BIOCHEMICAL SOCIETY TRANSACTIONS Substrate-binding to lactate dehydrogenase perturbs an intersubunit equilibrium sensed by engineered tryptophan-248 ADAM D. B. WALDMAN*, DAVID A. BARSTOW?, ANTHONY A. CLARKE*, ENRICO GRATTONS, TONY ATKINSON? and J. JOHN HOLBROOK* *Department of Biochemistry, University of Bristol Medical School, Briston BS8 1 T D , U . K . ; 1-P.H.L.S. Centre f o r Applied Microbiology and Research, Microbial Technology Laboratory, Salisbury SP4 OJG, U.K., and $Department of Physics, I l l 0 West Green Street, Urbana, I L 61801, U.S.A. The time-resolved fluorescence properties of tryptophan residues in proteins are very sensitive to small changes in local environment (Beechem & Brand, 1985). However, this potentially useful optical probe of protein internal mobility has not been widely used because most proteins have multiple tryptophan residues (average 3.6) with overlapping spectral bands. To circumvent this ambiguity the techniques of molecular genetics were used to construct a lactate dehydro- genase gene which lacks tryptophan codons (Waldman et al., 1987), and site-directed mutagenesis of this gene was used to reintroduce a single tryptophan at a specific site within the known primary and three-dimensional structure at which information on environment changes is sought. The method has revealed domain re-arrangements in the 0.1 ms timescale during recognition of the substrate pyruvate and the allosteric effector fructose 1,6-bisphosphate (Atkinson et al., 1987), and has partly characterized stable intermediates during the re-folding of this protein from guanidinium hydrochloride (Clarke et al., 1987). Molecular recognition depends upon fast internal motions in proteins and we report here the nanosecond proper- ties of a single-tryptophan construct based on Bacillus stearothermophilus lactate dehydrogenase, in which wild- type tyrosine-248 is replaced by tryptophan (and the three natural tryptophans at 80, 150 and 203 by tyrosine). The 248-substitution was directed by the mismatch of a 23mer oligonucleotide to the whole gene between the Hind111 and PstI sites, the mutant protein over-produced by pLDH41 in Escherichia coli TG2 cells and purified by affinity chromatography on oxamate-Agarose from pLDH41 in exactly the same manner as for other mutants (Clarke et al., 1986). The 248-site is normally tryptophan in other lactate dehydrogenases and is well placed to sense structural re-arrangements: being (1) close to the very stable Q-axis intersubunit contact and (2) just two residues away from isoleucine-250 which forms the hydrophobic surface which is tightly covered when the dihydropyridine ring of NADH binds in the active centre. Fig. 1 shows the modulation ratios and phase differences of the fluorescence of single-tryptophan-248 dehydrogenase measured with the multi-frequency phase fluorometer described by Alcala et al. (1987). In the apo-tetrameric protein (stabilized by fructose 1,6-bisphosphate) the single tryptophan fluorescence is analysed as two logarithmic decays (82% 5.0ns, 18% 1.411s) and in the very stable quaternary complex formed between the protein, regulator, NAD+ and oxalate, the proportions are reversed (21% 4.8 ns, 79% 0.61 ns). This frequency-domain result has been w s a Modulation frequency (MHz) Fig. 1 . Frequency-domain measurements qffluorescence lifetimes q~single-tryptophan-248 B. stearothermophilus lactate dehydrogenase Measured phase difference ( + ) and modulation ratio ( x ) of IOpM-enzyme solution in 0.1 M-triethanolamine HCI buffer, pH 6 and IOmM-fructose 1,6-bisphosphate at 25°C as a function of modulation frequency is compared to the predicted continuous curves calculated from the lifetime and pre-exponential factors given in the text. The results and pair of curves at the higher frequencies are from a solution containing in addition 0. I mM-NAD' and 1 mM-Na oxalate.