Hi everyone,

Just wanted to second the SAXS suggestion:

the approach should provide a very rigorous way of testing the candidate models apparent in the crystallographic lattice, and can assist with questions of mass.  (among the most common applications)

With regards to the mass question, test as many concentrations as you can to get a pulse on the strength of the interaction, and also use other complementary solution methods such as AUC or SEC-MALS.  Note that the concentrations tested in SAXS are several fold higher than those examined in AUC or SEC/SEC-MALS, and may not be physiologically relevant in that regard.

Indeed, a 12.5 kD monomer will have very weak scattering power (intensity varies as the square of molecular volume).

You can compensate for that by increasing the concentration or using synchrotron radiation.

The ‘rule of 100’ is a good guide – in your SAXS expt, start at a sample concentration where MW (kD) * concentration (mg/mL) ~100 and then adjust accordingly. 

(I’ve successfully studied molecules as small as a 8.5 kD Tudor domain and a ~7kD CHAPS micelle in this way.)

Hope that helps,

Kushol

Kushol Gupta, Ph.D.

Research Associate - Van Duyne Group

Department of Biochemistry and Biophysics

Perelman School of Medicine at The University of Pennsylvania

[log in to unmask] / 215.573.7260 / 267.259.0082 / www.stwing.upenn.edu/~kgupta

From: CCP4 bulletin board [mailto:[log in to unmask]] On Behalf Of David Briggs
Sent: Friday, December 12, 2014 12:37 PM
To: [log in to unmask]
Subject: Re: [ccp4bb] asymmetric homotrimer in the asu

Hi Hay, 

I think SAXS should be more than capable of discriminating between a 12.5 kDa monomer vs ~37.5 kDa trimer. 

Lysozyme is a useful standard used in SAXS (as with most structural biology!), and Lysozyme is only slightly larger than your proteins. 

Cheers,

Dave

On Fri Dec 12 2014 at 5:13:26 PM Hay Dvir <[log in to unmask]> wrote:

Tanner:

Thanks, GREAT reference on asymmetric homo oligomers!

SAXS sounds like a good idea for a bit larger particles. I'm afraid  it might be very difficult to get enough resolution to resolve oligomerization of a rather small 12.5 kDa protein like ours, but will look into it more closely.

Joes:

Thanks, we are aware of the serious problem of instability of asymmetric homo-oligomers which could essentially polimerize as you nicely explain and cite.  Indeed one of the hypothesis we aim to test if we get additional evidence about the the trimeric assembly concerns its known function to interact with another protein, which could potentially provide the complementary quaternary stability. Interface mutational analysis sounds like a good approach to take in such cases.

Thanks again an very best,

Hay

On Dec 12, 2014, at 5:39 PM, Tanner, John J. wrote:

Two thoughts on asymmetric oligomers.

1.  Here is a recent survey of asymmetric homodimers in the PDB.  I know you are looking for trimers, but at least this provides a precedent for asymmetric oligomers.  

Swapna LS, Srikeerthana K, Srinivasan N. Extent of structural asymmetry in

homodimeric proteins: prevalence and relevance. PLoS One. 2012;7(5):e36688. doi: 

10.1371/journal.pone.0036688. Epub 2012 May 22. PubMed PMID: 22629324; PubMed

Central PMCID: PMC3358323.

2. SAXS is a very effective method for determining whether assemblies observed in crystals are stable in solution, since it provides not only the oligomeric state, but also the quaternary structure.  The oligomeric state can be obtained from the volume of correlation (1) and Porod-Debye analysis (2).  The quaternary structure can be deduced by comparing the experimental SAXS curve to theoretical curves calculated from oligomer models identified by PISA or from manual inspection.   The FoXS server and CRYSOL are good tools for this. FoXS also allows ensembles of oligomers (MES) to be used in fitting the data (e.g. mixture of monomer + dimer). I believe ATSAS also has an ensemble program, but the name escapes me at this time.  We have used this approach to show that assemblies that are predicted to be stable by PISA are not found in solution (3 and unpublished results). 

1: Rambo RP, Tainer JA. Accurate assessment of mass, models and resolution by

small-angle scattering. Nature. 2013 Apr 25;496(7446):477-81. doi:

10.1038/nature12070. PubMed PMID: 23619693; PubMed Central PMCID: PMC3714217.

2: Rambo RP, Tainer JA. Characterizing flexible and intrinsically unstructured

biological macromolecules by SAS using the Porod-Debye law. Biopolymers. 2011

Aug;95(8):559-71. doi: 10.1002/bip.21638. Epub 2011 Apr 20. PubMed PMID:

21509745; PubMed Central PMCID: PMC3103662.

3: Luo M, Singh RK, Tanner JJ. Structural determinants of oligomerization of

δ(1)-pyrroline-5-carboxylate dehydrogenase: identification of a hexamerization

hot spot. J Mol Biol. 2013 Sep 9;425(17):3106-20. doi: 10.1016/j.jmb.2013.05.027.

Epub 2013 Jun 7. PubMed PMID: 23747974; PubMed Central PMCID: PMC3743950.

On Dec 12, 2014, at 4:56 AM, Jose Manuel Duarte wrote:

Dear Hay

Your post prompted me to respond, since I think the issue of symmetry is extremely important.

I would like to reinstate here what should be obvious to everyone: a stable asymmetric assembly of proteins in solution is essentially impossible (or at most very very unlikely), purely because of topological reasons.

This is beautifully explained in a classic paper now 50 years old: Monod, Wyman, Changeux (1965) "On the Nature of Allosteric Transitions: A Plausible Model". The reasoning there is that a homomeric protein in solution can only associate in 2 ways: isologous (binding with same surface patches in both monomers, necessarily through a 2-fold axis) or heterologous (binding through different surface patches in both monomers). The isologous case is clearly symmetric (C2). Whilst in the heterologous case  the monomers can either assemble infinitely or form a closed symmetry. The conclusion that follows is that stable homo-oligomers can only be symmetric.

I especially like this paragraph:

"On the basis of these considerations, it is reasonable to assume that, if an oligomeric protein possesses a wide range of stability, it consists of a closed structure where all the protomers use the same binding sets; which implies, as we have just seen, that the molecule should possess at least one axis of symmetry."

The paper really explains it a lot better than me, it can be found here: http://www.pasteur.fr/ip/resource/filecenter/document/01s-00004j-0er/monod-wyman-changeux-1965.pdf

The conclusion in any case is that asymmetry in homomers is, if not impossible, highly unlikely. So in my opinion asymmetric assemblies should be proposed with a lot of care, only if experimental data really is overwhelmingly clear. For instance I don't think that gel filtration or AUC would be good evidence enough: it really needs to be demonstrated that the interface that you see in the crystal is the one leading to oligomerisation (perhaps with a mutagenesis experiment?). Otherwise the interface in the crystal is most likely simply a crystal contact.

Jose

On 12/12/14 10:15, Hay Dvir wrote:

Dear Jeremy,

Indeed, we also incline to think of it as a monomer in solution, but still quite un-eased by the extensive interactions in the asu being merely as a result of a crystallization artifact. As you said, we may need to rely more heavily on biochemical analysis and since SEC wasn't clear we are turning now to LS (hope to able to post a more conclusive update).

Regardless of what our final conclusion would be for this case, we became rather generally interested to find other similar cases of homomeric assemblies related only by non crystallographic translation symmetry (or as Engin Qzka pointed out "improper NCS" is the conventional terminology). So to rephrase our question we are interested to learn about additional structures of homomoeric improper ncs assemblies.

I truly appreciate ANY open-minded or skeptic thought, profound or trivial that we get here! They all, definitely those made by Mark Garavito, contribute to shaping our mind around this riddle.

Thanks for commenting on the skepticism, I brought it up as part of the discussion but a glitch of my own coffee time haziness might have slipped in. Perhaps I should try some o-cha instead .. :)

cheers,

Hay

On Dec 12, 2014, at 3:05 AM, Jeremy Tame wrote:

Dear Hay

I suggest that you use analytical ultracentrifugation to determine the oligomeric state of the protein in solution.
Mass spectrometry and light scattering are also useful, but there are so many examples of gel filtration proving
erroneous it has questionable value as an analytical technique. For an example of a dimer interface predicted by
PISA to be real you could look at Yoshida et al, JMB 423, 351 (2012). The protein is in fact a monomer in solution.
PISA is a fantastic tool, but interfaces in crystals do not always reflect the solution state. My guess (with the
information I have) is that your protein is probably a monomer too.

With regard to Michael Garavito's reply requesting more information, I would like to comment that scepticism
is indeed an important god in the pantheon of science, but that that minor deity open-mindedness also deserves the
occasional nod. 10-fold crystal symmetry is one example, but the list of "impossible" things now become mainstream
is a long one (continental drift, Earth >100,000 years old, quantum mechanics....and so on). Bayes theorem cannot
help you discover the truth if you have set its prior probability to zero. But I haven't my morning o-cha yet either.

good luck
Jeremy


On Dec 11, 2014, at 9:27 PM, Hay Dvir wrote:

Dear all,

We have a structure of a rather tightly packed homotrimer protein in the ASU with no apparent crystallographic or non-crystallographic rotational symmetry between monomers.

Attempting to establish the biological assembly, we are very interested to hear about additional similar cases you might know of.

Thanks in advance,

Hay

---------------------------

Hay Dvir Ph. D.

Head Technion Center for Structural Biology

Technion Haifa 3200003, Israel

Tel: +(972)-77-887-1901

Fax: +(972)-77-887-1935

E-mail [log in to unmask]

Website http://tcsb.technion.ac.il