The high affinity membrane transport system for ribose in Escherichia coli is a member of a large family of membrane transporters found in all forms of living organisms. Homologous systems in humans include the cystic fibrosis related transmembrane conductance regulator (CFTR), multiple drug resistance (MDR) proteins, and the TAP proteins which load foreign antigens onto MHC Class I antigen-presenting molecules. Membrane transporters of this family have at least one, usually two, ATP-binding domains of about 250 amino acids having quite hydrophilic character. These ATP-binding cassettes (ABC) are clearly homologous from bacteria to humans, implying a common protein tertiary structure and similarity of mechanisms. They are associated with extremely hydrophobic domains of 250 to 300 amino acids, either as parts of the same protein or as separate subunits of a multimeric complex. The CFTR and MDR proteins each have two ABC's and two hydrophobic domains, producing proteins of 1300 to 1500 amino acids. In contrast, the ribose transporter of E. coli has an ATP-binding protein, RbsA, which has only two ABC's in a 501 amino acid protein (thus creating a protein with homologous N- and C-terminal halves). The hydrophobic part of the transporter, RbsC, is encoded on a separate gene. This simplifies production and purification of the constituents of the transporter.
We have produced and purified RbsA and RbsC in milligram quantities for structural analyses. We have also produced the N-terminal half molecule of RbsA, which constitutes a single ABC, and that molecule has produced large crystals. The structure has been determined by X-ray diffraction analysis, the first structure of an ABC from any member of this family to be solved. We are screening conditions for crystallizing the whole RbsA molecule and also RbsC. No suitable crystals for X-ray analysis have been obtained yet.
We have successfully reconstituted the transport system in lipid vesicles using the purified RbsA and RbsC and have trapped the whole transport complex using vanadate to inhibit the ATPase activity at the transition state. Mechanistic studies are being performed using these systems, and we are attempting to grow crystals of the whole complex.
In a related study, the repressor that regulates expression of the ribose transport system, RbsR, has been purified and crystallized in complex with an 18 base pair operator DNA fragment. The X-ray structure of this complex has been determined and will be compared to the structures of PurR and LacI, homologous repressor proteins. Dr. Howard Zalkin, of this department, and Dr. Richard Brennan, Oregon State University, solved the structure of holo-PurR (with hypoxanthine bound) in complex with its operator DNA and also apo-PurR corepressor-binding domain (lacking the DNA-binding domain). The structure of RbsR will be informative, since the apo-form of RbsR binds DNA most strongly, the opposite mechanism from PurR. Thus, comparison of the two structures will allow us to deduce the features which contribute to stabilizing the binding to the DNA.