[RASMB] temperature control

Peter Schuck pschuck at helix.nih.gov
Mon Apr 1 14:40:01 PST 2002


Jack, 
I'm not sure how to achieve proper temperature equilibrium outside the
centrifuge chamber.  I usually let the rotor sit at 0 rpm for at least 1
hour (even after the temperature reading indicates the desired temperature)
in order to achieve proper equilibration.  My feeling is that this is a
limiting factor in the accuracy of sedimentation velocity. 

With regard to the analysis:  Virtually all the analyses of sedimentation
velocity I know assume that the sedimentation process starts at time zero
at a uniform concentration (or with a concentration distribution as a
step-function for models of synthetic boundary experiments), and that it
proceeds without any other forces than sedimentation, diffusion and
chemical reactions.  As you mentioned, any convection will distort the
boundary shape and positions of all the following scans, which will lead to
errors in the apparent s-values.  Because most analyses are implying the
sedimentation from time 0, one gets errors even if loading only later scans
where sedimentation is stable.  That's true for ordinary Lamm equation
modeling, dcdt, g*(s) or ls-g*(s), c(s), as well as van Holde-Weischet, and
an exception would be a second moment method if put it a differential form. 

However, there is a way to calculate Lamm equation solutions using an
experimental scan as initial condition.  This means that one can take a
scan after the sedimentation is stable or temperature has been reached,
respectively, and calculate how from this time on the sedimentation
evolves, and fit all subsequent scans, without any reference to the prior
history.  This was an idea of Cox (1966), Science 152:359-361, and it is
implemented in sedfit.  Detailed instructions can be found at 
http://www.analyticalultracentrifugation.com/sedfit_help_SetInitialData.htm

Drawback of this procedure is that it only works for single-component
systems, i.e. a single species that sediments idealy or with
self-association.  Also, the time-invariant noise calculations for
interference data are not compatible, therefore it's only applicable for
absorbance data.  Also, there are some minor complications caused by the
fact that we can't see through the entire solution column because of the
artifacts at meniscus and bottom, but this is usually not problematic
(especially if the meniscus has cleared). 

Hope this helps, 
Peter

>
>This comes from a novice
>We have been doing some preliminary temperature studies using the AUC. The
>assembled rotor was put into a thermostated cabinet for about two hours at
>4 C. It was then transferred to the 4 C chamber of the AUC and the run was
>started. At the end of the run it was clear that the rotor was not
>initially at thermal equilibrium. It took about 60 minutes at speed to get
>there. 
>The problem was much less pronounced with runs at 25 C. At 40 C it took
>about 15 minutes at speed to achieve thermal equilibrium.
>My questions pertain to velocity runs:
>1. Does anyone have suggestions as to how one can avoid starting runs at
>one temperature and finishing at another short of letting the rotor
>equilibrate o/n in the AUC? Could we, for instance, immerse the assembled
>rotor in a thermostated water bath, dry it and then put it into the AUC?
>It is clear that heat transfer in a vacuum or in a refrigerator is very
>slow.
>2. This is a character defect: I would like to use the instrument without
>becoming
>an expert. I have tried to figure out whether velocity runs started at one
>temperature and finished at a second can yield any useful data once the
>desired temperature is reached. My intuitive feeling is that the shape of
>the boundary is dictated by the temperature. This does not correct itself
>as the run conditions change and become stable. If this reasoning is
>correct, it means that useful velocity sedimentation data cannot be taken
>unless one starts, runs and finishes at the same temperature. Is there a
>rule of thumb as to how much leeway is acceptable? If I would like to run
>at 40 C but start at 38 C how much error is likely introduced?
>
>Thanks for any help you can give
>jack kornblatt via m. judith kornblatt
>
>Judith Kornblatt
>Dept of Chemistry and Biochemistry
>Concordia University
>1455 de Maisonneuve Ouest
>Montreal, Qc H3G 1M8
>Tel: 1 514 848 3384   FAX: 1 514 848 2868
>
>
>_______________________________________________
>RASMB mailing list
>RASMB at rasmb-email.bbri.org
>http://rasmb-email.bbri.org/mailman/listinfo/rasmb


***********************************************************
Peter Schuck, PhD
Molecular Interactions Resource
Division of Bioengineering and Physical Science, ORS
National Institutes of Health
Bldg. 13 Rm. 3N17
13 South Drive 
Bethesda, MD 20892 - 5766
Tel: (301) 435-1950
Fax: (301) 480-1242
email: Peter_Schuck at nih.gov
***********************************************************



More information about the RASMB mailing list