With increasing SDS/protein ratios, covalent phosphorylation by ATP and Pi is abolished before ATP hydrolysis (Pi production) ceases. We have shown that the SDS-dependent profiles of the decline in covalent phosphorylation by either substrate are virtually identical, reflecting a common mechanism of detergent interaction, while ATP can be hydrolysed via a non-covalent phosphointermediate. Our studies support that the transfer of both terminal Pi from ATP, as well as Pi to its final binding site, is a multistep reaction involving electrostatic interaction with one or more amino acid side chains, including a Lys residue.The reaction cycle of the sarcoplasmic reticulum transmembranal ATPase has been well characterized to date (see minimum scheme by deMeis and Vianna [l] [4, 51, and by the use of site-directed mutagenesis [6, 71, have indicated an additional involvement of other amino acid residues in the phosphoryl-transfer reaction (see also [S]). The partly distant location of these residues from the catalytic site Asp351, with respect to the primary structure [9], stresses the importance of a well-preserved tertiary structure for this reaction step.We have reported recently [lo] that rising non-solubilizing SDS concentrations depress and then abolish steady-state phosphorylation of the ATPase by Pi (E2P). This has been found to be partly due to competitive inhibition of covalent Pi binding by SDS. We also indicated a surprisingly higher SDS resistance of overall ATP hydrolysis, compared to that of phosphorylation by Pi. Here we show that this discrepancy is not due to a higher stability towards SDS of Eyzp, the first covalent phosphointermediate in the forward mode of the ATPase reaction cycle, as observed with non-ionic detergents [ l l , 121. However, by filtration measurements ([I4C]ATP and [32P]ATP) we obtained higher levels of enzyme-adherent Pi (the sum of all phosphointermediates containing Pi covalently and non-covalently bound) which, depending on the concentrations of free Ca2+ in the assay, also exhibited less sensitivity to SDS-mediated degradation.
MATERIALS AND METHODSSarcoplasmic reticulum vesicles were isolated from rabbit skeletal muscle by the procedure introduced by Hasselbach and Makinose [13]. Further treatment of the vesicles before use included dialysis to reduce further the concentration of K + from previous steps in preparation [lo], adjustment of the protein concentration to 10 mg . ml-', and incubation with 50 pM of the Ca ionophore, A23187, at room temperature for several minutes.All experiments were performed in the absence of K+, at a final pH of 6.15.The integrity of the vesicular material with respect to unaltered protein and phospholipid compositions, in the presence of increasing non-solubilizing SDS concentrations, was previously established [lo].