Enzyme-enzyme interactions within human erythrocytes as suggested from prelytic release.

The molecular sieving property of human erythrocyte membrane in slightly hypotonic media has been utilized for studying the intracellular localization of some proteins (Cseke et al., 1978, FEBS Lett. 96 15-18). It is now shown that four proteins which appear to be uniformly distributed within the cell behave similarly irrespective of the membrane resistance differences caused by changes of temperature and metabolic energy supply. Five enzymes catalyzing consecutive reactions in the glycolytic pathway between triosephosphate formation and lactate production are released from erythrocytes in quantities deviating from those predicted on the basis of molecular sieving. The data are compatible with the assumption that these enzymes form complexes with each other under in vivo conditions and that complex formation is facilitated if glycolysis is working at high rate (37 degrees C).

On the molecular sieving property of the human erythrocyte membrane. Localization of some proteins within the cell.

The membrane of swollen erythrocytes was found to be more resistant to hypotonicity at 37 degrees C than at 0 degrees C either in the presence or absence of added glucose and even more resistant if ATP formation was stimulated by preincubation of the sedimented cells with adenosine. The membrane, irrespective of temperature and metabolism behaves as a molecular sieve for proteins, i.e. the smaller the molecular radius of a protein, the higher is the amount released by molecular sieving. This phenomenon allows one to discriminate between uniform distribution of proteins within the cell and complex formation of enzymes. Some glycolytic enzymes catalyzing reactions between triosephosphate and lactate production were found to be clustered near the surface of the cell but not bound to the membrane. The data indicate that some of these enzymes may form complexes under in vivo conditions.

Some structural features of rabbit muscle aldolase as derived from its limited proteolysis.

The peptides released during the limited tryptic proteolysis of rabbit muscle aldolase (Biszku et al., 1973) were located in the primary structure. The pattern of peptide liberation, peptide bond splitting and activity decrease in compatible with two structural models for the truncated tetrameric product, named aldolase-T. According to the more probable model aldolase-T has the structure A+A+B++B++. Subunits B++ are deprived of the segments comprising residues 1-27, 42-71 and 306-364 of the intact enzyme and are inactive. The fragment comprising residues 28-41 is non-covalently attached to these subunits. Subunits A+ are depleted only of peptides 1-27 and 324-332 and retain 70% activity. In these subunits the fragment comprising residue 333-364 remains non-covalently bound. The molecular weights of the truncated subunits, determined with polyacrylamide-gel electrophoresis in the presence of sodium dodecylsulfate support the above conclusions. Aldolase-T can be reversibly denatured at pH 2 or in 4 M urea. The recovery of enzymatic activity after decreasing urea or acid concentration indicates the non-covalent rebinding of fragment 333-364. This fragment is named the "T-peptide" of trypsin-treated aldolase. It is suggested that segments 1-27 and 324-364 are not necessary for the renaturation process. Since aldolase-T is a tetramer it seems that large parts of the N- and C-terminal regions of the enzyme are not involved in the intersubunit interactions. The C-terminal region of aldolase, starting around residue 324, appears to be necessary to the structure of the active site. In contrast to this, the N-terminal region up to residue 27 and probably to residue 60, is not part of the active center.