Adenovirus-Mediated Delivery of an Anti-V Antigen Monoclonal Antibody Protects Mice against a Lethal Yersinia pestis Challenge

Abstract

Pneumonic plague, caused by inhalation of Yersinia pestis, represents a major bioterrorism threat for which no vaccine is available. Based on the knowledge that genetic delivery of monoclonal antibodies (MAbs) with adenovirus (Ad) gene transfer vectors results in rapid, high-level antibody expression, we evaluated the hypothesis that Ad-mediated delivery of a neutralizing antibody directed against the Y. pestis V antigen would protect mice against a Y. pestis challenge. MAbs specific for the Y. pestis V antigen were generated, and the most effective in protecting mice against a lethal intranasal Y. pestis challenge was chosen for further study. The coding sequences for the heavy and light chains were isolated from the corresponding hybridoma and inserted into a replication-defective serotype 5 human Ad gene transfer vector (AdαV). Western analysis of AdαV-infected cell supernatants demonstrated completely assembled antibodies reactive with V antigen. Following AdαV administration to mice, high levels of anti-V antigen antibody titers were detectable as early as 1 day postadministration, peaked by day 3, and remained detectable through a 12-week time course. When animals that received AdαV were challenged with Y. pestis at day 4 post-AdαV administration, 80% of the animals were protected, while 0% of control animals survived (P < 0.01). Ad-mediated delivery of a V antigen-neutralizing antibody is an effective therapy against plague in experimental animals and could be developed as a rapidly acting antiplague therapeutic.

Yersinia pestis, the etiologic agent of plague, has been responsible for high mortality in epidemics throughout human history and remains a current threat as a potential biological warfare agent (35, 39). Y. pestis infection can result in bubonic, septicemic, or pneumonic plague, with the last being the most likely following a deliberate Y. pestis release (23, 35, 39, 47, 57) (http://www.bt.cdc.gov/agent/agentlist-category.asp). Pneumonic plague is highly contagious and is easily transmitted person to person through airborne droplets, resulting in a rapid onset of disease and a mortality rate of almost 100% if treatment is delayed more than 24 h postexposure (23, 39, 46, 57).

There are currently no plague vaccines available in the United States. Although several active vaccine candidates are being developed, most require multiple administrations to achieve protective immunity (1, 2, 4, 6, 15, 18, 27, 43, 46, 48, 49, 51, 53-55). In the context that it is improbable that nonmilitary populations will be prophylactically immunized against plague, vaccines requiring multiple administrations over weeks to months are not likely to be useful in response to a bioterror attack. However, several studies in experimental animal models have demonstrated the efficacy of passive antibody administration against plague (3, 16, 20-22, 33, 37). In combination with the capacity to effectively target antibiotic-resistant Y. pestis strains, the ability of passive immunotherapy to provide an immediate state of protection has increased interest in developing antibody-based therapeutics for plague.

The Y. pestis virulence (V) antigen has been identified as a potent protective antigen (PA) against plague and has consequently been evaluated as a subunit vaccine candidate and as a target for passive immunotherapy (2, 16, 20-22, 27, 33, 52, 54). V antigen has multiple roles during the course of Y. pestis infection. It is required for translocation of bacterial effector proteins into host cells via a type III secretion system and additionally is associated with increased interleukin 10 levels and decreased tumor necrosis factor alpha levels through an unknown mechanism (5, 36, 38, 41, 44). Transfer of immune sera from animals immunized with V antigen to naive animals confers immediate protection against Y. pestis challenge (16, 33). Additionally, passive transfer of an anti-V antigen monoclonal antibody (MAb) protects mice against a lethal challenge with Y. pestis (20-22).

Delivery of the coding sequences for MAbs with viral vectors has been effective against both infectious diseases and cancers and is an alternative platform to administration of purified antibodies (9, 10, 25, 26, 28, 45, 50). The rapid transgene expression kinetics from adenovirus (Ad) gene transfer vectors renders them applicable as antibody delivery vehicles for potential bioweapons. With this background, we generated an anti-V antigen MAb that neutralizes Y. pestis following passive transfer to experimental animals and constructed a replication-defective human Ad serotype 5 gene transfer vector expressing the coding sequences for this protective anti-V antigen MAb (AdαV). Following administration to mice, AdαV generates high-serum anti-V antigen antibody titers and, most importantly, protects mice against a lethal challenge with a fully virulent strain of Y. pestis.

MATERIALS AND METHODS

Purification of recombinant V antigen.

Recombinant V antigen from Yersinia pestis was produced as a reagent for screening reactivity of MAbs against V antigen. The V antigen coding sequence was inserted into the T7 promoter-driven prokaryotic expression plasmid pRSET (Invitrogen, Carlsbad, CA) to generate the pRSET-V plasmid, expressing V antigen as a histidine tag fusion protein. pRSET-V was transformed into the BL21(DE3)pLysS strain of Escherichia coli, and expression of the V antigen was induced with isopropyl-β-d-thiogalactopyranoside. The V antigen was affinity purified using a Ni-nitrilotriacetic acid (NTA) Superflow column (Qiagen, Valencia, CA) under native conditions. The purity of the protein was confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (NuPAGE system; Invitrogen), and its identity was confirmed by Western analysis with a rabbit anti-V antigen antibody (kindly provided by Sina Bavari, USAMRIID, Fort Detrick, MD).

Generation and screening of anti-V antigen MAbs.

Murine MAbs were raised against purified recombinant V antigen by the Northeast Biodefense Center (NBC) Monoclonal Antibody Core, Albert Einstein College of Medicine. Eight-week-old female BALB/c mice were immunized intraperitoneally (IP) with 100 μg of purified His-tagged V antigen in complete Freund's adjuvant and boosted 4 weeks later with 100 μg of the same antigen in incomplete Freund's adjuvant. Four weeks later, the mice had anti-V antigen antibody titers as measured by enzyme-linked immunosorbent assay (ELISA) that were in excess of 10⁶. After resting to let the titer drop, one mouse was boosted IP with 100 μg of the same antigen in saline on two successive days, and the spleen was removed 2 days after the second boost. The red blood cells were lysed, and approximately 3 × 10⁷ nucleated spleen cells were fused with the Ag8.653 cell line at a 3:1 ratio (12). The fused cells were plated onto 96-well plates in hypoxanthine-, aminopterin-, and thymidine-containing media. As soon as the wells were 25 to 50% confluent, the medium from each of the wells was screened by ELISA using the same His-tagged V antigen, and ∼480 wells were found to produce immunoglobulin G (IgG) antibodies that reacted with His-V antigen, using goat anti-mouse IgG isotype-specific antibody-alkaline phosphatase (Southern Biotechnology Associates, Birmingham, AL) as the secondary antibody. The positive wells were secondarily screened by ELISA for reactivity with His-tagged B7-H14-Ig (hB7) to identify any His binding antibodies. The medium from the 48 wells that were the most strongly positive by ELISA for His-V antigen but negative for His-hB7 was screened by Western analysis for reactivity with recombinant V antigen, and the hybridoma cells from 20 wells were subcloned in soft agar (13). The supernatants of ∼20 subclones from each well were checked for isotype and reactivity with V antigen. Totals of 12 IgG1, 4 IgG2a, and 4 IgG2b clones producing stable, cloned hybridomas were identified. Eleven hybridomas producing MAbs with the strongest V antigen reactivity by Western analysis were selected for analysis of in vivo protective efficacy. These hybridomas were adapted to serum-free AIM medium (Invitrogen, Grand Island, NY), and high concentrations of MAbs (approximately 1 mg/ml) were produced in Integra flasks (Wilson Wolf Manufacturing, New Brighton, MN).

For passive transfer studies, female BALB/c mice (n = 5/group) received 100 to 500 μl of supernatant administered IP 2 h prior to intranasal challenge with a fully virulent strain of Y. pestis (described below). Supernatants from clone 2C12.4 provided the best protection against the lethal challenge, and this clone was chosen for further analysis.

Ad vectors.

The coding sequences for the heavy and light chains of the protective anti-V antigen antibody 2C12.4 were obtained by RNA ligase-mediated rapid amplification of cDNA ends (GeneRacer kit; Invitrogen, Carlsbad, CA) using RNA extracted from the corresponding hybridoma cells. The isolated coding sequences were cloned into a replication-defective human Ad vector to generate AdαV. AdαV is a serotype 5, E1⁻ E3⁻ Ad gene transfer vector constructed to direct the expression of a full-size (heavy- and light-chain) murine MAb against the Yersinia pestis V antigen. The anti-V light-chain and heavy-chain sequences were separated by a furin cleavage site and the self-cleaving 2A peptide from foot-and-mouth disease virus to facilitate expression of both protein subunits from a single cytomegalovirus (CMV) promoter (9-11). The expression cassette in the AdαV vector contains (5′ to 3′) the CMV promoter/enhancer, the anti-V heavy-chain coding sequence, a 4-amino-acid furin cleavage site, the 24-amino-acid self-cleaving 2A peptide, the anti-V light chain, and the SV40 polyadenylation signal.

AdαPA, a similarly constructed gene transfer vector encoding an unrelated antibody against anthrax PA, was used as a negative control (9). The AdαPA vector contains (5′ to 3′) the CMV promoter/enhancer, the anti-PA light-chain coding sequence, the poliovirus internal ribosome entry site, the anti-PA heavy-chain coding sequence, and the SV40 polyadenylation signal. AdNull, a gene transfer vector containing no transgene, was used as an additional negative control (19).

AdαV, AdαPA, and AdNull were produced in 293 cells and purified by centrifugation twice through a CsCl gradient as previously described (42). The titer of each recombinant Ad preparation was determined spectrophotometrically and expressed as particle units (pu) (31).

Assessment of AdαV in vitro.

Expression and the specificity of the anti-V antibody from AdαV following infection of cells in vitro were assessed by Western analysis. A549 cells were infected with AdαV (5,000 pu/cell), and infected-cell supernatants were harvested at 72 h postinfection. Supernatants were concentrated using Ultracel YM-10 centrifugal filters (Millipore, Billerica, MA) and evaluated for the expression of anti-V antigen antibody by Western analysis under reducing and nonreducing conditions using a peroxidase-conjugated sheep anti-mouse IgG secondary antibody (Sigma, St. Louis, MO) and ECL reagent (Amersham). The specificity of the Ad-expressed anti-V antibody in infected-cell supernatants for the Y. pestis V antigen was determined by Western analysis with recombinant V as the target antigen. Negative controls included AdαPA-infected cell supernatants. Detection was with a peroxidase-conjugated sheep anti-mouse IgG secondary antibody and ECL reagent.

Assessment of AdαV in vivo.

Female BALB/c and male C57BL/6 mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and Taconic (Germantown, NY), respectively. The animals were housed under specific-pathogen-free conditions and used at 6 to 8 weeks of age. Male C57BL/6 mice were administered AdαV (10¹¹ pu) via the intravenous route. Naive mice injected with AdNull or AdαPA were used as negative controls. Ad vectors were diluted with saline to the specified dose. To assess the ability of AdαV to generate anti-V antigen antibodies in vivo, at various times following vector administration, serum was collected via the tail vein, centrifuged at 8,000 × g for 20 min, and stored at −20°C. Anti-V antigen antibody levels in mouse serum were assessed by ELISA using flat-bottomed 96-well EIA/RIA plates (Corning, New York, NY) coated with 0.5 μg recombinant V antigen per well in a total volume of 100 liters of 0.05 M carbonate buffer, pH 7.4, overnight at 4°C. The plates were washed with phosphate-buffered saline (PBS) containing 0.05% Tween 20 (PBS-Tween) and blocked with 5% dry milk in PBS for 1 h at 23°C. Serial serum dilutions were added to each well and incubated for 1 h at 23°C. The plates were washed four times with PBS-Tween, and 100 μl/well of 1:10,000-diluted peroxidase-conjugated sheep anti-mouse IgG1 (Sigma) in PBS containing 1% dry milk was added and incubated for 1 h at 23°C. The plates were washed four times with PBS-Tween and once with PBS. Peroxidase substrate (100 μl/well; Bio-Rad, Hercules, CA) was added and incubated for 15 min at 23°C, and then a stop solution of 2% oxalic acid (100 μl/well) was added. Absorbance at 415 nm was read with a microplate reader (Bio-Rad). Antibody titers were calculated with a log(optical density)-log(dilution) interpolation model and a cutoff value equal to twofold the absorbance of background.

To assess the ability of AdαV to protect mice against a lethal challenge with Y. pestis, male C57BL/6 mice (n = 5/group) were injected with AdαV. Mice injected with AdNull were negative controls. Four days after vector administration, each mouse was challenged intranasally with Y. pestis CO92. As a positive control, concentrated MAb 2C12.4 hybridoma supernatant containing approximately 500 μg MAb was administered IP 2 h prior to challenge. The Y. pestis challenge studies were carried out at the Public Health Research Institute (PHRI) at the International Center for Public Health (Newark, NJ) under BSL3 conditions. Y. pestis CO92 was grown aerobically in heart infusion broth (Difco; BD, Franklin Lakes, NJ) at 30°C and diluted in saline solution at doses from 10³ CFU to 10⁶ CFU as specified. The challenge dose was calculated as 2 × 10⁴ CFU, which corresponds to 363 50% lethal doses (LD₅₀). Fifty liters of bacterial suspension was used for intranasal infection of mice; the bacterial dose was controlled by plating on Yersinia-selective agar (Oxoid, Hampshire, United Kingdom) and counting colonies for CFU determination. Survival was monitored daily for 14 or 30 days as indicated.

Statistical analyses.

The data are presented as means ± standard errors of the means. Statistical analyses were performed using the nonpaired two-tailed Student t test, assuming equal variance. Survival evaluation was carried out using Kaplan-Meier analysis. Statistical significance was determined at P values of <0.05.

RESULTS

Novel V antigen-specific MAbs protect against Y. pestis challenge.

Hybridoma supernatants from a panel of clones generated by immunizing mice with purified recombinant V antigen were screened for V antigen specificity by Western analysis. As examples, Western analysis of clones 2C12.4, 2G9.8, and 1A10.14 were assessed for reactivity against bovine serum albumin (BSA) (Fig. 1, lanes 1, 3, and 5) or purified recombinant V antigen (Fig. 1, lanes 2, 4, and 6). The isotypes of all MAbs were determined; clones 2C12.4, 2G9.8, and 2G9.10 are IgG2b κ, and the three showed strong reactivity with V antigen. Clones 1F1.3 and 1A10.14 are IgG1 κ. Clone 1F1.3 was strongly reactive with V antigen, and clone 1A10.14 did not recognize V antigen.

FIG. 1.

Screening hybridoma supernatants for V antigen specificity. A panel of hybridoma supernatants was screened for V antigen recognition by Western analysis. To determine specificity, each supernatant was additionally tested for reactivity with BSA, an unrelated protein. A representative Western blot of hybridoma supernatants from three independent clones, 2C12.4, 2G9.8, and 1A10.14, is presented. Clones 2C12.4 and 2G9.8 demonstrate specificity for V antigen (lanes 2 and 4) but not BSA (lanes 1 and 3). Clone 1A10.14 does not react with either V antigen (lane 6) or BSA (lane 5).

Eleven MAbs screened by Western analysis were selected for studies of passive protection against a lethal Y. pestis challenge. Concentrated hybridoma supernatants (each containing approximately 500 μg antibody) were administered IP to mice 2 h prior to intranasal challenge with Y. pestis CO92, and survival was monitored for 14 days (Fig. 2A). Pooled V antigen-specific hyperimmune serum from mice immunized multiple times with AdsecV, an Ad gene transfer vector encoding V antigen, was used as a positive control (8). Negative controls for this experiment included PBS and clone 1A10.14, a MAb that did not react with V antigen. Shown in Fig. 2A is a representative experiment screening four V antigen-specific MAbs (1F3.3, 2G9.10, 2G9.8, and 2C12.4). In four independent replicate experiments, MAb 2C12.4 consistently mediated protection against Y. pestis. The dose-dependent efficacy of MAb 2C12.4 was also determined in passive protection experiments against Y. pestis challenge (Fig. 2B). Mice received 100, 250, or 500 μg of antibody IP and were challenged intranasally with Y. pestis at 2 h postadministration. The results of this single experiment indicate that MAb 2C12.4 protects animals in a dose-dependent manner, with all doses providing some degree of protection.

FIG. 2.

Survival of mice passively immunized with concentrated hybridoma supernatants prior to challenge with Yersinia pestis. (A) BALB/c mice (n = 5/group) received IP 500 μl of concentrated hybridoma supernatants (∼500 μl/g antibody) 2 hours prior to an intranasal challenge with a lethal dose of Yersinia pestis strain CO92. As a positive control, a group of mice received 400 μl of V antigen-specific hyperimmune serum. (B) BALB/c mice (n = 5/group) received IP 100, 250, or 500 μg of concentrated MAb 2C12.4 hybridoma supernatant 2 h prior to an intranasal challenge with a lethal dose of Yersinia pestis strain CO92. Negative controls included PBS or 500 liters of concentrated hybridoma supernatant from clone 1A10.14, which does not react with recombinant V antigen. After the challenge, the survival of the animals was monitored for 14 days.

Characterization of AdαV-expressed full-length anti-V antibody.

Based on the data demonstrating that antibody 2C12.4 was the most effective in protecting mice from a lethal challenge with Y. pestis, the coding sequences for the heavy and light chains of antibody 2C12.4 were isolated from the corresponding hybridoma line and cloned into a replication-defective serotype 5 human Ad gene transfer vector to generate AdαV. To examine expression of the anti-V antigen antibody by AdαV, A549 cells were infected with AdαV. AdαPA, an Ad vector expressing an unrelated antibody, was used as a positive control, and AdNull, an Ad vector with no transgene, was used as a negative control. At 72 h postinfection, cell supernatants were collected, and antibody expression was examined under nonreducing conditions to assess assembly of a full-size antibody and under reducing conditions to assess expression of the individual heavy and light chains (Fig. 3). Supernatants collected from AdαV- or AdαPA-infected cells demonstrated the presence of antibody heavy (50-kDa) and light (25-kDa) chains (Fig. 3A, lanes 2 and 4). No antibody was detected with uninfected or AdNull-infected cells (Fig. 3, lanes 1 and 3). When these supernatants were analyzed by Western blotting under native (nonreducing) conditions, samples from both AdαV- and AdαPA-infected cells contained a protein of the expected size for a completely assembled antibody (150 kDa; Fig. 3B, lanes 6 and 8). No antibody was detected with uninfected or AdNull-infected cells (Fig. 3B, lanes 5 and 7).

FIG. 3.

Expression of full-size anti-V antigen MAb 2C12.4 in cells infected with AdαV. A549 cells were infected with 5,000 pu per cell of AdαV, AdαPA (a vector expressing an irrelevant antibody) as a positive control, or AdNull as a negative control. At 72 h postinfection, cell supernatants were assessed for antibody expression by Western analysis with a horseradish peroxidase-conjugated sheep anti-mouse IgG antibody. (A) Antibody expression under reducing conditions. Lane 1, supernatant from uninfected cells; lane 2, supernatant from AdαPA-infected cells; lane 3, supernatant from AdNull-infected cells; and lane 4, supernatant from AdαV-infected cells. The heavy and light chains of the antibodies expressed by the Ad vectors have molecular weights (in thousands) of 50 and 25, respectively. (B) Antibody expression under native conditions. Lane 5, supernatant from uninfected cells; lane 6, supernatant from AdαPA-infected cells; lane 7, supernatant from AdNull-infected cells; and lane 8, supernatant from AdαV-infected cells. The molecular weight (in thousands) of a completely assembled antibody (150) is indicated.

The specificity of the Ad-expressed anti-V antibody was assessed by Western analysis. This experiment tested antibody expression in vitro in AdαV-infected cell supernatants and in vivo in serum from AdαV-administered mice. AdαPA was used as a negative control, and MAb 2C12.4 hybridoma supernatant was used as a positive control. When purified recombinant V antigen was probed with supernatants from AdαV-infected cells, sera obtained from AdαV-immunized mice, or the 2C12.4 hybridoma supernatant, a strong signal corresponding to the expected molecular weight of V antigen was detected (Fig. 4, lanes 3, 4, and 5). In contrast, supernatants from AdαPA-infected cells or sera from AdαPA-immunized mice did not recognize V antigen (Fig. 4, lanes 1 and 2). Thus, AdαV functions to direct expression of a fully assembled MAb that has specificity for V antigen in vitro and in vivo.

FIG. 4.

Specificity of the anti-V mouse MAb expressed by AdαV. A549 cells were infected with 5,000 pu per cell of AdαV or, as a negative control, AdαPA, and supernatants were collected at 72 h postinfection. In a parallel experiment, C57BL/6 mice received 10¹¹ pu of AdαV intravenously, and serum was collected at 72 h postadministration. The infected cell supernatants and sera were assessed for the ability to bind to purified recombinant V antigen by Western analysis. Lane 1, AdαPA-infected cell supernatant; lane 2, sera from AdαPA-injected mice; lane 3, AdαV-infected cell media; lane 4, sera from AdαV-injected mice; and lane 5, MAb 2C12.4 hybridoma supernatant used as a positive control. Purified recombinant V antigen has a molecular weight (in thousands) of 37.

Time-dependent AdαV-mediated in vivo expression of an anti-V antibody.

The time-dependent in vivo expression profile of the anti-V antibody was assessed with mice administered either AdαV or AdαPA. Serum anti-V antigen antibody levels were measured with a V antigen-specific ELISA over 12 weeks. Mice that received AdαV had high serum anti-V antibody levels as early as 1 day postadministration (Fig. 5). These antibody titers peaked at day 3 postadministration with a titer of 10⁵ and then gradually decreased over time, with a titer of 10² at the end of the 12-week experiment. The serum antibody concentration was determined to be 1.15 mg/ml on day 3 postadministration and 0.85 mg/ml on day 5 postadministration. Mice that received the AdαPA or AdNull control vectors did not have any measurable anti-V antibody titers at any time point.

FIG. 5.

Time course of serum anti-V antibody levels following administration of AdαV to C57BL/6 mice. Mice (n = 5/group) were administered AdαV (10¹¹ pu) intravenously. Naive mice and mice that received AdαPA (10¹¹ pu) intravenously were included as negative controls. Serum antibody levels were measured using an anti-V-specific ELISA from day 0 through 12 weeks postadministration.

Protection against intranasal challenge with Y. pestis CO92.

To determine the efficacy of Ad-expressed anti-V antibodies against Y. pestis, mice were immunized with AdαV or AdNull and were then challenged at 4 days postadministration (Fig. 6). Although both bubonic and pneumonic plague models were considered for the challenge experiments, an intranasal challenge route was selected to most closely mimic pneumonic plague, the most likely form of plague to be encountered in an act of bioterrorism. Concentrated MAb 2C12.4 hybridoma supernatants containing approximately 500 μg MAb were used as the positive control and were administered 2 h prior to challenge. All animals were challenged at the same time. Administration of AdαV to mice resulted in substantial protection against Y. pestis challenge (93.3% survival), with the protective effect comparable to that observed for positive-control mice that received MAb 2C12.4. In contrast, animals immunized with AdNull were not protected against the challenge (P < 0.01, AdαV versus AdNull). The data presented in Fig. 6 show the averages from three independent experiments.

FIG. 6.

Survival of C57BL/6 mice challenged with Yersinia pestis 4 days after AdαV administration. Mice (n = 7/group) were administered AdαV (10¹¹ pu) intravenously or, as a negative control, AdNull (10¹¹ pu). Four days after vector administration, the mice were challenged with a lethal dose of Yersinia pestis. As a positive control, concentrated MAb 2C12.4 hybridoma supernatant was administered IP to a separate group of mice 2 h before the challenge. After the challenge, survival of the animals was monitored for 14 days.

In a similarly designed experiment, mice were immunized with AdαV and were then challenged at 4 days postadministration with increasing doses of Y. pestis, ranging from 363 LD₅₀ to 9,090 LD₅₀ (Fig. 7). At each challenge dose, AdαV immunization conferred 100% protection.

FIG. 7.

Survival of C57BL/6 mice challenged with different doses of Yersinia pestis 4 days after AdαV immunization. Mice (n = 9/group) received AdαV (10¹¹ pu) intravenously and were then challenged at 4 days postadministration with increasing doses of Y. pestis, ranging from 363 LD₅₀ to 9,090 LD₅₀. Naive mice were included as negative controls. Survival was monitored for 14 days.

The relative timing of protection against challenge was determined by administering AdαV at various time points, both pre- and post-363 LD₅₀ Y. pestis challenge (Fig. 8). In two independent experiments, mice were immunized with AdαV 4 days prior to challenge, at the same time as the challenge, 6 h after challenge, or 24 h after challenge. After a 1-month time period, animals that received AdαV at 24 h following the challenge were partially protected (12.5% protection). However, the animals immunized with AdαV 4 days prior to challenge, at the time of challenge or 6 h after challenge survived (93.7% protection) without any signs of adverse health effects, demonstrating the potential utility of AdαV as both a pre- and a postexposure therapy. Additionally, the survival rates for these experiments were unchanged from day 7 through day 30 postchallenge, demonstrating robust protection with no evidence of vaccine breakthrough. The data shown in Fig. 8 represent the averages from two independent experiments.

FIG. 8.

Relative timing of protection conferred by AdαV against Yersinia pestis challenge. C57BL/6 mice (n = 8/group) were immunized intravenously with AdαV (10¹¹ pu) or, as a negative control, AdαPA (10¹¹ pu) at various time points pre- and postchallenge, including 4 days before challenge (−4 days), at the time of challenge (0 h), 6 h after challenge (6 h), and 24 h after challenge (24 h). All animals were challenged with 363 LD₅₀ of Y. pestis. After the challenge, survival was monitored for 30 days.

DISCUSSION

The ability to express full-size antibody molecules with Ad gene transfer vectors facilitates the use of this convenient platform as a therapy for potential agents of bioterrorism. We generated and characterized a protective MAb against the Y. pestis V antigen and demonstrated that a single administration of AdαV, an Ad gene transfer vector encoding the protective antibody, rapidly generates a high concentration of antibody in serum of immunized mice and most importantly, protects immunized mice against a lethal Y. pestis challenge as effectively as administration of a V antigen-specific antibody preparation. In the context that there is no vaccine available for treatment of Y. pestis, the strategy of genetic delivery of an anti-V antibody by AdαV as an antibody-based drug may be a useful strategy to develop for use in a Y. pestis bioterrorism attack.

Y. pestis therapeutics.

Aerosol transmission of virulent Y. pestis results in pneumonic plague, a rapidly fatal disease that is a threat related to the potential use of the organism as a biological weapon (23, 35, 39, 46, 57; http://www.bt.cdc.gov/agent/agentlist-category.asp). The currently recommended plague therapy is antibiotic treatment, but multidrug-resistant Y. pestis isolates have been identified, thus limiting the general utility of antibiotics (10, 11, 14, 23, 30, 40).

Both live attenuated and killed whole-cell vaccines against plague have been used in humans for decades, but none are very effective, and none are available in the United States (43, 48, 49). The live attenuated vaccines are all based on mutants of fully virulent strains, but they are not licensed for use in the United States, relating to significant safety and efficacy concerns (21, 24, 43; http://cdc.gov/ncidod/dvbid/plague/prevent.htm). A formaldehyde-killed vaccine has been developed, but this vaccine is also no longer licensed or available for use in the United States, relating to adverse side effects and the inability to provide complete protection against pneumonic plague (21, 43). Current vaccine development efforts are focused largely on recombinant Y. pestis proteins, with V antigen and the capsular F1 protein as the primary targets. Although these vaccine candidates demonstrate efficacy, most require multiple administrations and time for the development of protective immunity (2, 27, 48, 49, 54).

MAbs have several advantages over antibiotics and vaccines in the prevention of disease caused by many extracellular pathogens, including the ability to treat antibiotic-resistant pathogens and the immediate state of protection conferred upon administration (7, 56). There have been a variety of studies demonstrating the efficacy of polyclonal and MAbs as a treatment for Y. pestis infection (3, 16, 20-22, 33, 37). For example, a murine MAb raised against F1 protects mice in models of bubonic and pneumonic plague (3). Also, V antigen-specific MAbs protect mice in bubonic and pneumonic plague models when administered IP (20, 22). When anti-F1 and anti-V antigen MAbs are coadministered to mice, a synergistic effect is observed in response to Y. pestis challenge (20, 21).

Genetic delivery of MAbs.

The limitations of MAb therapy include complex and costly production and purification methods as well as the relatively short half-life of some antibody molecules. Genetic transfer of therapeutic antibodies is an effective strategy for generating specific antibodies in vivo and a consequently attractive strategy for a variety of therapeutic applications, including infectious diseases. It has been possible to express antibody molecules with a variety of viral vectors either in vitro or in vivo, including baculovirus, rhabdovirus, vaccinia virus, Ad, and adeno-associated virus (AAV) (9-11, 17, 26, 28, 29, 32, 34, 45, 50). The ability to directly administer viral vectors expressing MAbs to a host is an alternative technology to the standard practice of administering preparations of purified antibodies. With Ad or adeno-associated virus vectors, this technology has demonstrated efficacy against a variety of infectious diseases (9, 26, 28, 50). When administered to mice, an Ad vector encoding an antibody against the PA component of anthrax toxin is effective at protecting animals from an anthrax toxin challenge as early as 1 day postadministration (9). Similarly, an AAV vector encoding the same antibody protects immunized mice from anthrax toxin challenge throughout a 6-month time period (9). In other experiments, administration of an AAV vector encoding a human MAb that neutralizes human immunodeficiency virus to murine muscle results in significant serum levels of human immunodeficiency virus-neutralizing activity for over 6 months (28). Finally, we have demonstrated the efficacy of AAV-delivered anti-respiratory syncytial virus antibodies in reducing viral load in mice following challenge with wild-type respiratory syncytial virus (50). Collectively, these data indicate the convenience and efficacy of this strategy in producing both acute or durable protective immune responses and underscore the utility of genetic antibody delivery as an infectious disease therapeutic.