<!doctype html public "-//W3C//DTD W3 HTML//EN">
<html><head><style type="text/css"><!--
blockquote, dl, ul, ol, li { padding-top: 0 ; padding-bottom: 0 }
--></style><title>[RASMB] Re: RASMB digest, Vol 1 #360 - 2
msgs</title></head><body>
<div>Dear Tim</div>
<div><br></div>
<div>If the sample is heterogeneous (i.e. there are species present
that are not participating in the reversible equilibrium.) the larger
species will fractionate toward the bottom of the cell independently
of the<b> local</b> concentration. Remember, the composition of a
purely heterogeneous, non-interacting system will vary purely as a
function of the radius independently of the local concentration (cf.
Yphantis, Biochemistry,1964). At the lower<b> loading</b>
concentrations one will be collecting data at higher radii for any
given<b> local</b> concentration span, therefore, the (weight) average
composition will be weighted toward the higher molecular weight values
because the data are being collected at higher radii. So if you fit
your data to a single exponential*, you will get a value for sigma
that is higher at the lower loading concentrations (which is
determined from data at higher radii) than at the higher loading
concentrations (which is determined from data at lower radii). The
cut-off gradient and/or maximum concentration observable pretty much
determines where you can collect the data.</div>
<div><br></div>
<div>It sounds to me as though Satinder's system is either
non-interacting or a mixture of interacting and non interacting
components.</div>
<div><br></div>
<div>* By the way, fitting your data to a single exponential and
calling that sigma a "weight average", although it seems to
be a popular exercise, is not theoretically correct. What you get is
only an approximation to a weight average, and if there is significant
curvature in the ln(c) vs r**2 plot, the value you get can be quite
far from the correct weight average.</div>
<div><br></div>
<div>At 1:31 PM -0500 11/15/04, Timothy Mack wrote:</div>
<blockquote type="cite" cite>Dear RASMBers,<br>
<br>
I have a question concerning Satinder's results. Satinder, you
mentioned that sigma increased with concentration and with rotor
speed. For a reversibly homo-associating system I thought that the
sigma would remain constant with speed (assuming all data between the
meniscus to the bottom can used in the fitting) whereas for a
heterogeneous system the sigma would actually decrease as a function
of increasing speed. What causes the increase in sigma with rotor
speed? I realize that one should not put too much stock in
these diagnostic graphs (sigma vs speed and/or conc) but I just want
to make sure that I am thinking about it correctly so if someone could
set me straight on this issue I would greatly appreciate it.<br>
<br>
Regards,<br>
Tim<br>
<a
href="mailto:rasmb-request@server1.bbri.org"
>rasmb-request@server1.bbri.org</a> wrote: f<br>
<blockquote type="cite" cite><tt>Send RASMB mailing list submissions
to<br>
<x-tab> </x-tab></tt><a
href="mailto:rasmb@rasmb-email.bbri.org"><tt
>rasmb@rasmb-email.bbri.org</tt></a><tt><br>
<br>
To subscribe or unsubscribe via the World Wide Web, visit<br>
<x-tab> </x-tab></tt><a
href="http://rasmb-email.bbri.org/mailman/listinfo/rasmb"><tt
>http://rasmb-email.bbri.org/mailman/listinfo/rasmb</tt></a><tt><br>
or, via email, send a message with subject or body 'help' to<br>
<x-tab> </x-tab></tt><a
href="mailto:rasmb-request@rasmb-email.bbri.org"><tt
>rasmb-request@rasmb-email.bbri.org</tt></a><tt><br>
<br>
You can reach the person managing the list at<br>
<x-tab> </x-tab></tt><a
href="mailto:rasmb-admin@rasmb-email.bbri.org"><tt
>rasmb-admin@rasmb-email.bbri.org</tt></a><tt><br>
<br>
When replying, please edit your Subject line so it is more
specific<br>
than "Re: Contents of RASMB digest..."<br>
<br>
<br>
---------------------------------------------------------------------<span
></span>-------------<br>
The older archived RASMB emails can be found at:<br>
</tt><a
href="http://rasmb-email.bbri.org/rasmb_archives"><tt
>http://rasmb-email.bbri.org/rasmb_archives</tt></a><tt><br>
and current archives at<br>
</tt><a
href="http://rasmb-email.bbri.org/pipermail/rasmb/"><tt
>http://rasmb-email.bbri.org/pipermail/rasmb/</tt></a><tt><br>
Search All the Archives at:<br>
</tt><a
href="http://rasmb-email.bbri.org/rasmb_search.html"><tt
>http://rasmb-email.bbri.org/rasmb_search.html</tt></a><tt><br>
---------------------------------------------------------------------<span
></span>-------------<br>
<br>
<br>
Today's Topics:<br>
<br>
1. question about WinNONLIN (Satinder Singh)<br>
2. Re: question about WinNONLIN (Peter Schuck)<br>
<br>
--__--__--<br>
<br>
Message: 1<br>
Date: Thu, 4 Nov 2004 12:06:39 -0500 (EST)<br>
From: Satinder Singh</tt> <a
href="mailto:ss2337@columbia.edu"><tt><ss2337@columbia.edu></tt></a
><tt></tt></blockquote>
<blockquote type="cite" cite><tt>To:</tt> <a
href="mailto:rasmb@server1.bbri.org"><tt>rasmb@server1.bbri.org</tt></a
><tt><br>
Subject: [RASMB] question about WinNONLIN<br>
<br>
Hello.<br>
<br>
I am trying to determine the dissociation constant of a homodimer.
The<br>
monomer molecular weight (by both sequence & molecular weight) is
33,400<br>
Da. At 15,000 rpm, the sigma is 0.95. I ran the experiment at
15,000,<br>
20,000, 27,000, and 35,000 rpm with 4 different protein
concentrations<br>
(0.2, 0.4, 0.8, & 1.2 mg/ml) using A280 detection.<br>
<br>
Before concluding that I indeed had a monomer-dimer equilibrium, I
used<br>
the "sigma test" to see if there was evidence of
association. Indeed, the<br>
sigma increased as a function of protein concentration (for a
single<br>
rotor speed). It also increased as a function of rpm (for a single
protein<br>
concentration).<br>
<br>
The data are best fit to a monomer-dimer equilibrium, although the<br>
residuals with this model (sigma fixed) still look a bit skewed. The
SRV,<br>
however, is pretty good at 5.6384x10-3. The question I have is<br>
this: When I FIX sigma at the monomer molecular weight, I get a<br>
dissociation constant of 12 uM. However, if I let sigma float, sigma
falls<br>
to 0.8276, which corresponds to a molecular weight of 28,000 Da,
and I<br>
get a dissociation constant of 2.3 uM.<br>
<br>
If the data are fitted correctly and nothing has happened to the
protein<br>
sample during the experiment, i.e., aggregation, precipitation,
shouldn't<br>
the 2 dissociation constants be relatively close? Could anyone tell
me<br>
what I may be doing wrong?<br>
<br>
Thanks in advance for your help.<br>
<br>
Satinder<br>
<br>
<br>
<br>
<br>
<br>
--__--__--<br>
<br>
Message: 2<br>
Date: Thu, 04 Nov 2004 13:14:01 -0500<br>
To: Satinder Singh</tt> <a
href="mailto:ss2337@columbia.edu"><tt><ss2337@columbia.edu></tt></a
><tt><br>
From: Peter Schuck</tt> <a
href="mailto:pschuck@helix.nih.gov"><tt><pschuck@helix.nih.gov></tt
></a><tt><br>
Subject: Re: [RASMB] question about WinNONLIN<br>
Cc:</tt> <a
href="mailto:rasmb@server1.bbri.org"><tt>rasmb@server1.bbri.org</tt></a
><tt><br>
<br>
Hi Satinder,<br>
there's probably several possibilities, but it reminds me of a protein
we<br>
studied when there was a bit of low molecular weight impurities in
the<br>
prep. Sometimes, these can go undetected on SDS PAGE, but would
skew the<br>
sedimentation profiles in a way consistent to what you describe.
I would<br>
try a model fixing the molecular weight (and v-bar - I assume
that's<br>
known), floating the binding constant, and adding an unrelated species
of<br>
fixed molar mass (low Mw, perhaps a few kDa) into the global fit to
see if<br>
that can explain the data. (I'm not sure if WinNONLIN let's you
do this,<br>
I'm using SEDPHAT.) In order to verify this, if you still have
the sample,<br>
it might be useful to do a velocity run (after shaking up the sample,
that<br>
should work well enough even with a few mm column). In our case,
we saw<br>
the same species in a very slow boundary in a velocity experiment done
in<br>
parallel.<br>
Hope that helps,<br>
Peter<br>
<br>
<br>
<br>
At 12:06 PM 11/4/2004 -0500, you wrote:<br>
</tt><br>
<blockquote type="cite"
cite><tt
>--------------------------------------------------------------------<span
></span>--------------<br>
The older archived RASMB emails can be found at:<br>
</tt><a
href="http://rasmb-email.bbri.org/rasmb_archives"><tt
>http://rasmb-email.bbri.org/rasmb_archives</tt></a><tt><br>
and current archives at<br>
</tt><a
href="http://rasmb-email.bbri.org/pipermail/rasmb/"><tt
>http://rasmb-email.bbri.org/pipermail/rasmb/</tt></a><tt><br>
Search All the Archives at:<br>
</tt><a
href="http://rasmb-email.bbri.org/rasmb_search.html"><tt
>http://rasmb-email.bbri.org/rasmb_search.html</tt></a><tt><br>
---------------------------------------------------------------------<span
></span>-------------<br>
<br>
Hello.<br>
<br>
I am trying to determine the dissociation constant of a homodimer.
The<br>
monomer molecular weight (by both sequence & molecular weight) is
33,400<br>
Da. At 15,000 rpm, the sigma is 0.95. I ran the experiment at
15,000,<br>
20,000, 27,000, and 35,000 rpm with 4 different protein
concentrations<br>
(0.2, 0.4, 0.8, & 1.2 mg/ml) using A280 detection.<br>
<br>
Before concluding that I indeed had a monomer-dimer equilibrium, I
used<br>
the "sigma test" to see if there was evidence of
association. Indeed, the<br>
sigma increased as a function of protein concentration (for a
single<br>
rotor speed). It also increased as a function of rpm (for a single
protein<br>
concentration).<br>
<br>
The data are best fit to a monomer-dimer equilibrium, although the<br>
residuals with this model (sigma fixed) still look a bit skewed. The
SRV,</tt></blockquote>
<blockquote type="cite" cite><tt>however, is pretty good at
5.6384x10-3. The question I have is this: When<br>
I FIX sigma at the monomer molecular weight, I get a dissociation
constant<br>
of 12 uM. However, if I let sigma float, sigma falls to 0.8276,
which<br>
corresponds to a molecular weight of 28,000 Da, and I get a
dissociation<br>
constant of 2.3 uM.<br>
<br>
If the data are fitted correctly and nothing has happened to the
protein<br>
sample during the experiment, i.e., aggregation, precipitation,
shouldn't<br>
the 2 dissociation constants be relatively close? Could anyone tell
me<br>
what I may be doing wrong?<br>
<br>
Thanks in advance for your help.<br>
<br>
Satinder<br>
<br>
<br>
<br>
<br>
_______________________________________________<br>
RASMB mailing list<br>
</tt><a
href="mailto:RASMB@rasmb-email.bbri.org"><tt
>RASMB@rasmb-email.bbri.org</tt></a><tt><br>
</tt><a
href="http://rasmb-email.bbri.org/mailman/listinfo/rasmb"><tt
>http://rasmb-email.bbri.org/mailman/listinfo/rasmb</tt></a><tt><br>
</tt><br>
</blockquote>
</blockquote>
<blockquote type="cite" cite><tt><br>
<br>
***********************************************************<br>
Peter Schuck, PhD<br>
Division of Bioengineering & Physical Science<br>
National Institutes of Health<br>
Bldg. 13 Rm. 3N17<br>
13 South Drive<br>
Bethesda, MD 20892 - 5766<br>
Tel: (301) 435-1950<br>
Fax: (301) 480-1242<br>
email:</tt> <a
href="mailto:Peter_Schuck@nih.gov"><tt>Peter_Schuck@nih.gov</tt></a><tt
><br>
***********************************************************<br>
<br>
<br>
<br>
--__--__--<br>
<br>
_______________________________________________<br>
RASMB mailing list<br>
</tt><a
href="mailto:RASMB@rasmb-email.bbri.org"><tt
>RASMB@rasmb-email.bbri.org</tt></a><tt><br>
</tt><a
href="http://rasmb-email.bbri.org/mailman/listinfo/rasmb"><tt
>http://rasmb-email.bbri.org/mailman/listinfo/rasmb</tt></a><tt><br>
<br>
<br>
End of RASMB Digest<br>
</tt></blockquote>
</blockquote>
<div><br></div>
<div><br></div>
<x-sigsep><pre>--
</pre></x-sigsep>
<div><font face="Times" size="-1"
color="#0000FF"
>----------------------------------------------------------</font></div
>
<div><font face="Times" size="-1" color="#0000FF">Walter
Stafford</font></div>
<div><font face="Times" size="-1"
color="#0000FF">mailto:stafford@bbri.org</font></div>
<div><font face="Times" size="-1" color="#0000FF">direct
dial: 617-658-7808</font></div>
<div><font face="Times" size="-1" color="#0000FF">receptionist:
617-658-7700</font></div>
<div><font face="Times" size="-1"
color="#0000FF"
>----------------------------------------------------------</font></div
>
</body>
</html>