Re: ProtSeq PVDF blots

Steve Latshaw (latshaws@mis.finchcms.edu)
Thu, 14 May 1998 09:54:57 -0500

Deb,
I went back through some of the old sequences where we had peptides
with DP bonds and out of 9 sequences two had observable DP cleavage during
the first cycle and may have represented a failure to thoroughly dry the
sample prior to sequencing. None of them showed any significant increase in
the Pro background over what I would have expected. Glycosylated proteins
may be a double whammy providing both -OH and trapping water during drying,
I've always recommended deglycosylation to our clients because of the poor
success rate we've had with them.

Good luck,

Steve Latshaw
Chicago Medical School

At 11:38 AM 5/13/98 -0700, Deb McMillen wrote:
>Steve, Thanks for the acetic anhydride idea. I will look that up.
>
>Do you think then that under the anhyrdrous conditions, not even DP is
>breaking significantly (in a case where there is low S and T content).
>This particular protein, however, is loaded with S and T.
>
>On Wed, 13 May 1998, Steve Latshaw wrote:
>
>> Deb,
>> As I recall, a number of years ago George Tarr had studied
>> nonspecific background cleavage and had related it to the Ser and Thr
>> content of a protein. Under the anhydrous conditions of TFA treatment any
>> cleaving of the peptide bond would require a nucleophile and with no water
>> around S/T could provide its -OH. He did a modification of the sequencer
>> chemistry to include an acetic anhydride treatment after the initial
>> coupling step. It may be in one the Techniques in Protein Chemistry as it
>> was presented as a poster at one of the Protein Science meetings.
>>
>> Steve Latshaw
>> Chicago Medical School
>>
>>
>> At 01:58 PM 5/12/98 -0700, Deb McMillen wrote:
>> >Thanks (Gautam, David, Jacek and Len) for the responses to my question.
>> >
>> >I have now tried cleaning up the PVDF blot with 50% metoh with TEA. The
>> >signal is a lot cleaner, however, it is clear that the protein is still
>> >rapidly falling apart. Turns out from the sequence that it has 3 DP
>> >bonds. In addition, it has several other bonds that I understand are acid
>> >or base labile: 1 GP, 1 DG, 2TP and 2 NG bonds. Does anyone know if any
>> >or all of these bonds do break during Edman degradation under the
>> >conditions in the ABI 470? Are there any other bonds that are
>> >particularly labile under these conditions?
>> >
>> >Thanks for your help,
>> >Deb
>> >
>> >
>> >On Mon, 11 May 1998, Deb McMillen wrote:
>> >
>> >> Len, hello,
>> >> I have a (long) question for you.
>> >>
>> >> I have a few PVDF samples that have been very darkly stained with
>> >> Coomassie Blue. On sequencing, there is one major sequence, but the
>> >> background is very high--and gets too high to read the sequence very
>> >> rapidly. On one of these samples I know the sequence and will check for
>> >> acid/base labile peptide bonds to see if the protein is degrading rapidly
>> >> in the sequencer--recently I had a sample with 6 DP bonds and it broke
>> >> down quickly and gave a very large "background" which continued to
>> >> increase at each cycle.
>> >>
>> >> I remember though that you had a method for "cleaning up" PVDF blots. Can
>> >> it be on these very dark PVDF samples that the Coomassie Blue is hanging
>> >> on to cleaved-derivitized amino acids and that these then continue to come
>> >> off the PVDF during subsequent cycles, as well as hang on to more freshly-
>> >> cleaved amino acids?
>> >>
>> >> Do some brands of PVDF cause large background in the sequencer?
>> >> In both of these current cases I cannot get the client to increase their
>> >> protein/PVDF ratio--so there is quite a bit of PVDF in the reaction
>> >> chamber.
>> >>
>> >> And/or do proteins that are glycosylated typically show large background
>> >> on sequencing? On proteins that we have thought were glycosylated we
>> >> would see cycles with huge peaks at some amino acids (seemed to be a
>> >> pattern to the amino acids that were up) and then see fairly clean cycles
>> >> scattered in between--supposedly amino acids that were not glycosylated.
>> >>
>> >> Thanks for any help,
>> >> Deb McMillen
>> >>
>> >> Institute of Molecular Biology
>> >> University of Oregon
>> >> Eugene OR 97403
>> >>
>> >
>> >
>>
>
>