|Home | About | Journals | Submit | Contact Us | Français|
Due to the potential use of ricin and other fast-acting toxins as agents of bioterrorism, there is an urgent need for the development of safe and effective antitoxin vaccines. A candidate ricin subunit vaccine (RiVax) consisting of a recombinant attenuated enzymatic A chain (RTA) has been shown to elicit protective antitoxin antibodies in mice and rabbits and is currently being tested in phase I human clinical trials. However, evaluation of the efficacy of this vaccine for humans is difficult for a number of reasons, including the fact that the key neutralizing B-cell epitopes on RTA have not been fully defined. Castelletti and colleagues (Clin. Exp. Immunol. 136:365-372, 2004) recently identified a linear epitope on RTA, spanning residues L161 to I175, as a primary target of serum antibodies derived from humans who had been treated with ricin immunotoxin. While affinity-purified polyclonal IgG antibodies against this region of RTA were capable of neutralizing ricin in vitro, their capacity to confer protection against ricin challenge in vivo was not determined. In this report, we describe the production and characterization of GD12, a murine monoclonal IgG1 antibody specifically directed against residues 163 to 174 (TLARSFIICIQM) of RTA. GD12 bound ricin holotoxin with high affinity (KD [dissociation constant], 2.9 × 10−9 M) and neutralized it with a 50% inhibitory concentration of ~0.25 μg/ml, as determined by a Vero cell-based cytotoxicity assay. Passive administration of GD12 was sufficient to protect BALB/c mice against intraperitoneal and intragastric ricin challenges. These data are important in terms of vaccine development, since they firmly establish that preexisting serum antibodies directed against residues 161 to 175 on RTA are sufficient to confer both systemic and mucosal immunity to ricin. The potential of GD12 to serve as a therapeutic following ricin challenge was not explored in this study.
Recent bioterrorism incidents in the United States and abroad have alerted public health officials to the need for vaccines against pathogens and toxins previously deemed to be of little concern (2, 25). The development and implementation of vaccines for biodefense and emerging infectious diseases are inherently challenging, because phase III clinical efficacy trials of candidate vaccines are generally not feasible or ethical. To address this issue, the Food and Drug Administration (FDA) has implemented the “two-animal rule,” which enables candidate vaccines to advance toward licensure based on efficacy studies performed with two or more relevant animal models (8, 49). For compliance with this FDA policy, the animal models must mimic the pathophysiology of human disease, and the defined end point(s) of the efficacy studies must correlate with the desired effects for humans. However, even well-established animal models cannot completely substitute for human studies. Therefore, whenever possible, specific correlates of protection against select agents and emerging infectious diseases should be established in humans, and surrogate assays should be developed that can be used to estimate immunity in vaccinated human populations (35).
Ricin is a category B toxin, as classified by the Centers for Disease Control and Prevention (CDC). The toxin is naturally produced by the castor bean plant, Ricinus communis, which is cultivated on industrial levels around the world for the production of castor oil. Ricin constitutes up to 5% of the total dry weight of the castor bean and can be extracted from the mash through several simple enrichment steps. The toxin is a member of the so-called type II ribosome-inactivating proteins (RIPs); it consists of two subunits, RTA and RTB, each with a molecular mass of approximately 30,000 Da (31, 47). RTA is an RNA N-glycosidase whose substrate is a conserved adenine residue within the so-called sarcin/ricin loop of eukaryotic 28S rRNA. Ribosome progression is arrested upon cleavage of this residue by RTA (9). RTB is a bivalent lectin with specificity for glycoproteins and glycolipids containing β(1-3)-linked galactose and N-acetylgalactosamine residues (4). RTB mediates the attachment and internalization of ricin into host cells and facilitates retrograde transport of the toxin to the Golgi apparatus and endoplasmic reticulum (ER) (21, 38). Ricin in semipurified or purified form is extremely toxic to humans following injection, inhalation, or ingestion (3) and has been used as an agent of bioterrorism. Ricin was weaponized by the United States and other countries during World War II (24, 48); it has been used in assassinations; and it was recently uncovered in a number of government facilities, including a South Carolina postal facility, and packed in envelopes delivered to offices of the U.S. Senate (12, 40).
Because of the toxicity of ricin and the ease of its preparation, public health officials and defense agencies have made a concerted effort to develop a vaccine against it that could be administered to emergency first responders and military personnel (25). Although formaldehyde-treated ricin toxoid (RT) preparations are effective at eliciting protective immunity in rodents, they are not being considered for use in humans, because of concerns about residual toxicity (11). Therefore, the current emphasis is on the development of attenuated subunit vaccines (6, 16, 32, 45, 52). One of the most promising candidates is a recombinant derivative of RTA containing two point mutations: one in the enzymatic active site (Y80A) and the other in a residue (V76M) involved in eliciting vascular leak syndrome (42-45, 52). This vaccine, known by the trade name RiVax, is safe and immunogenic in mice and rabbits and, when administered intramuscularly, elicits serum antitoxin IgG antibodies capable of protecting animals against a systemic ricin challenge of 10 50% lethal doses (LD50s) (42, 44). RiVax has also been shown to elicit an antibody response capable of protecting mice against both intragastric (i.g.) and aerosol challenges (45). Based on these animal studies, a pilot phase I clinical trial of RiVax was undertaken in 2006 (52). The trial consisted of three groups of five healthy volunteers injected at monthly intervals with 10, 33, or 100 μg of the vaccine. The results of this study revealed that RiVax was well tolerated and resulted in dose-dependent seroconversion (52).
While RiVax was deemed safe and immunogenic, evaluation of the efficacy of this vaccine in humans remains challenging. For example, in the pilot phase I clinical trial noted above, there was no observed correlation between serum anti-RTA IgG titers and in vitro ricin-neutralizing activity (52). Specifically, two individuals with virtually identical serum anti-RTA IgG levels (4.73 ± 0.019 μg/ml versus 4.36 ± 0.16 μg/ml) had toxin-neutralizing titers that differed by >10-fold (1.4 ± 0 versus 0.13 ± 0.02). These data suggest that the polyclonal response to RTA consists of a mixture of neutralizing and nonneutralizing antibodies and that the ratio of the two types of antibodies can differ from individual to individual. This interpretation is supported by work from Maddaloni and colleagues, who identified both potent neutralizing monoclonal antibodies (MAbs) (e.g., RAC18) and a number of MAbs that bound to RTA with high avidity but failed to neutralize ricin in vitro or in vivo (23, 37). In fact, one MAb, designated RAC23, actually enhanced ricin toxicity in vivo. These data indicate that distinct neutralizing and nonneutralizing B-cell epitopes exist on RTA.
Castelletti and colleagues recently identified a linear B-cell epitope (L161 to I175) on RTA recognized by serum antibodies from 15 Hodgkin's lymphoma patients who had received ricin immunotoxin therapy (7). In two of the serum samples tested, the majority of the antiricin specific antibodies were directed against this epitope. Affinity-purified serum antipeptide IgG was capable of neutralizing ricin in vitro, suggesting that this epitope is an important target of anti-RTA neutralizing antibodies in vivo. However, the possibility that the observed activity was due to a minor contaminating population of antibodies directed against a similar or closely situated epitope cannot be excluded. Moreover, Castelletti and colleagues did not examine whether antibodies directed against L161 to I175 were actually capable of neutralizing ricin in an animal model of ricin intoxication.
In an effort to identify the epitopes on ricin that are the key targets of neutralizing antibodies in vivo with the goal of better understanding vaccine-induced immunity to ricin, we have produced and characterized a murine IgG1 MAb, referred to as GD12, that is specifically directed against the linear B-cell epitope on RTA described by Castelletti and colleagues as being immunodominant in humans. GD12 neutralized ricin with a 50% inhibitory concentration (IC50) of ~0.25 μg/ml, as determined by a Vero cell-based cytotoxicity assay. More importantly, passive administration of GD12 to mice was sufficient to protect the animals against both systemic (i.e., intraperitoneal [i.p.]) and mucosal (i.e., intragastric) ricin challenge, underscoring the importance of the epitope spanning residues 161 to 175 on RTA as a key target of neutralizing antibodies in vivo. However, because these studies focused on vaccine development, rather than on therapeutic applications of GD12, this MAb was not examined for the ability to rescue animals following toxin exposure.
Ricin, RTA, and RTB were purchased from Vector Laboratories (Burlingame, CA). The amount of ricin toxin maintained in the laboratory was less than 100 mg at any one time, and therefore it was exempt from select-agent registration, as defined by the CDC. Ricin toxoid (RT) was produced by treatment of holotoxin with paraformaldehyde (4%, vol/vol), as described previously (26). Ricin and RT were dialyzed against phosphate-buffered saline (PBS) at 4°C in 10,000-molecular-weight-cutoff Slide-A-Lyzer dialysis cassettes (Pierce, Rockford, IL) prior to use in cytotoxicity and mouse challenge studies. Paraformaldehyde (16%) was purchased from Electron Microscopy Sciences (Fort Washington, PA). General protein electrophoresis reagents, including Tween 20, molecular weight markers, and precast polyacrylamide gels, were purchased from Bio-Rad (Torrance, CA). Goat serum was purchased from Gibco-Invitrogen (Carlsbad, CA). Unless noted specifically, all other chemicals were obtained from the Sigma Company (St. Louis, MO). Vero cells, irradiated MRC-5 human lung fibroblast cells, and the murine myeloma cell line P3X63.Ag8.653 were purchased from the American Type Culture Collection (Manassas, VA). Cell culture media were prepared by the Wadsworth Center medium facility. Cell lines and hybridomas were maintained in a humidified incubator at 37°C under 5% CO2.
The studies described here utilized female BALB/c mice approximately 6 to 8 weeks of age, purchased from Taconic Labs (Hudson, NY). Animals were housed under conventional, specific-pathogen-free conditions and were treated in compliance with the Wadsworth Center's Institutional Animal Care and Use Committee (IACUC) guidelines.
Female BALB/c mice were primed with RT (250 μg) administered i.p. on day zero and were then boosted with RT (100 μg) on days 10 and 20. On day 24, the animals were euthanized, and total splenocytes were fused with the myeloma cell line P3X63.Ag8.653 (10, 26). Hybridomas were seeded into wells of 96-well microtiter plates containing a layer of irradiated MRC-5 feeder cells. Hybridomas were cultured in a 1:1 mixture of NCTC (Sigma Co.) and RPMI media containing 10% fetal calf serum and penicillin-streptomycin, and occasionally supplemented with 1% Opti-MAb (Gibco-Invitrogen). Hybridoma GD12 was cloned by limiting dilution and was transitioned to a serum-free, protein-free, antibiotic-free medium (CD Hybridoma; Gibco-Invitrogen). For cell cytotoxicity assays and mouse challenge studies, GD12 was prepared as a sterile-filtered, serum-free, protein-free hybridoma supernatant. MAbs R70 and TFTB-1 were prepared as described previously (27).
Nunc Maxisorb F96 microtiter plates (Thermo Fisher Scientific) were coated with ricin, RTA, RTB, or bovine serum albumin (BSA) (0.1 μg/well) in PBS (pH 7.4) and were incubated overnight at 4°C in a humidified chamber; they were then washed three times with PBS-Tween 20 (PBS-T; 0.05%, vol/vol) and were blocked with goat serum (2%, vol/vol, in PBS-T) for 1 h at room temperature before being probed with MAbs or hybridoma supernatants diluted in blocking solution. For the enzyme-linked immunosorbent assays (ELISAs), horseradish peroxidase (HRP)-labeled goat anti-mouse IgG-specific polyclonal antibodies (SouthernBiotech) were used as the secondary reagents, and 3,3′,5,5′-tetramethylbenzidine (TMB; Kirkegaard & Perry Labs, Gaithersburg, MD) was used as the colorimetric detection substrate. ELISA plates were analyzed with a SpectroMax 250 spectrophotometer, with Softmax Pro software (version 5.2; Molecular Devices, Sunnyvale, CA).
The RTA peptide array consisted of 44 12-mers, each overlapping its neighbors by 6 amino acids, collectively spanning the RTA sequence (Table (Table1).1). The rationale for using relatively short peptides for this study was based on our recent success in identifying linear B-cell epitopes on anthrax toxin using a library of peptides of this length (15). The peptides were synthesized, unbound, in 96 individual tubes, in a 96-well plate format, and were provided at a 2.5-μmol scale (1.8 mg per peptide, average) at >75% purity (New England Peptide, Gardner, MA). The peptides were solubilized in dimethyl sulfoxide, and aliquots were stored at −20°C. RTA peptide arrays were prepared by coating the wells of microtiter plates with individual peptides (1 μg/well), followed by overnight incubation at 4°C. The plates were washed three times with 0.025% Tween 20 in PBS, blocked with 1% goat serum for 1 h, and then incubated with individual MAbs for 4 h, all at 4°C. The plates were developed as described above for the ELISAs.
For denaturing conditions, ricin and RTA were diluted into Laemmli sample buffer containing 5% (vol/vol) β-mercaptoethanol; then they were boiled for 6 min prior to being subjected to sodium dodecyl sulfate (SDS)-12% polyacrylamide gel electrophoresis (PAGE). For nondenaturing conditions, ricin and RTA were simply diluted into Laemmli sample buffer (without β-mercaptoethanol) prior to SDS-PAGE. For Western blot analysis, proteins were transferred from the polyacrylamide gels to nitrocellulose membranes (pore size, 0.45 μm; Bio-Rad) by semidry electroelution. Nitrocellulose membranes were blocked with 2% (wt/vol) goat serum in PBS-Tween and were then incubated with GD12 (1 μg/ml) overnight at 4°C. Membranes were probed with goat anti-mouse IgG conjugated to HRP (0.4 μg/ml) (SouthernBiotech), developed using an enhanced chemiluminescent detection (ECL) kit (Bio-Rad), and then exposed to Kodak X-Omat film (Thermo Fisher Scientific).
Vero cells were cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal calf serum and were maintained in a humidified incubator (37°C, 5% CO2). For cytotoxicity assays, the cells were trypsinized, adjusted to approximately 0.5 × 105 to 1.0 × 105 cells per ml, seeded (100 μl/well) onto white 96-well plates (Corning), and allowed to adhere overnight. Vero cells were then treated with either ricin (10 ng/ml), ricin-MAb mixtures, or medium alone (negative control) for 2 h at 37°C. The cells were washed to remove noninternalized toxin or toxin-MAb mixtures and were then incubated for 40 h. Cell viability was assessed using CellTiter-Glo reagent (Promega) according to the manufacturer's instructions, except that the reagent was diluted 1:5 in PBS prior to use. Luminescence was measured with a SpectraMax L luminometer, as indicated above. All treatments were performed in triplicate, and 100% viability was defined as the average value obtained from wells in which cells were treated with medium only.
Systemic and mucosal ricin challenge studies were conducted as described previously (23, 34, 53). For systemic challenge studies, female BALB/c mice were injected i.p. with ricin (50 μg/kg) diluted in PBS (final volume, 0.4 ml). Thereafter, the animals were allowed food and water ad libitum. Hypoglycemia was used as a surrogate marker of intoxication (34). Blood (<5 μl) was collected from the tail veins of the animals at 18- to 24-h intervals. Blood glucose levels were measured with an Accu-Chek Aviva handheld blood glucose meter (Roche, Indianapolis, IN). In compliance with the end point specified by the Wadsworth Center's IACUC, mice were euthanized when they became overtly moribund and/or when blood glucose levels fell below 25 mg/dl. For statistical purposes, readings at or below the meter's limit of detection of ~12 mg/dl were set to that value.
Mucosal challenge studies were conducted as previously described (53). Ricin (5 mg/kg) was diluted into PBS and administered i.g. to female BALB/c mice by means of a 22-gauge, 1.5-in blunt-end feeding needle (Popper Scientific, New Hyde Park, NY). Animals were fasted for ~2 h prior to challenge and were then provided with food ad libitum ~1 h after challenge. Twenty-four hours later, the animals were euthanized by CO2 asphyxiation. Segments of the proximal small intestine were immersed in ice-cold cell lysis buffer (Cell Signaling, Beverly, MA) supplemented with protease inhibitors and were then homogenized on ice using a Tekmar Tissumizer (Thermo Fisher Scientific). Monocyte chemotactic protein 1 (MCP-1) levels in intestinal homogenates were determined by the BD cytometric bead array (CBA) flex set (BD Biosciences, San Jose, CA), as described previously (53).
For systemic challenge studies, individual MAbs were diluted into endotoxin-free PBS, and each MAb was administered in a final volume of 0.4 ml to BALB/c mice by i.p. injection. Mice were challenged with ricin by i.p. injection 24 h later. For mucosal challenge studies, MAbs were passively administered to mice by the “backpack tumor” method (5, 28, 33, 41). For the backpack tumor studies, ~2 × 106 hybridoma cells secreting the desired MAb (e.g., GD12, TFTB-1) were implanted subcutaneously into the backs of mice, using a 1-ml syringe and a 25-gauge needle. Approximately 10 days later, blood was collected from the mice by tail bleed, and serum MAb concentrations were determined by ELISA. The animals were challenged with ricin 24 h later, as described above.
Ricin (5 ng) was incubated with individual MAbs (50 to 300 ng) for 10 min at room temperature, and the mixture was then added to a cell-free reticulocyte lysate mixture containing ribosomes, amino acids, and ATP (Retic Lysate IVT; Ambion, Austin, TX). The cocktail was incubated at 25°C for 20 min before the addition of uncapped in vitro-transcribed template RNA (1 μg) encoding firefly luciferase (Luc) or Xenopus elongation factor 1 (XeF-1), each obtained from Promega (Madison, WI). The xef-1 template was used as a negative-control mRNA, since the translated XeF-1 protein is incapable of cleaving the luciferin substrate. The samples (25 μl) were incubated in microcentrifuge tubes for ~3 h in a 30°C water bath and were then transferred to white 96-well plates (Corning, Lowell, MA) before the addition of Bright-Glo luciferin substrate (Promega). Luminescence was measured with a SpectraMax L luminometer interfaced with SoftMax Pro software (version 5.2; Molecular Devices). All experiments were performed in triplicate.
Statistical analysis was carried out with Excel 2003 (Microsoft, Redmond, WA) and SigmaStat, version 3.5 (Systat Software, San Jose, CA). Passive protection studies were analyzed by one-way analysis of variance (ANOVA) and Tukey's posthoc tests. All other data were analyzed using Student t tests, unless indicated otherwise. The open-source molecular visualization software PyMol (DeLano Scientific LLC, Palo Alto, CA), accessed at www.pymol.org, was used for epitope modeling.
We sought to produce a MAb directed against the immunodominant linear epitope on RTA that had been identified by Castelletti and colleagues as a potential target of neutralizing serum IgG antibodies in humans who had received ricin immunotoxin therapy (7). Toward this end, we immunized groups of BALB/c mice with RT and screened their sera for antibodies capable of reacting with a peptide spanning RTA residues L163 to I174 (TLARSFIICIQM). The sera from the majority of RT-immunized animals reacted with this peptide (data not shown), indicating that this epitope is recognized by the BALB/c strain of mice. Sera from RiVax-immunized animals also demonstrated reactivity against this peptide, although less strongly than the RT serum samples (data not shown). B-cell hybridomas were therefore produced from the spleens of RT-immunized animals and were then screened by ELISA for those secreting antibodies reactive against both RTA and the L163-to-I174 peptide. Hybridoma GD12 was identified as producing an IgG1 having these properties and was therefore chosen for further study.
For confirmation of the epitope specificity of MAb GD12, a hybridoma supernatant was used to probe an RTA peptide array consisting of 44 overlapping 12-mers that collectively span the length of the RTA sequence (Table (Table1).1). GD12 preferentially reacted with a peptide (“F11”) corresponding to residues L163 to I174 (Fig. (Fig.1).1). It should be noted that despite numerous attempts, synthesis of the original peptide described by Castelletti and colleagues (i.e., L161 to I175) (7) proved unsuccessful, and we were therefore unable to determine the reactivity of GD12 with this specific epitope.
For comparative purposes, MAb R70 was also subjected to peptide array analysis. R70 (also known as UNIVAX 70/138) is the most potent neutralizing MAb identified to date and one of the few MAbs shown to be capable of protecting mice against a lethal dose of ricin (20, 27). R70 is proposed to recognize a linear epitope within a 26-amino-acid loop-helix-loop motif (Y91 to T116) on RTA, although the precise epitope has not been identified (19). Peptide array analysis revealed that R70 bound exclusively to a single peptide (“E12”), corresponding to residues N97 to F108 (NQEDAEAITHLF) on RTA (Fig. (Fig.1).1). The locations of the epitopes recognized by GD12 and R70 were modeled on the structure of RTA using PyMol software. The resulting image revealed that the MAbs recognized spatially distinct, solvent-exposed regions on the tertiary structure of ricin that are roughly equidistant from the active site of the toxin (Fig. 1A and B).
To further validate the specificity of GD12, we examined the reactivity of GD12 with ricin holotoxin, RTA, and RTB, and we then compared these profiles to those obtained previously with MAb R70 (27). As expected, GD12 bound to ricin holotoxin and RTA but not to RTB (Fig. (Fig.2A).2A). This profile was similar to that observed for R70, except that GD12 reacted slightly less well with RTA than with holotoxin (Fig. 2A and B). By Western blot analysis, GD12 recognized the reduced and nonreduced forms of RTA (Fig. (Fig.2C),2C), a result consistent with the MAb binding to a linear (or continuous) epitope on RTA. The actual affinity of GD12 for ricin was determined by BIAcore analysis. This analysis revealed that GD12 had a dissociation constant (KD) for ricin of approximately 2.9 × 10−9 M (Table (Table2),2), which is virtually identical to the value we reported for R70 (27).
We used a Vero cell cytotoxicity assay to assess the capacity of GD12 to neutralize ricin in vitro. Ricin (10 ng/ml) was incubated for 1 h with GD12 or R70 at a range of concentrations, and the mixture was then applied in triplicate to Vero cells grown in 96-well microtiter plates (see Materials and Methods). The viability of the Vero cells was determined 40 h later. GD12 protected Vero cells from the cytotoxic effects of ricin in a dose-dependent manner and had an estimated IC50 of ~0.25 μg/ml (Fig. (Fig.3).3). GD12 was approximately twice as effective at neutralizing ricin as R70, which had an IC50 of ~0.5 μg/ml. These data demonstrate that an antibody against the linear epitope encompassing L163 to I174 of RTA is capable of neutralizing ricin in vitro.
Because GD12 and R70 recognize different epitopes on RTA, we wanted to test whether a combination of the two MAbs would be more effective than either of the individual MAbs singly at neutralizing ricin in vitro. To assess this, we tested equivalent concentrations of GD12, R70, and a 1:1 mixture of GD12 and R70 in a Vero cell cytotoxicity assay. The mixture of GD12 and R70 did not result in enhanced neutralization activity relative to the activities of the MAbs tested individually (data not shown), indicating that GD12 and R70 do not act additively or synergistically.
We used an established mouse model of systemic ricin intoxication to assess the capacity of GD12 to neutralize ricin in vivo (23, 34). For these challenge studies, hypoglycemia was used as a surrogate marker of intoxication (34). MAb GD12 or R70 was passively administered to female BALB/c mice (5 to 40 μg/animal) by i.p. injection. Twenty-four hours later, the animals were challenged by i.p. injection with the equivalent of five times the LD50 of ricin (50 μg/kg), and blood glucose levels were subsequently assessed at 24-h intervals. Ricin-challenged control mice experienced a dramatic decline in blood glucose levels within 30 h and subsequently died or were euthanized (Table (Table3).3). Mice that received GD12 or R70 all survived ricin challenge. At early time points (24 and 48 h), the mean blood glucose levels of ricin-challenged, GD12- or R70-treated mice were not statistically different from each other. However, at 72 to 77 h postchallenge, mice treated with low doses of R70 (5.0 to 10 μg per animal) had blood glucose levels significantly lower than those of the control group (Table (Table3),3), suggesting that only partial protection, and/or delayed recovery, was achieved. In contrast, at the same time point, the mean blood glucose levels of animals treated with GD12 at any of the four doses were not statistically different from those of control animals, suggesting that GD12 confers full protection under these conditions.
Ricin is highly toxic to mucosal tissues, including both the respiratory and intestinal tracts (45, 53). It was therefore of interest to determine whether GD12 was sufficient to confer mucosal immunity to this toxin. We have previously established an in vivo model of intestinal intoxication in which mice are challenged by gavage with ricin toxin (5 mg/kg) and, 24 h later, MCP-1 levels in the animals' proximal small intestines are assessed as a surrogate marker of tissue damage and inflammation (53). While we speculated that secretory IgA (SIgA) is necessary to confer intestinal immunity to ricin (26, 53), we have recently shown that serum antitoxin IgG antibodies are in fact sufficient to confer protection in this oral challenge model (L. Neal, E. McCarthy, C. McGuinness, and N. Mantis, unpublished data). To assess the capacity of GD12 to confer intestinal immunity to ricin, the MAb was passively administered to mice by the “backpack tumor” method, and animals were challenged with ricin 10 to 12 days later. The backpack tumor model is a widely used and well-established method to assess the efficacy of an individual MAb at conferring immunity to mucosal pathogens and enterotoxins (1, 5, 13, 28, 33, 41). Thus, groups of mice were implanted with GD12 or TFTB-1 hybridoma “backpacks,” as described in Materials and Methods. TFTB-1 is an IgG1 MAb that binds ricin but lacks demonstrable neutralizing activity (27). Approximately 10 days later, blood was collected from the mice by tail bleed, and serum MAb concentrations were determined by ELISA. Serum TFTB-1 concentrations in these animals ranged from 1.0 to 216 μg/ml, whereas serum GD12 concentrations ranged from 1.7 to 1,022 μg/ml. Twenty-four hours following blood collection, the animals were challenged with ricin.
Prior to performing mucosal challenges, we first wanted to confirm that administration of GD12 by the backpack tumor method confers systemic immunity to ricin, as we observed when the MAb was administered by i.p. injection (Table (Table3).3). Therefore, groups of GD12 backpack mice were challenged with the toxin by i.p. injection. GD12 hybridoma backpack mice showed no outward signs of ricin intoxication and had virtually normal blood glucose levels 48 to 54 h after ricin challenge (Fig. (Fig.4A).4A). In contrast, mice bearing TFTB-1 backpacks and injected with ricin showed severe signs of intoxication (e.g., ruffled fur, hunched posture, lethargy, labored breathing) and experienced rapid and dramatic declines in blood glucose levels (from >100 mg/dl to <40 mg/dl) within 24 h (Fig. (Fig.4A).4A). These animals were indistinguishable from the control, non-backpack-bearing animals that had been challenged with ricin, and they were euthanized 30 h postchallenge.
To assess mucosal immunity to ricin, groups of control or hybridoma backpack-bearing mice were challenged with ricin by the i.g. route. In unchallenged animals, intestinal MCP-1 levels were approximately 30 pg/ml, whereas the levels in the intestines of ricin-challenged control (data not shown) or TFTB-1-treated mice were greater than 100 pg/ml (Fig. (Fig.4B).4B). MCP-1 levels in the intestinal homogenates of ricin-challenged, GD12 hybridoma backpack-bearing mice were indistinguishable from those observed in control, unchallenged mice. These data demonstrate that GD12 is sufficient to confer mucosal immunity to ricin, at least in a mouse i.g. challenge model.
It has been postulated that one mechanism by which anti-RTA MAbs may neutralize ricin is by interfering with the capacity of the toxin to arrest protein synthesis (23). We expected that both GD12 and R70 might affect the activity of RTA, because both MAbs bind epitopes located in close proximity to the active site (Fig. 1A and B). We used a cell-free in vitro translation assay to examine this hypothesis (see Materials and Methods). GD12 or R70 was incubated with ricin holotoxin or RTA for 10 min at room temperature and was then added to a rabbit reticulocyte lysate mixture in which luciferase (luc) mRNA was provided as the template. Luminescence was used as an indicator of in vitro translation and protein synthesis. In control reactions in which luc mRNA was combined with the cell-free in vitro translation mixture in the absence of ricin and MAbs, robust luciferase activity was detected (Fig. (Fig.5A).5A). This activity was dependent on luc mRNA, as evidenced by the lack of detectable luciferase activity when xef-1 mRNA was used as an irrelevant template control. The addition of ricin reduced luc-dependent luciferase activity by more than 5 log units, demonstrating the effect of the toxin on ribosome function. The addition of GD12 inhibited the enzymatic activity of ricin in a dose-dependent manner, although even at the highest concentration of MAb tested, only partial inhibition of activity was observed (Fig. (Fig.5A).5A). R70 also interfered with the enzymatic activity of ricin, but less effectively than GD12 (Fig. (Fig.5B).5B). Two MAbs directed against RTB, 24B11 and TFTB-1, had no affect on the enzymatic activity of the toxin (Fig. (Fig.5B).5B). These data demonstrate that in an in vitro cell-free system, both GD12 and R70 can interfere, at least partially, with the enzymatic activity of ricin.
The development and licensure of vaccines for biodefense is inherently challenging, because human efficacy trials of candidate vaccines are generally not feasible or ethical. Recognizing this roadblock, the FDA has implemented the two-animal rule, by which candidate vaccine development can proceed based on efficacy studies of two or more relevant animal models. For an animal model to be deemed relevant, the underlying pathogenesis and mechanisms of immunity in response to the select agent in question must be similar to those in humans. In addition, the animal model(s) must take into account the possibility that toxin or pathogen exposure can occur by multiple routes, such as injection, inhalation, and ingestion. In such situations, vaccines may need to elicit both systemic and mucosal immunity in order to be considered useful. Finally, whenever possible, identification of the underlying correlates of protection, as well as the development of surrogate markers of immunity, should be validated in humans.
The fact that RTA has been under investigation for more than 2 decades as a possible immunotoxin for cancer therapy has yielded some basic information regarding the human immune response to ricin (7, 50). Most notable is the recent report by Castelletti and colleagues, who identified an immunodominant linear B-cell epitope on RTA (residues 161 to 175) that was recognized by serum antibodies from 15 Hodgkin's lymphoma patients who had received RTA immunotoxin (7). While that study demonstrated that affinity-purified antipeptide antibodies were capable of neutralizing ricin in vitro, it did not determine whether the antibodies were capable of neutralizing ricin in vivo. This is not an academic point, given that Maddaloni and colleagues recently described several MAbs against RTA that were highly effective at neutralizing ricin in vitro yet afforded no protection in a mouse model of ricin challenge (23). In an effort to resolve this issue, we have produced and characterized GD12, a murine IgG1 MAb directed against virtually the same epitope on RTA as that described by Castelletti and colleagues. We found that GD12 bound the holotoxin with high affinity and neutralized ricin with an IC50 of ~0.25 μg/ml, as determined by a Vero cell-based cytotoxicity assay. Most importantly, we found that this MAb, when administered passively to mice, was sufficient to protect the animals against both systemic (i.e., intraperitoneal) and mucosal (i.e., intragastric) ricin challenge. These data, when considered in the context of the work by Castelletti and colleagues, establish the importance of residues 161 to 175 on RTA as a target of ricin-neutralizing antibodies in vitro and in vivo.
RTA is composed of three distinct subdomains, two of which are now known to be the targets of neutralizing MAbs (14). Subdomain 1 spans residues 1 to 117; subdomain 2 spans residues 118 to 210; and subdomain spans 3 residues 211 to 276. The epitope recognized by GD12 is localized entirely within subdomain 2. Specifically, the GD12 epitope maps to a long α-helix that forms the core of the subdomain. R70, on the other hand, recognizes subdomain 1 (19), although prior to this study the precise epitope recognized by this MAb was not known. We determined by peptide array analysis that R70 binds exclusively to a single 12-mer consisting of residues N97 to F108 (NQEDAEAITHLF). This epitope, like that recognized by GD12, is situated within an α-helix (14). There is evidence to suggest that antibodies directed against subdomain 3 may be poor neutralizers, or even potentiators, of ricin toxicity. For example, Maddaloni and colleagues identified several anti-RTA MAbs capable of enhancing the potency of ricin in vitro and in vivo. One of these MAbs, RAC23, is postulated to bind residues within subdomain 3 (23). In addition, we have identified several MAbs directed against linear epitopes on subdomain 3, and at least preliminarily, these MAbs appear unable to neutralize ricin in vitro (J. O'Hara and N. Mantis, unpublished data). Systematic identification of the B-cell epitopes on RTA that elicit both neutralizing and nonneutralizing antibodies may have immediate implications for vaccine design.
While this study clearly demonstrated the potential of GD12 to neutralize ricin in vitro and in vivo, the mechanism by which this is achieved remains unresolved. The epitope recognized by GD12 is immediately adjacent to two residues (E177 and R180) on RTA known to be involved in rRNA catalysis (17). For this reason, we suspected that the association of GD12 with RTA may interfere with the subunit's enzymatic activity, possibly by blocking access to the rRNA substrate or by physically distorting the active site. Indeed, GD12 was observed to reduce the capacity of RTA to inhibit protein synthesis in an in vitro cell-free assay. However, the relevance of this observation to the capacity of GD12 to neutralize ricin in whole cells or animals remains unclear. RTA interacts with its substrate only after a circuitous journey of the holotoxin from the plasma membrane. Specifically, following endocytosis, ricin undergoes vesicular retrograde transport to the ER. RTA and RTB dissociate within the ER, after which RTA is unfolded and delivered across the ER membrane by the Sec61 complex, a process known as retrotranslocation (21, 36, 39). If GD12 acts through interference with the enzymatic activity of RTA, then the MAb would have to remain associated with RTA throughout the retrograde and retrotranslocation processes. Alternatively, based on several recent reports documenting the mechanism by which MAbs neutralize Shiga toxins (Stx) (18, 46), we speculate that GD12 (and other anti-RTA MAbs) may interfere with the intracellular trafficking of ricin. For example, Krautz-Peterson and colleagues recently demonstrated that a MAb directed against the A subunit of Stx1 blocked the retrograde transport of the toxin into the Golgi apparatus and the endoplasmic reticulum (18).
As a biothreat agent, ricin can potentially be disseminated as an aerosol, or it could be used to contaminate food and water supplies. By either route, the toxin would first contact a mucosal surface in the host, an important consideration when a ricin vaccine is being developed and evaluated. Studies of animal models have established that both the respiratory and intestinal epithelia are sensitive to ricin intoxication (45, 53). In the gut, protection of the intestinal epithelium from enterotoxins is a function generally ascribed to SIgA (22, 51). However, we have shown in a separate study that serum antiricin IgG antibodies are themselves sufficient to protect the intestinal epithelium from ricin intoxication, at least in a mouse model (L. Neal et al., unpublished). In the present study, we have demonstrated, using the backpack tumor model, that GD12 was sufficient to protect the intestinal mucosa from ricin. While the underlying mechanism by which serum IgG confers mucosal immunity to ricin remains to be elucidated, this finding has potentially important implications for vaccine delivery, in that it suggests that parenteral immunization with a ricin subunit vaccine may be sufficient to elicit immunity in both systemic and mucosal compartments.
The results of a recent pilot phase I clinical trial of RiVax underscore the need to identify additional correlates of immunity to ricin (52). Of particular concern was the fact that total anti-RTA titers, as determined by ELISA, were poor predictors of the ability of individual serum samples from RiVax-immunized patients to neutralize ricin. While antitoxin titers in themselves provide a convenient means by which seroconversion can be assessed in a human cohort, they do not necessarily provide useful information regarding toxin-neutralizing activity. The only means, at present, by which to assess neutralizing activity is a cell-based cytotoxicity assay or passive protection studies of mice. Unfortunately, those types of assays are relatively insensitive and require significant amounts of serum. We propose, based in part on the results of the current study, that it may be possible to predict an individual's immune status relative to ricin from serum antibody reactivity against specific RTA-derived linear B-cell epitopes. For example, in preliminary studies, we have observed a significant correlation between serum antibody reactivity with three distinct linear epitopes of RTA (including L163 to I174) in mice and protection against systemic lethal toxin challenge (J. O'Hara, L. Neal, and N. Mantis, unpublished data). Microarray profiling of serum antibodies is being applied to the development and evaluation of effective diagnostics, therapeutics, and vaccines in the areas of cancer and emerging infectious diseases (29, 30). We propose that such technology is ideally suited to the field of biodefense, in which the ultimate licensure of vaccines must be achieved in the absence of clinical efficacy trials.
We thank the following individuals at the Wadsworth Center for assistance with this project: Helen Johnson (Animal Pathology Core) for tissue processing, Jane Kasten-Jolly (Immunology Core) for BIAcore analysis, and Karen Chave (Protein Expression Core) for monoclonal antibody purification.
This work was supported by grants from the National Institutes of Health (U01AI070624; to R.N.B.) and the Northeast Biodefense Center (U54-AI057158-Lipkin; to N.J.M.).
Editor: S. R. Blanke
Published ahead of print on 26 October 2009.