Hello,
While I am not a professional MS person, I have had the privilege of working with many MS scientists and some knowledge has rubbed off despite my best efforts to stay benighted. All the correct information in the post below is therefore credited to the MS community whereas all the errors are mine :-) First of all, I hesitate to recommend a specific service provider, simply because I have been spoiled by having MS experts in-house throughout most of my career. Some tips on identifying the right kind of provider can be derived from general considerations regarding protein MS: Generally speaking you can expect a 0.03% mass accuracy from an ESI MS experiment, provided that your equipment is in good shape and the operator knows their stuff. For the end-user-run MALDI experiment around 0.3% mass accuracy can be expected. With an expert operator and a better-than-average ESI MS system one can expect 0.01% mass accuracy (+/-1 Da in 10 kDa). So, assuming good instrument and skilled operator, for a 40 kDa protein a typical mass error should be no more than 4 Da in either direction. However, if you submit a complex mixture of closely related proteins, you can expect a considerable loss of both accuracy and precision! The more pure the protein, the better the result should be. The instrument resolution is just as important. For instance, with instrument resolution of 5000 or 10000 it was not possible to resolve pyroglutamate derivatives (-17 Da) of antibodies from their parents (see reference below). Instead, the presence of pyroglutamate was observed as a mass shift of the main species and the ratio of unmodified/modified protein was quantified on the basis of this shift. On the other hand, mass accuracy for *different* antibodies was less than +/- 4Da. For a crystallographer, this generally means that in most cases an ESI-MS experiment will correctly identify your construct, as compared to an unrelated protein, proteolysis product, etc. However, a point mutant of a construct may not always be easy to recognize since some mutations are only different in mass by a few Daltons, which would be below the expected mass accuracy threshold. If one is suspected it is better to run a peptide map and hope that one's coverage is sufficiently broad to catch the peptide that contains the suspect mutation. Notably, an L<>I mutation is undetectable by intact-protein MS or by peptide-map MS, unless one considers fragmentation patterns of individual AA's. In general, the deeper one begins to delve into the MS of proteins, the more weirdness and fun effects one will uncover. With sufficient precision, accuracy, and sensitivity one can observe such fine phenomena as amino-acid mis-incorporation (done by the ribosomes that is). It is always there, but the amounts of wrong products are typically so low that most people never worry about them - unless they're looking for subspecies in e.g. a protein/peptide sample submitted for FDA approval as a therapeutic product. Practically speaking, in the majority of cases a straightforward ESI-MS of a protein done by a trained professional using a well-tuned machine is enough. I have routinely used this kind of information to detect e.g. mutants after a round of PCR mutagenesis, before the sequencing data were received. If you also sequence the construct DNA, you're nearly guarranteed to be positive in your identification. In the absence of sequencing data, a peptide map (with good coverage!) of the protein can be almost as accurate. For complicate cases involving e.g. glycosylation, hyper-phosphorylation, or other complex PTM additional experimental data may be needed. Good luck, Artem Reference: <http://www.sciencedirect.com/science/journal/10440305> Journal of the American Society for Mass Spectrometry Volume <http://www.sciencedirect.com/science?_ob=PublicationURL&_tockey=%23TOC%2352 70%232006%23999829993%23622790%23FLA%23&_cdi=5270&_pubType=J&view=c&_auth=y& _acct=C000050221&_version=1&_urlVersion=0&_userid=10&md5=1398427748390e93e7f 496907d5b2ad2> 17, Issue 6, June 2006, Pages 867-872 _____ From: CCP4 bulletin board [mailto:[EMAIL PROTECTED] On Behalf Of Patrick Loll Sent: Wednesday, September 05, 2007 12:42 PM To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] MS for verification of protein constructs I wonder if anyone would care to share experiences/ideas/biases that relate to the use of mass spectrometry to verify the identity of protein constructs used for crystallization. Our experience with different MS facilities has been checquered. Specifically: What's the current thinking on the best approach to get masses for intact proteins of moderate size (say, 40 kD)? ESI-TOF? What kind of resolution should one hope to obtain in such cases (10E-04?) Any suggestions as to good facilities offering fee for service MS characterization are welcome (but should be shared off line, I think; continental US only). Thanks, Pat ---------------------------------------------------------------------------- ----------- Patrick J. Loll, Ph. D. (215) 762-7706 Associate Professor FAX: (215) 762-4452 Department of Biochemistry & Molecular Biology Director, Biochemistry Graduate Program Drexel University College of Medicine Room 10-102 New College Building 245 N. 15th St., Mailstop 497 Philadelphia, PA 19102-1192 USA [EMAIL PROTECTED]
