**3. Discussion**

In the present study, a neutralized monomer-based ricin vaccine was generated by reduction and alkylation of the disulfide bond linking RTA to RTB, in order to immunize a horse for the production of highly potent neutralizing anti-ricin antibodies.

After demonstrating in rabbits that immunization with monomerized ricin was as effective as immunization with native ricin in elicitation of anti-ricin neutralizing antibodies, a thorough characterization of the antigen was conducted. Previous studies demonstrated that reduction of disulfide bonds might abolish the capability to produce biologically active antibodies, as in the case of *Plasmodium falciparum* merozoite surface glycoprotein (gp195). Reduction and alkylation of gp195 triggered an inappropriate folding of the unbound subunits, resulting in a drastic conformational change and significantly altered antigenicity [24]. Thus it was essential to demonstrate that monomerized ricin, which would be repeatedly administered to the horse over a considerably long period of time, retains its structural activity thereby implying proper folding and potential antigenicity of the irreversibly-separated subunits. Indeed, the monomeric alkylated-RTB bound firmly to an α-lactose agarose column, and eluted only following 0.5 M galactose addition, confirming RTB lectin activity. The catalytic activity of the alkylated-RTA, namely protein synthesis arrest in the TnT assay, was not only retained but even was superior to that of native RTA. This is probably due to a limited, or non-complete, separation of the holotoxin subunits of native ricin following reduction, and the relative proximity between the reduced native subunits leading to sporadic reconstitution, in contrast to the complete, irreversible, separation of the monomerized-ricin subunits.

Although the alkylated-subunits were fully active in cell free systems, a >99.5% reduction in alkylated-ricin cytotoxicity was determined in a cell culture, in which case the binding of the monomeric B subunit to the cell surface cannot promote internalization of the detached A subunit. This dramatic change in ricin-induced cytotoxicity, together with an in vivo test in mice, in which no body weight loss nor any noticeable side effect were observed, allowed the safe usage of the monomerized ricin for immunizing a horse.

Following immunization, high and stable ELISA and neutralizing antibody titers, were reached, and F(ab')2-based anti-ricin antitoxin was produced from the hyperimmune horse plasma. It was previously claimed [5], that fractionation of anti-ricin IgG antibodies to F(ab')2 could severely affect the neutralizing capabilities of the latter in vivo. Preliminary experiments conducted by us (data not shown) have demonstrated that the F(ab')2 fragment possess the same potency as its IgG precursor both in vitro and in vivo. Indeed, the F(ab')2-based antitoxin was found highly potent. The ED50 of the antitoxin in cell culture (~0.7 nM) was comparable to the ED50 of rabbit derived polyclonal anti-ricin IgG fraction [4]. These neutralizing capabilities at the nanomolar range are a property of highly efficient antibodies. In line with this, the antitoxin was found highly protective in mice, following administration at 24 h post intranasal intoxication with a lethal challenge of ricin. The survival ratios were >60% (whether the antitoxin was administered intranasally or intravenously), much higher than those obtained with the previously characterized rabbit-derived antibodies (~35%) [20,22]. Furthermore, high survival rates (34%) were obtained at a very late treatment time point, namely 48 h following intoxication. In sharp contrast to this, negligible surviving ratios (4%) were obtained when mice were treated with rabbit-derived anti-ricin antibody at 48 h post exposure [21]. In fact, the survival percentages of intoxicated mice treated at 24 h post exposure (PE) with horse-derived antitoxin were similar to those obtained following a combinational treatment with rabbit-derived anti-ricin antibody and immunomodulatory drugs [20,22], while the survival ratios following treatment at 48 h with horse antitoxin, were comparable to those obtained for mice treated with rabbit antibody at 24 h PE. This difference in protection is also reflected in the fact that treatment with the horse-derived F(ab')2-based antitoxin induces a significant reduction in inflammatory parameters in comparison to rabbit-derived antibody-based treatment. Thus, while administration of horse antitoxin in itself induced a sharp attenuation in IL-6 levels (~90% reduction) at 72 h PE, rabbit antibody treatment had virtually no effect on IL-6 levels [25]. An additional outstanding effect of the horse antitoxin-based treatment shown in the present study was a significant reduction (by ~60%) of XO levels in the BALF of ricin-intoxicated mice. XO is an important source of reactive oxygen species (ROS) formation [26,27], which contributes significantly to pulmonary edema formation in diverse lung pathologies [28,29]. In contrast, we found that rabbit antibody-based treatment was not effective in reducing the levels of XO (data not shown). Edema markers were also dramatically reduced following treatment with the horse antitoxin (60% reduction in BALF protein content and 75% reduction in ChE levels at 72 h PE), in contrast to insignificant changes following treatment with rabbit antibody [25]. Indeed, when we previously abolished neutrophil-induced lung injury via total body irradiation (TBI), treatment at 48 h with rabbit antibody resulted in 42% survival rates, in parallel to sharp attenuation of IL-6 and edema marker levels [21].

In summary, in the present study we demonstrate that neutralized ricin antigen obtained by toxin monomerization and alkylation is completely safe for immunization, and induces high titers of potent ricin neutralizing antibodies that can be effectively used for passive immunization. F(ab')2-based ricin antitoxin produced from the hyperimmune plasma of a horse that was immunized with monomerized ricin, conferred high level protection following pulmonary intoxication, which is the most fatal exposure route for ricin intoxication. This horse antitoxin-based treatment displayed effective anti-edematous and anti-inflammatory activities, apparently via balancing cytokine levels toward inflammatory attenuation in the lungs of intoxicated animals. We note that residual damage marker levels in the F(ab')2-based ricin antitoxin-treated mice, nevertheless remained higher at 72 h following intoxication in comparison to the levels measured in naïve mice, suggesting that survival ratios may be further improved by combinational treatment with immunomodulatory drugs.
