Conquering the nerve gases: High-resolution mass spectrometry of chemical warfare agents

Nerve agent control

Nerve agents are banned under the Chemical Weapons Convention but that does not mean that we have seen the back of them. There is strong evidence that they have been deployed in the current civil war in Syria as well as in other recent conflicts. On top of that, stockpiles around the world are being decommissioned and research on nerve agents continues in the lab.

These activities have the potential to expose people to nerve agents, so reliable methods must be in place to detect them quickly. One of the conventional ways to do so is to carry out urine testing and mass spectrometric methods are at the forefront in this department. In particular, tandem methods involving GC/MS and LC/MS have proved to be successful. Most of them have been set up to look for specific metabolites of nerve agents, rather than a general non-targeted analysis, which could be seen as restrictive.

Sticking to mass spectrometry, scientists in the USA have taken a different approach. Guided by the success that high-resolution mass spectrometry (HRMS) has had in matching the performance of tandem methods, Elizabeth Hamelin and colleagues from the Centers for Disease Control and Prevention, Atlanta, and Oak Ridge Institute for Scientific Education have optimised an HRMS procedure for five nerve agents in urine. They described their method in Rapid Communications in Mass Spectrometry, in which one key advantage of HRMS was illustrated: the ability to detect extra compounds which are not in the target list.

High resolution

The team set up a panel of five urinary metabolites which are produced in humans from VX, Russian VX (also called VR), soman (GD), sarin (GB) and cyclosarin (GF). They were VX acid and VR acid, in which the thio substituents were replaced by a hydroxy group, and GD acid, GB acid and GF acid, in which the fluoro groups were replaced by hydroxy groups. A set of deuterated and ¹³C-labelled internal standards were also assembled to cover each metabolite. Urine samples were spiked with a mixture of the five nerve agent metabolites and the internal standards before being extracted by automated solid-phase extraction in a 96-well plate system on silica adsorbent. The extracts were taken up in aqueous acetonitrile for separation on liquid chromatographic systems linked either to a triple quadrupole mass spectrometer for MS/MS experiments or a high-resolution LC/MS system with a heated source. Both mass spectrometers were operated in negative electrospray ionisation mode.

For HRMS, the so-called balanced automated gain control setting was selected, in preference to the ultimate and high settings, as it provided the best balance between sensitivity and mass accuracy. The resolution itself was set to 50,000.

Urinary metabolites by HRMS

A total of 19 pooled urine samples were analysed on both instruments. The variabilities were less than 13.4% and 9.4% for HRMS and MS/MS, respectively, both values within the US FDA recommendations for analytical methods.

Using a mass extraction window of 10 ppm, the HRMS studies of blank urine revealed no contributing species or interferences for the five nerve agent metabolites. However, for GB acid, the background signal increased over time. This phenomenon remained unexplained but it was controlled by regular cleaning of the ion transfer tube in the mass spectrometer. In the MS/MS analysis of 72 urine samples from people with no known exposure to nerve agents that had been spiked with internal standard, one false positive result was recorded for GB acid, compared with no false positives at all for HRMS. Given that the retention factor was the same in both instruments, the difference was attributed to the mass spectrometer rather than the chromatographic system. The detection limits for MS/MS and HRMS were similar, in the range 0.14-0.70 ng/mL and the accuracies and precisions were both very high. Overall, the HRMS method performed as well as the MS/MS method for the measurement of the nerve agents. This is despite the fact that only one ion is measured in the HRMS procedure compared with two product ions in the MS/MS method.

However, measuring a single ion is restrictive because it does not provide the opportunity for compound confirmation from a second ion, for which a second mass spectrometric method needs to be devised. In addition, the isotopically labelled internal standards are not the best because they produce the same fragment ions as their unlabelled compounds. In an ideal world, more discriminatory standards would be used but the cost of their synthesis is prohibitive, say the research team. One big advantage of the HRMS method is the ability to detect other compounds which are not in the original target list. In this case, an ion corresponding to a metabolite which was not one of the five targets was detected. From its relative retention time, accurate mass and product ion, it was attributed to a metabolite of GA acid.

This study has shown that a HRMS method is suitable for the detection and measurement of nerve agents in human urine. Its attraction lies in the ability to screen and measure compounds in one method but its confirmatory powers are weak. As such, MS/MS will remain the method of choice until further work on metabolite confirmation is carried out.