I have run my protein (monomeric size 65.8kDa) down both a AcA44 Ultrogel (Biosepra)[4 degrees C] and a Hiload Superdex 200 (Pharmacia) [Room Temp]. Both samples were loaded in 20mM Tris-HCl, 150mM NaCl and 5% glycerol (both at pH 6 and pH 8). The protein ran as a monomer on the AcA44 but as a dimer on the Superdex. Does anyone know of any examples of this kind of anomalous behaviour on gel filtration columns (references?) and is this kind of event the 'norm'?
All standards I ran (Alcohol dehydrogenase, BSA,
Carbonic anhydrase and cytochrome c) were fine and gave a linear calibration curve.

Thankyou in advance.

The difference may not be the column or the temperature. Was the sample older when you ran it on the Superdex column?

The sample was fresh each time and part of a continuous purification procedure. The volume loaded was 1ml for a 300ml column in both cases.
The addition of 5% glycerol to the sample also had no effect.

There appear to be few "norms" regarding chromatography of proteins (see earlier discussions, for example: Gel filtration, Oct 13, 2000; Elution of proteins in RP-HPLC, June 2, 2000). You may indeed have an interesting result, but it contributes little unless you tell us more. Are you assuming, on basis of retention time, that you have dimers emerging on the Ultogel, or do you use a dynamic light scattering detector, etc. What protein do you have? Is it hydrophobic, what is its pI, shape? ? ?
Did you inject the same batch (out of the same container), more than once, reversing injection sequence on the two columns? Why did you use diff. temps? Do you know anything about the surfaces of the columns?

Analytical ultracentrifugation studies show that the protein is dimeric (sedimentation equilibrium experiment at 4 degrees C) with a Kd of approximately 250nM. All samples run on these columns were at 100uM concentrations (monomeric concentration).
I am assuming monomer/dimer solely on the basis of retention time.
The protein is quite hydrophobic and has a calculated pI of 5.33 (gel filtration carried out at pH 6.0 and 8.0 are identical).
We are not sure of the shape of the molecule but know from previous data that in the monomer there are two domains joined by a flexible linker.
The same batch of protein was used for both columns.
Different temperatures were used only because the Superdex is used for other room temperature purification procedures and we can only afford the one column at present.
The Ultrogel is composed of 3% acrylamide, 4% agarose, with a bead size of 60-160 microns (swollen)
Superdex 200 is a dextran/agarose composite with a bead size of 22-44 microns.
Hope this information is of help.

Thankyou for your interest.

Rereading your statements, I noticed that you do not mention the mobile phases used. With the information given so far one could imagine a lower adsorption contribution on the Superdex, giving a lower retention time, or, more likely, that your protein has a different shape on the two columns (cause: diff. temp, and? Mobile phase?). What happens if you change mobile phases, did you try "iso-osmotic" PBS (0.15M phosphate + 0.15 NaCl), etc? This question is relevant even if you really have dimers eluting on the Superdex.

Thanks for the reply.

I have not really played around too much with the mobile phase. The standard condition has been 20mM Tris pH 8.0, 150mM NaCl, 2mM DTT. I tried with slightly lower salt (50mM NaCl) and with a different buffer (50mM MES pH 6.0 with 150mM NaCl).
I have heardof someone in my department with the same Superdex matrix in a column at 4 degrees so I will see if there is a temperature dependence.
I'll let you know the outcome.

Thanks again

I have no experiences in protein and you columns, but your results looks me as effect of additional adsorption. In molecular weight distribution by Size Exclusion Chromatography (Gel Permeation Chromatography) one fundamental rule is avoiding effect of adsorption analyte to the column bed. If there is adsorption, then peaks are shifted to higher retention times (lower molecular weights). In my experiences, using polystyrene (hydrophobic) column and dimethylformamid with several percent of water as (nonhydrophobic) eluent, I am able to shift retention times of (hydrophobic) o-phtalic esters even over the total permeation time, while (nonhydrophobic) phtalic acids are eluted according the molecular weight. In one run I can observe molecular weight distribution of polyvinylchlorid and identify phtalic esters plasticisers.
Nice explanation you can find in Berek&Hunkler paper "Liquid chromatography of macromolecules under limiting conditions of adsorption" (L.Liq. Chrom. & Rel. Technol. 22(19) 2867-2878 (1999).
Similar chromatography in critical point of adsorption is presented in several papers by Trathnigg, Gorbunow, Pasch.

Thanks. I'll go hunting for references.

Running the sephadex at 4oC gave the same result as that at room temperature. It looks as though it may be the composition of the matrix causing this effect - less adsorption on the polyacrylamide column.