We have measured the Raman spectra of oxidized nicotinamide adenine dinucleotide, NAD+, and its reduced form, NADH, as well as a series of fragments and analogues of NAD+ and NADH. In addition, we have studied the effects of pH as well as deuteration of the exchangeable protons on the Raman spectra of these molecules. In comparing the positions and intensities of the peaks in the fragment and analogue spectra with those of NADH and NAD+, we find that it is useful to consider these large molecules as consisting of component parts, namely, adenine, two ribose groups, two phosphate groups, and nicotinamide, for the purpose of assigning their spectral features. The Raman bands of NADH and NAD+ are found generally to arise from molecular motions in one or another of these molecular moieties, although some peaks are not quite so easily identified in this way. This type of assignment is the first step in a detailed understanding of the Raman spectra of NAD+ and NADH. This is needed to understand the binding properties of NADH and NAD+ acting as coenzymes with the NAD-linked dehydrogenases as deduced recently by using Raman spectroscopy.

We report the Raman spectra of reduced and oxidized nicotinamide adenine dinucleotide (NADH and NAD+, respectively) and adenosine 5'-diphosphate ribose (ADPR) when bound to the coenzyme site of liver alcohol dehydrogenase (LADH). The bound NADH spectrum is calculated by taking the classical Raman difference spectrum of the binary complex, LADH/NADH, with that of LADH. We have investigated how the bound NADH spectrum is affected when the ternary complexes with inhibitors are formed with dimethyl sulfoxide (Me2SO) or isobutyramide (IBA), i.e., LADH/NADH/Me2SO or LADH/NADH/IBA. Similarly, the difference spectra of LADH/NAD+/pyrazole or LADH/ADPR with LADH are calculated. The magnitude of these difference spectra is on the order of a few percent of the protein Raman spectrum. We report and discuss the experimental configuration and control procedures we use in reliably calculating such small difference signals. These sensitive difference techniques could be applied to a large number of problems where the classical Raman spectrum of a "small" molecule, like adenine, bound to the active site of a protein is of interest. The spectrum of bound ADPR allows an assignment of the bands of the bound NADH and NAD+ spectra to normal coordinates located primarily on either the nicotinamide or the adenine moiety. By comparing the spectra of the bound coenzymes with model compound data and through the use of deuterated compounds, we confirm and characterize how the adenine moiety is involved in coenzyme binding and discuss the validity of the suggestion that the adenine ring is protonated upon binding. The nicotinamide moiety of NADH shows significant molecular changes upon binding.(ABSTRACT TRUNCATED AT 250 WORDS)

We have measured the frequency of the carbon-hydrogen stretching mode of the pro-R and pro-S C4-H bonds of NADH in solution and when bound to pig heart lactate (LDH) or mitochondrial malate (mMDH) dehydrogenases. This is achieved by specifically deuterating the C4 pro-R or pro-S hydrogens of NADH and determining the frequencies of the resulting C4-D stretches by Raman difference spectroscopy. We find that the frequencies of the two C4-D stretching modes for the two bonds are essentially the same for the unliganded coenzyme. On the other hand, the position of the pro-S-[4-2H]NADH stretch shifts upward by about 23-30 cm-1 in its binary complex with either lactate or malate dehydrogenase relative to that observed in solution, while that for the bound pro-R-[4-2H]NADH is relatively unchanged. The fact that the frequency of the pro-R hydrogen is not significantly affected during complex formation suggests that the rate enhancements for reaction of substrate with NADH brought about by both pig heart LDH and mMDH apparently do not involve either stabilization or destabilization of the pro-R hydrogen of NADH in enzyme-coenzyme binary complexes, in agreement with previous chemical studies. That these proteins are able to regulate the frequencies of the two C4-D bonds differentially, and hence the electronic distributions in these bonds, has important implications for the stereochemical reactions catalyzed by the NAD dehydrogenases, and this is discussed.(ABSTRACT TRUNCATED AT 250 WORDS)

The binding of the coenzymes NAD+ and NADH to lactate dehydrogenase causes significant changes in the Raman spectra of both of these molecules relative to spectra obtained in the absence of enzyme. The molecular motions of the bound adenine moiety of both NAD+ and NADH as well as adenine containing analogues of these coenzymes produce Raman bands that are essentially identical, suggesting that the binding of adenine to the enzyme is the same regardless of the nicotinamide head-group nature. We also have observed that the molecular motions of the bound adenine moiety are different from both those obtained when it is in either water, various hydrophobic solvents, or various other solvent compositions. Protonation of the bound adenine ring at the 3-position is offered as a possible explanation. Significant shifts are observed in both the stretching frequency of the carboxamide carbonyl of NAD+ and the rocking motion of the carboxamide NH2 group of NADH. These shifts are probably caused by hydrogen bonding with the enzyme. The interaction energies of these hydrogen-bonding patterns are discussed. The aromatic nature of the nicotinamide moiety of NAD+ appears to be unchanged upon binding. Pronounced changes in the Raman spectrum of the nicotinamide moiety of NADH are observed upon binding; some of these changes are understood and discussed. Finally, these results are compared to analogous results that were recently reported for liver alcohol dehydrogenase [Chen et al. (1987) Biochemistry 26, 4776-4784]. In general, the coenzyme binding properties are found to be quite similar, but not identical, for the two enzymes.

No abstract