(16k)Until recently when unusual amino acids or blocked N-termini appeared in our sequences, we would call on a full-time mass spectrometrist to help us out of our difficulties. Our perspective changed at the 1988 MPSA conference in Berlin when Peter Roepstorff presented mass data not only on peptides but also on proteins using a plasma desorption time-of-flight mass spectrometer (PDMS). This commercially available PDMS instrument appeared to be simple to operate, relatively inexpensive, and would not require the full-time attention of a mass spectrometrist. We decided that the time had arrived for us to invest in mass spectrometry in our own laboratory. However science moved faster than our funding. The classic 1989 Science paper by John Fenn (1) on ESMS drew our attention to a different type of analyzer. The decision to purchase the more expensive ESMS instead of the PDMS was based on specific needs in our laboratory. For us the approximate 10-fold increase in the accuracy of mass determinations is important. We also had a need to obtain sequence information from fragmentation data, run in the neutral loss mode (e.g. for the detection of phosphorylated peptides), and to examine directly liquid chromatographic eluates in the LCMS mode. Instrument developments in this field move quickly and a purchase today would require a re-evaluation of a new stable of possibilities.
The following table has been constructed from the literature, our own experience, and discussions with users.
(16k)
Although most of the statements are clearly debatable, we have tried to take a conservative position particularly in regard to instrument manufacturers' claims. We realize, however, that quite different results may be obtained depending upon the particular mass analyzer/detector that is being used as well as the level of experience of the investigator. In addition, the results are usually quite sample dependent and, in the case of mass accuracy, also dependent upon the mass of the compound being studied. For a more complete discussion of the problems attendant to the generalizations made in the following table, the reader is referred to the accompanying references (2-5). Despite its obvious and many limitations, we believe the following table nonetheless provides a useful starting point for the novice to qualitatively compare the wide array of MS ionization/detector modes that are currently available.
With some trepidation we were prepared to train ourselves to become part-time mass spectrometrists, however luck was on our side. A few months before delivery of our ESMS, Richard Johnson (former graduate student of MIT's Klaus Biemann) walked into Ken Walsh's office wondering if we knew of any job openings for a mass spectrometrist! Under his leadership the learning period has been shortened and we have been able to move into some advanced studies.
This somewhat typical laboratory history illustrates a few of the problems involved in introducing new technologies into a core facility. Timing, funding, convincing yourself and others that the instrumentation is important, risk taking, and a bit of luck all play important roles.
The grant application of course included a long list of potential applications, particularly since this was to be a shared instrument facility. The real surprise is the much longer list of applications for which the ESMS is now being used.
I. Synthetic Peptides
We check peptides on the ESMS before any amino acid or sequence analysis. Without automatic injection we have run 92 samples in an afternoon. A reliable time saver is Terry Lee's MacBioSpec program(6) which has several features including a program for calculating the theoretical mass of peptides and proteins. For small peptides we have found it important to keep the orifice voltage as low as possible to minimize partial fragmentation on entry. Synthetic chemists need an estimate of the purity of the intact peptide, and unintentional fragmentation in the MS falsely indicates a low yield of the whole peptide.
It is a joy to work with these people. They usually come in with umoles or nanomoles of material and we need only picomoles. I encourage them to bring in their peptides straight from the HPLC, still dissolved in the TFA/water/acetonitrile solvent which is quite compatible with ESMS. Dry samples can usually be dissolved in 1% formic acid, 50% methanol -but it is a good idea to add the formic first, followed by water and then methanol. Occasionally 5 mmolar NH4Ac in 50% to 100% methanol comes in handy and, in the case of very hydrophobic peptides, a cocktail containing chloroform is useful.
ESMS (and occasionally ESMS/MS to check sequence) may well be the fastest and cheapest method to check for the removal of blocking groups and the purity of the synthetic peptide. The quality of product supplied by the synthetic chemists to the end-users has improved dramatically because of the speed and ease of ESMS checking.
II. Peptides Isolated from Proteins
Everything in section I. applies, except that the investigators usually have smaller amounts of material available. Since signal-to-noise ratios become important, it is critical that peptides from an HPLC run are from clean systems and are collected in clean tubes. Most of our investigators rinse their tubes in formic acid, Milli-Q water, and methanol followed by drying before collecting their paltry picomoles of peptide.
With small amounts of sample (e.g. <50 pmoles for particular sequences), contamination and/or adsorption of peptide onto the container wall can be avoided by continuous and direct injection into the ESMS of the HPLC eluate in an LCMS run. Since most HPLC pumps have a higher flow rate than recommended for ESMS (2 to 20 Ill/min.) the outflow from the HPLC column is split with approximately 90% going to the UV monitor and collector, and the remaining 10% directed to the ESMS. The amount of data collected from a single LCMS run can easily total 2MB, so it is important to have a large hard drive and frequent downloadings onto floppies or removable disk cartridges.
Occasionally the investigator has a cDNA sequence and/or partial sequence from Edman degradation of the peptide. The mass determined from ESMS can be used to calculate the length of the peptide. When the mass of a peptide or protein determined by ESMS does not equal the calculated mass from translated cDNA sequence or previously determined amino acid sequence, we look for unusual or modified amino acids, attached residues (carbohydrate, phosphate, fatty acids, etc.) and/or errors in the reported sequence. Based on our experience, many proteins do not have the predicted mass. Revisions of sequence or identification of post-translational modifications are more a rule than an exception.
III. Example: Mass of a Protein and its N-Terminal Tryptic Peptides
Retinal recoverin is a recently identified and sequenced protein (7). Working with Alexander Dizhoor in Jim Hurley's research group, we found novel heterogeneity of fatty acylation at the N-terminus.
By ESMS, non-acylated recombinant recoverin from E. coli had a mass of 23,203.6 Da, consistent with the loss of initiator methionine, but native bovine recoverin had a mass of 23, 412.2 Da (standard error of the mean = 0.8), thus a difference of 208.6 Da. The N-terminal sequence of GNSKS matches a motif that would be recognized by N-myristoyl transferase, however myristic acid would result in an addition of 210 Da. Tryptic digestion followed by HPLC purification yielded four N-terminal tryptic peptides, two of which contained double bonds: C14:0-GNSK 614.3 Da; C14: 1-GNSK 612.3 Da; C14:2-GNSK 610.4 Da; C12:0-GNSK 586.4 Da with the myristoleate (C14:1) being the most abundant. The mass of the non-acylated GNSK peptide is 404.2 Da. The array of additional mass found in the four N-terminal peptides is due to heterogeneous fatty acid acylation.
The ESMS literature is growing very rapidly. Articles by Carr, et al. (2) and Edmonds and Smith (3) are good places to start. The latest innovations can best be found in poster sessions.
References
1. Fenn, J.B., Mann, M., Meng, C.K., Wong, S.F., Whitehouse, C.M., (1989) Science 246 64-71.
2. Carr, S.A., Hemling, M.E., Bean, M.F., Roberts, G.D. (1991) Anal. Chem. 63 2802-2824.
3. Edmonds, C.G., Smith, R.D., (1990) Methods in Enzymology 193 412-431.
4. Jennings, K.R., Dolnikowski, G.G., (1990) Methods in Enzymology 193 37-61.
5. Roepstorff, P., Klarskox, K., Andersen, J., Mann, M., Vorm, O., Etienne, G., Paello, J. (1991) Int. J. Mass Spectrom. 111 151-72.
6. MacBioSpec Program, available from Sciex Corp., Toronto, Canada.
7. Dizhoor, A.M., Ray, S., Kumar, S., Niemi, G., Sprecer, M., Brolley, D., Walsh, K.A., Philipov, P.P., Hurley, J.B., Stryer,L., (1991) Science 251 915-918.
Return to the The ABRF Home Page