Biology Reference
In-Depth Information
It is now clear that neither a protein's net amino acid composition, nor its distribution
into totally artificial classifications, dictate whether the protein will reside in a membrane
or not. If the net amino acid composition has little if anything to do with locating
a protein into a membrane, the driving forcemustresidewithintheaminoacid
sequence
.
In a protein's three dimensional structure, the more hydrophobic amino acids should
reside together, forcing that portion of the protein into the membrane, while the more
hydrophilic amino acids should organize into protein domains that will interact favor-
ably with water. Therefore, membrane proteins, analogous to membrane lipids (see
Figure 5.7), must also be amphipathic (
Figure 6.2
), having both hydrophobic and hydro-
philic regions.
The outstanding question is how the amino acid sequence relates to a protein's loca-
tion either into or out of the membrane. One direct possibility would be to determine the
precise location of each amino acid in a protein by X-ray crystallography. Unfortunately,
however, this has proven to be a very difficult task for most membrane proteins. There
are now more than 30,000 known water-soluble globular proteins and more than 20,000
membrane proteins. However, it is so much easier to obtain crystals from globular
proteins that
6,000 crystallography structures of these proteins have been determined
compared to a few dozen structures of integral membrane proteins. So, are there other
methods of identifying protein domains that prefer a location in the hydrophobic core
of a membrane? To answer this question, the amino acids cannot simply be lumped
into classes, but instead must be characterized individually. This approach involves
hydropathy plots. By this method each amino acid is assigned a unique Hydropathy
Index value that expresses its lipid/water partition. These values represent a measure
of the free-energy change accompanying the movement of an amino acid from a hydro-
phobic solvent to water. Some amino acids are very hydrophobic while others are very
hydrophilic. Still others are in between. Since the assigned Hydropathy Index values
are entirely empirical, a number of very different Hydropathy Index tables have been
suggested. Probably the major approach for this type of membrane study comes from
Kyte and Doolittle
[5]
. Their method is based on a hydropathy scale that measures
the relative hydrophilicity/hydrophobicity of each of the 20 common amino acids. The
values are an amalgam of experimentally derived observations taken out of
>
the
POLAR
HYDROPHOBIC
POLAR
FIGURE 6.2
Amphipathic nature of membrane integral proteins. The protein on the left would sit on one side of
the membrane (it would be either an ecto or endo protein) while the protein on the right would span the membrane
(it would be a trans-membrane protein).
