Biochemical characterization of detergent-resistant membranes: a systematic approach.

Lateral segregation of cholesterol- and sphingomyelin-rich rafts and glycerophospholipid-containing non-raft microdomains has been proposed to play a role in a variety of biological processes. The most compelling evidence for membrane segregation is based on the observation that extraction with non-ionic detergents leads to solubilization of a subset of membrane components only. However, one decade later, a large body of inconsistent detergent-extraction data is threatening the very concept of membrane segregation. We have assessed the validity of the existing paradigms and we show the following. (i) The localization of a membrane component within a particular fraction of a sucrose gradient cannot be taken as a yardstick for its solubility: a variable localization of the DRMs (detergent-resistant membranes) in sucrose gradients is the result of complex associations between the membrane skeleton and the lipid bilayer. (ii) DRMs of variable composition can be generated by using a single detergent, the increasing concentration of which gradually extracts one protein/lipid after another. Therefore any extraction pattern obtained by a single concentration experiment is bound to be 'investigator-specific'. It follows that comparison of DRMs obtained by different detergents in a single concentration experiment is prone to misinterpretations. (iii) Depletion of cholesterol has a graded effect on membrane solubility. (iv) Differences in detergent solubility of the members of the annexin protein family arise from their association with chemically different membrane compartments; however, these cannot be attributed to the 'brick-like' raft-building blocks of fixed size and chemical composition. Our findings demonstrate a need for critical re-evaluation of the accumulated detergent-extraction data.