compositions, and environments to be able to put these subsystem-based models
together and obtain a reliable model for the actual, more complex, target system.
We will show for one example – the light-harvesting complex of green plants
(LHCII) – some aspects of multiscale modeling of membrane protein systems and
some of the problems that need to be addressed if one wants to go beyond generic
CG models and retain a certain level of chemical specificity.
4.2 The Light-Harvesting Complex
The major light-harvesting complex (LHCII) of the photosynthetic apparatus in
green plants binds more than half of the plant’s chlorophyll (Chl) and is presumably
the most abundant membrane protein on Earth. It has become an intensely studied
model membrane protein for several reasons. Its structure is known in near-atomic
detail [309, 310] and much of its biochemistry has been elaborated in the past
decades [311]. Moreover, LHCII spontaneously self-organizes from its protein and
pigment components in vitro; therefore, recombinant versions of it can easily be
produced and modified almost at will [312]. The assembly of LHCII and the
concomitant folding of its apoprotein has been studied in some detail [313,
314]. Both processes occur spontaneously upon combining the unfolded apoprotein
and pigments in detergent solution. In vivo, the assembly of LHCII takes place in
the lipid environment of the thylakoid membrane and, most likely, is influenced by
the lipid and protein components of this membrane. This is difficult to analyze
experimentally because, so far, the self-organization of LHCII cannot yet be
achieved in a lipid membrane environment. Recently, the disassembly of LHCII
and the role of the bound/dissociating pigments in the falling apart of LHCII trimers
has also become the subject of increased interest. These pigments constitute about
one third of the total mass of LHCII and, according to the structure, significantly
contribute to the stability of the pigment–protein complex. The structural behavior
of LHCII has been analyzed by circular dichroism (CD), fluorescence, and electron
paramagnetic resonance (EPR) [312, 314–316].
One important aspect of LHCII that specifically relates to other aspects
discussed in the present review is the question of how the membrane environment
(lipid composition, membrane curvature, etc.) affects the association of LHCII
monomers to form trimers and the assembly of these trimers into the antenna
complex around the photosynthetic reaction centers. The nonbilayer-forming lipid
monogalactosyldiacylglycerol (MGDG) constitutes half of the thylakoid membrane. This membrane maintains its lamellar structure only with proteins inserted,
predominantly LHCII which, due to its concave shape, eases the curvature pressure
exerted by MGDG. It has been suggested that this curvature pressure is a driving
force for protein interaction in the membrane [317]; however, because it is not
known whether, e.g., the formation of supercomplexes of LHCII trimers eases or
increases curvature pressure, it is unclear whether MGDG (or other curvature
Computational Studies of Biomembrane Systems: Theoretical Considerations. . .
267
together and obtain a reliable model for the actual, more complex, target system.
We will show for one example – the light-harvesting complex of green plants
(LHCII) – some aspects of multiscale modeling of membrane protein systems and
some of the problems that need to be addressed if one wants to go beyond generic
CG models and retain a certain level of chemical specificity.
4.2 The Light-Harvesting Complex
The major light-harvesting complex (LHCII) of the photosynthetic apparatus in
green plants binds more than half of the plant’s chlorophyll (Chl) and is presumably
the most abundant membrane protein on Earth. It has become an intensely studied
model membrane protein for several reasons. Its structure is known in near-atomic
detail [309, 310] and much of its biochemistry has been elaborated in the past
decades [311]. Moreover, LHCII spontaneously self-organizes from its protein and
pigment components in vitro; therefore, recombinant versions of it can easily be
produced and modified almost at will [312]. The assembly of LHCII and the
concomitant folding of its apoprotein has been studied in some detail [313,
314]. Both processes occur spontaneously upon combining the unfolded apoprotein
and pigments in detergent solution. In vivo, the assembly of LHCII takes place in
the lipid environment of the thylakoid membrane and, most likely, is influenced by
the lipid and protein components of this membrane. This is difficult to analyze
experimentally because, so far, the self-organization of LHCII cannot yet be
achieved in a lipid membrane environment. Recently, the disassembly of LHCII
and the role of the bound/dissociating pigments in the falling apart of LHCII trimers
has also become the subject of increased interest. These pigments constitute about
one third of the total mass of LHCII and, according to the structure, significantly
contribute to the stability of the pigment–protein complex. The structural behavior
of LHCII has been analyzed by circular dichroism (CD), fluorescence, and electron
paramagnetic resonance (EPR) [312, 314–316].
One important aspect of LHCII that specifically relates to other aspects
discussed in the present review is the question of how the membrane environment
(lipid composition, membrane curvature, etc.) affects the association of LHCII
monomers to form trimers and the assembly of these trimers into the antenna
complex around the photosynthetic reaction centers. The nonbilayer-forming lipid
monogalactosyldiacylglycerol (MGDG) constitutes half of the thylakoid membrane. This membrane maintains its lamellar structure only with proteins inserted,
predominantly LHCII which, due to its concave shape, eases the curvature pressure
exerted by MGDG. It has been suggested that this curvature pressure is a driving
force for protein interaction in the membrane [317]; however, because it is not
known whether, e.g., the formation of supercomplexes of LHCII trimers eases or
increases curvature pressure, it is unclear whether MGDG (or other curvature
Computational Studies of Biomembrane Systems: Theoretical Considerations. . .
267