Phylogenetic analysis of membrane trafficking proteins: a family reunion and secondary structure predictions

Abstract

The realization that a highly conserved family of membrane proteins are localized to transport vesicles and selectively interact with proteins anchored at appropriate target sites of membrane fusion inspired a simple and compelling explanation of how proteins might be transferred and segregated within the cell, the "SNARE hypothesis". This model holds that vesicle and target membrane proteins (designated as v-SNARE and t-SNARE proteins, respectively) wind around one another to form a three-stranded coiled coil structure, termed the prefusion complex. While the molecular topology of the prefusion complex has not been established, the concept that phylogenetically diverse SNARE proteins may become interlocked in a stable coiled coil is particularly attractive, because such a tertiary fold would only be permitted between strictly matched binding partners. For this reason, we have performed a phenetic analysis of all known SNARE sequences to assess the evolutionary and structural relatedness of these ancient protein families. Our phylogenetic analysis and consensus structure predictions revealed that syntaxin and SNAP-25 homologs are significantly related and constitute a superfamily of t-SNARE proteins that fall naturally into four major classes with distinct architectural motifs. The synaptobrevins sorted into three different classes of v-SNARE proteins. Comparison of the consensus structure predictions within each lineage or class of SNARE proteins strongly implied that coiled coil domains may not be required for fusion complex assembly in simple eukaryotic cells. It is our hypothesis that SNARE proteins in the late secretory pathway of mammalian cells may have elaborated more complex secondary structures (coiled coils), at about the time metazoan organisms diverged from yeast, that provide a sterically rigid foundation for positioning a conserved binding domain, the amphipathic alpha-helix.