Segregation of micron-scale membrane sub-domains in live murine sperm.

Lipid rafts, membrane sub-domains enriched in sterols and sphingolipids, are controversial because demonstrations of rafts have often utilized fixed cells. We showed in living sperm that the ganglioside G(M1) localized to a micron-scale membrane sub-domain in the plasma membrane overlying the acrosome. We investigated four models proposed for membrane sub-domain maintenance. G(M1) segregation was maintained in live sperm incubated under non-capacitating conditions, and after sterol efflux, a membrane alteration necessary for capacitation. The complete lack of G(M1) diffusion to the post-acrosomal plasma membrane (PAPM) in live cells argued against the transient confinement zone model. However, within seconds after cessation of sperm motility, G(M1) dramatically redistributed several microns from the acrosomal sub-domain to the post-acrosomal, non-raft sub-domain. This redistribution was not accompanied by movement of sterols, and was induced by the pentameric cholera toxin subunit B (CTB). These data argued against a lipid-lipid interaction model for sub-domain maintenance. Although impossible to rule out a lipid shell model definitively, mice lacking caveolin-1 maintained segregation of both sterols and G(M1), arguing against a role for lipid shells surrounding caveolin-1 in sub-domain maintenance. Scanning electron microscopy of sperm freeze-dried without fixation identified cytoskeletal structures at the sub-domain boundary. Although drugs used to disrupt actin and intermediate filaments had no effect on the segregation of G(M1), we found that disulfide-bonded proteins played a significant role in sub-domain segregation. Together, these data provide an example of membrane sub-domains extreme in terms of size and stability of lipid segregation, and implicate a protein-based membrane compartmentation mechanism.