‘Applications of Mass Spectrometry into the UK Microbiology Services’

Haroun N. Shah, Head, Molecular Identification Services Unit. HPA, Centre for Infections:

Mass Spectrometry: Mass spectrometers provide an accurate measurement of the molecular weight (Mr) of a compound in addition to information on the fragmentation of the molecule that helps deduce its molecular structure. Figure 1 shows the structure of a respiratory quinone (menaquinone, MK-5). The ‘largest’ mass ion (580) corresponds to the Mr of the compound, the peak at 565 indicates the loss of its methyl group (CH3=15) while intense peaks at 225 and 187 provide evidence for the structure of a quinone. The regular abundance peaks of 68 mass units are from the side chain which extends from the 3 position of the ring structure. Such analysis has been the driving force in the era of chemotaxonomy (ca. late 1970s to early 1990s) and paved the way for the prevailing period of molecular phylogeny. However, like other forms of mass spectrometry these earlier instruments did not have the capacity to determine masses greater than 1000 daltons. Proteins, for example, may be several hundred thousand daltons and had to await the arrival of MALDI-TOF-MS before they could be analysed (Karas M., Hillenkamp F. 1988. Anal. Chem. 60. 2299; Tanaka et al., 1988 Rapid Commun. Mass Spectrom. 2,151).

MALDI-TOF-Mass Spectrometry: There are numerous variations and permutations of the basic technique (see van Baar, 2000). Two examples, MALDI-TOF-MS and SELDI-TOF-MS, that have potential applications in clinical laboratories are briefly discussed here.

In order for these technologies to make their way to the forefront of diagnostics, they must be at least semi-automated, rapid, robust, cost effective, high throughput, reproducible between laboratories and sufficiently versatile to enable each laboratory to adapt it and optimise it for its specific needs. The analysis of intact cells by MALDI-TOF-MS fulfils many of these criteria and is now being developed specifically for use in such laboratories. The principle of the technique is simple and takes place in nanoseconds (ns). A matrix material (an organic solvent) is essential for the formation of the charged ionic species and, its addition to the sample enables the formation of a crystalline lattice with the surface molecules of the sample. A brief pulse (2 ns) of nitrogen laser light (337 nm) is directed at a minute section of the sample in vacuo causing the molecules to vaporise and ionise almost instantaneously. This produces a complex mixture of ionic species (positive and negative charges) that are free to move in nanoseconds. Because there is a constant potential difference between the sample target plate and the detector (left and right respectively in Figure 2), the ions will travel according to the law of conservation of energy. Thus, the ions with the lower mass to charge ratio (m/z) will travel at a greater velocity and reach the detector faster than the larger m/z ions. The differences in m/z ratios will result in differences in the time of flight of each mass ion and is the basis of Time of Flight Mass Spectrometry. The resulting output of such analyses is a mass spectral profile that determines the molecular masses of ions in the original plume. In MALDI-TOF-MS, the ions carry single charges, hence for a given bacterial species, characteristic mass ions in the spectrum may be directly flagged up as unique biomarkers and this approach of characterising microbes by specific mass ions has been adopted by several workers.

Genetic diversity among microbial species may be marked, hence reliance on a few biomarkers for comparative analysis of strains within a species may lead to error. Consequently, we have endeavoured to utilise most of the significant mass ions within the spectrum (Shah et al. 2002 , Keys et al., 2004). Preliminary studies based upon the analysis of several hundred species has shown that using two matrix solutions, -cyano-4-hydroxy-cinnamic acid (-cyano) for gram negative species and 5-chloro-2-mercaptobenzothiazole (CMBT) for gram positive, unique profiles may be generated for each species (see Figure 3 below).

Surface Enhanced Laser Desorption/Ionisation Time of Flight Mass Spectrometry(SELDI-TOF-MS) utilises a ProteinChip array to capture various classes of proteins followed by analysis using a MALDI-TOF Mass spectrometry. SELD-TOF-MS is currently used to search for “biomarkers” for human diseases and many applications are related to various cancers as well as neurological disorders. However, its use in proteomic analysis of microorganisms is now being explored.

Our laboratory is developing the methodology to search for unique biomarkers of human pathogens with potential application as diagnostic markers, elucidating new pathogenic determinants and potential vaccine targets. We also aim to establish a global database for microbial identification based upon unique bacterial mass spectral profiles. Unilke MALDI-TOF-MS described above where the majority of mass ions are collected between 500 -10,000 daltons, using SELDI proteins up to 150,000 daltons have been analysed. Figure 5 shows the SELDI-TOF-MS profiles of different classes of proteins collected Clostridium innocuum.

Data analysis: The analysis of mass spectral data is highly complex due to the nature of the proteome of microorganisms. In order to identify potential biomarkers, advanced analysis software packages need to be employed. In collaboration with G Ball and L. Lancashire (NottinghamTrentUniversity), we are currently using Artificial Neural Networks Analysis (ANN, a form of data analysis that is loosely modelled on the human brain multiple neuron interconnections), to develop models for various pathogens. The output leads to a predicative value for each species. The sensitivity of the model depends upon the quality of the data and the nature of the species, but studies so far have indicated >95% success. Table 1 shows the results for N. gonorrhoeae (predicated value, <1.5). The red rings indicate the strains that did not conform while Figure 6 shows the results for a study of 106 strains of Staph. aureus. The ANNs model predicated a continuum of variants between MRSA (1.0) and MSSA (2.0) and subsequently identified 97 % of the population correctly.

.

Summary and Conclusions:

The parameters (both biological and analytical) that are likely to affect the stability and reproducibility of mass spectral patterns of bacterial species have been extensively studied and a set of standard operating procedures devised. Using these protocols, a database of some 4,000 mass spectral profiles has been established (Keys et al. 2004). However, it is well recognised that clinical isolates will differ in certain traits from these strains and ongoing studies with samples taken directly from a hospital laboratory are now being tested blindly against this database. Early indications are that some taxa will be readily identified but the challenge for the technology will depend upon the database carrying sufficient diversity to successfully match against its corresponding counterpart. Preliminary results are promising and indicate that MALDI-TOF-MS will have a significant role to play in the future of modern diagnostic microbiology. Since MALDI-TOF-MS analyses the surface-associated molecules while SELDI-TOF-MS profiles the intracellular/membrane-bound proteins of a cell, the combination of both methods should provide a powerful new high resolution tool for microbiology.

References:

Shah HN, Keys CJ, Schmid O, Gharbia SE. Matrix-Assisted Laser Desorption/Ionistaion Time of

Flight Mass Spectrometry and Proteomics; a New Era in Anaerobic Microbiology. Clinical

Infectious Diseases: 2002: 35 58-64.

Keys CJ, Dare DJ, Sutton H, Wells G, Lunt M, McKenna T, McDowall M, Shah HN (2004) Compilation of a MALDI-TOF mass spectral database for the rapid screening and characterisation of bacteria implicated in human infectious diseases. Infection, Genetics and Evolution: 4, 221-242.

van Baar, B. L., 2000. Characterisation of bacteria by matrix-assisted laser desorption/ionisation and electrospray mass spectrometry. FEMS Microbiol Rev 24, 193-219.

O. Schmid, G. Ball, L. Lancashire, R. Culak and H. Shah (2005). New approaches to identification of bacterial pathogens by surface enhanced laser desorption/ionization time of flight mass spectrometry in concert with artificial neural networks, with special reference to Neisseria gonorrhoeae. J Med Microbiol; 54: 1205-1211

1