What kind of autosampler are you using? That and the HPLC might=20
make a difference in some specific answers, however there are=20
some things you could try to isolate the problem:
1. Set up your 4 rep's, but for the first one use A buffer=20
instead of sample. Now try step 2.
2. Run a blank run (no injection) before your 4 rep's. If there=20
is no loss, it could be a column equilibration problem. You need=20
to be sure the column is in the exact same state (with respect to=20
how much B buffer it has seen and for how long) for each of the 4=20
rep's.
If step 1 didn't have the 20-30% loss and 2 did, look for air=20
leaks or other problems in the autosampler that could cause the=20
first injection to be short. If you are using an HP-1090 do an=20
injector wash.
You didn't say if this happens only in the first run of the day,=20
or how long the interval between samples is. Some columns will=20
'eat' a small amount of protein after extensive washing, then=20
give stable recoveries until they are stripped of that initial=20
amount of protein. This is most pronounced with the first=20
injection on a new column, but can happen to a lesser extent=20
after any long wash. If that is the case, the best suggestion I=20
have is to run a protein of similar molecular weight to your=20
sample an the beginning of each day. This will also, cycle the=20
column to be sure it is properly equilibrated for subsequent=20
injections.
If this doesn't help post more specifics on the equipment and how=20
you are running these samples.
Good luck.
--tks
Tim Slattery
Protein Biochemistry and Biophysics Department=20
Berlex Biosciences
tim_slattery@berlex.com
=20
______________________________ Reply Separator __________________________=
_______
Subject: autosampler help
Author: JHolt@US.PMC-Vacc.Com (Holt; John - PMC-US) at Internet
Date: 11/16/98 4:11 PM
I am hoping someone in the group can straighten out a newish autosampler=20
user ....
=20
When we inject four replicates of a test sample, the signal in the firs=
t
run is 20-30% lower than in the three that follow, which are in pretty go=
od=20
agreement. This happens whether the replicates are obtained by sampling =
the=20
same vial four times, or by sampling four separate vials once each.=20
Moreover, a blank run follows every test run, and shows no significant=20
peaks, so it's hard to see how some sort of cross contamination is=20
occurring.
=20
All insights gratefully received!
=20
John Holt
=20
Pasteur M=E9rieux Connaught