protein folding

Richard Engh engh at
Wed Apr 20 02:17:14 EST 1994

Simon Brocklehurst (Bioc) (smb18 at made the
useful point:
: It's might be worth remembering when thinking about this particular point that
: both X-ray crystallography and multidimensional NMR spectroscopy provide
: time-averaged models (with some qualitative information about flexibility).
: If time-averaged models are not appropriate descriptions of the structure
: (e.g. because the proteins, or bits of the proteins are "flexible") then 
: we shouldn't expect the two techniques to produce "results" that are
: in agreement.
:   The point, though, is that when we have characterised the structure of a
: protein, protein-protein complex, protein-nucleic acid complex etc in a 
: VARIETY of different environments e.g. solution + crystal, and the results are
: IN AGREEMENT, then surely we can bet that we have a good model for the
: "physiological" structure.

: Thus, at the moment it seems that X-ray structures of isolated globular
: proteins are likely to be pretty useful models, as are X-ray models of high
: affinity protein - "other molecule" complexes.  The jury is still out on
: how useful crystallography by itself will be for weakly associating
: systems.

One of the ambiguities we have been faced with regarding serpins is the
word 'flexibility'.  It may refer to cases of localized vibrational
modes which can be resolved as a time-average model in crystallography,
usually with the assumption of harmonic isotropic motion.  It is used
to refer to motions which occur at very different time scales which may
be either resolved or time-averaged in NMR.  And we also found no better
word to describe the reorganization of secondary structure which takes
place in serpins, which cannot be reasonably time averaged.

Maybe this is straying from the topic, but are there any suggestions or
existing conventions for these alternative forms of 'flexibility'?

More information about the Xtal-log mailing list