HIGH-AFFINITY BINDING OF ESCHERICHIA-COLI SECB TO THE SIGNAL SEQUENCEREGION OF A PRESECRETORY PROTEIN

The Escherichia coil cytosolic homotetrameric protein SecB is known to be involved in protein export across the plasma membrane. A currently prevalent view holds that SecB functions exclusively as a chaperone i nteracting nonspecifically with unfolded proteins; not necessarily exp orted proteins, whereas a contrary view holds that SecB functions prim arily as a specific signal-recognition factor-i.e., in binding to the signal sequence region of exported proteins. To experimentally resolve these differences we assayed for binding between chemically pure SecB and chemically pure precursor (p) form (containing a signal sequence) and mature (m) form (lacking a signal sequence) of a model secretory protein (maltose binding protein, MBP) that was C-terminally truncated . Because of the C-terminal truncation, neither p nor m was able to fo ld. We found that SecB bound with 100-fold higher affinity to p (K-d 0 .8 nM) than it bound to m (K-d 80 nM). As the presence of the signal s equence in p is the only feature that distinguished p from m, these da ta strongly suggest that the high-affinity binding of SecB is to the s ignal sequence region and not the mature region of p. Consistent with this conclusion, we found that a wild-type signal peptide, but not an export-incompetent mutant signal peptide of another exported protein ( LamB), competed for binding to p. Moreover, the high-affinity binding of SecB to p was resistant to 1 M salt, whereas the low-affinity bindi ng of SecB to m was not. These qualitative differences suggested that SecB binding to m was primarily by electrostatic interactions, whereas SecB binding to p was primarily via hydrophobic interactions, presuma bly with the hydrophobic core of the signal sequence. Taken together o ur data strongly support the notion that SecB is primarily a specific signal-recognition factor.