Detect to Protect

The false alarm in the Senate Office Building  last February reminds us that rapid and reliable analysis of chemicals and airborne microorganisms is required in public buildings as well as on the battlefield.  Biological agents present a particularly complex challenge, and a number of analytical methods have been evaluated for speed, reliability, ruggedness and automatability. 

Among detection methods for a building air sampling system mass spectrometry has the strength of being very 'broadband': where immunoassays or other selective techniques ask 'Is it there?'  MS can, quickly and with high sensitivity, ask 'What is there?'

Dr. Fenselau presented examples of detecting specific proteins from TMV, Sindbis virus, and bacillus cereus spores, which had been acid-solubilized from particulates isolated by an air extraction system.  Early work showed the limitations of library matching, and the advantages of database searching methods now familiar from proteomics.  Significant refinements were made in bioinformatics,  scoring the significance of the peptides tabulated from the genomes of different organisms, for example, e.g. 10 peptides from H. pylori's small genome had a p-value of 0.001, where 10 peptides from E. coli’s  big genome were far less significant.  Looking forward, significant analysis time could be saved by monitoring for peptides likely to come from organisms of interest: in silico cleavage of several bacillus strains selected several peptides unique to only one of the strains. 

Questions clarified that this work has primarily used MALDI ionization, which gives a fast quick answer.   A caution with electrospray ionization is that microfluidic systems can be unreliable in the desert at 135 C. 

Putting the Pieces Together for Proteomics Strategies

 A great number of proteomics tools are being developed.  Each proteomic strategy must be optimized for the type of sample (blood, tissue), the chemical nature of the proteins (hydrophobic, basic), and the information sought (quantitative comparison, inventory).  In addition to the mass spectrometric method careful attention is also required in selecting:  : 

   Reproducible fractionation of protein mixtures
   Enrichment of minor components
   Facile methods for quantitative comparisons (esp. in clinical samples)

 The question of  how cancer cells become resistant to drugs has been important enough medically to be targeted by traditional methods, one protein at a time, for over thirty years.  It is thus a priority to apply preoteomics technology to a global analysis of normal vs. drug resistant cells. 

Because 95% of the drugs target membrane proteins, we start with isolation of the membrane -- “the skin of the grape”.  For this application, the cell membrane is best isolated by first coating the cells with silica microbeads, crosslinking the membrane to the beads, then isolating the beads.  This avoids co-purification of intracellular membranes (golgi, ER, mitochondria).  Intense gel bands seen from the resulting membrane fraction are indetectably dilute in a whole-cell lysate. 

Metabolic labeling, mixing normal cells grown in unlabelled media with drug-resistant cells grown in defined media containing ¹³C₆ Lys and Arg,  provides the best internal control for the lossy processes of lysis, extraction, digestion, drying and resuspension.  However, patients or animals can’t be isotopically labeled, so such samples must be labelled and mixed after extraction.  Of the available labelling strategies  (ICAT, GIST, iTRAQ, 2MEGA), the preferred is ¹⁸O water.  It is best performed by re-binding the peptides to trypsin after digestion in ¹⁶O water, because peptides are easier to redissolve than proteins, and it uses less ¹⁸O water.   Note that some peptides exchange in a few minutes; others require over 90 minutes; Arg is generally faster than Lys. 

Solution IEF of proteins before digestion was immensely valuable.  IEF was found to concentrate as well as to fractionate the sample before digestion and LC-MS analysis, and, significantly, to remove salts.   IEF was performed using immobilized ionophore membranes, with a caution that brands vary (some tore easily).  6 LC chromatograms of the contents of 6 chambers (between membranes with pI 3, 5, 6.5, 8, 11) were reproducible, and contained comparable numbers of proteins (30 – 60).  Comparing IEF-LC-MS/MS with LC-MS/MS, it identified 281 vs. 54 peptides (167 vs. 24 proteins) using 200-fold less sample.