A resonant mirror biosensor was used to study cyclic
nucleotide–receptor interactions. In particular, a
novel method was developed to determine inhibition constants
(Ki) from initial rates of ligate association
to immobilized ligand. This approach was applied to the
comparison of cyclic nucleotide-binding properties of the
wild-type isolated B domain of the cAMP-dependent protein
kinase type Iα regulatory subunit and its Ala-334-Thr
(A334T) variant that has altered cyclic nucleotide specificity.
A cUMP-saturated form of the B domain was used for all
measurements. Under the conditions used, cUMP did not affect
the kinetics of B domain association to immobilized cAMP.
Triton X-100 was required to stabilize the protein at nanomolar
concentrations. The association and dissociation rate constants
for wild-type and A334T B domains yielded equilibrium dissociation
constants of 11 and 16 nM. Heterogeneity of ligate and immobilized
ligand, mass transport effects, and other factors were evaluated for
their influence on biosensor-determined kinetic constants.
Biosensor-determined relative inhibition constants
(K′i =
KicAMP/Kianalog)
for 16 cyclic nucleotide analogs correlated well with those determined
by a [3H]cAMP binding assay. Previously published
K′i values for the B domain in the
intact regulatory subunit were similar to those of the isolated B domain.
The K′i values for the wild-type and
A334T B domains were essentially unchanged except for dramatic enhancements
in affinity of cGMP analogs for the A334T B domain. These observations
validate the isolated B domain as a simple model system for studying
cyclic nucleotide–receptor interactions.