Again, I am indebted to the ABRF for their assistance. Every time I am at=
my=20
wits end, somebody from this group brings me back with knowledge, the ben=
efit of
experience or even friendship and moral support. I want to thank you ALL =
for=20
that. I just wanted to follow up my sequencing question with a patchwork =
of=20
ideas and answers that were sent to me. We have a huge list of things to =
check=20
on our instrument. We've only done a few things so far so we don't have a=
ny=20
answers, but I feel confident that we'll figure this one out with everyth=
ing=20
I've learned from the experts out there. Below I've attached all of the e=
mails I
received. I cropped off all names so that I wouldn't inadvertently offend=
=20
anybody. Sometimes it's nice to see answers as well as questions, so here=
they=20
are...
1) This isn't a fix for your problem, but more a diagnostic suggestio=
n. If=20
you see very little lag along with a low RY, then the problem is likely d=
ue=20
to washout and all wash and reagent delivery steps should be examined as=20
possible problem sources. If your lag is terrible and you have a low RY,=
=20
then the chemistry is at fault and either a bad reagent(s) is/are present=
or=20
the optimization is at fault. If the optimization seems to be at fault, =
try=20
systematically changing the amount of R2 and R3 used up and down a little=
and=20
see what impact on RY is seen.
2) You do not say which sequencer you now have, but I take that it must=
be
either a 492 or 494. I can offer you the following hints, as to what we =
do
if we have problems - we have both a 477 and a 494. I take it that the
Error Log does not show any discrepancies.
What I have tried to teach our new users is to start from the one end of =
the
instrument, where the results appear.
Set up the sequencer to run a number of PTH standards and with this you c=
an
see how a Blank is, - if necessary put a bottle of acetonitrile in the PT=
H
position and run this to see if there are any spurious peaks - you will s=
ee
if your retention times are stable and if you have any variation in the p=
eak
heights. for example Lys should always be 10 -20% higher than Leu.
If this is OK, this means that you can then go over to the Cartridge side=
of
the instrument. We usually set up a series as follows:
1. Glass filter
2. Glass filter with 25 =B5l water
3. Glass filter with 25 =B5l Biobrene solution (dissolved in the same wa=
ter)
4. B-lac standard
The first 3 are run with your Filter Precycle programme - in this way you
get a Blank, a Standard and a Normal run showing the baseline with any
impurities.
The B-lac should give you >93% r.y. and about 6 pmol if you are loading 1=
0
pmol. Characteristic of it, is that Ile in cycle 2 is often higher than =
Leu
in 1. If you have a problem with S2, then the T-Q-T sequence will drop
drastically. If you have a problem with your A and B solvents, you will =
see
this on Met and Lys, strangely you will probably not see this in your
standard, but I think that this is because you are loading 10 pmols of ea=
ch
PTH-aa. A r.y. should only be made on L, I and V. W may be very low, bu=
t
Tyr should be visible in cycle 20. Also these 3 aa's should be on or jus=
t
on either side of your r.y. graph.
If all this is ok, then it may be that your peptides are washing out. I =
do
not know if they are from a digest or are synthetic. We use less Biobren=
e
than in the packet - we dissolve in 1100 =B5l water and divide it out in =
to
55=B5l portions and freeze. Then take out a tube and this does 2 filters=
with
25 =B5l Biobrene on each. If you had Biobrene after the peptide on a gla=
ss
filter, then your yield will drop drastically.
I would also recommend that you invest, just once maybe, in a bottle of t=
he
new R2 - N-methylpiperidine. Under constant use, it will last a good mon=
th,
otherwise you will have difficulties talking to PE about r.y. when you ar=
e
not using all their chemicals. I actually used the DIEA mixture on the 4=
77
about 6 years back and it gave a couple of % better r.y., but I think tha=
t
there was a large chemical top.
I know that a colleague at Novo Nordisk has a 494 and is doing sequencing
from peptide resins and on the peptides isolated from the resin and I do =
not
think that they are experiencing any difficulties. But I can ask them fo=
r
their programmes and you can then import them into your own Cycles and
Gradients.
3) (Note: I asked this person an extra question about covalently coupli=
ng=20
peptides to membranes for sequencing)=20
For the easy part. My memory hadn't served me well with the manufacturers=
of the
covalent coupling kit. It's Millipore. I used their Sequelon ArylAmine pr=
oduct=20
GEN92033. They have chemistries to accomodate coupling either through sid=
e chain
amines or carboxyls, and a new one that I'm not familiar with (SequeNet).=
Once=20
attached, the peptide sequences normally except with blank cycles at each=
=20
attached amino acid. Unless there is an attachement site near the C-termi=
nal=20
however, you'll get the same sample washout on the residues remaining.
The issue with the Met's was to get an idea if you were experiencing an=20
oxidation problem, and the point of monitoring lag was to see if there wa=
s=20
really an indication that the chemistry was being inefficient because of =
bad=20
reagents but your description of "all the residues coming out at once" so=
unds=20
like something else is going on, almost as if you hydrolyzed the peptide.=
I have
no experience with sythetic peptides attached to their resins so can't be=
gin to=20
know what to tell you to be cautious about, but to diagnose a machine pro=
blem=20
(R3 valve diaphram problem or other hardware issue) I'd run PE's favorite=
test=20
protein, B-Lactoglobulin, for 15 cycles on Biobrene and see what you get.=
If you
see the same phenomenon with that protein let me know.
4) With regard to the wash and delivery steps, what you see should be=20
similar to what is described in the tuning guides for the chemistry you a=
re=20
using. It has been a long time since I ran a 477A, but I remember one fa=
irly=20
old guide that described everything in detail. As each chemistry=20
modification came out, the technical notes for the modification would=20
describe any changes that needed to be made. The attached PDF file conta=
ins=20
portions of the NMP and microcartridge technical notes.
You can try decreasing the S1 and S2 solvent deliveries and see if y=
ou=20
get a positive effect on RY. As I recall, solvents were adjusted for=20
"delivery to midpoint", where the filter is wet, but no liquid has passed=
=20
through. You might want to verify the appearance of solvent deliveries w=
ith=20
someone in another lab who is currently running the sequencer.
5) Jeni, Is your "new" instrument another old 477--or is it a 494? Th=
is
sounds a lot like a problem Sandie from Texas reported last week --I
wonder if they are related? If they are, it might actually point to a
chemical problem and maybe you could both compare notes on lot numbers of
reagents used. Otherwise, it is just a coincidence that you have the sam=
e
instrument problem--but it could be due to the same thing. One thing I
wonder about is is the sequencer fine when it starts out and then as you
progress through several cycles you have a venting problem that prevents
delivery of something somewhere in the system, hence lousy ry. Is the
size of your injections the same as you proceed? Does your ry pickup
again each time you restart a sample? So some line is vented that isn't
vented from run to run? If you run many standard cycles in a row do you
get the same phenomenon? If you stop this sample when ry is low and open
the system up, and restart does ry go back up? Have you watched
deliveries to be assured that TFA is delivered for the cleavage step?
Does the transfer to the flask look good? We've run, many years ago, all
150 oof the peptides that we made in our 470 (usually with one grain of
resin if possible) to check synthesis before cleavage--always seemed to
work for us then.
6) If you've checked out all the simple and obvious possibilities and s=
till
can't get past the first few cycles, then I suggest you "babysit" the
cycle, i.e. sit and watch every delivery to make sure all valves are
functioning properly. Usually you wont be able to get a field service tec=
h
to do this. Once, several years ago, we had a similar problem. It turned
out that the power supply was not delivering full voltages which are
required to hold some of the valves in the open position.The way the thin=
g
works is that a high voltage is required to open the valve, and then the
voltage is dropped to hold it in the open position. Somehow, the valve (I
can't remember which one it was) was not staying open long enough. I
replaced the power supply and that solved the problem. You will be able t=
o
determine this rather easily just by watching. BTW, I spent a good deal o=
f
my life watching and listening to valves open and close.
7) It certainly sounds like a coupling problem and you are running out =
of
vapor phase after a few cycles because the liquid phase is too dilute to
replace the vapor phase quickly enough as cycles proceed. You might
consider switching to a new bottle at the first sign of drop off. That wa=
y
you'll be sure what corrected the problem. We use PE's N-methylpiperidine
concoction. You can test the column by running 10 cycles of PTH stds to s=
ee
if the column is holding up. Also, you may wish to change your A buffer i=
n
case peroxides may be building up from the THF, but in that case usually
all cycles will be bad and not just the later ones.
Janelle Lauer-Fields
Department of Chemistry & Biochemistry
Florida Atlantic University
777 Glades Road
Boca Raton, Florida 33431
561-297-2094/Fax 561-297-2759