[RASMB] estimation of Vbar from sedimentation velocity data

[mitrana at mail.utexas.edu]

Hi Matthew

I would think that sedimentation equilibrium rather than velocity would be a
better way to estimate the v-bar. I guess you could do this and see if you turn
up with a similar value if you float the v-bar and fix the MW to 25,000 while
fitting to an ideal single component model. Just putting in my two cents.
Mitra

Quoting Matthew Hockin <mhockin at gmail.com>:

> I am working with 25 kDa linear polymer Polyethylenimine.  What I am trying
> to do is use sedimentation velocity to characterize the product obtained
> upon conjugating a protein to the polyethylenimine.  There are, so far as I
> can tell, no published values for the partial specific volume of this
> polymer.  I have done the following experiments and would like feed back as
> to what can done to strengthen my analysis.
> 1) SV experiments of the polymer alone in 250 mM NaCl pH 7.0 (the polymer is
> a very strong buffer thus no other components)
>
> 3 concentrations were analyzed spanning 0.64-0.1 mg/ml using interference
> optics at 52000 RPM.
>
> In an attempt, perhaps feeble, to determine the vbar I began by fitting each
> concentration data set using c(S) in sedfit and with known calculated buffer
> density and viscosity corrected to the experimental temperature (22.5 c),
> and adjusted the partial specific volume to obtain a reasonable MW estimate
> after examining the peak MW distribution in the c(S) plot.
>
> Landing on a value of 0.85 for vbar, the fits are exceptionally good as
> measured by examination of the residual plot (< 0.03 variance) as well as by
> examination of the graphical residual bit map, uniform grey (actually the
> best uniform grey fit I have ever seen).  These fits yield a single peak at
> experimentally determined S values of 0.6 at all concentrations.  Every
> concentration yields a best fit experimental S value of 0.6 using frictional
> ratios of 2.34, which is reasonable given that this polymer is supposed to
> act as a ridgid rod in solution. Prior to the centrifuge the polymer has
> been gel filtered over high resolution superdex 200 columns and
> characterized with respect to the average number of titratable nitrogens.
>  The gel filtration is used simply to narrow the polydispersity while the
> titrations provide a lower limit to the chain length and this data is
> consistent with a polymer of approximately 25 kDa.
>
> I realize this is not the easiest way to approach this value but do not have
> access to the instrumentation required to determine it.  It seems reasonable
> in that my ultimate goal is to use this estimated vbar value to calculate a
> weight average vbar for a protein-polymer conjugate that I am making.  The
> protein vbar can of course be calculated.  I would ultimately like to
> demonstrate that the actual MW of the conjugate as well as stoichiometry is
> correct, and determine if the conjugate behaves as a monomer in solution.
>  Initial analysis of the conjugate using the weight average vbar seem to
> confirm the monomeric nature as well as support a MW addition of 25,000 more
> or less.  The protein vbar is calculated to be 0.7475 and fits to the
> conjuate are relatively poor at this value but improve when it is increased
> to 0.78 which is the weight avg vbar.
>
> You would be helping me quite a lot if you could provide either a criticism
> of this approach (there are likely many) and a suggestion for a means to
> either validate this or approach it in a different way.  Analysis of the
> conjugate is somewhat complicated by the presence of free polymer and thus
> two species with quite different frictional ratios so the best fits are
> obtained using the c(S)ffo  model in sedfit, the fit range ffo values for
> each peak in the c(s) distribution correlate with the anticipated component
> represented in each peak, i.e. free polymer S experimental at 1.3 has ffo of
> >2 whereas the anticipated conjugate experimental S value of 2.6 has ffo
> values spanning 1.6-1.8.  The corresponding MW plots show anticipated values
> of approx 28 kDa for polymer and 60-80 Kda for the presumed conjugate.
>  Analysis of the free protein yields excellent fits with a single peak and
> an ffo values of 1.4-1.6 and a MW of 43000 (actual 41406)
>
> Sorry for such a long question for my first posting to this board. I hope to
> have provided enough information so that a useful discussion can arrise.  I
> thought I understood that I could use the buoyant MW relationship to
> estimate vbar but seem not to understand how that would work in this
> situation... Thanks a lot
>
> Matthew
>


-- 
Mitra S. Rana
Graduate Student
Institute for Cellular and Molecular Biology
2500 Speedway, UT-Austin
Austin, TX-78712