The transmembrane topology of the <i>sn</i>‐glycerol‐3‐phosphate permease of <i>Escherichia coli</i> analysed by <i>phoA</i> and <i>lacZ</i> protein fusions

The Escherichia coli glpT gene encodes a transport protein that mediates uptake of sn‐glycerol‐3‐phosphate, a member of a class of bacterial organophosphate permeases which transport substrates by antiport with inorganic phosphate, and a two‐dimensional model for the permease is tested.

Abstract

The Escherichia coli glpT gene encodes a transport protein that mediates uptake of sn-glycerol-3-phosphate. This permease is a member of a class of bacterial organophosphate permeases which transport substrates by antiport with inorganic phosphate. The glpT gene product, probably an oligomer of a single polypeptide chain, is thought to span the cytoplasmic membrane several times, as predicted by the hydropathic profile. Protein fusions, in which varying lengths of the amino-terminal end of the permease is attached to alkaline phosphatase (phoA) and to beta-galactosidase (lacZ) were constructed. On the assumption that phoA fusions only exhibit high enzymatic activity when fused to extra-cytoplasmic regions of the target protein, whereas lacZ fusions will only be active when the beta-galactosidase portion is attached to cytoplasmic domains of the target protein, the activities of the fusions were used to test a two-dimensional model for the permease. The model proposes that GlpT contains 12 transmembrane segments divided by a larger cytoplasmic region. Despite some limitation caused by hot-spot sites of transpositions, the TnphoA approach was consistent with the model. In contrast, we feel that the enzymatic activity of lacZ fusions is only a limited parameter for studying the topology of a complex membrane protein.