Message-Id: <v03007802aef2fadda7ab@[131.215.4.147]>
Date: Fri, 3 Jan 1997 11:06:02 -0800
From: Gary Hathaway <hathaway@cco.caltech.edu>
Subject: re: high sensitivity MS/MS sequencing
To: Recipients of ABRF List <abrf@aecom.yu.edu>
Ken Williams wrote:
We are coming under ever increasing pressure to offer MS/MS sequencing
on very low/sub-pmol amounts of "unknown" proteins separated by SDS PAGE -
Note: this was originally posted directly to Ken. He has requested I post
it to the bulletin board, so you can complain to him!
Ken,
I am responding to your inquiry although I do not consider myself either
an expert, nor as having a great deal of experience in this area. Still,
since I have posed the same or similar questions to a number of people in
and outside of the ABRF, I can offer some info albeit anecdotal and second
hand.
1.) not all peptides break well in low energy collision experiments and
thus incomplete sequences are usually mode o'day. However, tryptic
fragments almost always contain a high pK lys or arg at the C-terminus
which helps sensitivity and gives somewhat easier(?) spectra (mostly Y-ion
series) to interpret.
2.) data are generally collected much quicker than they are interpreted. On
the '97 peptide, I collected data for about 20 minutes on my TOF instrument
which requires tuning the reflector and collecting about 20 spectra. I
spent the better part of a day analyzing it. I must admit I had analyzed
only a few peptides previous to this and so I wouldn't say my experience is
a good gauge of efficiency in this area. Point is, analysis time to data
collection time is high for this technique and so for cost recovery, you
may have to institute an hourly charge if you haven't already. This will
surely cause problems unless you or someone else highly visible on your
campus is outspoken and can make the risks for this analysis known and
acceptable.
3.) most of the people who responded to my inquiry as to what is needed to
setup an MS/MS operation said that a minimum of two, and better three
individuals were necessary with one person doing sample prep., one computer
expert and the other a good protein chemist. For myself, I think two people
should be able to handle the job depending on throughput desired.
4.) some who responded decried the hype certain authors have added to this
area, raising expectations beyond what good people can routinely deliver.
The feeling that these so called experts have done protein chemistry a
disservice was expressed. My guess is that for routine identification of a
known protein, 3-4 picomoles in a gel slice are required. For de novo
sequencing, about double that.
On a personal note, I use a MALDI-TOF with collision cell. I believe this
setup has distinct advantages over on-line systems. Resolution on this
instrument will soon be extended to 15,000 fwhm at mass 8500(!) allowing
greater resolution of mixtures for timed-ion selection. Analysis time is
extended to near infinity and mixtures of peptides can be sequenced. With
the newer polycrystal layer matrix techniques (Xiang and Beavis, RCMS 8,
199-204 (1994), there is no longer the stumbling block of sample cleanup
following digestion. I believe this is a largely overlooked major
development. I now use an X/Y robotic station which I've set to split hplc
fractions to collect 5-10% of an hplc peak fraction directly to the mass
spec. stage, or with low flow, all of the peak. With Randy Nelson's
(Dogruel et al. Anal. Chem. 67, 4343-4348, (1995)) or (Parker et al., Prot.
Science 5, 2329-2332, (1996)) covalently coupled trypsin and other enzymes
will allow separation of complex mixtures of proteins with direct in-situ,
accelerated digestion (1 hr., 50C) and analysis with NO SAMPLE HANDLING!
One cannot overestimate the increase in sensitivity which will accrue from
reaching the "nirvana" of no sample handling. With microbore or capillary
HPLC, nanogram amounts of proteins can be collected directly and digested
for identification without the losses inherent to in-gel digests.
Finally, it is regrettable that historically, both chemical sequencing and
mass spectrometry have been compromised largely as a result of the
insistence of biochemists and molecular biologists to hang on to
polyacrylamide gel electrophoresis as the final step in protein
purification. What is needed, is a two dimensional method which is
compatible with the analytical methodology. I think wherever possible, we
should encourage the use of 2D HPLC or develop some other technologies
which get around the gel problem (Kassel et al., Techniques in Protein
Chemistry VI (John Crabb, ed.) pp 39-46).
I think this overly long note should be sufficient to pay my "lurker dues"
for the rest of the year.