INSULIN-RESPONSIVE TISSUES CONTAIN THE CORE COMPLEX PROTEIN SNAP-25 (SYNAPTOSOMAL-ASSOCIATED PROTEIN-25) A-ISOFORM AND B-ISOFORM IN ADDITION TO SYNTAXIN-4 AND SYNAPTOBREVIN-1 AND SYNAPTOBREVIN-2

SNAP-25 (synaptosomal-associated protein 25), syntaxin and synaptobrev in are the three SNARE [soluble NSF attachment protein receptor (where NSF = N-ethylmaleimide-sensitive fusion protein)] proteins that form the core complex involved in synaptic vesicle docking and subsequent f usion with the target membrane. The present study is aimed at understa nding the mechanisms of fusion of vesicles carrying glucose transporte r proteins with the plasma membrane in human insulin-responsive tissue s. It describes the isolation and characterization of cDNA molecules e ncoding SNAP-25 A and B isoforms, syntaxin 4 and synaptobrevins (also known as vehicle-associated membrane proteins) from two major human in sulin-responsive tissues, skeletal muscle and fat. The DNA and deduced amino acid sequences of SNAP-25 revealed perfect identity with the pr eviously reported human neural SNAP-25 A and B isoforms. Our results i ndicate the presence of both isoforms both in insulin-responsive tissu es and in in vitro cultured 3T3-L1 cells, but suggest a differential p attern of gene expression: isoform A is the major species in adipose t issue, and isoform B is the major species in skeletal muscle. The pres ence of SNAP-25 protein in 3T3-L1 cells was demonstrated by immunofluo rescence microscopy using an anti-SNAP-25 monoclonal antibody. Immuno- precipitation experiments using the same monoclonal antibody also reve aled the presence of SNAP-25 protein in plasma membrane fractions from rat epididymal fat pads. The syntaxin 4-encoding region from skeletal muscle contains five nucleotide differences from the previously repor ted placental cDNA sequence, two of which result in amino acid changes : Asp-174 to Glu and Val-269 to Ala. The synaptobrevin 1 cDNA from ske letal muscle contains two nucleotide differences when compared with th e corresponding clone from neural tissues, one of which is silent and the other resulting in the amino acid change Thr-102 to Ala. The cDNA sequence of the protein from fat is identical with that of human synap tobrevin 1 from neural tissues. Furthermore, we have confirmed the pre sence of syntaxin 4 in fat and of synaptobrevin 2 in skeletal muscle b y PCR amplification and Southern hybridization analysis. Using the yea st two-hybrid system, an interaction was observed between the full-len gth cytoplasmic domains of syntaxin 4 and synaptobrevin 2, a vesicle m embrane SNARE previously shown by others to be associated with vesicle s carrying the GLUT4 glucose transporter protein, but no interaction w as seen with synaptobrevin 1. Flow cytometry of low-density microsomes isolated from fat cells was used to demonstrate the binding of syntax in 4 to a subset of vesicles carrying GLUT4 protein; whereas SNAP-25 o n its own bound poorly to these vesicles, the syntaxin 4-SNAP-25 compl ex gave a strong interaction.