Structural measurements on membrane PsbH protein

**Zbyněk Halbhuber⁺, Norbert Müller^*, Ivana Kutá Smatanová^+,$, Julie Wolfová⁺, Dalibor Štys^+,$,#**

+ Institute of Physical Biology, University of South Bohemia, Zámek 136, 373 33 Nové Hrady, Czech Republic

* Johannes Kepler Universität, Altenbergerstrasse 69, Linz, Austria

$ Institute of Landscape Ecology, Academy of Sciences of the Czech Republic, Zámek 136, 373 33 Nové Hrady, Czech Republic

# Institute of Microbiology, Academy of Sciences of the Czech Republic, Opatovický mlýn, 37901 Třeboň, Czech Republic

Introduction

The psbH protein is one of key components for assembly of Photosystem II [1]. In higher plants it is one of the proteins expressed in etiolated and illuminated leaves on the same level, which indicates that it function may be considered separately from the rest of the multiprotein complex. It was thus chosen as model small protein with dominant transmembrane helix for assessment of several methods for structure determination.

Experiments

Preparation of psbH

The PsbH protein of cyanobacterium Synechocystis sp. PCC 6803 was expressed as a fusion protein with glutathione-S transferase (GST) in E. coli [2]. We isolated the 15N labeled PsbH protein in concentration of 1.1 mg/ml in presence of detergent octyl glucoside (OG). We also isolated non-labeled protein for preliminary lipid titration experiments measured by circual dichroism (CD) spectrometer.

Reconstitution and CD measurements

The liposomes were prepared by reverse-phase evaporation technique from the thylakoid membrane lipids; sulphoquinovosyl diaglyceride (SQDG), digalactosyl diglyceride (DGDG), monogalactosyl diglyceride (MGDG) and phosphatidyl glycerol (PG). The most favourable lipid, which induced complex protein folding, detected as formation of the negative band approx. 222 nm in CD spectra, seemed to be PG. Very similar changes were observed at higher concentration also in SQDG, however folding of clearly different nature was achieved upon titration by DGDG. This indicates that the protein folding may not be directly related to specific binding of lipids, rather we observe two different types of folding in lipid bilayers of two different properties.

NMR measurements

The CD measurements revealed folding of the PsbH protein in detergent micelles after addition of sufficiant amount of lipid. We added to each protein sample appropriate amount of the lipid to reach optimal protein/ lipid ratio. Unfortunately NMR measurements showed a huge decrease of signal and recording of the remaining 15N signals into the narrow area (figure 1). This would indicate very rigid lipid-protein micelles, which relax to fast to be recorded.

Micelle destabilisation using sonication or temperature increase led to only partial improvement, therefore we added into the sample new detergents; CHAPS and digitonin. The simple addition of CHAPS or digitonin did not destabilize micelles suficiently and we had to remove lipids by dialysis. After dialysis signal recovered, moreover the new peaks indicated the further protein folding. The combination of digitonin and octyl glucoside was determined as the most effective combination to induce apparent protein folding (figure 2).

Figure 1 PsbH protein after addition of PG

Figure 2 PsbH OG with digitonin

Crystallisation of fusion protein and X-ray diffraction

Fusion protein GST-psbH was crystallized using standard procedures. Sufficiently large crystals were obtained. The diffraction was, however, not measurable due to problems of crystal cryoprotection. Experiments continue on improvement of crystal quality of both the fusion protein and the protein itself.

[1] Komenda J., Lupínková L. & Kopecký J., Eur. J. Biochem., 269 (2002) 610–619.