The effect of salt stress on
glycine betaine-binding activity has been investigated in periplasmic fractions released from Rhizobium meliloti 102F34 by cold osmotic shock. Binding activity was monitored by three techniques: equilibrium dialysis, filter procedure, and detection of 14C
ligand-protein binding by direct non-denaturing
polyacrylamide gel electrophoresis (PAGE) followed by autoradiography. The three methods demonstrated the existence of a strong
glycine betaine-binding activity, but only in periplasmic fractions from cells grown at high osmolarity. The non-denaturing PAGE of such periplasmic
shock fluids mixed with [methyl-14C]
glycine betaine showed only one radioactive band, indicating the involvement of one
glycine betaine-
binding protein. To determine the possible implication of this
binding protein in
glycine betaine uptake, transport activity was measured with cells submitted to cold osmotic shock. No significant decrease of transport activity was noticed. This lack of effect could be explained by the small quantity of
periplasmic proteins released as judged by the low activity of
phosphodiesterase, a periplasmic marker
enzyme, observed in the
shock fluid. The specificity of binding was analysed with different potential competitors: other betaines such as
gamma-butyrobetaine,
proline betaine, pipecolate
betaine,
trigonelline and
homarine, or
amino acids like
glycine and
proline, did not bind to the
glycine betaine-
binding protein, whereas
glycine betaine aldehyde and
choline were weak competitors. Optimum pH for binding was around 7.0, but approx. 90% of the
glycine betaine-binding activity remained at pH 6.0 or 8.0. The calculated binding affinity (KD) was 2.5 microM. Both
glycine betaine-binding activity and affinity were not significantly modified whether or not the binding assays were done at high osmolarity. A 32 kDa osmotically inducible periplasmic
protein, identified by SDS-PAGE, apparently corresponds to the
glycine betaine-
binding protein.