Cooperative binding of anti-tetanus toxin monoclonal antibodies: Implications for designing an efficient biclonal preparation to prevent tetanus toxin intoxication

Vaccine 36 (2018) 3764–3771

Contents lists available at ScienceDirect

Vaccine journal homepage: www.elsevier.com/locate/vaccine

Cooperative binding of anti-tetanus toxin monoclonal antibodies: Implications for designing an efficient biclonal preparation to prevent tetanus toxin intoxication Ivana Lukic a, Ana Filipovic a, Aleksandra Inic-Kanada b, Emilija Marinkovic a, Radmila Miljkovic a, Marijana Stojanovic a,⇑ a

Department of Research and Development, Institute of Virology, Vaccines and Sera – TORLAK, Vojvode Stepe 458, 11152 Belgrade, Serbia OCUVAC – LBCE, Institute of Specific Prophylaxis and Tropical Medicine, Center for Pathophysiology, Infectiology and Immunology, Medical University of Vienna, Kinderspitalgasse 15, 1090 Vienna, Austria b

a r t i c l e

i n f o

Article history: Received 23 January 2018 Received in revised form 7 May 2018 Accepted 11 May 2018

Keywords: Antibodies Cooperative effect Protection Tetanus toxin

a b s t r a c t Oligoclonal combinations of several monoclonal antibodies (MAbs) are being considered for the treatment of various infectious pathologies. These combinations are less sensitive to antigen structural changes than individual MAbs; at the same time, their characteristics can be more efficiently controlled than those of polyclonal antibodies. The main goal of this study was to evaluate the binding characteristics of six biclonal equimolar preparations (BEP) of tetanus toxin (TeNT)-specific MAbs and to investigate how the MAb combination influences the BEPs’ protective capacity. We show that a combination of TeNTspecific MAbs, which not only bind TeNT but also exert positive cooperative effects, results in a BEP with superior binding characteristics and protective capacity, when compared with the individual component MAbs. Furthermore, we show that a MAb with only partial protective capacity but positive effects on the binding of the other BEP component can be used as a valuable constituent of the BEP. Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction Tetanus is a severe and often fatal disease that can develop after exposure to the tetanus toxin (TeNT),1 a neurotoxin produced by the anaerobic bacterium Clostridium tetani. Mandatory vaccination against tetanus was introduced worldwide and was a crucial measure that led to a significant decline of cases during the second half of the 20th century [1]. Nevertheless, the prevalence of the disease is not negligible, particularly in the developing world [2]. Tetanus cases are mostly reported in elderly patients, as immunity to tetanus disappears gradually over time [3]. In addition, tetanus outbreaks related to injuries seen during natural disasters such as earthquakes and tsunamis have been documented by the WHO [4]. TeNT intoxications can be efficiently treated with various polyclonal antibody (PoAb)2-based therapies [4,5]. The WHO recommends treatment with human immunoglobulin preparations [6]. However, polyclonal preparations of animal origin, consisting of TeNT-binding (F(ab)2) fragments, are still used when human⇑ Corresponding author. 1 2

E-mail address: [email protected] (M. Stojanovic). TeNT – tetanus toxin. PoAbs – polyclonal antibodies.

https://doi.org/10.1016/j.vaccine.2018.05.058 0264-410X/Ó 2018 Elsevier Ltd. All rights reserved.

derived preparations are not available. The manufacture and use of animal-derived therapeutic products has several problems: (i) a long immunization procedure, (ii) batch-to-batch variation in the therapeutic efficacy [7], (iii) potential patient hypersensitivity, and (iv) the risk of acquiring certain zoonosis [8]. In contrast, humanderived products significantly reduce the risk of reactogenicity. They also carry certain difficulties, such as a requirement for intensive pathogen-focused control and a tedious large-scale production process. Currently, monoclonal antibodies (MAbs)3 are considered the reagent of choice for tetanus prevention/treatment [9]. It has already been shown that a high affinity toward TeNT does not automatically result in a protective effect against TeNT intoxication, since high affinity is only one of the requirements [10–12]. To be considered protective, an antibody must not only recognize TeNT with sufficiently high affinity but also prevent the initial step in TeNT intoxication (i.e., the interaction of TeNT with gangliosides exposed on the surface of neurons) [13]. This is in line with recommendations for MAbs, which are expected to provide protection against botulinum toxin, a molecule that closely resembles TeNT, both structurally

3

MAb – monoclonal antibody.

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

and functionally [14]. MAb-based therapeutic products have the advantage over PoAbs of being highly purified and wellcharacterized, and they are generally associated with a lower risk of side effects due to cross-reactivity. However, the homogenous specificity of single MAb-based preparations can also be a problem. Oligoclonal MAb preparations are less sensitive to structural changes in the antigen (conformational, glycosylation-associated, alterations in the primary structure) than the corresponding single MAb-based preparations [15]. In addition, their characteristics can be more efficiently controlled than PoAb preparations [16,17]. Further, the presence of multiple MAbs that can simultaneously bind one antigen allows for better opsonization and, consequently, faster removal of the target antigen [18–20]. When more than one MAb binds simultaneously to a flexible antigen such as a protein, a phenomenon known as a cooperative binding can be observed. Positive cooperative binding is one of the main reasons for the superiority of oligoclonal combinations over single MAb-based preparations, assuming that two MAbs can bind to non-overlapping epitopes of the same antigen [17]. The binding of the first MAb induces a conformational change that affects the epitope of the second MAb, resulting in stronger binding to its paratope. The idea of using well-defined oligoclonal MAbbased preparations instead of single MAbs has already been considered for the treatment of infections caused by Clostridium botulinum [21–23] and Bacillus anthracis [24]. In both cases, better protection was achieved by using oligoclonal preparations than single MAbs. The aims of this study were to evaluate the binding characteristics and protective capacity of TeNT-specific MAb-based biclonal equimolar preparations (BEPs),4 to determine how these binding characteristics correlate with the observed in vivo effects, and to investigate whether these preparations possess any advantages over single MAb-based preparations.

3765

well, 4 °C, overnight), the selected MAbs were biotin-labeled (MAb-B),6 and their saturating concentrations were determined by testing their binding to the TeNT-coated plates at various concentrations. After determining the saturating concentrations, microtiter plates (MaxiSorp, Nunc) were coated with TeNT (1 lg mL 1, 50 lL per well, 4 °C, overnight), blocked with BSA/PBS (1% w/v) for 2 h at a room temperature (RT), and washed with Tween 20/PBS (0.05% v/v, 4  200 lL per well). The MAb-containing samples were then added to the wells (50 lL per well) and incubated for 1 h at RT, followed by another washing step (0.05% Tween 20/ PBS; 4  200 lL per well). The extrAvidin-peroxidase/OPD system (Sigma-Aldrich, Germany) was used for reaction visualization. The reaction was stopped by the addition of 2 M H2SO4 (50 mL per well), and the absorbance was read at 492/620 nm (A492/620). Two types of samples were prepared: 1) samples containing only one MAb at its saturating concentration, and 2) samples containing two out of four selected MAbs, each at their saturating concentrations. The AI for the defined MAb pair was calculated according to the following equation: AI = {[AMAb1+MAb2 (AMAb1 + AMAb2)/2]/(AMAb1 + AMAb2)/2}  100, where AMAb1+MAb2 represents the A492/620 value for the sample containing two MAbs, each at saturating concentration, while AMAb1 and AMAb2 represents the A492/620 value for the samples containing only one MAb at saturating concentration [26]. By convention, an AI value below 20 implies that the two MAbs recognize the same or closely linked epitopes and cannot bind the antigen simultaneously due to steric hindrance. For MAbs with AI values between 20 and 40, simultaneous binding is possible, but a certain degree of steric hindrance may still occur. An AI>40 implies that the two MAbs can bind to the antigen simultaneously without any obstacles [26].

2.3. Cooperative binding of anti-TeNT MAbs 2. Material and methods 2.1. TeNT and anti-TeNT MAbs TeNT was purified using hydrophobic chromatography, as described in our previous work [13]. The source of TeNT was a supernatant obtained after filtration of a C. tetani culture through a 0.2-mm filter, part of the standard TTd vaccine manufacturing process at the Institute of Virology, Vaccine and Sera – Torlak (Belgrade, Serbia). Four murine TeNT-specific MAbs, designated as MAb33, MAb39, MAb51 and MAb71, were selected for this study. The selected MAbs belong to the same IgG subclass, have similar circulatory half-lives, possess sufficient affinity for TeNT to permit its neutralization in solution, and are all capable of providing a certain degree of protection against TeNT intoxication [13,25]. These MAbs were used to prepare six BEPs: MAb33/MAb39, MAb33/MAb51, MAb33/MAb71, MAb39/MAb51, MAb39/MAb71 and MAb51/ MAb71. 2.2. Determination of the additivity indices for selected anti-TeNT MAbs Determination of the additivity indices (AIs)5 for selected MAb pairs was carried out by measuring their simultaneous binding to TeNT by ELISA, as described below. First, MaxiSorp plates (Nunc; Roskilde, Denmark) were coated with TeNT (1 lg mL 1, 50 lL per 4 5

BEP – biclonal equimolar preparation. AI – additivity index.

The mutual influence of selected MAbs after binding to TeNT was evaluated through pair-wise analyses. The MAb being tested (its biotin-labeled form, MAb-B, in concentrations ranging from 0.03 mg mL 1 to 1 mg mL 1) was incubated (1h at RT) with TeNT (0.5 mg mL 1) or with TeNT that had been pre-incubated (1h at RT) with one of the three other selected MAbs (0.5 mg mL 1 TeNT + MAb at 0.5 mg mL 1; TeNT/MAb). After incubation, a saturated solution of ammonium sulfate (pH = 7) was added (up to a final saturation of 40%) to the TeNT/MAb/MAb-B samples, which were incubated overnight at 4 °C. To separate precipitated immunocomplexes, the samples were centrifuged at 20,000  g for 30 min. The concentration of free MAb-B in the supernatant was determined by direct ELISA, as follows: MAb-B-containing supernatants (50 ml per well) were added to MaxiSorp plates (Nunc) and incubated for 1 h at RT. Plates were then blocked with 1% BSA/PBS w/ v (200 ml per well, 2 h at RT) and washed with 0.05% Tween 20/ PBS v/v (4  200 ll per well). Reactions were visualized by means of the extrAvidin-peroxidase/OPD system, and the absorbance was read at 492/620 nm (A492/620). Serial dilutions of the MAb-B (at concentrations ranging from 1 mg mL 1 to 0.03 mg mL 1) were directly adsorbed to microtiter plates after being processed in an identical fashion (blocking with 1% BSA in PBS, washing) to generate a standard curve. The binding data are presented as Klotz plots, with the percentage of bound TeNT plotted against free MAb-B concentrations in logarithmic form. The concentrations of free MAb33-B, MAb39-B, MAb51-B and MAb71-B in solution, when 50% of TeNT was bound (MAb33f,50%, MAb39f,50%, MAb51f,50%, and MAb71f,50%, respectively), 6

MAb-B – biotin-labeled MAb.

3766

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

were calculated after sigmoidal dose-response curve fitting (software GraphPad PRISM 5.0).

Table 1 Additivity indices for MAb33, MAb39, MAb51 and MAb71. BEPa

2.4. Potency of TeNT-specific MAb-based preparations: in vitro determination In vitro determination of the potency of the anti-TeNT MAb(s) preparations was based on their inhibitory effect on the binding of TeNT to GD1b. Potency was defined as the amount of MAb(s) needed to achieve 50% inhibition of TeNT binding to GD1b (IC50) and was calculated relative to the 1st International Standard for Tetanus Immunoglobulin (NIBS code TE-3; NIBS, Potters Bar, UK; anti-TeNT standard). Potency was determined according to a previously described procedure, with slight modifications [13]. Briefly, GD1b (10 mg mL 1 in ethanol, 50 ll per well) was adsorbed onto PolySorp microtiter plates (Nunc) after overnight evaporation at RT. TeNT (at a final concentration of 25 mg mL 1) was incubated in test tubes for 1 h at RT with either selected MAbs or MAb-based BEPs (at final antibody concentrations ranging from 0.05 mg mL 1 to 30 mg mL 1). The binding of TeNT to GD1b was determined by adding equine tetanus antitoxin (ETA; Institute of Virology, Vaccine and Sera – Torlak), followed by biotin-labeled anti-horse IgG (Sigma-Aldrich). ETA was added at a concentration of 0.1 Lf mL 1 (50 ll per well, 1 h, RT). The IC50 values for selected MAbs and MAb-based BEPs were calculated by using as a reference for 100% binding the sample containing 25 mg mL 1 TeNT (without any MAbs added). IC50 values (IC50 sample) were calculated from the plots, which represent the percent inhibition of TeNT binding to GD1b with respect to log MAb(s) concentration after sigmoidal dose-response curve fitting (GraphPad PRISM 5.0 software/nonlinear curve fitting/doseresponse curve, variable slope). The potency of the anti-TeNT standard (IC50 standard) determined with the same procedure was 0.25 ± 0.10 IU mL 1. The relative potencies of the MAbs and their BEPs were calculated as ‘‘IC50 standard/IC50 sample” and were expressed in IU mg 1. In addition, each BEP was tested in the same manner to measure its inhibitory effect on TeNT binding to GD1b, at a final concentration of 10 mg mL 1 (the concentration of each MAb was 5 mg mL 1, so the total MAb concentration was 10 mg mL 1). The calculated inhibitory percentages were compared with those reported for preparations containing only one MAb at concentrations of 10 mg mL 1 [13]. 2.5. Determination of the affinities of MAb-based BEPs Determination of the MAb affinity constants (KA) was performed as described in our previous work [13]. Serial dilutions of MAb-based BEPs were prepared for the analyses (antibody concentrations ranging from 3 mg mL 1 to 30 mg mL 1). 2.6. Animals To determine the therapeutic effectiveness of MAb-based BEPs, we used eight- to ten-week-old outbred Swiss white mice weighing 30 ± 1 g fed sterilized granular diet and water ad libitum. Animals were handled in strict accordance with good animal practices, as defined by the Serbian Code of Practice for the Care and Use of Animals for Scientific Purposes (published in Sluzˇbeni Glasnik No. 41/9), the Guide for the Care and Use of Laboratory Animals of the Institute of Virology, Vaccines and Sera – Torlak (2133/1, 21.04.2011), and the Basel declaration, committed to the 3R principle (replace, reduce, refine). All procedures were carried out in strict accordance with Serbian laws and European regulations on animal welfare and were approved by the committee at the Institute of Virology, Vaccines and Sera – Torlak and by the Ethics Committee for the Welfare of Experimental Animals of the

MAb33/MAb39 MAb33/MAb51 MAb33/MAb71 MAb39/MAb51 MAb39/MAb71 MAb51/MAb71 a b

AIb 60.52 ± 2.01 46.59 ± 1.30 67.88 ± 3.22 52.82 ± 5.78 71.97 ± 5.85 101.23 ± 7.27

BEP – biclonal equimolar preparation. AI – additivity index, mean AI ± SE of three independent measurements.

Republic of Serbia (Approval No. 011-00-00510/2011-05/5). Animal tests were planned and carried out with extreme care, and every effort was made to minimize animal suffering. The experiments were carried out following a double-blind design. Mice were observed daily by trained animal care staff, and animals requiring care were referred to the attending veterinarian for immediate care. Terminal euthanasia was carried out by lethal CO2 overdose. We did not observe any unexpected animal deaths.

2.6.1. In vivo protective capacity of anti-TeNT MAb-based BEPs The protective effect of anti-TeNT MAb-based BEPs was evaluated in vivo. Mice (10 animals per group, six groups in total) were injected intraperitoneally (i.p.) with 0.5 mL of a solution containing 2LD50 TeNT + 10 mg of specific MAb-based BEP in PBS. Positive and negative controls (10 animals per group) were i.p. injected with 0.5 mL of a solution containing either 2LD50 TeNT or PBS, respectively. None of the mice belonging to the positive control group survived the challenge (four days after the TeNT injection, the survival rate was 0%). The survival rate for the age-matched negative control group was 100%, with no visible signs of pathology. All solutions were incubated for 1 h at RT prior to the i.p. injection. The day of the TeNT injection was designated as day 0. The animals were observed daily for the next 16 days. The severity of systemic tetanus symptoms in live mice was graded as follows: 0 – no symptoms of tetanus; 1 – slight stiffness, visible only when the mouse was suspended from its tail; 2 – obvious limping, but limbs were used to walk; 3 – obvious limping, limbs moved but were not functional; and 4 – extensive stiffness, including rigid limbs. Grade 5 corresponded to mice that were found dead. Mice with a score of 4 on the day of observation were euthanized and recorded the next day as pathology score 5.

2.6.2. Therapeutic potential of MAb51/MAb71 Mice were injected i.p. with 0.5 mL of a solution containing 2LD50 TeNT in PBS. Two and six hours later, they were injected i. v. with 10 mg of MAb51/MAb71 BEP in PBS (10 mg MAbs in 100 mL PBS) via the tail vein. There were 10 animals per group (2h and 6 h), plus two control groups (positive controls injected i.p. with 2LD50TeNT in 0.5 mL PBS only, and non-treated agematched negative controls). The severity of the systemic tetanus was monitored and graded as described above.

2.7. Statistical analysis The data are presented as the mean value ± standard error (SE).7 The statistical significance of the differences recorded among experimental groups was evaluated by one-way repeated measures 7

SE – standard error.

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

ANOVA followed by Bonferroni’s multiple comparison test. Correlation between variables was evaluated by Pearson’s bivariate correlation analyses and determination of Pearson’s correlation coefficient (Pcc) (software: IBM SPSS Statistics 20). The statistical significance of the difference in binding of a particular MAb-B to TeNT alone versus TeNT complexed with one of the three other MAbs was determined by Paired-sample t-test. A probability (P) value of 0.05 was taken as the limit of significance.

3767

3. Results 3.1. Anti-TeNT MAbs can bind simultaneously to different TeNT epitopes Pair-wise determination of the AIs for selected MAbs was performed to evaluate whether they could simultaneously bind to TeNT. The AI values for all specified MAbs pairs were above 40

Fig. 1. Pair-wise evaluation of the cooperative binding of MAb33, MAb39, MAb51 and MAb71. Each MAb was assessed for binding to TeNT alone or to TeNT preincubated with one of the other three selected MAbs in solution. The percentage of bound TeNT was plotted against the logarithmic concentration of free MAb-B. The concentration of free MAb33-B, MAb39-B, MAb51-B and MAb71-B in solution, when 50% of TeNT was bound to a particular MAb-B (MAb33f,50%, MAb39f,50%, MAb51f,50%, and MAb71f,50%, respectively), was calculated after sigmoidal dose-response curve fitting (13–15 points used for fitting) and is presented in the table. The data used for fitting originated in three independently performed experiments where each sample was assessed in triplicate (each dot represents the mean logarithmic concentration of free MAb-B vs. the mean percentage of bound TeNT). The statistical significance of the differences in binding of a given MAb-B to TeNT and to TeNT complexed with one of the other three MAbs was determined by Paired-sample t-tests, where the percentages of TeNT bound in the presence of the same total concentration of a particular MAb were compared (*P < 0.05, ** P < 0.005, ***P < 0.001).

3768

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

Table 2 Capability to prevent TeNT-GD1b binding, affinity for TeNT and potency calculated for BEPs consisted of specified MAbs. BEPa MAb33/MAb39 MAb33/MAb51 MAb33/MAb71 MAb39/MAb51 MAb39/MAb71 MAb51/MAb71

Inhibition of TeNT binding to GD1b (%)b,e 79.4 ± 4.0 92.6 ± 0.9 72.2 ± 4.5 65.1 ± 3.6 55.6 ± 4.1 93.2 ± 1.8

Affinity(108 M

1 c,e

In vitro potencyd,e (IU mg

)

5.4 ± 0.7 (R > 0.92) 21.9 ± 2.8 (R > 0.91) 8.8 ± 1.4 (R > 0.86) 5.8 ± 0.6 (R > 0.98) 8.6 ± 0.9 (R > 0.94) 32.3 ± 2.1 (R > 0.98)

1

MAb)

145.6 ± 5.9 407.3 ± 17.6 223.1 ± 12.4 140.2 ± 10.9 115.5 ± 14.4 695.3 ± 15.5

a

BEP – biclonal equimolar mixture. Final concentration of each MAb was adjusted to 5 mg mL 1 with a total MAb concentration of 10 mg mL 1. R values for linearization are indicated in parenthesis. d In vitro determined potencies for each individual MAb under the same conditions were as follows: 133.52 ± 0.72 IU mg 358.17 ± 1.50 IU mg 1 for MAb51, 104.06 ± 0.83 IU mg 1 for MAb71. e Results are presented as mean value ± SE of three independent measurements. b

c

1

for MAb33, 122.31 ± 0.56 IU mg

1

for MAb39,

Table 3 Survival rate, severity of tetanus and recovery time in mice challenged i.p. with 2LD50 TeNT preincubated for 1 h with assigned BEP. BEPa MAb33/MAb39 MAb33/MAb51 MAb33/MAb71 MAb39/MAb51 MAb39/MAb71 MAb51/MAb71 a b c d

Survival rate (%)b

Maximal pathology scoreb,c

Recovery time (day)b,c,d

100 100 100 80 60 100

0.7 ± 0.3 0.2 ± 0.1 0.8 ± 0.1 1.2 ± 0.1 2.0 ± 0.3 0.9 ± 0.1

4.8 ± 0.6 1.6 ± 0.7 4.4 ± 0.2 8.0 ± 0.4 9.7 ± 0.7 2.6 ± 0.5

BEP – biclonal equimolar preparation. Cumulative data of two independent experiment each comprising 10 mice per group. Results are presented as mean value ± SE calculated for mice that survived challenge (up to 20 mice per group, depending on survival rate). The day of TeNT challenge was taken as a day 0.

(Table 1), indicating that the TeNT epitopes recognized by MAb33, MAb39, MAb51 and MAb71 are positioned in a way that allows the simultaneous binding of the antibodies.

MAb39, MAb33/MAb51, MAb33/MAb71, and MAb51/MAb71) were higher than the potencies of their individual MAb components. BEP MAb51/MAb71 showed the strongest in vitro potency. Bivariate correlation analysis revealed a significant positive correlation between BEP’s potencies and their affinity to TeNT (Pcc = 0.981,

3.2. Anti-TeNT MAb-based BEPs possess superior binding characteristics than single MAb preparations The mutual influence of selected MAb pairs when binding to TeNT was evaluated by determining whether the binding characteristics of one MAb changed if TeNT was first preincubated with the other selected MAb. Preliminary analysis resulted in nonlinear Scatchard plots, which was an indicator that cooperative binding might occur (Supplement 1) [27,28]. The MAbf,50% values were determined after dose-response sigmoidal curve fittings for a MAb bound to TeNT alone or to TeNT preincubated with the other MAb (Fig. 1). MAbf,50% values correlate with the dissociation constants (Kd); higher values imply negative cooperative binding, while lower values imply positive cooperative binding [27]. The data presented in Fig. 1 show (i) a significant positive effect of MAb39, MAb51 and MAb71 on the binding of MAb33 to TeNT; (ii) a positive effect of MAb51 and MAb71 on the binding of MAb39 to TeNT; (iii) a significant positive effect of MAb71 and a significant negative effect of MAb33 on the binding of MAb51 to TeNT; and (iv) a significant negative effect of MAb39 on the binding of MAb71 to TeNT. All BEP preparations could inhibit the interaction of TeNT with GD1b (Table 2). The BEPs MAb33/MAb51 and MAb51/MAb71 were the most potent inhibitors of TeNT binding to GD1b. Preincubation of TeNT with MAb33/MAb51 and MAb51/MAb71 almost completely inhibited TeNT-GD1b interactions (92.6 ± 0.9% and 93.2 ± 1.8% inhibition, respectively), while the MAb39/MAb71 combination was the weakest inhibitor (55.6 ± 4.1%). The binding affinities of all tested BEPs to TeNT were above 1  108 M 1 (Table 2). Furthermore, the in vitro potencies of all four tested BEPs (MAb33/

Fig. 2. Pathology scores of mice challenged with a lethal dose of TeNT (2LD50) preincubated with TeNT-specific BEPs consisting of MAb33, MAb39, MAb51 or MAb71 (total MAb 10 mg). Plots summarize the results of two independently performed experiments (10 mice per group and per experiment, i.e., 20 mice per group in total). The mean pathology scores ± SE for mice belonging to the same group at a specified post-challenge day are presented. The day of challenge was designated as day 0. The mean pathology scores were calculated considering (i) the severity of the pathology when mice were alive at the moment of observation and (ii) mortality at the moment of observation/euthanized the previous day (for grading system, see Material and Methods, Section 2.6.1). Black arrows indicate days when mouse mortality occurred. The survival rate at day 15 for each group is indicated in brackets.

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

3769

P = 0.001), as well as their inhibitory effects on TeNT-GD1b interactions (Pcc = 0.816, P = 0.048) (Supplement 2). 3.3. BEP’s protective effects positively correlate with the inhibition of TeNT-GD1b interactions Preincubation of TeNT with four out of six tested BEPs (MAb33/ MAb39, MAb33/MAb51, MAb33/MAb71, and MAb51/MAb71) prevented TeNT-induced death in all i.p. challenged mice (Table 3). The severity of tetanus-related pathology during the postchallenge period is depicted in Fig. 2. The best outcome was seen when TeNT was preincubated with MAb33/MAb51: the survival rate in this group was 100%, with no visible signs of TeNT-related pathology (maximal pathology score: 0.2 ± 0.1) and complete recovery in 1.6 ± 0.7 days (the shortest for all tested BEPs). The most severe pathology was seen in mice challenged with TeNT preincubated with MAb39/MAb71 (the survival rate in this group was 60%), although mobility was not seriously impaired with mice that recovered from the challenge (maximal pathology score 2.0 ± 0.3, recovery time 9.7 ± 0.7 days). The protective effects of the different BEPs (expressed as the survival rate after challenge with TeNT preincubated with the BEPs) showed a significant positive correlation with their inhibitory effects on TeNT-GD1b interactions (Pcc = 0.837, P = 0.038). The affinity and potency of BEP influenced the intensity of tetanus severity and recovery time (Supplement 3). 3.4. The therapeutic efficacy of TeNT-specific MAb-based BEPs is higher than that of single MAb preparations The therapeutic efficacy of MAb71-containing BEPs (with MAb71 accounting for 50% of the total MAbs) was significantly higher than that of MAb71 administered alone at twice the BEP amount [13] (Supplement 4). The maximal average pathology scores in mice challenged with 2LD50 TeNT preincubated with MAb71-containing BEPs were significantly lower than in mice injected with TeNT/MAb71 alone: 0.9 ± 0.1 for TeNT//MAb51/ MAb71 (P < 0.005), 2 ± 0.3 for TeNT//MAb39/MAb71 (P < 0.05), and 0.8 ± 0.1 for TeNT//MAb33/MAb71 (P < 0.005). The BEP containing MAb51/MAb71 was more efficient at preventing tetanus than the MAb71/MAb33 BEP because the recovery time was significantly shorter (P < 0.05; Table 2), although maximal pathology scores and survival rates were similar. The BEP containing MAb51/MAb71 showed a better therapeutic effect than MAb71 and MAb51 administered alone and in amounts equal to the total MAbs present in the BEP. Intravenous injection of this BEP 2 h or 6 h after TeNT challenge (Fig. 3) resulted in 100% survival rates, with maximal pathology scores of 1.9 ± 0.3 and 2.6 ± 0.2, respectively. This outcome was comparable to that seen when MAb51 was injected alone (Supplement 4). Nevertheless, the recovery time in mice treated with the MAb51/MAb71 BEP (2h: 9.8 ± 1.0 days, 6 h: 11.7 ± 0.3 days) was significantly shorter than that seen with MAb51 alone (P < 0.05 in both cases). 4. Discussion Here we demonstrated that MAb-based biclonal equimolar preparations recognizing non-overlapping TeNT epitopes showed improved binding characteristics when compared with their individual MAb components. Consequently, this could lead to better protection against TeNT intoxication. Oligoclonal preparations of anti-TeNT MAbs or their antigenbinding fragments have already been evaluated, and the results suggested improved efficacy when two or more TeNT-specific MAbs were used [10,11,29,30]. Nevertheless, our study is the first to investigate in vitro the mechanisms by which one MAb may

Fig. 3. Pathology scores of mice challenged with a lethal dose of TeNT (2LD50) and treated 2 h (h) or 6 h (j) later with BEP containing MAb51/MAb71 (total MAb 10 mg, i.v.). Pathology score changes in mice challenged with 2LD50TeNT preincubated with BEP containing MAb51/MAb71 (total MAb 10 mg) are indicated by the dotted line. Plots summarize the results of two independently performed experiments (10 mice per group and per experiment, i.e., 20 mice per group in total). The mean pathology scores ± SE for mice belonging to the same group at specified postchallenge days are presented. The day of challenge was designated as day 0. The mean pathology scores were calculated considering (i) the severity of the pathology when mice were alive at the moment of observation and (ii) mortality at the moment of observation/euthanized the previous day (for grading system, see Material and Methods, Section 2.6.1).

affect the binding characteristics of the second MAb and how this property correlates with the in vivo protective effect of TeNTspecific oligoclonal preparations. It has been demonstrated in vivo that TeNT-specific MAbs can exert a synergistic effect in oligoclonal mixtures. Competitive inhibition studies showed that MAbs generated in mice immunized with tetanus toxoid bound to approximately 20 different epitopes. This finding suggested that efficient in vivo TeNT neutralization may require the simultaneous binding of at least two MAbs specific for non-overlapping epitopes, an approach that may translate in up to 200-fold better protection than that induced by the individual MAbs [11]. In another study, 11 murine TeNT-specific MAbs were characterized in vitro and their protective potential evaluated. A few cases of synergism between MAbs were documented when they were administered in vivo in the form of biclonal preparations. However, the in vitro binding characteristics of the biclonal TeNTspecific preparations were not investigated [30]. Another study with similar design and outcome was performed with human anti-TeNT antibodies [29]. Furthermore, a group carried out pairwise analyses of scFv binding to TeNT with the aim of selecting scFv that would be suitable to engineer TeNT-specific chelating recombinant antibodies (CRAbs; chimeric bispecific TeNT-binding molecules consisting of tandemly linked scFvs that bind to distinct non-overlapping epitopes). Positive cooperative binding was observed for several non-competing scFv pairs, leading the authors to hypothesize that this phenomenon might influence the protective effect of CRAbs. Unfortunately, in vivo experiments to confirm/rule out this hypothesis were not performed [10]. The MAbs used in our study differed primarily in their capacity to prevent the initial step in TeNT intoxication, namely the formation of TeNT-GD1b complexes [13]. Consequently, these MAbs have different protective potential against TeNT intoxication. Pair-wise analyses and AI determinations demonstrated that all of the selected MAbs could bind TeNT simultaneously, which is the main

3770

I. Lukic et al. / Vaccine 36 (2018) 3764–3771

requirement for MAbs to be included in a biclonal preparation. Due to the flexibility of the TeNT molecule, binding of a particular MAb may result in conformational changes that (i) include the ganglioside binding site and interfere with formation of TeNT-GD1b complexes [31,32], (ii) encompass epitopes recognized by other MAbs and influence their binding (positive or negative cooperative binding) [10], or (iii) result in the generation of neo-epitopes [33]. In our study, the in vitro calculated potencies of the majority of the BEPs were higher than the potencies of their component MAbs. This strongly suggests that net positive cooperative binding occurs in the majority of cases and results from complex interactions, which occur when TeNT is exposed to MAbs able to simultaneously bind it. Downward-curved Scatchard plots imply that binding of one MAb promotes the generation of a (neo)epitope, which is than recognized by the second MAb with higher affinity [27,28]. Our experiments also revealed that negative cooperative binding occurred in some cases. For example, MAb33 did not exert any positive influence on the binding of other MAbs to TeNT. MAb33 did not influence TeNT-MAb39 or TeNT-MAb71 interactions but, in fact, exerted a negative influence on the binding of MAb51 to TeNT. However, MAb39, MAb51 and MAb71 exerted a strong positive impact on TeNT-MAb33 interactions, resulting in improved protective effects by all the MAb33-containing preparations that significantly exceeded that of the corresponding single MAb-based preparations. The most significant improvement in binding characteristics and protective capacity was achieved by combining MAb51 and MAb71, which exerted mutual positive effects on TeNT binding. The significant correlations between the protective capacity of TeNT-specific MAb-based BEPs and their binding characteristics are in agreement with previously reported data [13]. It was shown that the protective capacity of TeNT-specific single MAbs with sufficiently high affinity to allow TeNT neutralization in solution (higher than 1  108 M 1) was mainly influenced by their ability to prevent the initial contact of TeNT with neurons (the establishment of TeNT-GD1b interactions). For a MAb that binds to an epitope within the ganglioside-binding site of TeNT or in its close proximity, the capacity of inhibiting TeNT-GD1b interactions is mainly determined by its affinity. However, MAbs recognizing epitopes outside the ganglioside-binding site have also been shown to be protective [11,14]. For these MAbs, TeNT conformational changes induced by their binding and that affect the gangliosidebinding site must be considered [34]. Moreover, if antibody preparations capable of establishing high affinity interaction(s) with TeNT are administered in excess over TeNT (e.g., 10 mg MAbs vs 1 ng kg 1 bw) [35], it would be expected that their fine epitope specificity would dominantly determine their protective capacity. Our results indicate that MAbs that do not show any significant protective effect when administered alone, such as MAb39 and MAb71, may still be a valuable component of TeNT-specific oligoclonal preparations if they exert positive effects on the binding of other BEP MAb constituents. In vitro screening of oligoclonal TeNT-specific preparations based on their affinity/potency and inhibitory effect on TeNT-GD1b interactions may permit the rational selection of preparations with high in vivo TeNT neutralization potential. Thus, in vivo testing would be rationalized and the number of oligoclonal formulations producing negative outcomes would be minimized.

Acknowledgements This research has been supported by Ministry of Education, Science and Technological Development of the Republic of Serbia [Grant no. 172049].

Authors’ contribution Conceived and designed experiments MS, IL; Performed experiments IL, AF, RM, EM; Analyzed data IL, MS, AIK, AF, RM, EM; Graphical presentation of the results IL, AF, AIK; Writing the manuscript MS, AIK, IL.

Conflict of interest Authors declare no competing interests.

Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.vaccine.2018.05. 058. References [1] Centers, for, Disease, Control, and, Prevention. Tetanus. In: Atkinson W, Wolfe S, Hamborsky J, editors. Washington, DC: Epidemiology and Prevention of Vaccine-preventable Diseases. Public Health Foundation; 2012. [2] Global, regional, and national age-sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2015;385:117–71. [3] Taylor AM. Tetanus. Cont Educ Anaesth Crit Care Pain 2006;6:101–4. [4] WHO. Current recommendations for treatment of tetanus during humanitarian emergencies; 2010. [5] Hassel B. Tetanus: pathophysiology, treatment, and the possibility of using botulinum toxin against tetanus-induced rigidity and spasms. Toxins 2013;5:73–83. [6] Tetanus vaccines: WHO position paper, February 2017 – Recommendations. Vaccine; 2017. [7] Saylor C, Dadachova E, Casadevall A. Monoclonal antibody-based therapies for microbial diseases. Vaccine 2009;27(Suppl 6):G38–46. [8] EMEA. Note for guidance on production and quality control of animal immunoglobulins and immunosera for human use. The European agency for the evaluation of medicinal products evaluation of medicines for human use; 2002. [9] Lipman NS, Jackson LR, Trudel LJ, Weis-Garcia F. Monoclonal versus polyclonal antibodies: distinguishing characteristics, applications, and information resources. ILAR J 2005;46:258–68. [10] Scott N, Qazi O, Wright MJ, Fairweather NF, Deonarain MP. Characterisation of a panel of anti-tetanus toxin single-chain Fvs reveals cooperative binding. Mol Immunol 2010;47:1931–41. [11] Volk WA, Bizzini B, Snyder RM, Bernhard E, Wagner RR. Neutralization of tetanus toxin by distinct monoclonal antibodies binding to multiple epitopes on the toxin molecule. Infect Immun 1984;45:604–9. [12] Matsuda M, Kamei M, Sugimoto N, Ma Y, Hashizume S. Characteristics of toxin-neutralization by anti-tetanus human monoclonal antibodies directed against the three functional domains [A], [B] and [C] of the tetanus toxin molecule and a reliable method for evaluating the protective effects of monoclonal antibodies. Eur J Epidemiol 1992;8:1–8. [13] Lukic I, Marinkovic E, Filipovic A, Krnjaja O, Kosanovic D, Inic-Kanada A, et al. Key protection factors against tetanus: anti-tetanus toxin antibody affinity and its ability to prevent tetanus toxin – ganglioside interaction. Toxicon 2015;103:135–44. [14] Cheng LW, Stanker LH, Henderson 2nd TD, Lou J, Marks JD. Antibody protection against botulinum neurotoxin intoxication in mice. Infect Immun 2009;77:4305–13. [15] Both L, Banyard AC, van Dolleweerd C, Wright E, Ma JK, Fooks AR. Monoclonal antibodies for prophylactic and therapeutic use against viral infections. Vaccine 2013;31:1553–9. [16] Logtenberg T. Antibody cocktails: next-generation biopharmaceuticals with improved potency. Trends Biotechnol 2007;25:390–4. [17] Diamant E, Torgeman A, Ozeri E, Zichel R. Monoclonal antibody combinations that present synergistic neutralizing activity: a platform for next-generation anti-toxin drugs. Toxins 2015;7:1854–81. [18] Abboud N, Chow SK, Saylor C, Janda A, Ravetch JV, Scharff MD, et al. A requirement for FcgammaR in antibody-mediated bacterial toxin neutralization. J Exp Med 2010;207:2395–405. [19] Daeron M. Fc receptor biology. Annu Rev Immunol 1997;15:203–34. [20] Datta-Mannan A, Witcher DR, Tang Y, Watkins J, Wroblewski VJ. Monoclonal antibody clearance. Impact of modulating the interaction of IgG with the neonatal Fc receptor. J Biol Chem 2007;282:1709–17. [21] Nowakowski A, Wang C, Powers DB, Amersdorfer P, Smith TJ, Montgomery VA, et al. Potent neutralization of botulinum neurotoxin by recombinant oligoclonal antibody. Proc Natl Acad Sci USA 2002;99:11346–50.

I. Lukic et al. / Vaccine 36 (2018) 3764–3771 [22] Marks JD. Deciphering antibody properties that lead to potent botulinum neurotoxin neutralization. Movement Disord: Off J Movement Disord Soc 2004;19(Suppl 8):S101–8. [23] Fan Y, Garcia-Rodriguez C, Lou J, Wen W, Conrad F, Zhai W, et al. A three monoclonal antibody combination potently neutralizes multiple botulinum neurotoxin serotype F subtypes. PLoS One 2017;12:e0174187. [24] Froude JW, Stiles B, Pelat T, Thullier P. Antibodies for biodefense. mAbs 2011;3:517–27. [25] Seatovic S, Inic-Kanada A, Stojanovic M, Zivkovic I, Jankov RM, Dimitrijevic L. Development of sandwich enzyme-linked immunosorbent assay for determination of tetanus toxoid concentration. J Immunoassay Immunochem 2004;25:31–44. [26] Friguet B, Djavadi-Ohaniance L, Pages J, Bussard A, Goldberg M. A convenient enzyme-linked immunosorbent assay for testing whether monoclonal antibodies recognize the same antigenic site. Application to hybridomas specific for the beta 2-subunit of Escherichia coli tryptophan synthase. J Immunol Methods 1983;60:351–8. [27] Wilkinson KD. Quantitative analysis of protein-protein interactions. Methods Mol Biol (Clifton, NJ) 2004;261:15–32. [28] Bordbar A, Saboury A, Moosavi-Movahedi A. The shapes of Scatchard plots for systems with two sets of binding sites. Biochem Educ 1996;24:172–5.

3771

[29] Lang AB, Cryz Jr SJ, Schurch U, Ganss MT, Bruderer U. Immunotherapy with human monoclonal antibodies. Fragment A specificity of polyclonal and monoclonal antibodies is crucial for full protection against tetanus toxin. J Immunol 1993;151:466–72. [30] Yousefi M, Khosravi-Eghbal R, Reza Mahmoudi A, Jeddi-Tehrani M, Rabbani H, Shokri F. Comparative in vitro and in vivo assessment of toxin neutralization by anti-tetanus toxin monoclonal antibodies. Human Vaccines Immunother 2014;10:344–51. [31] Fitzsimmons SP, Clark KC, Wilkerson R, Shapiro MA. Inhibition of tetanus toxin fragment C binding to ganglioside G(T1b) by monoclonal antibodies recognizing different epitopes. Vaccine 2000;19:114–21. [32] Chen C, Fu Z, Kim JJ, Barbieri JT, Baldwin MR. Gangliosides as high affinity receptors for tetanus neurotoxin. J Biol Chem 2009;284:26569–77. [33] Roguin LP, Retegui LA. Monoclonal antibodies inducing conformational changes on the antigen molecule. Scand J Immunol 2003;58:387–94. [34] Sheppard AJ, Cussell D, Hughes M. Production and characterization of monoclonal antibodies to tetanus toxin. Infect Immun 1984;43:710–4. [35] Gill DM. Bacterial toxins: a table of lethal amounts. Microbiol Rev 1982;46:86–94.