Central neuroprotection demonstrated by novel oxime countermeasures to nerve agent surrogates

Abstract

Oximes remain a long-standing element of the therapy for nerve agents, organophosphates (OPs) that poison by inhibiting the enzyme acetylcholinesterase (AChE), resulting in hypercholinergic activity both centrally and peripherally. Oximes, such as the pyridinium oxime pralidoxime (2-PAM) in the United States, can reactivate the inhibited AChE and restore cholinergic function. However, there are several drawbacks to the current oximes; one of them, the inability of these oximes to effectively enter the brain, is the subject of study by several laboratories, including ours. Our laboratory invented a platform of substituted phenoxyalkyl pyridinium oximes that were tested against highly relevant surrogates of the nerve agents, sarin and VX. Using high sublethal dosages of the OPs, the novel oximes were observed to attenuate seizure-like behavior in rats and to reduce the levels of glial fibrillary acidic protein (an indicator of glial scarring) to control levels, in contrast to levels observed with 2-PAM or no oxime therapy. Using lethal levels of surrogates, some novel oximes protected against lethality compared with 2-PAM, shortened the time to cessation of seizure-like behavior (from 8+ to 6 h), and protected the brain neurons. Therefore, some of these novel oximes are showing exceptional promise alone or in combination with 2-PAM as therapeutics against nerve agent toxicity.

Introduction

Organophosphorus anticholinesterases have been known since the late 1930s.¹ They were developed initially as insecticides, with a recognition that many of them were more acutely toxic than the insecticide that they largely displaced in usage, dichlorodiphenyltrichloroethane. These insecticidal organophosphates (OPs) span a wide range of acute toxicity levels; for example, as indicated by rat oral LD₅₀ values, parathion and phorate have acute toxicities of 14 and 2 mg/kg, respectively, while malathion has acute toxicity of 1200 milligrams per kilogram.² Many of the other OPs have acute toxicity levels in between these extremes and many have had wide usage in agricultural fields as well as commercial and domestic buildings and surroundings. While OPs have been replaced to a certain extent by newer insecticides, they are still widely used and, therefore, readily available. Accidental poisonings with OP insecticides have routinely occurred, especially with the more toxic chemicals, and in developing countries, they are regrettably used as an agent of suicides.³ They are also a potential threat agent for terrorist attacks because of the high toxicity and ready availability of some of these insecticides.

Because the OP chemistry was known to exert a toxic action on a neural target, acetylcholinesterase (AChE), common in both arthropod pests and non-target organisms, including humans, the OP chemistry was developed into chemical warfare agents in the Second World War, with the G agents arising in Germany and the V agents arising in Great Britain. While they were not used in World War II, they remain a threat agent for military use, rogue governments, and terrorists. The terrorist attacks with the nerve agent sarin in Tokyo and Matsumoto in 1995 illustrated how these could be used against civilians to cause death and permanent neurological damage,⁴ while more recently the attacks with sarin on citizens in Syria in 2013 and 2017,⁵ the assassination of the brother of the North Korean dictator in a Kuala Lampur airport using the nerve agent VX in 2017,⁶ and the assassination attempt on a former Russian spy with the new agent Novichok in 2018⁷ illustrate the danger that is inherent in these highly toxic chemicals.

As has been well known for over half a century, these neurotoxic OPs inhibit the important neural enzyme AChE in synapses and neuromuscular junctions. The anticholinesterase (anti-ChE) OPs are typically pentavalent, with a doubly bonded oxygen and three additional groups, one of which is the “leaving group.” The phosphorus bonds with the serine hydroxyl group in AChE’s active site, freeing the leaving group. This phosphorylation of the active site serine by the OP is persistent, lasting hours to days. With the reduction or elimination of AChE’s normal function, that is, the very rapid hydrolysis and, therefore, destruction of the neurotransmitter acetylcholine, persistent hyperstimulation of cholinergic pathways ensues, with, at lethal levels of exposure in mammals, respiratory system shutdown resulting from bronchiolar constriction, excess bronchiolar secretion, paralysis of the respiratory muscles, and inhibition of the respiratory control centers in the medulla/pons.⁸ However, at high sublethal exposure levels in animal models, the hypercholinergic activity activates the excitotoxic glutamatergic pathways in the brain eliciting seizures, which, if prolonged and recurrent, result in permanent neuropathology.^9–12

Organophosphate therapy

For many years, the therapy for OP anti-ChE poisoning has used the repeated administration of the muscarinic receptor antagonist atropine to suppress the hypercholinergic activity that is responsible for respiratory failure; therefore, atropine is greatly beneficial in combatting some of the aspects of respiratory failure. Nucleophilic oximes (=N–OH) can attack phosphorus and through a transphosphorylation reaction can bond to the inhibiting OP moiety and remove it from the active site, thereby reactivating the AChE for further normal function.³ The third therapeutic used in severe OP poisoning is an anticonvulsant benzodiazepine, such as diazepam, to dampen the seizures and ideally to prevent seizure-induced brain damage. However, an additional concern here is that benzodiazepines typically are not effective in suppressing the recurrent seizures that high-dosage OPs produce in animal models and, therefore, the victim remains susceptible to brain damage.¹³

The original oxime developed and approved was the pyridinium oxime pralidoxime (2-PAM) and it is still the approved oxime therapeutic in the United States, while additional oximes, such as HI-6 (1-(2-hydroxyiminomethylpyridinium)-3-(4-carbamoylpyridinium)-2-oxapropane dichloride), are approved in some other countries.^3,14 Most of the effective oxime reactivators that have been synthesized are pyridinium or bispyridinium oximes, because the pyridinium ring moiety imparts greater efficacy to the oxime.

There are a few problems with the current oximes and many additional oxime molecules have been synthesized and tested by a number of laboratories to attempt to overcome some of these difficulties to produce a more effective therapeutic, although to date, no others have been approved for use.¹⁴ One problem is that the oxime-phosphate removed from the inhibited AChE can be a potent AChE inhibitor, so it can inhibit noninhibited AChE or can reinhibit reactivated AChE, thus worsening the poisoning. The oximes are strong nucleophiles, so some oximes themselves can inhibit AChE independent of OPs. The current oximes are not sufficiently broad-spectrum to reactivate AChE inhibited by the wide variety of OP anti-ChE chemistries. The typical oximes have a relatively short circulating half-life and, therefore, are unlikely to be very effective in a single administration, either when a very high dosage of OP is experienced because ample OP would be available for reinhibition of AChE, or in the case of a slowly absorbed OP, such as VX, that is most likely to be encountered by dermal exposure with the resultant prolonged rate of inhibition/reinhibition of AChE. And lastly, an oxime deficiency that our laboratory, as well as other laboratories, is trying to address is the inability of pyridinium oximes to appreciably penetrate the blood–brain barrier (BBB). The permanent positive charge pyridinium oximes possess makes them likely to be excluded from the brain, since the BBB is effective in excluding charged compounds.¹⁵ Because the vast majority of oximes synthesized to date do not act within the brain, they do not dampen the hypercholinergic activity elicited by OPs in the brain and do not suppress the pathway that leads to seizures and neuropathology.

Therefore, there remains a pressing need for an AChE reactivator that is effective within the brain. If the hypercholinergic activity, the immediate effect of AChE inhibition by OPs, is dampened sufficiently so that the excitotoxic glutamatergic pathways are not hyperstimulated, then it might be possible to prevent or sufficiently diminish the excitotoxicity and/or seizures such that neuropathology and damage to the brain function can be reduced or eliminated following high sublethal exposure to OPs. This perspective is the goal of the research programs, including ours, that are seeking a centrally acting AChE reactivator.

Summary of research programs seeking a brain-penetrating AChE reactivator

Several laboratories throughout the world have been or are engaged in identifying an effective reactivator that can cross the BBB. In some cases, these new reactivators are theoretically able to cross the BBB, but experimental evidence of their central efficacy against OP inhibition in vivo, in contrast to strictly peripheral action, has not yet been obtained. Amidine oximes have been synthesized, which were uncharged, and potentially zwitterionic, and theoretically would not be excluded by the BBB.^16–18 However, the in vivo studies conducted on these compounds delivered the oximes very shortly after OP administration, so that peripheral reactivation cannot be ruled out as the major action of these compounds. A prodrug derivative of 2-PAM, pro-PAM, was reported in 1976 as the uncharged dihydropyridine derivative of 2-PAM, with the potential to be activated to 2-PAM in the brain.^19,20 However, pro-PAM was considered not efficacious enough to be worthy of appreciable further study. Sugar oximes were synthesized with the theory that they could be transported into the brain on the glucose transporter, and some efficacy was observed.^21–23 A tertiary oxime, monoisonitrosoacetone, has the potential to cross the BBB because it is uncharged,²⁴ but it has not been developed further. Also, uncharged are 3-hydroxypyridine aldoximes.^25–28 A zwitterionic hydroxyiminoacetamido alkylamine, RS194b, has shown efficacy in reducing AChE inhibition and attenuating the signs of toxicity elicited by sarin and paraoxon in macaques, in addition to the remediation of AChE inhibition elicited by a sarin analogue in mice.^29–33

Substituted phenoxyalkyl pyridinium oximes

Our laboratories have invented a platform of substituted phenoxyalkyl pyridinium oximes with the hypothesis that adding a lipophilic chain to the pyridinium ring might provide characteristics to a pyridinium oxime that would allow for its entry into and accumulation within the brain at sufficient concentrations to reverse some or all of the AChE inhibition induced by the nerve agent and insecticidal OP chemistries (Fig. 1). Over 100 of these oximes were synthesized by the late Howard Chambers, and they were screened initially as brain ChE reactivators in in vitro assays using rat and bovine brain homogenates as the source of ChE (the term ChE will be used since our studies use acetylthiocholine as a substrate without the addition of specific inhibitors of butyrylcholinesterase, and the latter could contribute to the overall activity observed). The alkyl chain length had a range of 3–12 carbons and the efficacious ones for ChE reactivation had a chain length of 3, 4, or 5. Coincidentally, the oximes with a chain length of 6 showed some efficacy as paraoxonase enhancers.³⁴ The longer chain lengths were not effective as ChE reactivators.

Figure 1.

Generic structure of the mesylate salts of the novel substituted phenoxyalkyl pyridinium oximes, including the three substitutions for oximes 15, 20, and 55; n = 5 for oxime 15 and n = 4 for oximes 20 and 55.

Surrogates of sarin, VX, tabun, and soman were also synthesized in our laboratories by substituting a group, usually a 4-nitrophenyl group, for the presumed leaving group of the nerve agent; these surrogates are less toxic than their respective nerve agents and, therefore, safer for laboratory personnel and do not require the use of Surety Laboratories.³⁵ However, these surrogates are highly relevant for ChE reactivation studies since they leave ChE phosphylated by the same chemical moiety as the corresponding nerve agent for each. Two sarin surrogates were used, phthalimidyl isopropyl methylphosphonate (PIMP) and nitrophenyl isopropyl methylphosphonate (NIMP). NIMP was originally described by Ohta et al.³⁶ PIMP was used for our initial screening; the phthalimidyl group imparted some instability to the ester such that it hydrolyzed in aqueous solution within 15 minutes. This property created an advantage for in vitro screening in that it reduced or eliminated the opportunity for reinhibition of the reactivated ChE. The stable surrogate NIMP was used for subsequent in vivo testing. The VX surrogate was nitrophenyl ethyl methylphosphonate (NEMP), originally described by Fukuto et al.³⁷ The soman surrogate substituted 4-nitrophenol for the fluoride leaving group; however, greater inhibition was seen with the surrogate plus oxime assays than the surrogate alone, presumably from the anti-ChE potency of the oxime-phosphate formed, so that few tests were conducted with the soman surrogate. The tabun surrogate was synthesized by substituting 4-nitrophenol for the cyanide, which is assumed to be the leaving group; however, this surrogate did not appreciably inhibit ChE, so additional studies were not pursued. In addition, paraoxon, the active metabolite of the highly toxic insecticide parathion, was screened so that insecticidal chemistry was also part of our experimental paradigm to test for broad-spectrum efficacy of our novel oximes. The nerve agent surrogates are phosphonates and paraoxon, along with many of the other insecticidal OPs, is a phosphate; therefore, our studies have incorporated two distinct types of phosphorus esters.

Initial in vitro screens, using concentrations of the OP that yielded about 80% inhibition in 15 min, and a screening concentration of oxime of 0.1 mM for 30 min, yielded a range of reactivation efficacies from surrogate inhibition of rat or bovine ChE from about 15% to about 75%, all of which were below 2-PAM or trimedoxime bromide-4, 80–99%.^38,39 While enhancing lipophilicity was the objective of the invention of this platform of oximes, the efficacy in these in vitro screens did not display a relationship with lipophilicity.

The more promising of the novel oximes, as judged from the in vitro screens, were placed into initial in vivo screens in male rats with a high sublethal dosage of NIMP or NEMP (0.3 or 0.325 mg/kg, respectively, administered intraperitoneally in dimethyl sulfoxide (DMSO)). The selected paradigm provided peak brain ChE inhibition in 1 h post OP administration, at which time the oxime was administered intramuscularly (IM) in DMSO at a screening dosage of 0.1 millimoles per kilogram. We have done all our in vivo testing using equimolar concentrations or dosages to better compare the efficacies among the novel oximes and to 2-PAM. The concept behind this time point of peak brain ChE inhibition for oxime administration is that the OP would be in the process of being eliminated and, therefore, there would be diminishing levels of the OP available to reinhibit reactivated ChE. Oxime administration provided too soon following OP challenge has the potential to leave substantial OP in circulation, which could reinhibit the reactivated peripheral ChE, thereby lowering the amount of OP available to enter the brain; thus any reduction in brain ChE inhibition observed following premature oxime administration could be erroneously interpreted as central reactivation. Enhancement of brain ChE activity with this paradigm would support the idea of brain penetration by the oxime. In these experiments, several of the novel oximes provided increased brain ChE activity at 30 min or 2 h post oxime administration (up to 30%), while 2-PAM provided no increase.^38,39 These results strongly suggested that some of the novel oximes could cross the BBB.

The most promising five of these novel oximes were placed into 24-h survival trials with male rats using the sarin and VX surrogates, NIMP and NEMP, respectively (administered subcutaneously in Multisol, a biocompatible vehicle). LD₉₉ levels of the three OPs were determined in male rats (0.6 and 0.65 mg/kg, respectively) in which the animals did not survive with only atropine (0.65 mg/kg, IM in saline) as therapy but some animals did survive with both atropine and oxime (novel or 2-PAM, at the human equivalent dosage of three autoinjectors, 0.146 mmol/kg IM in Multisol). Novel oximes were tested alone or in combination with an equimolar dosage of 2-PAM. In most cases, the novel oximes alone provided 25% or higher survival than 2-PAM alone, with some improvement in survival by most of the binary combinations. In addition, observations of signs of toxicity over the initial 8 h of the experiment indicated that there was an attenuation of the seizure-like behavior observed by some of the novel oximes that was not observed with 2-PAM. These results indicated that the novel oximes had the ability to not only promote survival but also to promote the attenuation of signs of toxicity that were probably centrally mediated, at least in part, and, therefore, these results supported the concept that our novel oximes could enter the brain.⁴⁰

In addition to the behavioral observations on novel oxime attenuation of signs of toxicity, studies were conducted to assess structural neuroprotection. Initially, the high sublethal dosage of NIMP or NEMP was administered to male rats and slides from select brain regions were stained for glial fibrillary acidic protein (GFAP), an astrocyte marker indicative of astrogliosis as a result of damage^41–43 on days 2, 4, and 7 following challenge.⁴⁴ This study provided the evidence of neuroprotection from our lead oxime, oxime 20, in the piriform cortex and the dentate gyrus where the damage score for the animals receiving OP plus novel oxime therapy was the same as the vehicle control, indicating protection, while the damage score for the animals receiving 2-PAM therapy was the same as the OP alone, indicating no protection. These data demonstrate that oxime 20 has a neuroprotective action in the brain, while, as expected, 2-PAM does not.

The next study was a quantification of undamaged neurons in the CA1 region of the hippocampus through the neuronal marker NeuN. Rats were challenged with a lethal dosage of either NIMP or paraoxon and then provided atropine plus one of the three downselected novel oximes, oximes 15, 20, and 55.⁴⁵ Similar to the earlier GFAP results, damage scores for animals treated with oximes 20 and 55 were not statistically different from vehicle controls, while damage scores for 2-PAM and oxime 15 were not statistically different from those for NIMP with only atropine. While this was some-what surprising, in that, oxime 15 had shown some efficacy earlier in attenuating seizure-like behavior as had oxime 20, but not oxime 55, the results here showed oximes 20 and 55 to be effective. A possible explanation for these results comes from the pharmacokinetic parameters of the novel oximes, where oxime 15 has the shortest half-life of the three (2.5 h), and oxime 20 has the intermediate (5.2 h) and oxime 55 has the longest (13.8 h) half-lives, and these are related to the lipophilicities of the three oximes (Table 1). As a point of comparison, 2-PAM chloride administered IM in rats displayed a plasma half-life of 34 min, so considerably shorter than our two efficacious oximes.⁴⁶ Our explanation of these results is that the most quickly cleared oxime 15 is able to reduce hypercholinergic activity (and, therefore, glutamate-mediated excitotoxicity) quickly enough to dampen seizures but does not reside in the bloodstream long enough to provide protection of brain architecture from the OP insult, whereas the opposite is true of oxime 55 that takes longer to reach its peak circulating concentration but resides in the bloodstream long enough to afford neuroprotection. Oxime 20, an intermediate, can provide both remedies. Therefore, oxime 20 remains our lead compound at this point.

Table 1.

Pharmacokinetic parameters of novel oximes following single intramuscular administration (50 mg/kg) in male Sprague-Dawley rats

Oxime	Cmax (ng/mL)	T1/2 (h)	(Kow)
MSU 15	2925 (NA)	2.51 (NA)	0.056
MSU 20	2547 (640)	5.20 (0.59)	0.352
MSU 55	1227 (654)	13.8 (0.31)	1.461

Note: Data represent mean (SD) of three rats; taken from Ref. 45. Only two rats were included in the oxime 15 group.

K_ow, octanol/water partition coefficient; NA, not applicable.

Since one of the deficiencies in the current oxime therapy is the short plasma residence time of 2-PAM requiring repeated administrations, the longer half-lives of our oximes, especially oximes 20 and 55, suggest that that they would be effective for a longer period of time following OP poisoning and, therefore, might be effective with fewer repeated administrations, thus reducing the likelihood of adverse side effects.

A question arises on how a pyridinium oxime can enter the brain since it has a permanent positive charge. A study on the p-glycoprotein (p-GP) export transporter in a model cell culture system indicated that some of our oximes that showed the greatest efficacy in vivo in reducing rat brain ChE inhibition following OP challenge were the poorest p-GP substrates, suggesting that they are less likely to be exported from the brain once they diffuse in.⁴⁷ Conversely, some of the novel oximes that were not, or very poorly, effective in reducing rat brain ChE inhibition following OP challenge were the best p-GP substrates, indicating that they are probably readily exported from the brain and would not achieve sufficiently high concentrations in the brain to be therapeutic.

Conclusion

In conclusion, our novel substituted phenoxyalkyl pyridinium oximes have shown a variety of results that indicate that they have potential, either alone or in combination with 2-PAM, to be more effective antidotes than what is currently available. A summary of some of these results has recently appeared.⁴⁸ These oximes have shown that they can reduce the level of brain ChE inhibition following OP challenge. They have some broad-spectrum efficacy, being effective with two highly relevant nerve agent surrogates for sarin and VX and also insecticidal chemistry, paraoxon, which is representative of a number of insecticidal OPs. They can provide survival of lethal-level challenges of the nerve agent and insecticidal chemistries. They attenuate seizure-like behavior induced by the nerve agent and insecticidal chemistries. They provide neuroprotection through the dampening of the astroglial response and by preserving neuronal structure. They have a longer residence time in the blood, perhaps related to their lipophilicity, which might make them practical for use with OP poisoning scenarios that result in slow OP absorption. Therefore, our novel oxime platform continues to show very promising results for future continued development, and these results are consistent with the overall goals of the NIH CounterACT program that has supported the last 5 years of this project.