Dr Lupas,
Your posting was very interesting. It would appear that this is another
example. I would say that your example was *more* convincing than that of
Heinemann and Hahn's, since the sequence motif clearly conserved in the
permuted sequence. Are the DNA sequences known? I would assume so, given
the location of the papers.
We (that's Ponting and Russell) commented that circular permutations within
proteins would be aided by the occurence of tandem repeats, since it is
easier to extract a single stretch of DNA/protein than to take one piece out
and move it to another location. In other words, it requires fewer steps.
One could apply this thinking to your example, viz. one SLH repeat is:
H1 = helix 1, S= strand, H2 = helix 2 (by your prediction):
Consider a hypothetical ancestor of SLH-type domains that contains *four*
repeats of helix-strand-helix (labelled a-d):
H1a-Sa-H2a--H1b-Sb-H2b--H1c-Sc-H2c--H1d-Sd-H2d
(--Omp----)
(-----------Slp--------)
(----------------XynX-------------)
(--------------OlpB-----------------)
One could arrive at "homologues" of the four proteins you discussed by merely
cutting the ancestral protein as shown above, and obviously adding bits of
sequence and mutating as appropriate. For the evolution of OlpB, this would
avoid having to, for example, cut off the first helix and stick it on the
end: one would only need to take one section out of the ancestral protein as
shown.
I was particularly intrigued by Heinmann and Hahn's comment that
traditional methods of sequence comparison might not be able to detect
permutations within the database. Who knows how many of these things are
out there and un-discovered?
And one thing that strikes me is how permutations can go un-noticed, or at
least not spotted as "circular permutations" explictly. I think this applied
to the Heinmann and Hahn example, since the original paper showed in a Figure
that the sequence was "permuted", but the word "circular permutation" never
appeared (to the best of my knowledge) in the manuscript. The description
was "only in F. succinogenes the order of these domains is reversed" (Eur. J.
Biochem, 204, 3, 13-19, 1992). Again, one wonders how many more examples
there are.
To all other molbio.proteins readers, I ask for a general discussion of
the meaning and significance of these things within nature. I would ask:
a) how do they fold? b) what functional significance do they have, and c) does
anyone know how many of these things are out there?
Yours, etc.
Robert B. Russell Biomolecular Modelling, Imperial Cancer Research Fund
44 Lincoln's Inn Fields, P.O. Box 123, London, WC2A 3PX, U.K.
Tel: 44 171 269 3583 FAX: 44 171 269 3479 e-mail: russell at icrf.icnet.uk
WWW http://bonsai.lif.icnet.uk/people/rob/rob.html
