625th MEETING, LONDON 347 Table 1. Effect of lipid-to-protein-ratio on the fluorescence properties of reconstituted Ca ?+, Mg' + -A Tl'ase ATPase was labelled and reconstituted as described in the text. Fluorescence quenching is expressed as F/F,,, where F and 6, are the fluorescence intensities for the donor in the presence and absence of acceptor, respectively. The buffer was 40 mM-Hepes/ KOH, 0 . 1 M-KCI, 5 mM-MgSO,, 0.1 mM-EGTA, pH 7.0,25"C. F/Fi,* Polarixationt Lipid/protein (mol/mol) DOPC DPPC DOPC DPPC I40 0.70 0.49 0.27 1 0.332 250 0.76 0.59 0.2 I8 0.332 470 0.74 0.49 0.2 I 7 0.334 *Fluorescence excited at 475 nm and detected at 520 nm at an ATP concentration o f 0.6 p~. tFluorescence excited at 450 nm and detected at 520 nm at a molar ratio of FITC/EITC/ATPase of 0 . 2 : 0 . 8 : I at an ATPase concentration o f 0.3 p M .labelling ratio, as expected since energy transfer between FITC groups will increase in efficiency with increasing labelling ratio and energy transfer is a depolarizing event. As shown in Table 1, fluorescence polarization is, however, relatively independent of molar ratio of lipid to ATPase, again arguing against the formation of monomeric ATPase species at high dilutions in lipid. Fluorescence polarization is greater for the ATPase reconstituted into gel phase lipid than in liquid crystalline phase lipid (Table 1 ), presumably reflecting restricted motion for the ATPase in gel phase lipid.The experiments reported here suggest that the ATPase is present in reconstituted systems in oligomeric form, relatively independent of lipid-to-protein ratios. These oligomers are envisaged as dynamic crystalline arrays, continually forming and breaking up within the membrane (Napier et a [.,