A more than very late answer to the first question asked by Dr.
Brocklehurst ages ago...
I would like to discuss the idea that the most hydrophobic
amino-acids are those present in the transmembrane (TM) domains of membrane
proteins, and especially in membrane proteins with a single TM domain.
These domains are supposed to interact only with the hydrophobic core of
the membrane bilayer.
A very interesting paper deals with the distribution of amino-acids
within the putative TM domains of proteins : Jones D.T., Taylor W.R. &
Thornton J.M. "A mutation data matrix for transmembrane proteins" FEBS
Lett. 339 (1994) 269-275. The authors extracted documented TM sequences
from the SWISS-PROT database.
One can calculate from table I of this paper a "TM usage ratio" for
amino-acids, by dividing their frequency of occurence in single TM segments
by that in a general set of proteins.
The most "overused" amino-acids in single TM domains, i.e. the "most"
hydrophobic ones, are thus :
Ile 2.46 ("TM usage ratio")
Val 2.37
Leu 1.93
Ala 1.48
Phe 1.38
and the least represented amino-acids, i.e. the most hydrophilic, are :
Glu 0.08
Asp 0.11
Gln 0.16
Lys 0.20
Asn 0.26
Obviously this classification is not very different from the ones derived
from the physico-chemical properties of amino-acids, as discussed so far.
One may object that this view is biased, because the vast majority
of these TM domains found in databases are putative, i.e. derived from
prediction algorithms relying on these very same physico-chemical
properties of amino-acids. How many "hard" structures do we have for this
class of proteins ? - I know of only one. Nevertheless, I tend to believe
that we have sufficient biochemical data for many of these proteins to
warrant the assignment of rather long stretches of hydrophobic amino-acids
as transmembranal.
Hope this starts the discussion again.
Pierre
=============================================
Pierre Hubert
INSERM U. 338 - 5 rue Blaise Pascal
67084 Strasbourg Cedex France
tel (33) 88 45 67 20 - fax (33) 88 60 08 06
hubert@neurochem.u-strasbg.fr
=============================================