optimize this cala24-asp,asp61-gly defect, with only five laying on the c subunit and 13 in the

a subunit, all near the aarg210 TOSidue, which is required for proton translocation (further

discussed below) (119, 120). Early models of the organization of Fo suggested that the

ring of c subunits were situated on the periphery, surrounding the centrally located a and

b subunits, which rotated in the center of the ring (121). This model was proved wrong

by high-resolution NMR data (122) and cross-linking experiments (123, 124) which

indicated that the oligomer of c subunits are closely packed with a lipid filled core less

than 25 A+ wide. The individual c subunits are packed front-to-back such that the second

helix of each is situated towards the exterior and the first helix is located on the interior,

which renders the casp61 exposed to the lipid environment. The uncommonly high pKa

(7. 1) of the casp61 carboxyl side chain is likely due to this hydrophobic environment (125).

Furthermore, scanning force and cryoelectron microscopy demonstrated that Fo is

asymmetrically arranged in the membrane (27, 126, 127). For these reasons, the a and b

subunits are thought to be situated to the periphery of the ring of c subunits.

 High resolution structures of membrane-bound proteins were nonexistent for many

decades past structural determination of soluble proteins and still prove difficult to this

day due to their highly hydrophobic nature. The membrane intrinsic c subunit of E coli,

which was solved by NMR in an organic solvent (chloroform-methanol-water) in the

1990's, was one of the earliest high-resolution structures of a transmembrane helical

protein (122, 128-130). Notable, the c subunit could be reconstituted from the organic

solvent mixture with complete preservation of function; therefore, it was clearly not

irreversibly denatured (113). As predicted two decades prior, the c subunit folds as a

hairpin of two extended co-helices with the casp61 Of the second helix packed less than 5 A~