Background

Biological membranes fulfill a battery of essential functions in living cells. These functions are defined and regulated by specialized membrane proteins that are embedded in the hydrophobic matrix of a lipid-bilayer. This lipid matrix is not just an inert solvation space for membrane proteins but can adopt an active role in controlling their structure and function by specific lipid-protein interactions (Mourtisen and Bloom, 1993; Cantor, 1997; Mondal et al., 2014). In particular, a special lipid class called non-bilayer lipids can have a strong impact on membrane protein structure and function (van den Brinck-van der Laan et al., 2004). Non-bilayer lipids are abundant in biomembranes in general and in photosynthetic thylakoid membranes in particular (Boudiere et al., 2014). A concept that explains lipid-protein interactions in biomembranes and the role of non-bilayer lipids is the lateral membrane pressure hypothesis (Cantor, 1997; van den Brinck-van der Laan et al., 2004; Anishkin et al., 2014). This hypothesis predicts that the characteristic hydrostatic pressure profile along the z-axis of the lipid bilayer is dependent on the physical and chemical nature of the lipid mixture and that this pressure controls and modifies the structure of membrane-integral protein complexes (Cantor, 1997; van den Brinck-van der Laan et al., 2004; Anishkin et al., 2014). Studying lipid-protein interactions is challenging partially because native biomembranes are complex, composed of multi-component systems with different proteins and many different lipid types (diversity of headgroups and fatty acids). However, proteoliposomes are a system that allows examination of lipid-protein interactions under compositionally well-defined conditions. Proteoliposomes are composed of large unilamellar vesicles (LUVs) made of a lipid bilayer in which isolated membrane proteins have been incorporated (Rigaud et al., 1995; Rigaud and Levy, 2003). They have been used for studying energy-transducing membrane proteins (Rigaud et al., 1995), including photosynthetic membrane proteins (McDonnel and Staehelin, 1980; Moya et al., 2001; Yang et al., 2006; Natali et al., 2016; Crisafi and Pandit, 2017; Tietz et al., 2020). Different methods exist to integrate isolated detergent-solubilized membrane protein complexes into the lipid bilayer matrix of LUVs (Rigaud et al., 1995; Rigaud and Levy, 2003). Here, a detergent-based reconstitution protocol (Rigaud et al., 1995; Rigaud and Levy, 2003) was optimized to study lipid-protein interactions of photosynthetic light-harvesting complex II (LHCII) isolated from spinach leaves (Tietz et al., 2020). LHCII is a ~25 kDa-sized trimeric protein complex embedded in the thylakoid membranes of plants and green algae that binds 42 chlorophylls and four carotenoids (Barros and Kühlbrandt, 2009). It is the most abundant membrane protein complex on earth and optimized for highly efficient harvesting of sunlight. The LHCII-proteoliposomes preparation consists of four steps (Rigaud et al., 1995; Rigaud and Levy, 2003; Tietz et al., 2020): (i) preparation of LUVs with defined lipid composition and diameter, (ii) destabilization of the LUVs by doping them with a maximal amount of detergent without formation of mixed lipid-detergent micelles, (iii) incorporation of detergent-solubilized LHCII in the destabilized LUVs, and (iv) detergent removal by polystyrene beads (Biobeads) and purification by gel filtration. Compared to other protocols, the procedure introduced here generates LHCII-proteoliposomes with a very low protein density (lipid/protein ratio ~60,000), specifically allowing the study of lipid-protein interactions. At the same time, the structural and functional integrity of the LHCII is highly preserved during the proteoliposome preparation, as indicated by CD- and low-temperature fluorescence spectroscopy (Tietz et al., 2020). The protocol is versatile since it is expected to work for other membrane protein complexes and lipid mixtures as well.