Direct measurement of the free energies of transfer of hydrophobic membrane
-spanning alpha-helices from water to membranes is important for the determ
ination of an accurate experiment-based hydrophobicity scale for membrane p
roteins. An important objective of such a scale is to account for the prese
ntly unknown thermodynamic cost of partitioning hydrogen-bonded peptide bon
ds into the membrane hydrocarbon core. We describe here the physical proper
ties of a transmembrane (TM) peptide, TMX-1, designed to test the feasibili
ty of engineering peptides that spontaneously insert across bilayers but th
at have the important property of measurable monomeric water solubility, TM
X-1, Ac-WNALAAVAAALAAVAAALAAVAAGKSKSKS-NH2, is a 31-residue sequence with a
21-residue nonpolar core, N- and C-caps to favor helix formation, and a hi
ghly polar C-terminus to improve solubility and to control directionality o
f insertion into lipid vesicles. TMX-1 appeared to be soluble in water up t
o a concentration of at least 1 mg/mL (0.3 mM), However, fluorescence spect
roscopy, fluorescence quenching, and circular dichroism (CD) spectroscopy i
ndicated that the high solubility was due to the formation of molecular agg
regates that persisted at peptide concentrations down to at least 0.1 mu M
peptide. Nevertheless, aqueous TMX-1 partitioned strongly into membrane ves
icles with apparent mole-fraction free-energy values of -7.1 kcal mol(-1) f
or phosphatidylcholine (POPC) vesicles and -8.2 kcal mol(-1) for phosphatid
ylglycerol (POPG) vesicles. CD spectroscopy of TMX-1 in oriented multilayer
s formed From either lipid disclosed a very strong preference for a transme
mbrane a-helical conformation. When TMX-1 was added to preformed vesicles,
it was fully helical. A novel fluorescence resonance energy transfer (FRET)
method demonstrated that at least 50% of the TMX-1 insered spontaneously a
cross the vesicle membranes. Binding and insertion were found to be fully r
eversible for POPC vesicles but not POPG vesicles. TMX-1 was thus found to
have many of the properties required for thermodynamic measurements of TM p
eptide insertion. Importantly, the results obtained delineate the experimen
tal problems that must be considered in the design of peptides that can par
tition spontaneously and reversibly as monomers into and across membranes.
Our success with TMX-1 suggests that these problems are not insurmountable.