Post-VX exposure treatment of rats with engineered phosphotriesterases

The biologically stable and highly toxic organophosphorus nerve agent (OP) VX poses a major health threat. Standard medical therapy, consisting of reactivators and competitive muscarinic receptor antagonists, is insufficient. Recently, two engineered mutants of the Brevundimonas diminuta phosphotriesterase (PTE) with enhanced catalytic efficiency (kcat/KM = 21 to 38 × 106 M−1 min−1) towards VX and a preferential hydrolysis of the more toxic P(−) enantiomer were described: PTE-C23(R152E)-PAS(100)-10-2-C3(I106A/C59V/C227V/E71K)-PAS(200) (PTE-2), a single-chain bispecific enzyme with a PAS linker and tag having enlarged substrate spectrum, and 10-2-C3(C59V/C227V)-PAS(200) (PTE-3), a stabilized homodimeric enzyme with a double PASylation tag (PAS-tag) to reduce plasma clearance. To assess in vivo efficacy, these engineered enzymes were tested in an anesthetized rat model post-VX exposure (~ 2LD50) in comparison with the recombinant wild-type PTE (PTE-1), dosed at 1.0 mg kg−1 i.v.: PTE-2 dosed at 1.3 mg kg−1 i.v. (PTE-2.1) and 2.6 mg kg−1 i.v. (PTE-2.2) and PTE-3 at 1.4 mg kg−1 i.v. Injection of the mutants PTE-2.2 and PTE-3, 5 min after s.c. VX exposure, ensured survival and prevented severe signs of a cholinergic crisis. Inhibition of erythrocyte acetylcholinesterase (AChE) could not be prevented. However, medulla oblongata and diaphragm AChE activity was partially preserved. All animals treated with the wild-type enzyme, PTE-1, showed severe cholinergic signs and died during the observation period of 180 min. PTE-2.1 resulted in the survival of all animals, yet accompanied by severe signs of OP poisoning. This study demonstrates for the first time efficient detoxification in vivo achieved with low doses of heterodimeric PTE-2 as well as PTE-3 and indicates the suitability of these engineered enzymes for the development of highly effective catalytic scavengers directed against VX.


Introduction
Despite ban by the Chemical Weapons Convention, highly toxic organophosphorus nerve agents (OP) were repeatedly used in military conflicts, terrorist attacks or assassination attempts. The use of sarin in the Syrian civil war, the supposed VX poisoning of Kim Jong Nam 2017 as well as the homicidal attacks of Sergei Skripal 2018 and of Alexei Navalny 2020 with novichok underline the persistent threat of these vicious compounds (Berlinger 2017;Dewan and Alkashli 2017;Halasz et al. 2020;John et al. 2018;Steindl et al. 2021;Vale et al. 2018).
The high toxicity of OP results from covalent binding to and subsequent inhibition of the pivotal enzyme acetylcholinesterase (AChE). This leads to a synaptic overflow of acetylcholine (ACh), followed by overstimulation and desensitization of muscarinic and nicotinic cholinergic receptors causing a broad spectrum of clinical signs including miosis, salivation and cardiac arrhythmia (Aldridge and Davison 1953;Aldridge and Reiner 1972;Holmstedt 1959). Ultimately, central and peripheral respiratory depression may lead to death by asphyxia (Thier mann et al. 2013;Yanagisawa et al. 2006).
For the past 60 years the standard therapy of OP poisoning has been a combination of the competitive muscarinic receptor antagonist atropine and an oxime as AChE reactivator (Cannard 2006;Thiermann et al. 2013). However, multiple in vitro and in vivo studies have shown that this treatment has limited effectiveness against different OP and cannot prevent cholinergic signs Worek and Thiermann 2013). This triggered extensive research on enzyme-based scavengers to detoxify OP via hydrolysis to less toxic products in the blood compartment, thus preventing the distribution into target tissues and reducing toxicity (Masson and Nachon 2017;Worek et al. 2016a).
Phosphotriesterase (PTE) from Brevundimonas diminuta has emerged as a promising candidate for a catalytic bioscavenger (Masson and Rochu 2009;Worek et al. 2016a). Several research groups described PTE variants with improved catalytic activity, broadened substrate spectrum and stereopreference towards the more toxic P(−) nerve agent enantiomer (Bigley et al. 2013(Bigley et al. , 2015Cherny et al. 2013;Goldsmith et al. 2017;Worek et al. 2016a). Hereby, a main focus was to improve the detoxification of highly toxic and biologically stable V-agents such as VX (Benschop and de Jong 1988;Masson and Nachon 2017;Reiter et al. 2015).
In fact, post-exposure therapy of VX poisoned guinea pigs with the engineered PTE mutant C23 (k cat /K M = 5 × 10 6 M −1 min −1 ) ensured survival and prevented systemic cholinergic toxicity , indicating that PTE variants may have the potential for use as pre-and post-exposure prophylaxis as well as post-exposure treatment (Worek et al. 2016b). Another improved mutant, the PTE variant C23AL detoxified VX with k cat /K M = 12 × 10 6 M −1 min −1 , but in vivo data showed slower kinetics than the expected breakdown of VX extrapolated from in vitro data (Wille et al. 2016).
Novel variants were generated to improve catalytic efficacy and stability, e.g. by replacing two unpaired Cys residues by the more inert amino acid Val  or by creating a highly effective PTE heterodimer with broadened substrate spectrum ). These two promising mutants were optimized with regard to extended plasma half-life with an additional C-terminal PAS-tag (Schlapschy et al. 2013) and showed efficient detoxification in vitro towards a broad substrate spectrum of structurally diverse nerve agents (Köhler et al. 2021).
Consequently, we have now set out to test the efficacy of the two novel PTE mutants, PTE-2 and PTE-3 (Table 1), in comparison with wild-type PTE (PTE-1) in a post-exposure therapy model of VX exposed rats.

Chemicals
The OP VX (O-ethyl S-(diisopropylaminoethyl)phosphonothioate; > 98% by 1 H and 31 P NMR) and VR (O-isobutyl S-(diethylaminoethyl)phosphonothioate; > 98% by 1 H and 31 P NMR) were made available by the German Ministry of Defence. VX and VR stock solutions (1% v/v in acetonitrile) were stored at room temperature. For each in vivo test, a fresh VX working solution (36 µg mL −1 ) was prepared in saline and put on ice until use. For the stereoselective analytics of VX in rat blood samples a VR working solution (20 ng mL −1 ) was prepared in acetonitrile as internal standard.

Animals
Male Wistar rats (approximately 250-370 g; corresponding to an age range of 7-9 weeks) were purchased from Charles River (Sulzfeld, Germany). In total, n = 65 animals were used, five died during preparation. Prior to experiments, the animals were housed at least 7 days for acclimatization under controlled conditions (12 h light/ dark cycle [light from 7:00 a.m. to 7:00 p.m.], 20-24 °C, 45-65% humidity) with standard diet (Ssniff Spezialdiäten, Soest, Germany) and tap water ad libitum in eurostandard Makrolon cages type IV with heightened lids (Tecniplast, Hohenpeißenberg, Germany). For enrichment a house (Zoonlab, Castrop-Rauxel, Germany), embedding (Rettenmeier&Söhne, Rosenberg, Germany), nestlets and wood wool (both from Zoonlab) were provided and replaced once a week. Every attempt was made to minimize discomfort and pain and to reduce the number of animals used in the study. All experiments were in conformity with the German Animal Welfare Act of 18th May 2006 (BGBI. I S. 1206, 1313), the European Parliament and Council Directive of 22nd September 2010 (2010/63/EU) and carried out according to the Basel declaration as well as the 3R concept (replace, reduce, refine). The study was ethically approved by the institutional animal protection committee .

Experimental procedure for PTE plasma concentration profile determination
For PTE administration and blood sampling as shown in the time line (Fig. 1a), rats were repetitively anesthetized with short-time isoflurane inhalation (IsoFlo 100% w/w, Abbott Laboratories, Burgdorf, Germany; oxygen as carrier gas), 5% for induction of anesthesia and 2% for maintenance. In advance to vascular cannulation, local anesthetic creme (25 mg g −1 + 25 mg g −1 Prilocaine-Lidocaine; Pierre Fabre, Freiburg, Germany) was applied. To ensure an adequate depth of anesthesia, reflexes were checked regularly. Defined PTE doses were injected into the respective lateral tail vein using a peripheral intravenous inserted catheter (24 Gauge i.v. indwelling cannula, 0.7 × 19 mm; B. Braun, Melsungen, Germany) and flushed with saline before removal. Subsequently, blood samples were taken at various time points from the lateral tail vein over a period of 48 h (10 min, 20 min, 30 min, 60 min, 120 min, 240 min, 480 min, 960 min, 1440 min and 2880 min) with a maximum of five samples per animal. Sterile hypodermic needles were used to prick the vein and the blood samples were collected in heparinized Minivettes (POCT, Sarstedt, Nürmbrecht, Germany).

Experimental groups for PTE plasma concentration profiles
The animals (n = 24) were divided into three groups by simple randomization. PTE-1 was administered at a dose of 1.0 mg kg −1 , the doses of PTE-2 and PTE-3 were adjusted according to the molecular weight to achieve equal stoichiometries (note that PTE-2 is a single-chain heterodimer with two distinct active sites) (Table 1).

Experimental procedure for the therapeutic efficacy study
Rats were anesthetized by i.m. injection of a mixture of medetomidine (0.15 mg kg −1 ; Zoetis, Berlin, Germany), midazolam (0.4 mg kg −1 ; Ratiopharm, Ulm, Germany) and fentanyl (0.1 mg kg −1 ; Albrecht, Aulendorf, Germany) to reduce suffering from VX challenge. Reflexes were checked for depth of anesthesia throughout the whole experiment and anesthesia re-dosing was applied as required with 1/4-1/3 of the initial dose. Next, the animals were placed on a heatable operating table in supine position, a rectal thermistor was inserted and an electrocardiogram and a pulse oximeter were attached.
Eyes were covered with panthenol eye cream (Jenapharm Mibe, Brehna, Germany) to protect from dehydration. After preparation of the left a. carotis (for blood sampling) and the right v. jugularis (for PTE injection), catheters (24 G × ¾ [0.7 × 19 mm] Introcan; B. Braun) were inserted and fixed. The animals were monitored (blood pressure, heart rate, oxygenation, body temperature) throughout the whole experiment. Clinical signs were observed and heparinized blood samples (250 µL replaced by Ringer's solution) were collected. The first blood sample was taken before and 11 further blood collections were scheduled between 3 and 180 min after subcutaneous injection of 25 µg kg −1 (~ 2LD 50 ) VX prepared in saline (Misik et al. 2015;Myhrer et al. 2015).
For AChE activity determination, 50 µL whole blood was diluted (1:20) in distilled water, whereas 100 µL whole blood was used for VX analysis. The remaining blood sample was centrifuged and plasma was used for PTE analysis. All samples were snap-frozen in liquid nitrogen and stored at −80 °C until use. Directly after the end of the experiment tissue samples of diaphragm and medulla oblongata were taken and their wet weight determined. For the time line see Fig. 1b.

Experimental groups for the therapeutic efficacy study
The animals (n = 36) were divided into six groups by simple randomization: -Solvent control group (saline i.v. 5 min after saline s.c.; n = 6).  The time-dependent absorbance of samples from the calibration curve and plasma were calculated in mA min −1 . Predetermined calibration curves were used for the calculation of the enzyme level in plasma samples. The corresponding circulation half-life PTE t 1∕2 was obtained using a onecompartment model with a mono-exponential decay of the calculated plasma enzyme levels versus the blood collection time points, and the maximal (initial) plasma concentration of PTE variants (C max ) was estimated by retrograde extrapolation of the mono-exponential curve fit using Prism Version 4.03 (GraphPad, San Diego, CA, USA).

AChE activity assay
Diaphragm and medulla oblongata were removed after the end of the experiment or death of the animal, snap-frozen in liquid nitrogen and stored at −80 °C until the determination of the AChE activity. Diaphragm was mixed with a tenfold volume of TRIS/HCl buffer (50 mmol L −1 , pH 7.4), supplemented with EDTA (5 mmol L −1 ) and 1% v/v Triton X-100 followed by homogenization of the mixture with a T 25 digital ULTRA-TURRAX (IKA, Staufen, Germany). Medulla oblongata was mixed with a tenfold volume of phosphate buffer (0.1 mol L −1 , pH 7.4) containing 1% v/v Triton X-100. This mixture was subsequently homogenized with a glass-Teflon Potter (B. Braun). Afterwards, the homogenate was centrifuged (Microfuge 22; Hettich, Tuttlingen, Germany) at maximum speed for 5 min. The supernatant was used to measure the AChE activity according to a modified Ellman assay (Worek et al. 1999). The total protein concentration was determined with a bicinchoninic acid assay (Protein Quantification Kit; Interchim Uptima, Montlucon, France) using bovine serum albumin as standard (Smith et al. 1985).
For analysis of all in vivo AChE activities polystrol cuvettes were filled with phosphate buffer (0.1 mol L −1 , pH 7.4), 0.3 mmol L −1 DTNB and 0.02 mmol L −1 ethopropazine. After adding the sample the assay was started by adding 0.45 mmol L −1 ATCh and the absorbance change was recorded at 436 nm at 37 °C (Cary 50; Varian, Darmstadt, Germany). Erythrocyte AChE activity was referenced to the hemoglobin concentration of the respective blood dilution (Worek et al. 1999) and was calculated as mU µmol −1 Hb. Diaphragm and brain AChE activity were referenced to the total protein concentration determined by the bicinchoninic acid method (Smith et al. 1985) and presented as mU mg −1 .

Analysis of VX in blood samples
VX enantiomers in blood samples were quantified by LC-MS/MS as described before (Reiter et al. 2008(Reiter et al. , 2011 with slight modifications. 100 µL thawed, heparinized rat whole blood was diluted with 300 µL water and mixed with 37.5 µL 1 mol L −1 perchloric acid, followed by addition of 7.5 µL 1 mol L −1 potassium acetate, and vortexed. The precipitated proteins were separated by centrifugation for 10 min at 21,290 × g at 4 °C. The supernatant was mixed with 50 µL of 0.5% w/v ammonium hydroxide and 5 µL of internal standard solution (20 ng mL −1 VR working solution prepared in acetonitrile) and centrifuged again for 3 min at 21,290 × g at 4 °C. The sample was applied to a preconditioned (1 mL methanol, 1 mL purified water; Milli-Q, Merck Millipore, Schwalbach, Germany) SPE cartridge (Strata-X PRP, 30 mg, 1 mL; Phenomenex, Aschaffenburg, Germany) on a vacuum manifold. The column bed was rinsed with 1 mL water and carefully dried under maximum vacuum. The sample was eluted with 500 µL acetonitrile and collected in a cup filled with 50 µL purified water as keeper. The organic phase was evaporated in a vacuum rotation concentrator (50 min at room temperature) with 11,360 x g and the residue was reconstituted with 100 µL purified water.
The analytical LC-MS/MS system for the quantification of VX enantiomers consisted of a Prominence HPLC system (Shimadzu Deutschland, Duisburg, Germany) and a QTRAP 6500 mass spectrometer (Sciex, Darmstadt, Germany).
Gas 1, gas 2, and curtain gas were set at 35, 40, 45 arbitrary units. The entrance potential was set to 5.0 V, cell exit potential was set to 12 V. Dwell time was 750 ms and resolution was "unit" for both quadrupoles.

Calculation of VX half-life
The theoretical half-life of VX hydrolysis VX t 1∕2 was calculated by Eq. (1) according to Worek et al. (2016a) using in vitro k cat /K M and the maximal plasma concentration of PTE variants (C max ):

Data analysis
Data are shown as mean ± standard deviation (SD). Data analysis and statistical comparisons were performed using Prism Version 5.04.
For the analysis of the diaphragm and medulla oblongata AChE activity differences between groups, a one-way repeated measure analysis of variance (ANOVA) with Bonferroni post-hoc test multiple comparison test was used. A p < 0.05 value was considered to be statistically significant.

PTE plasma concentration profile in vivo
As a prerequisite to study the protective effect of the newly engineered PTE enzymes, PTE-2 and PTE-3 in comparison to wild-type PTE-1, against VX exposure in rats their pharmacokinetics (PK) after i.v. injection was studied in the same species (Fig. 1a). PTE concentrations were analyzed according to the one compartment model, resulting in plasma half-life values PTE t 1∕2 of 39 min, 165 min and 94 min for PTE-1, PTE-2, and PTE-3, respectively (Fig. 2). As expected, the application of PASylation technology (Schlapschy et al. 2013) led to a considerable extension of the PK for PTE-2 and PTE-3. This effect was most pronounced for the single-chain heterodimeric enzyme PTE-2 which has its both subunits linked by a 100 residue PAS linker ).

Calculated VX hydrolysis
From our previous studies, it was known that PTE-2 and PTE-3 show high catalytic efficiencies (k cat /K M ) towards VX hydrolysis in vitro (Table 2), which are by three orders higher than the one of wild-type PTE (PTE-1), whose k cat /K M value of 0.02 × 10 6 M −1 min −1 for VX was determined in the present study. The maximum PTE plasma concentrations (C max ), which were obtained from the PK analysis, allowed us to estimate half-lives of VX hydrolysis  in vivo VX t 1∕2 which were below 10 s for the engineered enzymes but above 1 h for wild-type PTE (Table 2).

Clinical signs and survival
Subcutaneous injection of 25 µg kg −1 (~ 2LD 50 ) VX (Misik et al. 2015;Myhrer et al. 2015) resulted in signs of chewing and an early onset of local muscle fasciculations directly at the injection site in all animals of the VX control, PTE-1 and PTE-2.1 groups. Chewing was observed in one single animal each of the PTE-2.2 and PTE-3 groups and local fasciculations were seen directly at the VX injection site in two animals of the PTE-2.2 group and in five animals of the PTE-3 group (data not shown). All degrees of convulsions, from uncontrolled muscle movements of head to forelimbs, hind legs and the entire body, were noted. Furthermore, respiratory distress was observed including the whole spectrum from labored breathing to respiratory depression. Over time, all animals of the VX control group and of the PTE-1 therapy group developed increasingly severe signs of a cholinergic crisis which persisted until death (Fig. 4). All animals of the VX Clinical signs and survival of VX exposed rats without treatment and with post-exposure treatment of PTE-1, PTE-2 and PTE-3. Time of survival is given in minutes and each line represents an individual animal. Time of death is shown by grey, hatched lines. The signs are subdivided into convulsions (local, generalized) and respiratory distress (labored breathing, respiratory depression). 25 µg kg −1 VX s.c. (VX); 25 µg kg −1 VX s.c. followed after 5 min by 1.0 mg kg −1 PTE-1 i.v. (PTE-1); 25 µg/kg VX s.c. followed after 5 min by 1.3 mg kg −1 PTE-2 i.v. (PTE-2.1); 25 µg kg −1 VX s.c. followed after 5 min by 2.6 mg kg −1 PTE-2 i.v. (PTE-2.2); 25 µg kg −1 VX s.c. followed after 5 min by 1.4 mg kg −1 PTE-3 i.v.  control group died between 27 and 71 min after subcutaneous VX injection (mean: 46 ± 17 min). The animals of the PTE-1 therapy group survived only slightly longer, between 32 and 97 min (mean: 54 ± 19 min) after s.c. VX injection (Fig. 3). In contrast, all rats treated with PTE-2.1, PTE-2.2 and PTE-3 survived until the end of the observation period of 180 min (Fig. 3). Only one of four VX poisoned rats treated with PTE-2.1 showed a longer-lasting dyspnea until the end of the observation period, whereas convulsions persisted in three of four rats (Fig. 4). Notably, none of the animals treated with PTE-2.2 and PTE-3 showed convulsions or respiratory distress as severe clinical signs throughout the experiment (Fig. 4).

Tissue AChE activity
VX poisoning resulted in an almost complete inhibition of medulla oblongata and diaphragm AChE activity (Figs. 6, 7). Therapy with PTE-1 had no protective effect whereas administration of PTE-2 and PTE-3 resulted in a partial preservation of AChE activity in these tissues. This effect was statistically significant in the PTE-2.2 and PTE-3 groups compared to VX and PTE-1 (Figs. 6, 7).

VX blood concentration
The quantification of VX enantiomers in whole blood samples revealed a rapid increase of (+)-VX and a delayed increase of the more toxic (−)-VX enantiomer (Fig. 8). In the VX control and PTE-1 treatment groups the concentration of both enantiomers increased during the further course of the experiment (Fig. 8). The analysis of (+)-VX in the VX control group showed a sharp increase in concentration up to 4.4 nmol L −1 within the first 6 min after VX administration and a maximum concentration of 5.4 nmol L −1 . The (−)-VX concentration showed a delayed increase with 1.5 nmol L −1 at the sampling time (6 min) and a lower maximal concentration of 3.4 nmol L −1 (Fig. 8). In contrast, administration of PTE-2 and PTE-3 5 min after VX challenge resulted in an ongoing hydrolysis, resulting in lower concentrations of (−)-VX (and (+)-VX) (Fig. 8). PTE-2.2 and PTE-3 treatment groups showed (+)-VX peak concentrations of 2.0 nmol L −1 and 1.9 nmol L −1 , respectively, at 3 min with a subsequent decrease in concentration. The (−)-VX concentration exhibited only a very slight increase up to a maximum of 0.3 nmol L −1 for the PTE-2.2 treatment group and 0.2 nmol L −1 for the PTE-3 treatment group at 3 min, with subsequent concentration below the quantification limit (Fig. 8).

Discussion
This study demonstrates the efficacy of post-exposure therapy of VX poisoned animals with the new PTE mutants PTE-2 and PTE-3. PTE-3 is a hyperactive stabilized homodimeric enzyme  carrying two PAS(200) tags (Schlapschy et al. 2013) whereas PTE-2 is a single-chain version with two distinct engineered active sites ) whose two subunits are linked by a PAS(100) spacer and which is equipped with one PAS(200) tag (Table 1).
Efficient OP detoxification requires a sufficiently high catalytic activity, which can be determined in vitro (k cat /K m ). Previous studies have postulated that a k cat /K m of at least 50 × 10 6 M −1 min −1 is needed for successful therapy of VX poisoning at a reasonable enzyme dose of 1 mg kg −1 Worek et al. 2016a). Lower degradation rates or lower enzyme doses may still ensure survival but not prevent severe toxic effects.
Knowledge of the k cat /K m values and the PTE plasma concentration enables the calculation of VX degradation half-lives (Table 2)  ). Our results demonstrate that the recombinant wild-type PTE from Brevundimonas diminuta (PTE-1), with an in vitro k cat /K m value of 0.02 × 10 6 M −1 min −1 , was ineffective in vivo to treat VX poisoning. Throughout the experiment the PTE-1 treatment group revealed increasingly severe toxic signs and circulatory depression (data not shown) similar to the VX control group (Fig. 4). Our findings for PTE-1 are in line with prior data from Masson and Rochu (2009) and Kolakowski et al. (1997), who determined a negligible protective effect of wild-type PTE against V-type nerve agents.
In contrast, the novel engineered enzyme PTE-3, with an in vitro k cat /K m of 38.1 × 10 6 M −1 min −1 , was able to prevent both mortality and severe toxicity in vivo. In comparison, the bispecific variant PTE-2, with an in vitro k cat /K m of 20.7 × 10 6 M −1 min −1 , was able to prevent mortality but not severe toxicity (convulsions, respiratory distress) at the initial dose of 1 mg kg −1 (Fig. 4). However, administration of a doubled PTE-2 dose was successful with regard to preventing severe toxic signs and ensuring survival.
The therapeutic success of the PTE-2 and PTE-3 enzymes is mainly based on their high catalytic activities. Available k cat /K m values and PTE plasma concentrations allow a first approximation of the degradation half-time (Table 2). These data indicate a negligible VX degradation in a relevant time window by PTE-1, whereas PTE-2 (PTE-2.2) and PTE-3 show extremely short t 1/2 values of 3.1 and 2.4 s, respectively. For a more detailed analysis, the extended circulation of the PASylated enzymes has to be taken into account, which further boosts the OP degradation as will be discussed below. These theoretical considerations are supported by analyses of the VX enantiomer concentrations in vivo (Fig. 8).
There was a gradual increase and persistence of both (−)and (+)-VX in the VX control and PTE-1 groups, whereas in the PTE-2 and PTE-3 treatment groups (−)-VX levels were at or below the limit of quantification shortly after administration of PTE (Fig. 8). Notably, the (+)-VX level also decreased after PTE administration in these groups, which corresponds to previous in vitro data on the stereoselective degradation of VX by PTE-2 (BdPTE-7) and PTE-3 (BdPTE-4), showing a preferential but not exclusive hydrolysis of (−)-VX (Köhler et al. 2021).
The s.c. VX injection resulted in a delayed appearance of (−)-VX in the blood, relevant (−)-VX concentrations were first observed in the 6 min samples of the VX control and PTE-1 groups. This is in accordance with the slightly delayed inhibition of erythrocyte AChE activity and the emerging first signs of VX poisoning in the VX control group between 11 to 26 min (Fig. 4).
The difference in catalytic efficiency and therapeutic efficacy between the PTE variants is also reflected by AChE activities in the diaphragm and medulla oblongata (Figs. 6, 7). While a virtually complete AChE inhibition was observed in the VX control and PTE-1 groups, enzyme activity was partially preserved in the PTE-2 and PTE-3 groups. This indicates that distribution of VX from the central blood compartment into target tissues could not be fully prevented but was reduced substantially depending on the catalytic activity of the different PTE variants and/or concentrations, i.e. PTE-3 > PTE-2.2 > PTE-2.1 >> PTE-1.
Efficient detoxification requires not only a high catalytic activity of PTE but also a long biological half-life. This is of special importance in the case of VX and related agents since V-agents exhibit a high biological stability and long persistence in vivo (Goldsmith et al. 2017;Reiter et al. 2011Reiter et al. , 2015van der Schans et al. 2003). In fact, in the case of VX poisoning current standard therapy with atropine and oxime must be administered over a prolonged period of time to preserve a sufficient level of active AChE and to ensure survival (Joosen et al. 2010). Hence, extension of the residence time of PTE in blood is the second key for a successful treatment. The PTE variants PTE-2 and PTE-3 were modified using PASylation technology (Binder and Skerra, 2017) in different ways, as explained above (Table 1). Plasma half-life of the unmodified wildtype PTE-1 was only 39 min whereas PTE-2 and PTE-3 showed a half-life of 165 min and 94 min (Fig. 2), respectively, which demonstrates that the PAS-tag has the ability to extend plasma half-life considerably. In fact, even more prolonged circulation can be expected when using longer PAS polypeptides, e.g. with 600 residues (Schlapschy et al. 2013;Binder and Skerra 2017;Gebauer and Skerra 2018).
The present study was based on anesthetized rats and s.c. administration of VX, which was due to animal welfare considerations and refinement at this stage of in vivo investigation. Further studies in conscious animals with percutaneous VX exposure are needed to verify the results and to more precisely evaluate required PTE doses and, potentially, the need for repeated PTE administration. Such studies are also needed to define the window of opportunity, i.e. the maximum time between agent exposure Fig. 6 Medulla oblongata AChE activity of the VX control group and the PTE-1, PTE-2, PTE-2.2 and PTE-3 treated rat groups. Data are presented as % of pre-exposure control AChE activity as mean values ± SD using box plots with min to max whiskers. (*p < 0.05 versus VX control group; # p < 0.05 versus PTE-1 group)

Fig. 7
Diaphragm AChE activity of the VX control group and the PTE-1, PTE-2, PTE-2.2 and PTE-3 treated rat groups. Data are presented as % of pre-exposure control AChE activity as mean values ± SD using box plots with min to max whiskers. (*p < 0.05 versus VX control group; # p < 0.05 versus PTE-1 group) and PTE administration without risking severe signs of toxicity.