Oligosaccharide analysis methods

mike quesenberry quese_m at JHUVMS.HCF.JHU.EDU
Mon Feb 20 07:57:50 EST 1995

Dear Glycoscience Interest Group,

I would like to offer comment on the following topic:

>E.g. The group I have just joined use mainly HPLC methods with fluorescent
>detection. We have a Dionex as well, but that is considered
>too variable. I have been use to mainly radioactive detection in
>conjunction with P4 and Dionex. Which method should prevail?

I've not made enough extensive use of different methods sufficient to make
comparisons, but I have worked with the Dionex BioLC system with PAD II
detection.  As for variability, I'm not sure whether you're referring to
retention times or peak size.

With regard to retention times it has been my experience that this is
reproducible once our system is well-equilibrated, and if it is re-equilibrated
following each injection.  I think this means that the user should be aware
what the sample may be doing to the column.  In the manuals which accompany
the columns are explicit instructions for regenerating the columns, so that
means that 'something' can accumulate and interfere with column performance.
What to do?  If a so-called 'wash step' is included with each column run,
time between injections increases for re-equilibration.  Without it, slow
degradation of performance may occur which is hard to notice.  Frozen
aliquots from reliable standards are a must here.  The other problem with
retention time variability may stem from solvents, which always have to be
'the SAME'.  Beware something growing in your acetate solution!  Other than
that, I often suspect our autosampler's performance, which is the system
formerly recommended by Dionex (I heard there's now a rift between those two
companies, though, but that may just be a rumor--or is that rumour?).  One
thing to do is make many sequential injections of the same standard (but I
use individual sample vials since the autosampler tray gets hot enough, and
the humidity--another variable?--can be dry enough to make me suspect some
evaporation could account for irreproducible results).

With respect to peak size, a 'non-interfering' internal standard is good,
and don't forget to polish your PAD!  And, of course, don't overpower the
column.  The different sensitivities available on the PAD could lead to the
temptation to overload the column (? right?--this also gets to the necessity
of column regeneration if this occurs), and that could potentially lead to
great variability.

I look forward to other comments on this topic.

Until then, Adios and Happy eTrails!

--Mike Q.

More information about the Glycosci mailing list