the retinal could be distinguished by making use of light. Spectra were
measured both before and after a laser flash. The differences between the
two spectra were used to identify changes in vibrations, which arise from
light-induced structural changes of the retinal and surrounding protein.
These difference spectra were combined for comparison with isolated
retinal and isotopic labeling of the retinal to assign the vibrations to specific
bonds. For example, the spectrum shows a carbon–carbon stretch mode
for the dark state at 1201 cm−^1 , shifted to 1167 cm−^1 in the light, consistent
with the trans-to-cisisomerization. Also, the identification of the protona-
tion state of the Schiff base could be made, based upon a shift of a bond
from 1646 to 1629 cm−^1 when the sample was deuterated.
Structural studies
The first indications of the structure of rhodopsin came from the sequence
of the gene encoding bacteriorhodopsin. This sequence showed the pres-
ence of seven long stretches of hydrophobic amino acid residues connected
by short segments of hydrophilic residues. If these residues were assumed
to form an αhelix, then each helix would be long enough to span the
cell membrane. Thus, the sequence of bacteriorhodopsin was used to pre-
dict the presence of seven transmembrane helices. The arrangement of
these helices was not determined by this analysis but mutagenesis studies
showed that residues from the C and E helices influenced the functional
properties.
A critical development was the elucidation of the three-dimensional
structure using electron microscopy. Bacteriorhodopsin is highly concen-
trated in the cell membrane and forms a crystalline array. The availability
of two-dimensional crystals provided the opportunity to use electron micro-
scopy. When the electron beam probes a sample, the image mapped out is
a projection of the protein onto a plane. To determine a three-dimensional
image, images are obtained for the sample rotated through all possible
angles. By knowing how much the sample has turned and how the image
has changed for each angle, it is possible to infer the three-dimensional
object.
A major limitation of images derived from biological samples by electron
microscopy is that the atoms that compose proteins – carbon, nitrogen, and
oxygen – have very comparable numbers of electrons. Not only can these
atoms not be distinguished from each other but these atoms are low in
contrast; that is, it is difficult to identify the protein from the surround-
ing matrix holding the sample. Samples can be labeled with heavy atoms,
but then the images are limited by the quality of the labeling and are
much too course for determination of a protein’s structure. Increasing the
strength of the signal by using intense electron beams does not help as
this results in rapid radiation damage of the sample.
380 PART 3 UNDERSTANDING BIOLOGICAL SYSTEMS USING PHYSICAL CHEMISTRY
