THREE-DIMENSIONAL STRUCTURES OF BIOLOGICAL MACROMOLECULES 
JOHANN Deisenhofer, PH.D., Investigator 
Dr. Deisenhofer's laboratory studies the three- 
dimensional structures of biological macromole- 
cules with the methods of x-ray crystallography. 
The aim of these studies is to understand folding, 
structural stability, and function of macromole- 
cules. Of particular interest are protein-protein in- 
teractions, the structure of membrane-spanning 
and membrane-associated proteins, photochemical 
energy conversion, energy transfer, and electron 
transfer. An ideal system to study all these aspects 
of structure and function has been the photosyn- 
thetic reaction center (RC) from the purple bacte- 
rium Rhodopseudomonas viridis. Experiments to 
extend structural studies to other systems are 
under way. 
I. Reaction Center from Rhodopseudomonas 
viridis. 
The RC was the first integral membrane protein 
whose structure was determined to atomic resolu- 
tion. It is a complex of four protein subunits (cyto- 
chrome, L, M, and H) of about 300 amino acids 
each. Associated with these protein subunits are 14 
cofactors: 4 hemes, 4 bacteriochlorophyll-ft, 2 bac- 
teriopheophytin-&, 2 quinones, 1 nonheme iron, 
and 1 carotenoid. The structure and arrangement 
of the subunits L and M and of their associated co- 
factors show a high degree of local twofold sym- 
metry 
The RC performs the first steps of the conversion 
of light energy into chemical energy during bac- 
terial photosynthesis: a photon is absorbed by 
the "special pair," a closely associated pair of bac- 
teriochlorophylls. From the excited special pair 
an electron is transferred to one of the bac- 
teriopheophytins within ~3 ps. From there the 
electron is passed on within —200 ps to the qui- 
none Qa. The electron ends up at the second 
quinone, Qb. While being transferred from the spe- 
cial pair to Qa, the electron crosses the membrane 
bilayer. Qb picks up two electrons and two protons 
and dissociates from the RC. Additional molecular 
systems in the bacterial membrane transfer elec- 
trons and protons back through the membrane; the 
electrons are recycled to the RC, and the protons 
build up a gradient that can be used by the bacte- 
rial cell to produce, for example, ATP. 
Crystallographic refinement of the atomic model 
of the RC at 2.3 A resolution has been completed; 
the R value, computed from 10,288 model atoms 
(including 201 waters) and from 95,000 unique re- 
flections is 0.193. The refinement of the RC struc- 
ture was done in collaboration with Drs. Hartmut 
Michel and Irmgard Sinning (Max-Planck-Institut 
fur Biophysik, Frankfurt, FRG) and with Dr. Otto 
Epp (Max-Planck-Institut fiir Biochemie, Martins- 
ried, FRG). 
A. Visualization of the detergent micelles. Before 
crystallization the RC molecules had to be solubi- 
lized, using the detergent A^,A^-dimethyldodecyl- 
amine-A^-oxide (LDAO). Crystallization had been 
carried out in the presence of LDAO and of the 
small amphiphile heptane-l,2,3-triol. These mole- 
cules turned out to be disordered in the crystal and 
therefore could not be located using x-ray diffrac- 
tion methods (an exception is one completely or- 
dered LDAO molecule per RC). 
To determine the distribution of the detergent in 
the crystal, a neutron diffraction study at 15 A reso- 
lution was carried out. The method of contrast vari- 
ation, using different ratios of H^O and D^O in the 
crystal mother liquor, led to a clear picture of deter- 
gent micelles formed around the central part of the 
RC molecule, where the protein surface is highly 
hydrophobic. This confirmed the assumptions, 
based on structural and functional considerations, 
of the position of the RC in the membrane. It is also 
the first successful visualization of a detergent mi- 
celle in a membrane protein crystal. This work was 
done in collaboration with Dr. Michel and Drs. 
Michel Roth and Anita Lewit-Bentley (Grenoble, 
France). 
B. Search for structural changes. One question 
that could not be answered by the structure analy- 
sis of the native RC is whether the complex un- 
dergoes structural changes during the various 
stages of light absorption and electron transfer. Ex- 
periments to answer this question were done by 
treatment of RC crystals with oxidizing (potassium 
ferricyanide) and reducing (ascorbate) agents and 
collection of sets of x-ray diffraction intensities. Re- 
finement against these data, starting from the re- 
fined native model, did not show any significant 
structural changes. 
To achieve conditions closer to the ones under 
which the RC works in vivo, attempts are being 
made to measure x-ray data from crystals under illu- 
Continued 
573 
