Domains and motifs

L.A. Moran lamoran at gpu.utcc.utoronto.ca
Wed Jan 15 11:48:04 EST 1997


In article <E40M45.Cy9 at utcc.utoronto.ca>, Kevin Gardner <gardner at bloch.
med.utoronto.ca> wrote:

>Whoa --- speaking as a structural biologist who has worked on DNA-binding
>domains such as these, I don't agree with this assessment.  The
>Cys(2)His(2) canonical zinc finger and the GAL4/Cys(6) both fall into
>a domain by your criteria that you initially listed: discrete, 
>independently-folding regions with globular structures.  They're connected
>to other domains by linker regions, and both have multiple regions of
>secondary structure (2 strands of beta sheet and a helix for the
>Cys(2)His(2), 2 helices for GAL4).  Yes, these domains require metal
>ion(s) in order to assume their folded, active state.  
>
>Given this above information, I'd be interested in hearing how these
>would fail to be considered domains.  Yes, they are small domains 
>(as asked for), but I believe they meet the appropriate criteria.

I readily admit that the literature is confusing on this issue and many
workers refer to helix-turn-helix "domains" and zinc finger "domains".
The purpose of my posting was to point out that the protein structure
experts often refer to these as "motifs" because the word "domain" has
a different meaning. You will notice in the more recent literature and
in textbooks that the word "motif" is gaining in popularity - rightly
so in my opinion. (see the titles of Lee et al. (1989), Berg (1990), 
and Chan et al. (1993)).

In Coleman's review of zinc proteins from 1992 he refers to zinc finger
motifs in some cases but to domains in other cases so even five years
ago there was some confusion in the literature.

Let's see what Creighton has to say in his textbook. In chapter 8 he talks
about DNA binding proteins and the various binding motifs including the
zinc finger motif (p.358). He doesn't refer to these regions as domains
because he defines domains quite differently,

     "The folded structures of most small proteins are roughly 
      spherical and remarkably compact, with very irregular surfaces.
      The structures of most proteins that have more than about 200
      residues appear to consist of two, three, or more structural 
      units, usually referred to as domains. The domains of a protein
      molecule interact to varying degrees, but less extensively than
      do structural elements within domains. Often a single segment of
      a polypeptide chain links the domains, and each domain consists
      of a single stretch of polypeptide chain. ... The definition of
      a domain is not rigorous, and the division of a structure into
      domains is a subjective process that is done in different ways 
      by different people. Other terms and subdivisions, such as
      subdomain and folding unit, are also encountered in the literature.
      Nevertheless, the presence of domains in many protein molecules
      is clear to all observers (Figures). Domains are most evident
      by their compactness, which can be expressed quantitatively as 
      the ratio of the surface area of a domain to the surface area
      of a sphere with the same volume ..."

                                            Creighton (1993) p.219

As a general rule of thumb you should be able to recognize domains in a
space-filling model of a protein. They form separate "blobs" that are
quite obvious. If you can't pick out a specific region easily by looking
at the gross morphology of the protein then it isn't a domain by Creighton's
definition.

Zinc fingers are not domains by this definition. A typical zinc finger
contains three regions of secondary structure (two beta regions and an
alpha helix). This is typical of many other supersecondary structures
or motifs that have been recognized in proteins (ie: beta-alpha-beta units).
There doesn't seem to be any convincing reason to call a zinc finger
a domain but not other types of supersecondary structure. And there
doesn't seem to be any logical reason to degrade the term "domain" when
we have a perfectly good alternative in "motif".



Larry Moran


Berg, J.M. (1990) Zinc Fingers and Other Metal-Binding Domains. J. Biol.
   Chem. 265,6513-6516.

Chan, Y-L., Suzuki, K., Olvera. J. and Wool, G. (1993) Zinc finger-like
   motifs in rat ribosomal proteins S27 and S29. Nucl. Acid. Res. 21,649.

Creighton, T.E. (1993) Proteins: Structures and Molecular Properties.
   W. H. Freeeman, New York 

Coleman, J.E. (1992) Zinc Proteins: Enzymes, Storage Proteins, Transcription
   Factors, and Replication Proteins. Ann. Rev. Biochem. 61,897-946.

Lee, M.S., Gippert, G.P., Soman, K.V., Case, D.A. and Wright, P.E. (1989)
   Three-Dimensional Solution Structure of a Single Zinc Finger DNA-Binding
   Domain. Sci. 245,635-637.




More information about the Methods mailing list