protein expression was performed by lysing cells in PBS with 1% NP-40; nuclei were separated via centrifugation at 10,000 ⫻ g for 10 min. Gag and Env protein contents were analyzed by Western blotting using sera pooled from HIV-positive donors. Cell lines and primary cells. In addition to the CEFs described above, a number of cell lines were employed in this study. These included murine C2C12 myoblasts (ATCC CRL-1772), murine 3T3 fibroblasts (ATCC CRL-1658), African green monkey kidney BSC-40 and Vero cells (ATCC CCL-81), human rhabdomyosarcoma TE671 cells (ATCC CRL 8805), human melanoma MelJuSo cells, human embryonic kidney 293 cells (ATCC CRL 1573), human cervical carcinoma HeLa cells (ATCC CCL-2), human U937 macrophages (ATCC CRL1593.2), human Jurkat T cells, and human kidney epithelial MRC-5 cells (ATCC CCL-171). All cell lines were obtained from the American Type Culture Collection, with the exception of BSC-40 (provided by Bernard Moss, NIH), Jurkat (obtained from Barney Graham, NIH), and MelJuSo (obtained from Markus

7024

ZHANG ET AL.

Thali, University of Vermont). Adherent cell lines were maintained in Dulbecco’s modified Eagle medium with 10% FCS; suspension cells were maintained in RPMI 1640 with 10% FCS. Primary human peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque separation and were maintained in RPMI 1640 medium with 10% FCS, 2 mM L-glutamine, penicillin G (100 U/ml), and streptomycin (100 ␮g/ml). After poxvirus infection, PBMCs were cultured in 2% RPMI 1640 medium. Human monocytes were isolated from the PBMC fraction by negative depletion with a monocyte isolation kit (Miltenyi Biotec, Auburn, CA). Isolated monocytes (CD45⫹ CD14⫹) were ⬎95% pure, as indicated by flow cytometry. Monocytes were cultured in RPMI with 10% FCS and 2% granulocyte-macrophage colony-stimulating factor. After 4 days of maturation in culture, monocytes were infected with MVA205 or vCP205, using the same methods as those used for the adherent cell infection process. Virus infection efficiency assay. Determination of the efficiency of infection was performed using viruses expressing GFP, namely, MVA-GFP (11) and vCPGFP (expressing the GFP gene under control of the vaccinia virus H6 promoter; originally called vCP1405). For this analysis, cells were infected at an MOI of 10 PFU/cell, as measured by CEF plaque titration. Cells were harvested at 24 h postinfection, washed three times with cold PBS containing 0.1% bovine serum albumin, fixed with 2% formaldehyde, and analyzed using a BD LSR II flow cytometer. Gates were determined using an identical population of uninfected cells. Measurement of virus-induced apoptosis. HeLa cells were infected with MVA205, vCP205, or vCP1452 at an MOI of 10 PFU/cell. Cells were harvested at 12 and 24 h postinfection, and DNA fragmentation was analyzed with an Apoptotic DNA Ladder kit (Roche Applied Science, Indianapolis, IN), using the manufacturer’s instructions. A second method used a cell death detection enzyme-linked immunosorbent assay (ELISA; Roche) according to the manufacturer’s instructions. Flow cytometry was used to measure apoptosis, using annexin V and propidium iodide (PI) staining. Cells were detached from plates with Accutase (eBioscience, San Diego, CA) at 12 or 24 h postinfection with vCP205 or MVA205 or at the same time following treatment with camptothecin. Cells were washed in cold PBS, stained with annexin V-fluorescein isothiocyanate or annexin V-allophycocyanin and PI, and analyzed on a BD LSR II flow cytometer or a BD FACSCanto flow cytometer. Real-time PCR. To examine the quantity of viral DNA present in cells infected with vCP205 or MVA205, HeLa cells were infected at an MOI of 10 PFU/cell in six-well plates. The cells were harvested at 0, 2, 4, 8, 12, 24, 48, and 72 h postinfection. Each experiment was performed in triplicate. Cells were washed once with cold PBS, and DNAs were extracted using a QIAamp DNA Mini kit (QIAGEN, Valencia, CA). The following real-time PCR primers located within the HIV gag region were used: 2463-2486, AGAGAAGGCTTTCAGCCCAGA AGT; and 2639-2616, TGCACTGAATGCACTCTATCCCAT. These primers amplified a 176-bp fragment. Quantitative real-time PCR was performed in a DNA Engine Opticon 2 system (Bio-Rad Laboratories, Hercules, CA). Serial dilutions of 20 ng/ml of pHIV32 plasmid at 1:10 dilution increments were used to generate standard curves at the same time in the same plate. Antigen presentation in human myeloid DC. Myeloid DC were derived from leukapheresis blood products from healthy HIV-seropositive donors, obtained under an Internal Review Board-approved protocol (RV149) and prepared as previously described (33). The isolated peripheral blood monocytes were treated with recombinant human interleukin-4 (R&D Systems, Minneapolis, MN) and recombinant human granulocyte-macrophage colony-stimulating factor (Fisher Clinical Services, Rockville, MD) to drive DC differentiation. The DC were exposed to the vectors at various MOI within a working range of vector/DC ratios of 1 to 5 PFU/cell. At 2 h postinfection, the DC were washed and then resuspended in medium to a concentration of 106/ml. Samples from each experiment were set aside for a functional DC enzyme-linked immunospot (DC-ELISPOT) assay, and the remaining DC were incubated at 37°C for 2 to 6 h for later expression analysis. Infection rates were determined using labeling with antivector antibodies (rabbit anti-canarypox virus [Sanofi Pasteur]) and an anti-vaccinia virus monoclonal antibody (MAb; clone TW2.3) (29). HIV transgene expression was measured by intracellular labeling with rabbit anti-Env (LS086-4-D51; a generous gift from Florence Boudet at Sanofi Pasteur) and an anti-p24 MAb (Dako, Glostrup, Denmark) and then was analyzed by flow cytometry (33). Functional effects of canarypox virus and MVA infection and HIV gene expression were assessed after vector loading of the DC in a DC antigen presentation assay. The DC-ELISPOT assay uses autologous DC and responder PBMCs from healthy HIV-seropositive donors to detect antigen-specific gamma interferon (IFN-␥), as previously described (15). Briefly, 96-well ELISPOT plates were precoated with anti-IFN-␥ MAb (Mabtech AB, Sweden) and then loaded with cells. The cells included infected (or mock-infected) DC distributed in triplicate wells, with or without PBMCs at a 1:10 ratio of DC to PBMCs. In some exper-

J. VIROL.

FIG. 1. HIV protein expression by vCP205 and MVA205. (A) Western blot of HIV antigen expression in infected CEFs. Cells were harvested 24 h following infection with test viruses. Cells were harvested (C) and supernatants pelleted (S) for detection of virus-like particle production. The blot was probed using HIV-positive patient sera. The positions of molecular mass markers are shown on the left. (B) HeLa cells were infected and treated as described for panel A.

iments, the DC were titrated and used across a 3-log dilution to verify dose effects (1:10, 1:100, and 1:1,000 ratios of DC to responders). The plates were incubated for 20 to 24 h at 37°C, washed, incubated with a biotinylated antiIFN-␥ MAb (Mabtech), washed again, and then developed with a horseradish peroxidase system (Vectastain ABC kit; Vector Labs, CA). The data were collected and analyzed using simple statistics (linear regression and paired Student’s t test).

RESULTS Expression of HIV proteins by MVA and canarypox virus recombinant vectors bearing identical HIV gene expression cassettes. In order to define vector-specific effects on antigen production and on host cells, we constructed an MVA recombinant virus that incorporates an HIV expression cassette identical to that in vCP205. vCP205 is a gag/pro/env HIV vaccine candidate vector that has been evaluated extensively in the laboratory and in human trials (6, 12, 24). This vector incorporates the HIV-1LAI gag and protease coding sequences under the control of the vaccinia virus I3L promoter and a gp120 gene from HIV-1 MN fused to the coding sequence for the transmembrane domain of gp41 from HIV-1LAI under the control of the vaccinia virus H6 promoter. The detailed construction of this vector has been described elsewhere (19, 20). The gag/pro and env expression cassette from vCP205 was placed within the deletion 3 region of MVA, using homologous recombination and marker selection. The resulting virus, MVA205, contains the identical HIV gene segments under the same poxvirus promoters as those in vCP205. MVA205 was

VOL. 81, 2007


7025

FIG. 2. Gag-Env pseudovirion production and release by recombinant poxviruses in mammalian adherent cell lines. Each cell line shown was infected at an MOI of 10 with vCP205 or MVA205. Cellular supernatants were harvested at multiple time points following infection, and p24 antigen was measured by an antigen capture ELISA. p24 antigen production by MVA205 is indicated by black squares, and antigen production by vCP205 is shown by open circles. Error bars represent standard deviations for triplicate wells.

plaque purified and grown to a high titer (1 ⫻ 108 PFU/ml). HIV protein expression was verified in CEFs by Western blotting (Fig. 1A). Both constructs produced roughly equivalent amounts of Gag protein in CEFs, although there was more intracellular cleavage of Pr55Gag in cells infected with vCP205. Gag-Env pseudovirion particles were released from CEFs infected with MVA205 and vCP205 (Fig. 1A, supernatant lanes). Unexpectedly, the amount of gp120 produced by vCP205 in CEFs was lower than that produced by MVA205. Protein expression and particle release by HeLa cells appeared to be basically equivalent for the two vectors by Western blotting (Fig. 1B). The magnitudes and patterns of Gag and Env protein expression in these cells appeared to be identical, validating the construction of heterologous vectors bearing the same gene expression cassette. In order to compare more precisely the temporal patterns and magnitudes of antigen expression by MVA205 and vCP205, we measured p24 antigen release in cellular supernatants over time. The full-length gag gene expressed by both vectors allows efficient pseudovirion particle formation and

release, and the levels and timing of supernatant antigen reflect cellular production (19). In CEFs, the patterns and magnitudes of p24 production were nearly identical (Fig. 2, CEF panel). However, substantial differences were observed in a series of mammalian cell lines. Production and release of antigen by MVA205 were more efficient than those by vCP205 in mammalian cells, with the exception of 293 cells (Fig. 2). The peak of antigen production by the MVA vector was delayed compared with that for canarypox virus in each mammalian cell line tested. This is best illustrated with the murine myoblast cell line C2C12, where peak particle production was seen at 24 h with vCP205 and at 48 h with MVA205 (Fig. 2, C2C12 panel). These data indicate that both the magnitude and temporal pattern of protein expression differ for MVA and canarypox virus, despite the use of identical promoter-gene expression cassettes. Expression of antigen by a vaccine vector in immortalized epithelial or muscle cell lines may not reflect that in more relevant cells. To address this issue, we extended this analysis to human T-cell and monocytic cell lines and to primary

7026

ZHANG ET AL.

J. VIROL.

FIG. 3. Gag-Env pseudovirion production and release from suspension cells and primary cells. Suspension cells were infected at an MOI of 10 with vCP205 or MVA205, and supernatants were collected at the indicated times for p24 antigen measurement. Note that the y-axis scale is different from that in Fig. 2. p24 production by MVA205 is shown by black squares, and vCP205 production is indicated by open circles. Error bars represent standard deviations for triplicate wells.

human cells. The production of antigen was significantly lower in these cells than in adherent mammalian cell lines. However, the magnitudes of antigen expression and release by MVA205 were higher than those by vCP205 in Jurkat T cells, the monocyte-derived U937 cell line, and human lung MRC-5 cells (Fig. 3, top panels). Antigen production and release from PBMCs were low (2 to 3 ng/1 ⫻ 106 cells), but these cells also demonstrated greater antigen production by MVA205 than by vCP205. Human monocyte-derived macrophages demonstrated a marked difference in antigen production, and the later peak of antigen production in these cells by MVA205 was clearly demonstrated (Fig. 3, macrophage panel). Taken together, these data demonstrate that the higher level of antigen production and later peak of antigen production/release seen with MVA205 in cell lines are also seen in primary human cells. In order to be certain that the release of antigen in these assays was an accurate reflection of the magnitude and temporal sequence of cellular protein production and that the delayed peak of protein production by MVA represented ongoing protein production in infected cells, we quantified cellular and supernatant p24 production over time for two of the mammalian cell lines. In HeLa cells, cell-associated p24 production by vCP205 peaked by 18 to 24 h postinfection, while production by MVA205 was still rising at these times (Fig. 4, top panels). Intracellular levels of p24 generated by MVA205 peaked at 36 h postinfection, and antigen release continued to rise until 48 h postinfection. TE671 cells demonstrated the same earlier and lower peak of p24 expression for vCP205 (Fig. 4, bottom panels) and the same extended production for MVA205. Thus, there was a longer duration of protein expression, and the total amount of antigen produced was two- to threefold higher, in these cells when the same genes were expressed by MVA205 than the case with vCP205.

We considered the possibility that differences in the efficiencies of MVA versus canarypox virus in entering and productively infecting the cell lines and primary cells tested may have been responsible for some of the observed differences in the magnitudes of antigen expression. To test this, we utilized MVA and canarypox virus vectors expressing GFP and monitored the efficiency of infection by flow cytometry. Cells were infected at an MOI of 10 PFU/cell of either virus, and the percentage of cells expressing GFP was assessed at 24 h postinfection. Both canarypox virus-GFP and MVA-GFP were able to efficiently infect each adherent cell line (Fig. 5A to J). Although there were modest differences noted in the expression of GFP in CEF and 293 cells, where these differences were seen the higher levels were with canarypox virus and thus could not account for enhanced antigen production by MVA. Most cell lines and primary cell preparations were infected equally by either vector. Not surprisingly, the efficiencies of infection of suspension cells (Jurkat and U937 cells) and of primary cells (PBMCs and macrophages) were markedly lower (10 to 20% of cells infected) than those seen with adherent cell lines. Overall, the ability of both vectors to infect cells at equivalent or near-equivalent levels ruled out differential entry/infection as the reason for enhanced antigen expression by MVA. Differential apoptosis induction by canarypox virus and MVA vectors. Apoptosis induction by live vector vaccines may have important effects on the resulting immune response. While rapid apoptosis may limit antigen expression and thus adversely impact specific immunity to vector-encoded antigens, apoptotic responses may also be desirable for antigen uptake and presentation through the alternative class I presentation pathway (antigen cross-priming). We examined apoptosis by MVA205 and vCP205 by first using a traditional DNA fragmentation assay with vector-infected HeLa cells (Fig. 6A). At 12 h postinfection with either virus, little DNA fragmentation

VOL. 81, 2007


7027

FIG. 4. Cellular versus supernatant antigen expression. Cells were infected at an MOI of 10 with either virus and harvested at the indicated times for p24 antigen quantitation for cells (left) and the corresponding pelleted supernatants (right). Black squares indicate MVA205, and white circles indicate vCP205. Error bars represent standard deviations for triplicate wells. (Top) HeLa cells. (Bottom) TE671 (human rhabdomyosarcoma) cells.

was observed. However, at 24 h postinfection, there was marked apoptosis induction, appearing approximately equal for both viruses (Fig. 6A). We included a second recombinant canarypox virus, vCP1452, in this analysis as a comparator. This recombinant expresses the same gag, pro, and env genes as vCP205 and MVA205 but was also engineered to express the vaccinia virus E3L and K3L genes, resulting in diminished activation of the RNA-dependent protein kinase, enhanced antigen expression, and diminished apoptosis in mammalian cells (20). As noted previously, apoptosis induction was markedly less prominent with vCP1452. We next performed a more quantitative assessment of apoptosis induction by MVA205 versus vCP205 with an ELISA-based assay for detection of histone-complexed DNA fragments in the cytoplasm. At 12 h postinfection, MVA205 and vCP205 demonstrated comparable levels of apoptosis in this assay. By 24 h, both MVA205 and vCP205 induced significant apoptosis above the negative control level and above the levels produced by vCP1452. The total amount of apoptosis induced by vCP205 was greater at 24 h than that of MVA205 by this assay. We then sought to define the role of the vector itself versus the role of gag and env gene expression in apoptosis induction by these poxvirus vectors in mammalian cells. To do this, we used annexin V and PI staining of either empty vector or vector plus insert 12 and 24 h following infection. Annexin V staining detects phosphatidylserine translocation to the outer membrane, an early marker of apoptosis (31), while PI staining reveals later stages of apop-

tosis or necrosis. Therefore, in our flow cytometry analysis, the lower right quadrant (annexin staining) represents early apoptosis while the upper right quadrant (annexin plus PI) represents late apoptotic events, and total apoptosis is indicated by the total of these two quadrants. In mock-infected HeLa cells, low levels of early and late apoptosis were observed (Fig. 6C, negative control). Camptothecin treatment resulted in a significant increase in total apoptosis observed compared to that for negative control cells (23.2% versus 5.6%). We then examined MVA and canarypox virus vectors lacking any HIV inserts. MVA itself generated modestly increased apoptosis 12 and 24 h following infection compared with that of control cells (Fig. 6C). In sharp contrast, MVA205 generated markedly higher levels of apoptosis at both 12 and 24 h. At the 24-h time point, MVA205 generated 41.9% apoptotic cells, in contrast to 14.6% apoptotic cells for the MVA empty vector. Apoptosis by canarypox virus empty vector was higher than that by MVA at 24 h (23.4%) but demonstrated the same pattern of greater apoptosis when expressing the gag-env insert (39% apoptotic cells). A potentially important difference was noted between vCP205 and MVA205 at the earlier time point. By 12 h, a large percentage of vCP205-infected cells were already double positive, indicating that this vector induced very rapid apoptosis that had progressed to late stages by 12 h. We included vCP1452, designed to decrease apoptosis, as a further control. This vector induced less apoptosis at both 12 and 24 h than the empty canarypox virus vector, despite expressing HIV gag and

FIG. 5. Comparison of efficiencies of infection by canarypox virus and MVA. Each of the cell types used in this study was infected with vCP-GFP and MVA-GFP, and the percentages of cells expressing GFP were quantified by flow cytometry. Gray histograms represent uninfected cells, and white histograms represent green fluorescence 24 h following infection. Bars and corresponding numbers indicate percentages of the infected cell populations expressing GFP. Suspension cells with lower infection efficiencies are shown in panels K and L. Results are representative of experiments performed three times. 7028

VOL. 81, 2007


7029

FIG. 6. Measurement of virus-mediated apoptosis. Three methods were used to assess apoptosis in virus-infected HeLa cells. (A) DNA fragmentation was assessed in cells infected at the indicated times. The DNA marker was lambda DNA digested with HindIII and EcoRI. The positive control was a kit positive control from the manufacturer (Roche), and the negative control sample was from uninfected cells. (B) ELISA for detection of histone-bound DNA in cytoplasm. MVA205, vCP205, and vCP1452 were evaluated in HeLa cells for apoptosis at 12 and 24 h. Blank, no added DNA; Neg Ctr, DNA from uninfected cells; Camp, positive control from camptothecin-treated HeLa cells. (C) HeLa cells were infected with the indicated viruses and harvested at 12 or 24 h for annexin V and PI staining by flow cytometry. Gates were set on single-stained uninfected cell populations. Numbers represent percentages of cell populations in each quadrant.

env genes. Taking all of these apoptosis data together, we concluded that vCP205 and MVA205 both induced substantial levels of cellular apoptosis in mammalian cells and that the HIV gene insert had a surprising and robust effect on the levels of apoptosis observed. Furthermore, apoptosis induction by vCP205 was more profound at early time points and thus may have contributed to the limitation of late antigen production by this vector. To verify that the HIV gene insert was responsible for apoptosis induction by these recombinant poxvirus HIV vaccine vectors, we compared apoptosis induction by MVA205 or

vCP205 to that by recombinant vectors expressing other genes. MVA205 infection of HeLa cells led to substantially more apoptosis at 24 h than did infection with MVA expressing T7 polymerase or MVA expressing GFP (Fig. 7E versus C and F). The difference was most pronounced in the double-positive (PI plus annexin V) population. Similarly, vCP205 induced a much higher level of apoptosis that had progressed to late stages by 24 h than did empty vector or a canarypox virus expressing GFP (Fig. 7G to I). These data confirm that the HIV gene insert exerts substantial effects on apoptosis induction by vCP205 and MVA205.

7030

J. VIROL.

ZHANG ET AL.

FIG. 8. Quantitative real-time PCR for viral DNA. Quantitation of viral DNA was performed using primers within the HIV gag gene compared with a purified DNA standard. Time zero represents input viral DNA. Gray squares represent DNA copies in MVA205-infected cells, and white circles represent DNA copies in vCP205-infected cells.

FIG. 7. Effect of insert on virus-induced apoptosis. HeLa cells were infected with the indicated viruses and harvested at 12 or 24 h for annexin V and PI staining by flow cytometry. Gates were set on singlestained uninfected cell populations. (A) Negative control, (B) camptothecin, (C) MVA/T7, (D) MVA empty vector, (E) MVA205, (F) MVA-GFP, (G) canarypox virus empty vector, (H) vCP205, (I) vCP-GFP. Numbers represent percentages of the cell populations in each quadrant.

Quantitation of the differential block to viral DNA replication in HeLa cells. MVA is able to replicate its viral DNA in mammalian cells but is arrested at a late (assembly) stage, while avipoxviruses have been reported to be arrested at a stage prior to DNA replication (47, 49). However, detailed quantitation of avipoxviral DNA levels over time following infection of human cells has not been reported. We considered the possibility that the extended duration and magnitude of heterologous gene expression by MVA relative to those for canarypox virus may be linked with the ability to replicate the viral genome and proceed with late-stage transcription. To test this possibility, we developed a real-time PCR assay for measuring the number of copies of gag present in cells infected with MVA205 or vCP205. Serial dilutions of the transfer plasmid used to insert the gag gene into the poxviruses were used to generate a standard curve for absolute quantitation of DNA copies. Cells (1 ⫻ 106) were infected with 1 ⫻ 107 PFU of either virus, and the inoculum DNA was measured at time zero. After 1 h, the cells were washed, and subsequent PCR data were derived from DNAs extracted from the infected cells. For MVA205, the number of DNA copies of gag representing the poxvirus genome increased from 12 to 24 h, to a level approximately 50-fold higher than the input, and subsequently declined (Fig. 8, filled squares). The copy number for gag in the vCP205-infected cells did not increase after infection but declined rapidly between 12 and 24 h after infection. These data confirm that avipoxvirus vectors fail to replicate their DNAs in HeLa cells, while the DNA copy number in MVAinfected cells increases at least until 24 h postinfection. Although ongoing foreign antigen expression by MVA may not be linked directly to the kinetics of DNA replication, this

finding further highlights the difference in the stages of replication arrest of MVA and canarypox virus in mammalian cells. Differential antigen production in human DC by MVA and canarypox virus vectors. The most relevant cell type for studying vector-specific antigen expression is not known with certainty, as antigen cross-presentation may occur from uptake of dying cells or apoptotic bodies by professional antigen-presenting cells (APCs). However, the ability of vectors to infect and produce antigens directly in DC may be desirable. We therefore compared antigen production by vCP205 and MVA205 in myeloid DC derived from multiple human donors. In a series of direct comparison experiments (MOI ⫽ 5), the vCP205 and MVA205 constructs consistently infected DC at the same rates (59.0% versus 60.2%) (Fig. 9A, white bars). The vector-loaded DC viability for each construct was comparable and stable (range, 80 to 90%). However, expression of the HIV antigens Env and Gag was consistently and significantly higher for MVA205 than for vCP205 (Fig. 9A, filled bars) (Env expression, 49.9% ⫾ 7.7% versus 5.0% ⫾ 2.5% [P ⬍ 0.0001]; Gag expression, 62.8% ⫾ 13.2% versus 10.5% ⫾ 1.8% [P ⬍ 0.001]). The different expression levels translated into differences in IFN-␥ production, as assessed by ELISPOT assay. Figure 9B to D show IFN-␥ ELISPOT data for three representative donors. MVA205, whether loaded onto autologous DC or dropped into the well directly, resulted in more IFN-␥ production in each experiment. There was a strong positive correlation between IFN-␥ production in the antigen presentation assay and HIV Env expression (r ⫽ 0.78; P ⬍ 0.0001) or HIV Gag expression (r ⫽ 0.80; P ⬍ 0.0001), as determined by the Spearman correlation (data not shown). These results suggest that the ability of a vector to sustain antigen production and achieve higher levels of antigens in DC could have functional effects that may impact vector-induced immune responses. DISCUSSION Live viral vectors represent central components of many current HIV vaccine regimens. The poxvirus family includes a number of vectors that are under evaluation as HIV vaccine candidates. In order to understand how these vectors should best be used in vaccination regimens, it is essential to analyze

VOL. 81, 2007


7031

FIG. 9. Infection of human DC and in vitro antigen presentation. (A) Expression of HIV Env and Gag proteins in DC loaded with vCP205 (left) or MVA205 (right), as assessed by flow cytometry. The percentage of cells infected at an MOI of 5 is shown by detection of the respective vector. Data shown are the means for six experiments plus standard deviations. (B to D) IFN-␥ ELISPOT analysis of T-cell responses following coculture with DC infected with vCP205, MVA205, or a vector controls (CPpp and MVAp580, respectively). Results from experiments with blood cells from three representative donors are shown. DC were infected at an MOI of 5 unless noted otherwise.

both vector-specific influences on antigen expression and the effect of the vector on host cell responses, such as the induction of apoptosis. In this study, we analyzed the pattern and magnitude of antigen expression by an avipoxvirus vector in direct comparison with those of an attenuated orthopoxvirus vector. Our study was unique in that it utilized a common gag/pro/env gene expression cassette placed within these disparate vector backbones. This allowed us to make conclusions regarding the timing and magnitude of antigen production characteristic of each vector backbone and to compare the effects of each virus on cellular apoptosis when utilized as an HIV antigen expression vehicle. Antigen production by MVA in a variety of cell lines was found to be greater than that of canarypox virus. This result was consistently seen in the majority of human cell lines tested, including PBMCs, macrophages, and DC. Interestingly, a longer period of protein production seemed to account for most of the observed differences. Cellular production of antigen by canarypox virus diminished after 18 to 24 h, while production of antigen by MVA205 continued until 36 to 48 h postinfection. The difference in the magnitudes of protein production was unrelated to the ability to infect target cells, as both vectors infected a wide variety of cells at similar efficiencies. The differences in antigen production in our cell culture experiments were in the range of two- to fivefold. It will be interesting to determine if these magnitudes of difference in the expression of a common antigen result in important differences in immune responses in different animal species, especially humans. The results from infected DC in the in vitro model presented here suggest that this difference may indeed

result in significant benefits in antigen presentation. However, a number of factors in addition to antigen expression by the vector have the potential to significantly influence the resulting immune response in immunized animals or humans. Among these, apoptosis induction has been studied intensively and may be critical. Poxviruses have developed a number of means to prevent cellular apoptosis (18, 40). These include mechanisms to avoid cytotoxic T-lymphocyte (CTL)-mediated killing by down-regulation of major histocompatibility complex class I molecules, expression of viral proteins that interfere with IFN and tumor necrosis factor signaling pathways, and expression of molecules that directly interfere with caspase activation. Attenuated poxviruses such as MVA retain some antiapoptotic genes that are essential for their replication. MVA carries E3L, a doublestranded RNA binding protein that inhibits protein kinase R activation (14). Deletion of E3L generates an MVA that is highly apoptotic in human cells and that fails to replicate in CEFs (21, 28). Recently, the F1L gene product has also been shown to contribute to the ability of MVA to inhibit apoptotic responses (21). The abundance of antiapoptotic factors produced by poxviruses suggests that for viral replication to succeed, it is essential for the viruses to potently disable cellular apoptotic responses. In the context of a vaccine vector, however, complete avoidance of apoptosis may not be desirable. Apoptosis may enhance immune responses through the formation of apoptotic bodies and the uptake and presentation of antigen through the alternative class I antigen presentation pathway (43). Enhanced apoptosis induction by MVA compared to that by vaccinia virus in human DC has been proposed

7032

ZHANG ET AL.

as an explanation for its greater immunogenicity, presumably through enhanced antigen cross-presentation (11). Apoptosis induction is also likely to play an important role in the immunogenicity of avipoxvirus-based vectors. vCP1452 was designed to enhance the immunogenicity of canarypox virus ALVAC vectors through increased antigen expression and avoidance of cellular apoptosis (20). In cell culture systems, this vector successfully diminished apoptosis and resulted in greater levels of HIV antigen expression. Cellular immune responses elicited by vCP1452 in human trials were disappointing, however, and were substantially lower than those seen in earlier trials employing vCP205 (6, 23). While there are other plausible explanations for the apparently lower level of responses generated by vCP1452, it is possible that the reduced apoptosis generated in host cells infected by vCP1452 contributed to the reduction in immunogenicity seen in human trials with this vaccine candidate vector. Thus, effective antigen production in the host must be considered together with the ability to elicit apoptosis as a key factor contributing to vector-induced immunity. The fact that gag/pro and env gene expression dramatically increased vector-induced apoptosis in our study is a novel finding that should be considered in strategies to modify the immunogenicity of poxvirus-based vectors. Perhaps this effect should not be surprising, because numerous studies have documented the proapoptotic effects of HIV gp120 (2, 3, 27, 30, 32, 41). The effects of expression of the Gag protein on apoptosis are less certain, but we have observed apoptotic nuclear changes in mammalian cells upon expression of the Gag protein (unpublished observations). We found that the proapoptotic effects of gag and env gene expression in either vector were profound, making the more subtle differences seen in MVA and canarypox virus empty vectors less important. These findings highlight the important role that the genetic insert may play in altering apoptosis. For the design of the ideal level of apoptosis by a recombinant vaccine vector, both the vector backbone and the insert effects must be considered. We note that insertion sites differed between the two vectors and that indirect effects on neighboring genes could have contributed somewhat to our findings. An advantage of viral vectors is their ability to enter mammalian cells and endogenously produce relevant intracellular proteins for subsequent antigen presentation. Viral vector characteristics that allow efficient entry, early protein production, and sufficiently prolonged cell survival to allow antigen presentation might be important, especially in the context of direct infection of professional APCs. In this direct head-tohead comparison of two different viral backbones with identical gene inserts, we found that expression levels of HIV proteins were directly correlated to ELISPOT responses in an in vitro assay of antigen presentation. Despite similar infection rates, very different expression levels of relevant antigens were observed within myeloid DC. Therefore, simple “targeting” of APCs with viral vectors may be insufficient to prime the immune response. We propose that longer cell survival of DC infected with MVA but not canarypox virus allowed sustained transgene expression and successful antigen presentation and recognition by T cells to occur and that somewhat delayed apoptosis may have been beneficial. Whether DC serve as the primary APCs (direct priming) or serve as a protein/peptide source for indirect priming (cross-priming), the amount of

J. VIROL.

antigen/protein produced is likely to be important. Understanding and balancing the dynamics of vector insert expression and vector-induced apoptosis will be important to optimize future live vector HIV vaccine development. These cell culture-based results provide the basis for direct, head-to-head immunogenicity comparisons of avipoxvirus and MVA vectors bearing identical inserts in animal models. It will be interesting to determine whether the differences in antigen production and presentation documented in this study using cell culture models bear a direct relationship to the cellular immune responses in animal models and in humans. ACKNOWLEDGMENTS This work was supported by grants R01 AI52007 and R21AI65312 (P.S.). This work was supported in part by the cooperative agreement DAMD17-98-2-8007 between the U.S. Army Medical Research and Materiel Command, the Henry M. Jackson Foundation for the Advancement of Military Medicine, and the Military Infectious Disease Research Program. We are grateful to investigators at Aventis for reagents and advice and to Linda Wyatt and Bernard Moss for key reagents for generating MVA recombinant viruses. The views and opinions expressed herein are those of the authors and do not purport to reflect the official policy or position of the Department of Defense. REFERENCES 1. Amara, R. R., F. Villinger, J. D. Altman, S. L. Lydy, S. P. O’Neil, S. I. Staprans, D. C. Montefiori, Y. Xu, J. G. Herndon, L. S. Wyatt, M. A. Candido, N. L. Kozyr, P. L. Earl, J. M. Smith, H. L. Ma, B. D. Grimm, M. L. Hulsey, J. Miller, H. M. McClure, J. M. McNicholl, B. Moss, and H. L. Robinson. 2001. Control of a mucosal challenge and prevention of AIDS by a multiprotein DNA/MVA vaccine. Science 292:69–74. 2. Amendola, A., G. Lombardi, S. Oliverio, V. Colizzi, and M. Piacentini. 1994. HIV-1 gp120-dependent induction of apoptosis in antigen-specific human T cell clones is characterized by ‘tissue’ transglutaminase expression and prevented by cyclosporin A. FEBS Lett. 339:258–264. 3. Bagetta, G., M. T. Corasaniti, W. Malorni, G. Rainaldi, L. Berliocchi, A. Finazzi-Agro, and G. Nistico. 1996. The HIV-1 gp120 causes ultrastructural changes typical of apoptosis in the rat cerebral cortex. Neuroreport 7:1722– 1724. 4. Barouch, D. H., S. Santra, M. J. Kuroda, J. E. Schmitz, R. Plishka, A. Buckler-White, A. E. Gaitan, R. Zin, J. H. Nam, L. S. Wyatt, M. A. Lifton, C. E. Nickerson, B. Moss, D. C. Montefiori, V. M. Hirsch, and N. L. Letvin. 2001. Reduction of simian-human immunodeficiency virus 89.6P viremia in rhesus monkeys by recombinant modified vaccinia virus Ankara vaccination. J. Virol. 75:5151–5158. 5. Belshe, R. B., G. J. Gorse, M. J. Mulligan, T. G. Evans, M. C. Keefer, J. L. Excler, A. M. Duliege, J. Tartaglia, W. I. Cox, J. McNamara, K. L. Hwang, A. Bradney, D. Montefiori, and K. J. Weinhold. 1998. Induction of immune responses to HIV-1 by canarypox virus (ALVAC) HIV-1 and gp120 SF-2 recombinant vaccines in uninfected volunteers. AIDS 12:2407–2415. 6. Belshe, R. B., C. Stevens, G. J. Gorse, S. Buchbinder, K. Weinhold, H. Sheppard, D. Stablein, S. Self, J. McNamara, S. Frey, J. Flores, J. L. Excler, M. Klein, R. E. Habib, A. M. Duliege, C. Harro, L. Corey, M. Keefer, M. Mulligan, P. Wright, C. Celum, F. Judson, K. Mayer, D. McKirnan, M. Marmor, and G. Woody. 2001. Safety and immunogenicity of a canarypoxvectored human immunodeficiency virus type 1 vaccine with or without gp120: a phase 2 study in higher- and lower-risk volunteers. J. Infect. Dis. 183:1343–1352. 7. Cao, H., P. Kaleebu, D. Hom, J. Flores, D. Agrawal, N. Jones, J. Serwanga, M. Okello, C. Walker, H. Sheppard, R. El-Habib, M. Klein, E. Mbidde, P. Mugyenyi, B. Walker, J. Ellner, and R. Mugerwa. 2003. Immunogenicity of a recombinant human immunodeficiency virus (HIV)-canarypox vaccine in HIV-seronegative Ugandan volunteers: results of the HIV Network for Prevention Trials 007 Vaccine Study. J. Infect. Dis. 187:887–895. 8. Carroll, M. W., and B. Moss. 1995. E. coli beta-glucuronidase (GUS) as a marker for recombinant vaccinia viruses. BioTechniques 19:352–356. 9. Carroll, M. W., and B. Moss. 1997. Host range and cytopathogenicity of the highly attenuated MVA strain of vaccinia virus: propagation and generation of recombinant viruses in a nonhuman mammalian cell line. Virology 238: 198–211. 10. Cebere, I., L. Dorrell, H. McShane, A. Simmons, S. McCormack, C. Schmidt, C. Smith, M. Brooks, J. E. Roberts, S. C. Darwin, P. E. Fast, C. Conlon, S. Rowland-Jones, A. J. McMichael, and T. Hanke. 2006. Phase I clinical trial

VOL. 81, 2007

11.

12.

13.

14.

15. 16.

17.

18. 19.

20.

21.

22. 23.

24.

25.

26.

27.

28.


safety of DNA- and modified virus Ankara-vectored human immunodeficiency virus type 1 (HIV-1) vaccines administered alone and in a prime-boost regime to healthy HIV-1-uninfected volunteers. Vaccine 24:417–425. Chahroudi, A., D. A. Garber, P. Reeves, L. Liu, D. Kalman, and M. B. Feinberg. 2006. Differences and similarities in viral life cycle progression and host cell physiology after infection of human dendritic cells with modified vaccinia virus Ankara and vaccinia virus. J. Virol. 80:8469–8481. Chen, X., M. T. Rock, J. Hammonds, J. Tartaglia, A. Shintani, J. Currier, B. Slike, J. E. Crowe, Jr., M. Marovich, and P. Spearman. 2005. Pseudovirion particle production by live poxvirus human immunodeficiency virus vaccine vector enhances humoral and cellular immune responses. J. Virol. 79:5537– 5547. Cox, W. I., J. Tartaglia, and E. Paoletti. 1993. Induction of cytotoxic T lymphocytes by recombinant canarypox (ALVAC) and attenuated vaccinia (NYVAC) viruses expressing the HIV-1 envelope glycoprotein. Virology 195:845–850. Davies, M. V., H. W. Chang, B. L. Jacobs, and R. J. Kaufman. 1993. The E3L and K3L vaccinia virus gene products stimulate translation through inhibition of the double-stranded RNA-dependent protein kinase by different mechanisms. J. Virol. 67:1688–1692. de Bruyn, G., A. J. Rossini, Y. L. Chiu, D. Holman, M. L. Elizaga, S. E. Frey, D. Burke, T. G. Evans, L. Corey, and M. C. Keefer. 2004. Safety profile of recombinant canarypox HIV vaccines. Vaccine 22:704–713. Egan, M. A., W. A. Pavlat, J. Tartaglia, E. Paoletti, K. J. Weinhold, M. L. Clements, and R. F. Siliciano. 1995. Induction of human immunodeficiency virus type 1 (HIV-1)-specific cytolytic T lymphocyte responses in seronegative adults by a nonreplicating, host-range-restricted canarypox vector (ALVAC) carrying the HIV-1 MN env gene. J. Infect. Dis. 171:1623– 1627. Evans, T. G., M. C. Keefer, K. J. Weinhold, M. Wolff, D. Montefiori, G. J. Gorse, B. S. Graham, M. J. McElrath, M. L. Clements-Mann, M. J. Mulligan, P. Fast, M. C. Walker, J. L. Excler, A. M. Duliege, and J. Tartaglia. 1999. A canarypox vaccine expressing multiple human immunodeficiency virus type 1 genes given alone or with rgp120 elicits broad and durable CD8⫹ cytotoxic T lymphocyte responses in seronegative volunteers. J. Infect. Dis. 180:290–298. Everett, H., and G. McFadden. 2002. Poxviruses and apoptosis: a time to die. Curr. Opin. Microbiol. 5:395–402. Fang, Z. Y., I. Kuli-Zade, and P. Spearman. 1999. Efficient human immunodeficiency virus (HIV)-1 Gag-Env pseudovirion formation elicited from mammalian cells by a canarypox HIV vaccine candidate. J. Infect. Dis. 180:1122–1132. Fang, Z. Y., K. Limbach, J. Tartaglia, J. Hammonds, X. Chen, and P. Spearman. 2001. Expression of vaccinia E3L and K3L genes by a novel recombinant canarypox HIV vaccine vector enhances HIV-1 pseudovirion production and inhibits apoptosis in human cells. Virology 291:272–284. Fischer, S. F., H. Ludwig, J. Holzapfel, M. Kvansakul, L. Chen, D. C. Huang, G. Sutter, M. Knese, and G. Hacker. 2006. Modified vaccinia virus Ankara protein F1L is a novel BH3-domain-binding protein and acts together with the early viral protein E3L to block virus-associated apoptosis. Cell Death Differ. 13:109–118. Franchini, G., S. Gurunathan, L. Baglyos, S. Plotkin, and J. Tartaglia. 2004. Poxvirus-based vaccine candidates for HIV: two decades of experience with special emphasis on canarypox vectors. Expert Rev. Vaccines 3:S75–S88. Goepfert, P. A., H. Horton, M. J. McElrath, S. Gurunathan, G. Ferrari, G. D. Tomaras, D. C. Montefiori, M. Allen, Y. L. Chiu, P. Spearman, J. D. Fuchs, B. A. Koblin, W. A. Blattner, S. Frey, M. C. Keefer, L. R. Baden, and L. Corey. 2005. High-dose recombinant canarypox vaccine expressing HIV-1 protein, in seronegative human subjects. J. Infect. Dis. 192:1249–1259. Gupta, K., M. Hudgens, L. Corey, M. J. McElrath, K. Weinhold, D. C. Montefiori, G. J. Gorse, S. E. Frey, M. C. Keefer, T. G. Evans, R. Dolin, D. H. Schwartz, C. Harro, B. Graham, P. W. Spearman, M. Mulligan, and P. Goepfert. 2002. Safety and immunogenicity of a high-titered canarypox vaccine in combination with rgp120 in a diverse population of HIV-1-uninfected adults: AIDS Vaccine Evaluation Group protocol 022A. J. Acquir. Immune Defic. Syndr. 29:254–261. Hanke, T., R. V. Samuel, T. J. Blanchard, V. C. Neumann, T. M. Allen, J. E. Boyson, S. A. Sharpe, N. Cook, G. L. Smith, D. I. Watkins, M. P. Cranage, and A. J. McMichael. 1999. Effective induction of simian immunodeficiency virus-specific cytotoxic T lymphocytes in macaques by using a multiepitope gene and DNA prime-modified vaccinia virus Ankara boost vaccination regimen. J. Virol. 73:7524–7532. Hel, Z., J. Nacsa, W. P. Tsai, A. Thornton, L. Giuliani, J. Tartaglia, and G. Franchini. 2002. Equivalent immunogenicity of the highly attenuated poxvirus-based ALVAC-SIV and NYVAC-SIV vaccine candidates in SIVmac251infected macaques. Virology 304:125–134. Herbein, G., U. Mahlknecht, F. Batliwalla, P. Gregersen, T. Pappas, J. Butler, W. A. O’Brien, and E. Verdin. 1998. Apoptosis of CD8⫹ T cells is mediated by macrophages through interaction of HIV gp120 with chemokine receptor CXCR4. Nature 395:189–194. Hornemann, S., O. Harlin, C. Staib, S. Kisling, V. Erfle, B. Kaspers, G.

29.

30. 31.

32.

33.

34.

35. 36. 37.

38. 39. 40. 41. 42.

43. 44.

45. 46. 47. 48. 49.

50.

7033

Hacker, and G. Sutter. 2003. Replication of modified vaccinia virus Ankara in primary chicken embryo fibroblasts requires expression of the interferon resistance gene E3L. J. Virol. 77:8394–8407. Ignatius, R., M. Marovich, E. Mehlhop, L. Villamide, K. Mahnke, W. I. Cox, F. Isdell, S. S. Frankel, J. R. Mascola, R. M. Steinman, and M. Pope. 2000. Canarypox virus-induced maturation of dendritic cells is mediated by apoptotic cell death and tumor necrosis factor alpha secretion. J. Virol. 74:11329– 11338. Kapasi, A. A., G. Patel, N. Franki, and P. C. Singhal. 2002. HIV-1 gp120induced tubular epithelial cell apoptosis is mediated through p38-MAPK phosphorylation. Mol. Med. 8:676–685. Koopman, G., C. P. Reutelingsperger, G. A. Kuijten, R. M. Keehnen, S. T. Pals, and M. H. van Oers. 1994. Annexin V for flow cytometric detection of phosphatidylserine expression on B cells undergoing apoptosis. Blood 84: 1415–1420. Lawson, V. A., K. A. Silburn, P. R. Gorry, G. Paukovic, D. F. Purcell, A. L. Greenway, and D. A. McPhee. 2004. Apoptosis induced in synchronized human immunodeficiency virus type 1-infected primary peripheral blood mononuclear cells is detected after the peak of CD4⫹ T-lymphocyte loss and is dependent on the tropism of the gp120 envelope glycoprotein. Virology 327:70–82. Marovich, M. A., J. R. Mascola, M. A. Eller, M. K. Louder, P. A. Caudrelier, R. El-Habib, S. Ratto-Kim, J. H. Cox, J. R. Currier, B. L. Levine, C. H. June, W. B. Bernstein, M. L. Robb, B. Schuler-Thurner, R. M. Steinman, D. L. Birx, and S. Schlesinger-Frankel. 2002. Preparation of clinical-grade recombinant canarypox-human immunodeficiency virus vaccine-loaded human dendritic cells. J. Infect. Dis. 186:1242–1252. Montefiori, D. C., J. T. Safrit, S. L. Lydy, A. P. Barry, M. Bilska, H. T. Vo, M. Klein, J. Tartaglia, H. L. Robinson, and B. Rovinski. 2001. Induction of neutralizing antibodies and Gag-specific cellular immune responses to an R5 primary isolate of human immunodeficiency virus type 1 in rhesus macaques. J. Virol. 75:5879–5890. Moss, B. 1992. Poxvirus expression vectors. Curr. Top. Microbiol. Immunol. 158:25–38. Moss, B. 1994. Replicating and host-restricted non-replicating vaccinia virus vectors for vaccine development. Dev. Biol. Stand. 82:55–63. Najera, J. L., C. E. Gomez, E. Domingo-Gil, M. M. Gherardi, and M. Esteban. 2006. Cellular and biochemical differences between two attenuated poxvirus vaccine candidates (MVA and NYVAC) and role of the C7L gene. J. Virol. 80:6033–6047. Pincus, S., J. Tartaglia, and E. Paoletti. 1995. Poxvirus-based vectors as vaccine candidates. Biologicals 23:159–164. Pitisuttithum, P. 2005. HIV-1 prophylactic vaccine trials in Thailand. Curr. HIV Res. 3:17–30. Pogo, B. G., S. M. Melana, and J. Blaho. 2004. Poxvirus infection and apoptosis. Int. Rev. Immunol. 23:61–74. Pugliese, O., M. Boirivant, and M. Viora. 1997. Apoptosis induction by human immunodeficiency virus type 1 (HIV-1) gp120 peptides. Viral Immunol. 10:95–102. Ranasinghe, C., J. C. Medveczky, D. Woltring, K. Gao, S. Thomson, B. E. Coupar, D. B. Boyle, A. J. Ramsay, and I. A. Ramshaw. 2006. Evaluation of fowlpox-vaccinia virus prime-boost vaccine strategies for high-level mucosal and systemic immunity against HIV-1. Vaccine 24:5881–5895. Reimann, J., and R. Schirmbeck. 1999. Alternative pathways for processing exogenous and endogenous antigens that can generate peptides for MHC class I-restricted presentation. Immunol. Rev. 172:131–152. Russell, N. D., B. S. Graham, M. C. Keefer, M. J. McElrath, S. G. Self, K. J. Weinhold, D. C. Montefiori, G. Ferrari, H. Horton, G. D. Tomaras, S. Gurunathan, L. Baglyos, S. E. Frey, M. J. Mulligan, C. D. Harro, S. P. Buchbinder, L. R. Baden, W. A. Blattner, B. A. Koblin, and L. Corey. 2007. Phase 2 study of an HIV-1 canarypox vaccine (vCP1452) alone and in combination with rgp120: negative results fail to trigger a phase 3 correlates trial. J. Acquir. Immune Defic. Syndr. 44:203–212. Sancho, M. C., S. Schleich, G. Griffiths, and J. Krijnse-Locker. 2002. The block in assembly of modified vaccinia virus Ankara in HeLa cells reveals new insights into vaccinia virus morphogenesis. J. Virol. 76:8318–8334. Singh-Sandhu, D., and J. Tartaglia. 2003. Poxviruses as immunization vehicles, p. 1297–1316. In S. A. Plotkin (ed.), Vaccines, 4th ed. Saunders, Philadelphia, PA. Tartaglia, J., M. E. Perkus, J. Taylor, E. K. Norton, J. C. Audonnet, W. I. Cox, S. W. Davis, J. van der Hoeven, B. Meignier, M. Riviere, et al. 1992. NYVAC: a highly attenuated strain of vaccinia virus. Virology 188:217–232. Tartaglia, J., S. Pincus, and E. Paoletti. 1990. Poxvirus-based vectors as vaccine candidates. Crit. Rev. Immunol. 10:13–30. Taylor, J., B. Meignier, J. Tartaglia, B. Languet, J. Van der Hoeven, G. Franchini, C. Trimarchi, and E. Paoletti. 1995. Biological and immunogenic properties of a canarypox-rabies recombinant, ALVAC-RG (vCP65) in nonavian species. Vaccine 13:539–549. Wyatt, L. S., B. Moss, and S. Rozenblatt. 1995. Replication-deficient vaccinia virus encoding bacteriophage T7 RNA polymerase for transient gene expression in mammalian cells. Virology 210:202–205.