Based on sensitivity of the protein to proteinase K digestion we now suggest that the N-terminus of subunit VIII is similarly oriented, implying that this subunit does not span the membrane. Despite this, however, subunit VIII cannot be extracted from the membrane even after treatment with 0.1 M Na2CO3 at pH 11.5, showing that the protein is integrally embedded in the membrane. A similar behaviour was displayed by another low molecular weight protein of the complex, subunit VII, which faces the matrix side. A model for the topology of these subunits in the membrane is discussed with respect to the structure of the complex and their involvement in quinone binding.,~?ey words: Membrane topology; Ubiquinol-cytochrome c ~.xidoreductase; Epitope tagging; Integral membrane protein; ( ~accharomyces cerevisiae) tein [8], have shown that at least this part of subunit VIII is exposed to the intermembrane space [9]. In addition, hydrophobicity plots of subunit VIII from S. cerevisiae, K. lactis, S. pombe and the homologous subunit from bovine heart predict the presence of a conserved hydrophobic stretch potentially capable of spanning a lipid bilayer (Fig. 1 and [10]), thus implying that the N-terminus of the protein faces the mitochondrial matrix.In order to verify the accuracy of these predictions, we have tagged the N-terminal half of the protein with two different epitopes (c-myc [11] and HA [12]). As indicated in Fig. I, these epitopes were not placed at the N-terminus of the protein, since the first 25 amino acids may contain a non-cleaved import signal [13]. Using limited protease treatment we have determined the position of the c-myc and the HA epitope with respect to the inner membrane. The outcome of this is combined with carbonate extraction experiments on the entire bca complex in order to further model the topology and subunit arrangement of this large respiratory enzyme.