[RASMB] AUC measurements in CsCl?

Fri Jun 7 06:20:30 PDT 2019

On Thu, Jun 06, 2019 at 06:26:39PM -0700, John Sumida wrote:
> Dear Peter,
> 
>  
> 
> Thank you for your paper.  It is useful to know that it is possible to measure species of this size using direct boundary approaches, also useful and interesting that a high dilution is necessary to avoid electrostatic effects – this is probably because of the charge on the nucleic acid? 

Yes, the charge in a low ionic strength will cause strong non-ideality
effects, especially for large DNA molecules. The good news is that in a
CsCl buoyant density experiment 1) the ionic strength is very high, and 2)
you are running very dilute anyway, and 3) the molecules are not moving,
they are at equilibrium, and all you care about is the peak position
and the area under the curve, so you can safely ignore this concern.

> I wonder if at this dilution, 0.005%wt/v, if there would be sufficient signal for a 10MDa species to use the absorbance optics.  I would assume so since nucleic acids absorb strongly at 260nm, but my assumptions oft go awry.  Perhaps another reason to go down the CsCl road.

Yes, more than enough. All of the signal would be concentrated in a
couple peaks, which would be very sharp due to the low diffusion if
the molecules aren't too much sheared. Attached is an image of T7 DNA
and T7 virus mixed together in such an experiment (in this case it was
done with the FDS). You can see how sharp the peaks are - almost like
mass spec expt. :)

Regards, -b.
> 
>  
> 
> Best regards,
> 
>  
> 
> John Sumida
> 
> Molecular Analysis Facility
> 
> University of Washington
> 
>  
> 
> From: Prevelige, Peter Edward, Jr [mailto:prevelig at uab.edu] 
> Sent: Thursday, June 06, 2019 5:58 PM
> To: John Sumida <jpsumida at uw.edu>
> Cc: Borries Demeler <demeler at gmail.com>; RASMB <rasmb at list.rasmb.org>
> Subject: Re: [RASMB] AUC measurements in CsCl?
> 
>  
> 
> Maybe worth noting that it was important in the M&S experiment that the DNA was sheared.  https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1307605/
> 
> Sent from my iPad
> 
> 
> On Jun 6, 2019, at 7:44 PM, John Sumida <jpsumida at uw.edu <mailto:jpsumida at uw.edu> > wrote:
> 
> Dear Borries,
> 
> Thank you for your response which is 100% on the mark with regards to the
> expected deliverables.  
> 
> While the work being proposed is not an exact repeat of the Meselson & Stahl
> experiment (one of my favorites as well) it is along the same vein.  The
> work is attempting to use the methods outlined in the Flory Vinograd work
> (which I've attached) which uses bromodeoxyuridine as a mass label (an
> interesting idea ) instead of N15; and yes, my collaborators are seeking to
> know the relative quantity of the species detected and the sedimentation
> value is less important.  
> 
> Interesting also that in the Messelson and Stahl experiment they used whole
> cell lysate, this was not significant to me when I read this paper back in
> the day, but now this seems particularly useful?
> 
> I do not have the 3mm cells here, do you think this experiment is still
> doable in a standard two channel SV cell?
> 
> Just so I am clear on the reason to use CsCl in this particular case is the
> ability to detect small amounts of DNA, the large masses involved, which
> would require a long slow speed spin, and avoid complications from
> heterogeneity that may be existant at a given buoyant density and which may
> be difficult to interpret?
> 
> Thanks so much for this information, I am eager to get started.
> 
> Best regards,
> 
> John Sumida
> Molecular Analysis Facility
> University of Washington
> 
> In many spheres of human endeavor, from science to business to education to
> economic policy, good decisions depend on good measurement.
> Ben Bernanke
> 
> 
> 
> 
> 
> -----Original Message-----
> From: Borries Demeler [mailto:demeler at gmail.com] 
> Sent: Thursday, June 06, 2019 4:36 PM
> To: John Sumida <jpsumida at uw.edu <mailto:jpsumida at uw.edu> >
> Cc: RASMB <rasmb at list.rasmb.org <mailto:rasmb at list.rasmb.org> >
> Subject: Re: [RASMB] AUC measurements in CsCl?
> 
> Hi John,
> Accurate molar mass measurement of any solute by SV or SE methods requires
> an accurate vbar. This varies depending on ionic strength, ie.
> hydration - so, do you know that? And if so, how? Cesium will bind to the
> DNA changing its vbar (it is a very dense molecule). SE will also be
> challenging since equilibrium will take forever to be achieved at lower
> speed, and the steep gradient you are likely going to see at the bottom of
> the cell will be distorted by refractive effects and not give you a lot of
> points. Moreover it would be nearly impossible to resolve any heterogeneity
> in the samples from size. So, I wouldn't use SE.
> 
> For SV experiments to additionally get a molar mass you need to be able to
> measure the diffusion coefficient accurately. Large molecules like this have
> very little diffusion signal. You would have to run for a very long time at
> a very slow speed to get sufficient accuracy.
> 
> Everything else requires making assumptions and estimates. So I think your
> traditional approaches for SV and SE will not be satisfactory.
> 
> It sounds like your collaborators are asking for the classic experiment
> performed by Meselson & Stahl in 1958 (paper attached, one of my favorites,
> I make my students read it). The different amounts of incorporated label
> impart slightly different molar masses that can be separated based on the
> different densities using buoyant density equilibrium gradients. It is the
> CsCl gradient that will make the two populations buoyant at different places
> in the cell. As you can see in the paper, this works quite well, and even
> better with modern equipment.
> 
> For this to work well in an AUC, I recommend using 3 mm centerpieces.
> It minimizes the CsCl gradient signal and hence attenuates the refractive
> effect from the gradient. Use optically pure CsCl so you don't have
> background at 260 nm where the DNA will absorb. This *should* work, though I
> haven't tried it. While it will not give you accurate molar masses it should
> resolve the two populations of labeled and unlabeled DNA. As an added bonus,
> any heterogeneity from shearing is pretty much masked, since all molecules
> will go where they have the same density, which differs for labeled and
> non-labeled molecules. Speeds between
> 40-50 krpm should work well. Let us know how it works out. You won't need a
> lot of DNA to make this measurement, since all the DNA will migrate to a
> small spot and concentrate there to make a big peak.
> Follow the paper to get the CsCl concentration correct. This is important
> and needs to be matched to the DNA samples so you get the best separation.
> 
> In the end you still don't know absolute molar masses but I don't think you
> actually need that. I bet your collaborators only care about the relative
> amount of each population. You could also try this by standard velocity
> analysis, but there the result will also encounter concentration dependent
> non-ideality, but may resolve the boundaries from the two different
> densities. Heterogeneity may be reducing your confidence here.
> Meselson and Stahl, though an old paper, still seems like the most elegant
> way to solve this problem. Let us know how it works out.
> 
> Good luck!
> 
> -Borries
> 
> Thu, Jun 06, 2019 at 02:49:44PM -0700, John Sumida wrote:
> 
> 
> 
> Dear RASMB,
> 
>  
> 
>  
> 
>  
> 
> First I have to thank all those who curate this list server as over 
> 
> the last
> 
> 10 years as a manager of the lab I run, it has and continues to prove 
> 
> a hugely valuable resource for me and the people I serve.
> 
>  
> 
>  
> 
>  
> 
> My question today has to do with the analysis of DNA and it stems from 
> 
> a user interest in performing an AUC equilibrium experiment in the 
> 
> presence of CsCl.  My experience to date has been with nanomaterials 
> 
> and proteins so DNA is a new material for me, however I am under the 
> 
> impression that running AUC samples in presence of CsCl is 
> 
> problematical because of the gradient CsCl itself will generate.
> 
>  
> 
>  
> 
>  
> 
> The estimated mass of the species of interest is in the range of 10 
> 
> mega Daltons; this is also a size regime that is new to me.
> 
>  
> 
>  
> 
>  
> 
> The objective of the experiment is to characterize nucleic acids 
> 
> extracted from cells which will have a label increasing the mass of 
> 
> the nucleic acid depending on the number of times it has been replicated.
> 
>  
> 
>  
> 
>  
> 
> Two questions:
> 
>  
> 
> Is CsCl necessary to determine the size distribution of molecular 
> 
> species
> 
> 10MDa?
> 
>  
> 
> Can a CsCl experiment be performed using equilibrium measurements?
> 
>  
> 
>  
> 
>  
> 
> Thank you for considering my question.
> 
>  
> 
>  
> 
>  
> 
> Best regards,
> 
>  
> 
>  
> 
>  
> 
> John Sumida
> 
>  
> 
> Molecular Analysis Facility
> 
>  
> 
> University of Washington
> 
>  
> 
>  
> 
>  
> 
> In many spheres of human endeavor, from science to business to 
> 
> education to economic policy, good decisions depend on good measurement.
> 
>  
> 
> Ben Bernanke
> 
>  
> 
>  
> 
>  
> 
>  
> 
>  
> 
> 
> 
> 
> 
> _______________________________________________
> 
> RASMB mailing list
> 
> RASMB at list.rasmb.org <mailto:RASMB at list.rasmb.org> 
> 
> http://list.rasmb.org/listinfo.cgi/rasmb-rasmb.org
> 
>  
> 
> <Flory Vinograd BrdU mtDNA 1973.pdf>
> 
> _______________________________________________
> RASMB mailing list
> RASMB at list.rasmb.org <mailto:RASMB at list.rasmb.org> 
> http://list.rasmb.org/listinfo.cgi/rasmb-rasmb.org
> 
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ss1.png
Type: image/png
Size: 37680 bytes
Desc: not available
URL: <http://list.rasmb.org/pipermail/rasmb-rasmb.org/attachments/20190607/3c3c081c/attachment-0002.png>