The absorption and steady-state fluorescence of
N-acetyl-l-tyrosinamide (NAYA), chosen to model
the tyrosine
in proteins, was measured in four solvents in the presence of
N,N-dimethylacetamide, N-methylacetamide,
and urea, chosen as ligands that model the amide of a protein backbone.
The data were fit to a multilinear
mathematical model to resolve the overlapping spectra of NAYA with and
without ligand. The ligand binding
constant was between 0.18 and 4 M-1,
increasing with solvent polarity; lack of a strong dependence of
binding
constant on N-methylation suggests that the carbonyl oxygen
is responsible for the hydrogen bonding. The
emission spectrum of NAYA was essentially identical in all solvents and
with and without added ligand. In
contrast, the excitation spectrum shifted by up to 10 nm, depending on
both solvent and ligand; this shift is
described as a sum of three terms: a blue-shift due to hydrogen
bonding to the phenolic oxygen which is
proportional to ligand donor acidity, a red-shift due to hydrogen
bonding to the phenolic hydrogen which is
proportional to ligand acceptor basicity, and a blue-shift proportional
to solvent dielectric effect. The extinction
coefficient varies by up to 40%, depending on solvent and complex
formation. The fluorescence quantum
yield of the hydrogen-bonded complex varies between 0.102 and 0.044,
increasing slightly with N-methylation,
but more dependent on solvent. In methanol, acetone, and dioxane,
the amide-model complex has a 3−10
times lower fluorescence quantum yield than that of NAYA in pure
solvent; in water the complex has a
higher quantum yield. Complex formation did not explain all of the
fluorescence quenching by ligand in
water and in dioxane, suggesting that the ligand also causes dynamic
quenching of the excited state in these
solvents. Many of the experimental findings are in good agreement
with semiempirical molecular orbital
calculations.