Ion mobility spectrometry (IMS), which as a means of compound analysis/identification, has been with us since the early 1970s, is currently at the heart of a number of IMS-based products for the trace detection of narcotics, explosives, toxic industrial chemicals (TICs) and chemical warfare agents.
In addition, since year 2000, IMS has seen rapid acceptance in the pharmaceutical manufacturing sector as a means of reducing cleaning verification time relative to other methods-such as high performance liquid chromatography 1.
Most recently, there has been interest in the development of IMS instrumentation and applications for the detection of biological (warfare) agents and, in the consumer industry, the detection of specific bacterial pathogens in processed and unprocessed foods.
This article details how IMS techniques can, rapidly and without recourse to specialized test kits and reagents, fingerprint and differentiate between bacterial strains by the analysis of whole bacterial cells.
How It Works
IMS measures the characteristic speed at which ions move through air under an applied electric field. For example, in a typical Smiths Detection IONSCAN® analyzer (Figure 1), clean, dry air transports thermally volatized sample material into an ionization/reaction region where ions (and/or clusters of ions) are formed. The ions are then “gated” into a drift chamber where (restricted to narrow path of drift by a series of magnetic focusing rings) they are accelerated, under an applied electric field, toward a collector electrode.
The brief opening of the gate represents the start of the “scan period” -which ends just before the next opening of the gate. Ions travel at different rates based on their size, charge to mass ratio, and shape. They are identified by measuring the time required for them to reach the collector electrode (ion drift time) relative to that of an internal calibrant. Also aiding identification is the assessment of how much “charge competition” takes place between the sample material’s ions and those of the calibrant or of the calibrant and a reagent.
The data from several successive scans are combined to improve signal to noise ratio and to produce a 2D “segment.” A series of segments, each with characteristic ion peak patterns for the strain, can then be displayed: either as a stacked series of segments (a 3D plasmagram, see Figure 2) or as a 2D plasmagram for the average of all segments.
Plasmagrams and Analysis
The 3D plasmagram shown in Figure 2 was obtained from the analysis of Listeria monocytogenes cells. The x-axis shows the drift time of the detected ions in milliseconds and the y-axis the amplitude of the ion peaks in digital units (du). The z-axis stacks the individual segments, in this case obtained at 1s intervals, over a 60 segment analysis period, to show how the segment profiles change over time. The profile changes, due to the heating of the desorber and the tendency for smaller molecules to be released and detected first, is very apparent when one looks at a number of the segments that contributed to the overall 3D plasmagram. Figure 3, parts A, B, C and D, shows four of the 60 segments from Figure 2.
For the analysis of Listeria monocytogenes the bacterial cells (obtained from the America type culture collection-ATCC, Manassas, VA, USA) were cultured using standard techniques and microgram quantities were placed on a Teflon® membrane onto the desorber of the IONSCAN.
In other analyses, using different strains and conditions of growth and varying the desorber temperature, different, but reproducible, plasmagrams were produced (in both positive and negative modes). Such findings indicate, strongly, the potential use of characteristic IMS bacterial plasmagram fingerprints for the identification of and differentiation between specific bacterial strains and species, including pathogens.
Furthermore, the analysis of more than 200 different strains/species of bacteria showed that no two had the same combined positive and negative mode IMS fingerprints: suggesting, again strongly, that combined mode fingerprinting has the potential of strain classification and differentiation.
Interestingly, in many of the 200 analyses performed, many of the cell components that gave rise to specific ion peaks in the plasmagrams were unknown. It was not, for instance, known if the detected ion peaks were from intact volatile cell constituents or from products resulting from the thermal decomposition of larger molecules during desorption.
Important clues though were unearthed during the analysis of Serratia marcescens-wild-type strains of which produce a red pigment, called prodigiosin, but only at low growth temperatures and in the absence of glucose. See Figure 4, parts A, B and C.
The prodigiosin could of course have degraded into a number of components, only one of which attracted charge and reached the detector. Not only was this unlikely it was ruled out by measuring the Ko value.
Measured Ko values can be used to estimate the molecular mass of an ion by the following relationship: Molecular mass = a/Ko + b, where a and b are constants relative to the IMS system, including the drift tube design and temperature. Using IONSCAN derived a and b values determined for typical narcotic molecules at a drift tube temperature of 225°C, the measured Ko value of prodigiosin (1.0252) in the studies detailed above, at a drift tube temperature of 190°C, corresponded to an estimated molecular mass of 378. This strongly indicates that the peak observed (B25) did in fact result from the ionization of intact molecules.
More Results
Other recent studies into the use of IMS for fingerprinting bacterial strains focused on distinguishing between Listeria monocytogenes and other, non-pathogenic species of Listeria. cells from growth on Listeria selective agar were analyzed directly by IMS and in parallel by standard methods. There was excellent correlation.
Specifically, the preliminary IMS Listeria test results were based on a simple algorithm making use of two ion peaks, designated L-1 and L-2, seen in negative mode plasmagrams of cells desorbed at 270°C. L-1 was detected in all species of the genus, while L-2 was found only in Listeria monocyogenes.
This noticeable difference was exploited and programmed into an IONSCAN for the automatic screening of Listeria samples. Specifically, the algorithm employed used the Ko values of 1.0266 for L-1 and 1.0412 for L-2 and IMS ion peak detection criteria including: drift time interval, ion peak amplitude, ion peak shape and the number of consecutive segments in which L-1 and/or L-2 were detected.
The L-1 and L-2 peaks had characteristic Ko values arising from distinct molecules which can be separated by desorption temperature. When a sample is first introduced into an IMS instrument it is most likely at ambient temperature and will require a finite period of time to reach the temperature of the desorber. See Figure 5, parts A and B.
Conclusion
The results of the studies detailed above and other similar investigations, into the use of IMS for fingerprinting bacterial strains, are very encouraging: and clearly demonstrate the potential of using ion mobility spectrometry in microbial identification and strain typing. Applications for such fingerprinting include diagnostic clinical and food microbiology.
IMS analysis for strain typing requires no reagents or sample preparation. Nor does it require pure or at least enriched cultures as samples.
1 P. Miller, “Ion Mobility Spectrometry Speeds Cleaning Verification Time,” A2C2, vol. 5, no.7, (July/August 2002) p21.