Separation of subcellular organelles by two-phase partition is thought to reflect differential partition of the organelles between the two phases or between one of the phases and the interface. Studies by Fisher and colleagues [Fisher & Walter (1984) Biochim. Biophys. Acta 801, 106-110] suggest that cell separation by phase partition is a dynamic process in which the partition changes with time. This is mainly due to association of the cells with sedimenting droplets of one phase in the bulk of the other. Rat liver organelle partition was studied to determine whether the same dynamic behaviour is observed. Partition was clearly time-dependent during 24 h at unit gravity, and was also affected by altering the volume ratio of the two phases and the duration of phase mixing. These results indicate that, as with cells, the partition of organelles between phases is a dynamic process, and is consistent with the demonstration that organelles adhere to the phase droplet surfaces. Optimization of the volume ratio between phases may lead to significant processing economies. Organelle sedimentation in the upper phase was significantly faster than in the isoosmotic sucrose. Theoretical modelling of apparent organelle sizes indicates that aggregation occurs in the poly(ethylene glycol)-rich upper phase. This phenomenon is likely to limit the use of this technique in organelle separations unless means can be found to decrease aggregation.
