230
G. BOLDT et aI.
acids (Engelman and Zaccai, 1980; Trewhella et ai., 1983) and partially deuterated
a-helices (Trewhella et al., 1986; Popot et al., 1989). This technique was also successful in the localization of perdeuterated or partially deuterated retinal within
the protein (Jubb et ai., 1984; Heyn et ai., 1988). Heavy-atom labelled retinal analogues incorporated into the protein confirmed these results by X-ray diffraction
(Biildt et ai., 1991). In early neutron diffraction experiments the hydration of purple membranes was studied (Zaccai and Gilmore, 1979) and more recently the
location of the proton channel in bR was determined by neutron diffraction in
combination with H20/D20 exchange experiments (Papadopoulos et ai., 1990).
In 1990, cryo-electron microscopy provided an additional breakthrough providing a structure of bR at 3.5 A resolution in the X-Y plane and 7 A in the Zdirection. This picture allowed for the first time, to construct a molecular model
of bR including all the information from other methods mentioned above (Henderson et aI., 1990). This model was further improved by cryo-electron microscopy data of Grigorieff et ai. (1996) and Kimura et al (1997).
2.2
Structural Investigations on Photocycle Intermediates
For a complete understanding of the function of a protein like bR it is desirable
to detect structural changes in space and time with high resolution parallel to the
working cycle. Since the intensity scattered from a single molecule is too low, taking into account the limits set by radiation damage, an ensemble of molecules has
always to be considered. In this situation information about structural changes
during the working cycle can be obtained in two ways, either by trapping intermediate states or by time-resolved detection of the scattered intensity after excitation. Both alternatives were successfully carried out in the case of bR.
2.2.1
Trapping of the M-State in Wild-Type bR
After the observation that a proton is releaced by bR at the extracellular side during the transition from L to M, the M-intermediate was considered to be a strategic state for the pumping process. It seemed that the knowledge of the M -state
structure would give some insight into the mechanism of proton translocation.
One of the first trapping experiments of the M-intermediate at low temperatures
was performed by Glaser et ai. (1986) using electron diffraction. These experiments showed no intensity changes in the resolution region from 60-5 A and
small changes between 5 and 3.5 A. Therefore neutron diffraction experiments
were undertaken by Dencher et ai. (1989) with the aim of observing changes in
the distribution of water molecules in comparison to the ground state. As it was
known that GuaHCI at high pH slows down the decay of the M state, a stack of
several PM-films was soaked in a buffer containing GuaHCI at pH 9.4 and illuminated at +8°C. The films became yellow indicating the complete transformation
to the M state, which was then preserved at liquid-nitrogen temperatures in a
cryostat. Neutron diffraction patterns of the ground state and the M-intermediate
showed clear differences in the reflection intensities of up to 9 % in L IAII/LI in
G. BOLDT et aI.
acids (Engelman and Zaccai, 1980; Trewhella et ai., 1983) and partially deuterated
a-helices (Trewhella et al., 1986; Popot et al., 1989). This technique was also successful in the localization of perdeuterated or partially deuterated retinal within
the protein (Jubb et ai., 1984; Heyn et ai., 1988). Heavy-atom labelled retinal analogues incorporated into the protein confirmed these results by X-ray diffraction
(Biildt et ai., 1991). In early neutron diffraction experiments the hydration of purple membranes was studied (Zaccai and Gilmore, 1979) and more recently the
location of the proton channel in bR was determined by neutron diffraction in
combination with H20/D20 exchange experiments (Papadopoulos et ai., 1990).
In 1990, cryo-electron microscopy provided an additional breakthrough providing a structure of bR at 3.5 A resolution in the X-Y plane and 7 A in the Zdirection. This picture allowed for the first time, to construct a molecular model
of bR including all the information from other methods mentioned above (Henderson et aI., 1990). This model was further improved by cryo-electron microscopy data of Grigorieff et ai. (1996) and Kimura et al (1997).
2.2
Structural Investigations on Photocycle Intermediates
For a complete understanding of the function of a protein like bR it is desirable
to detect structural changes in space and time with high resolution parallel to the
working cycle. Since the intensity scattered from a single molecule is too low, taking into account the limits set by radiation damage, an ensemble of molecules has
always to be considered. In this situation information about structural changes
during the working cycle can be obtained in two ways, either by trapping intermediate states or by time-resolved detection of the scattered intensity after excitation. Both alternatives were successfully carried out in the case of bR.
2.2.1
Trapping of the M-State in Wild-Type bR
After the observation that a proton is releaced by bR at the extracellular side during the transition from L to M, the M-intermediate was considered to be a strategic state for the pumping process. It seemed that the knowledge of the M -state
structure would give some insight into the mechanism of proton translocation.
One of the first trapping experiments of the M-intermediate at low temperatures
was performed by Glaser et ai. (1986) using electron diffraction. These experiments showed no intensity changes in the resolution region from 60-5 A and
small changes between 5 and 3.5 A. Therefore neutron diffraction experiments
were undertaken by Dencher et ai. (1989) with the aim of observing changes in
the distribution of water molecules in comparison to the ground state. As it was
known that GuaHCI at high pH slows down the decay of the M state, a stack of
several PM-films was soaked in a buffer containing GuaHCI at pH 9.4 and illuminated at +8°C. The films became yellow indicating the complete transformation
to the M state, which was then preserved at liquid-nitrogen temperatures in a
cryostat. Neutron diffraction patterns of the ground state and the M-intermediate
showed clear differences in the reflection intensities of up to 9 % in L IAII/LI in