[RASMB] Sw Isothem Fitting in SEPHAT using Fluorescence Data

[Cole, James]

Dear David-
You can  fit binding isotherms using a variety of packages that implement nonlinear least squares. I used Kaleidagraph to fit your data  to a simple 1:1 binding model (see attached). Given that the Kd appears to be much larger than the RNA concentration, you can use the simple approximation that [RNA]free = [RNA]added. Then, the data are described by simple hypberbolic binding function. If you don't make this approximation you end up with a quadratic function that you can also fit using a nonlinear least squares algorithm.  I assume that the concentrations are in micromolars. The two fitted parameters are the sedimentation coefficient of the complex (3.5 + m1) and the Kd (m2). As you can see, the fit is pretty bad.
.
Your curve is much steeper than the fit line. This means there's something wrong with the data. You cannot have cooperativity for a 1:1 binding interaction. First, it seems strange that a very weak interaction (Kd ~ 50 uM) would be slow on the AUC timescale. We have studied many protein-RNA interactions and we've never see one in slow exchange. If it is really in slow exchange, did you wait long enough after making the samples to allow them to reach equilibrium before doing AUC?

Finally, I would suggest that you  directly analyze the velocity data using either  Sedanal, sedphat or ultrascan using a 1:1 binding model. This is the most rigorous method to characterize the system. We have found that it works very well for protein-RNA interactions using either absorbance or fluorescence data. We have described the protocol  in: Wong, C.J., Launer-Felty, K. & Cole, J.L. (2011). Analysis of PKR-RNA interactions by sedimentation velocity. Meth Enzymol 488, 59–79.

Good luck,
Jim Cole



On May 3, 2012, at 4:17 PM, David Chiovitti wrote:

> Hi everyone,
>
> I am studying a tRNA-protein binding interaction using the AVIV AU-FDS system.  I am using a fixed amount of fluorescently labelled tRNA and titrating in the protein.  From the c(s) plots I have determined that it is a slow exchange between the protein and the tRNA.  The tRNA give a peak at approximately 3.5 S.  As I titrate in the protein the 3.5 S peak decreases and a new peak at 4.9 S appears.  This is consistent with one tRNA being bound by one protein.  For each titration I have used a weighted vbar value based on the amount of protein and tRNA.  I have been using the vbar value of 0.530 for the tRNA.
>
> For each of the titrations I have integrated the peaks to determine a weighted signal value and put them into a text file according to the formatting listed on the SEPHAT website.  I open the file up with no issue in SEDPHAT and select the hetero-assocation model.  I enter the molecular weight and S values of each of the components and float the logKa value, which I set to 5.  When I run and fit the data I get a curve that doesn't even come close to fitting the data points.  I'm really at a loss to what I'm doing wrong.
>
> Here is the isothem data:
> 0.82 0.1 3.6
> 2.74 0.1 3.48
> 7.4  0.1 3.50
> 22.7 0.1 4.0
> 66.6 0.1 4.94
> 200 0.1 4.90
>
> The first value being the concentration of the protein, the second the concentration of the tRNA, and the third is the Sw value.
>
> Other than SEDPHAT is there another program I could use to fit this Isotherm?
>
> Thanks,
> David Chiovitti
>
> Samuel Lunefeld Research Institute
> Mount Sinai Hospital Toronto
> _______________________________________________
> RASMB mailing list
> RASMB at rasmb.bbri.org
> http://rasmb.bbri.org/cgi-bin/mailman/listinfo/rasmb

-------------- next part --------------
A non-text attachment was scrubbed...
Name: David_fit.tiff
Type: image/tif
Size: 480174 bytes
Desc: David_fit.tiff
URL: <http://list.rasmb.org/pipermail/rasmb-rasmb.org/attachments/20120503/dcdb5b78/attachment.bin>