In this study we demonstrate the potential of combining fluorine-19 nuclear magnetic resonance (NMR) spectroscopy with molecular genetics. We are using the membrane-bound enzyme D-lactate dehydrogenase of Escherichia coli as a model system to characterize interactions between proteins and lipids. We have labeled D-lactate dehydrogenase with 4-, 5-, and 6-fluorotryptophans and obtained high-resolution fluorine-19 NMR spectra showing five resonances, in agreement with the five tryptophan residues expected from the DNA sequence. The five 19F resonances in the spectra have been assigned to the specific tryptophan residues in the primary sequence of D-lactate dehydrogenase by site-directed oligonucleotide mutagenesis of the cloned gene. We observe large differences in the relative fluorine-19 chemical shifts of each tryptophan residue when labeled by different isomers of fluorotryptophan. We have determined by NMR methods that two tryptophans are exposed to the solvent and that none of the tryptophan residues are within 10 A of the lipid phase. On the basis of 19F NMR spectroscopy of the labeled tryptophan residues, the conformation of D-lactate dehydrogenase is similar in aqueous solution and in the presence of a variety of lipids and detergents. This result indicates that the presence of lipids or detergents is not required to maintain the tertiary structure of this membrane-bound enzyme. In contrast, Triton X-100 induces a change to an abnormal conformation of the enzyme as judged from both NMR spectroscopy and the effect of temperature on the maximal velocity of the enzyme in the presence of this detergent.

No abstract

The structure and function of the membrane-bound D-lactate dehydrogenase of Escherichia coli have been investigated by fluorine-19 nuclear magnetic resonance spectroscopy of 5-fluorotryptophan-labeled enzyme in conjunction with oligonucleotide-directed, site-specific mutagenesis. 5-Fluorotryptophan has been substituted for nine phenylalanine, tyrosine, and leucine residues in the enzyme molecule without loss of activity. The 19F signals from these additional tryptophan residues have been used as markers for sensitivity to substrate, exposure to aqueous solvent, and proximity to a lipid-bound spin-label. The nuclear magnetic resonance data show that two mutational sites, at amino acid residues 340 and 361, are near the lipid environment used to stabilize the enzyme. There are a number of amino acid residues on the carboxyl side of this region that are strongly sensitive to the aqueous solvent. The environment of the wild-type tryptophan residue at position 469 changes as a result of two of the substitution mutations, suggesting some amino acid residue-residue interactions. Secondary structure prediction methods indicate a possible binding site for the flavin adenine dinucleotide cofactor in the carboxyl end of the enzyme molecule. These results suggest that the membrane-bound D-lactate dehydrogenase may have the two-domain structure of many cytoplasmic dehydrogenases but with the addition of a membrane-binding domain between the catalytic and cofactor-binding domains. This type of three-domain structure may be of general significance for understanding the structure of membrane-bound proteins which do not traverse the lipid bilayer of membranes.

A tryptophan-requiring strain of Escherichia coli can go through two doublings of optical density after L-tryptophan is replaced in the medium by 4-fluorotryptophan, during which the fluoro analog displaces approximately 75% of the L-tryptophan in cell protein. One doubling occurs in the presence of 5- or 6-fluorotryptophan, with 50-60% replacement of L-tryptophan by analog. When beta-galactosidase is induced at the time of addition of analog, it reaches 60% of the control specific activity in the presence of 4-fluorotryptophan, 10% of normal in the presence of 5- or 6-fluorotryptophan. Lactose permease activity is 35% of the control in the presence of 4- and 6-fluorotryptophan, less than 10% in the presence of 5-fluorotryptophan. D-Lactate dehydrogenase shows a specific activity twice that of the control in the presence of 4-fluorotryptophan, one-half with 5- or 6-fluorotryptophan. Thus fluorotryptophan can be incorporated into proteins and affect their activities, although the nature and magnitude of the effect cannot be predicted for any given enzyme. Such substituted proteins should be useful for the study of protein structure and function by 19F nuclear magnetic resonance and other techniques.