The observation of structural transitions of a single protein molecule.

Coherent neutron scattering measurements of an amorphous, in vivo deuterated C-phycocyanin are compared with a calculation of the individual protein molecule's coherent static structure factor. Both show the significant features associated with known structure factors of several amorphous materials, most notably, an unusually sharp first diffraction peak occurring near 1.4 A(-1). We show that in the protein, such a peak results from the product of a form factor associated with correlations of atoms within individual amino acids and a structural term expressing inter-amino-acid correlations. The measurement, interpreted through behavior of the first diffraction peak, indicates that inter-amino-acid correlations - a measure of the protein's medium-range structure - undergo transitions which are primarily related to hydration rather than to temperature.

Spatial arrangement of sigma-factor and core enzyme of Escherichia coli RNA polymerase. A neutron solution scattering study.

By means of neutron solution scattering we determined the position and orientation of core enzyme and sigma-factor within the Escherichia coli RNA polymerase holoenzyme with the aim of improving existing models. The individual components, core enzyme (E) and sigma-factor (sigma), were highlighted by deuterium labeling and their center-to-center distances determined in the monomeric and the dimeric holoenzyme. The following distance parameters were obtained: dE1-sigma 1 = 8.6(+/- 1) nm, dE1-E2 = 11.5(+/- 1) nm, d sigma 1-sigma 2 = 12.0(+/- 0.7) nm, dE1-sigma 2 = 9(+/- 3) nm. Using a triangulation procedure the position of the sigma-factors, sigma 1 and sigma 2, were determined with respect to the mass center of the core enzyme molecules, E1 and E2, assuming a symmetrical arrangement of the holoenzyme molecules in the dimer (C2 symmetry). In addition, the orientation of the sigma-factor with respect to core enzyme was estimated by means of model calculations. The obtained model of holoenzyme depicts the sigma-factor as buried in a groove of core enzyme, probably between the large subunits beta' and beta.

Low-frequency collective modes in dry and hydrated proteins.

We have observed Brillouin-like low frequency collective modes in the scattering of 1 A neutrons from a fully in vivo deuterated protein. These modes are tentatively interpreted as due to short-lived coherent excitations propagating with velocities between 2,000 and 4,000 m/s in elements of the secondary structure and patches of closely associated water.

Spatial arrangement of DNA-dependent RNA polymerase of Escherichia coli and DNA in the specific complex. A neutron small angle scattering study.

In this paper we demonstrate that neutron small angle scattering is a suitable method to study the spatial arrangement of large specific protein-DNA complexes. We studied the complex of DNA-dependent RNA polymerase of Escherichia coli and a 130 base-pair DNA fragment containing the strong promoter A1 of bacteriophage T7. Contrast variation of the complex with deuterium allowed us to "visualize" either RNA polymerase, or DNA, or both components in situ. From the corresponding scattering curves information was derived about: (1) Conformational changes of RNA polymerase and DNA by complex formation: comparison of the scattering profiles of the isolated and complexed components showed that by specific complex formation the cross-section of RNA polymerase decreases, while the DNA fragment does not undergo a gross conformational change. (2) The spatial arrangement of RNA polymerase and DNA in the specific complex from the cross-sectional radii of gyration of the complex the normal distance dn between the centre of gravity of the RNA polymerase and the axis of the DNA fragment was derived as 5.0 (+/- 0.3) nm. On the basis of these and footprinting data a low resolution model of the RNA polymerase-promoter complex is proposed. The main feature of this model is the positioning of RNA polymerase to only one side of the DNA.

NMR studies of a complex of deuterated calmodulin with melittin.

Completely deuterated calmodulin ([2H]CaM) has been prepared by expressing the chicken gene for CaM in Escherichia coli grown in 2H2O on a deuterated medium. The structural and dynamic properties of a 1:1 CaM/melittin (Mel) complex have been investigated by proton NMR. The spectrum of bound Mel is obtained directly from the spectrum of the [2H]CaM X Mel complex and is found to resemble strongly the spectrum of the helical species in methanol rather than that of the random coil species in water. The spectrum of bound CaM is obtained indirectly from the difference spectrum between [1H]CaM X Mel and [2H]CaM X Mel. Many changes are observed between free and bound CaM and they are distributed in both halves of the molecule, indicating that the binding of Mel affects the structure in both parts of the molecule. The rates of exchange of the amide protons of [2H]CaM with 2H2O were compared to those of [2H]CaM X Mel. The results showed that most, but not all, of the protons exchanged more slowly in the complex; after 40 hr, the residual peaks number 7 in CaM and greater than 20 in the complex. Again, changes in rates in CaM due to binding of Mel occurred in both halves of the molecule. The relative rates of amide proton exchange in CaM and its complex with Mel prove to be a sensitive criterion of differences in conformational stability and/or structure.

Deuterium incorporation into Escherichia coli proteins. A neutron-scattering study of DNA-dependent RNA polymerase.

Neutron small-angle scattering studies of single protein subunits in a protein-DNA complex require the adjustment of the neutron scattering-length densities of protein and DNA, which is attainable by specific deuteration of the protein. The neutron scattering densities of unlabelled DNA and DNA-dependent RNA polymerase of Escherichia coli match when RNA polymerase is isolated from cells grown in a medium containing 46% D2O and unlabelled glucose as carbon source. Their contrasts vanish simultaneously in a dialysis buffer containing 65% D2O. An expression was evaluated which allows the calculation of the degree of deuteration and match point of any E. coli protein from the D2O content of the growth medium, taking the 2H incorporation into RNA polymerase amino acids to be representative for all amino acids in E. coli proteins. The small-angle scattering results, on which the calculation of the degree of deuteration is based, were confirmed by mass spectrometric measurements.

Retinal location in purple membrane of Halobacterium halobium: a neutron diffraction study of membranes labelled in vivo with deuterated retinal.

Purple membranes were prepared by growing Halobacterium halobium in a medium containing nicotine (which inhibits biosynthesis of retinal) and the oxidation products of fully deuterated beta-carotene. This allowed the in vivo incorporation of deuterated retinal into the membranes. The labelled membranes were crystalline and isomorphous with native membrane as determined by X-ray diffraction, and their optical absorption spectra were very similar. Neutron diffraction data for the two dimensional in-plane lattice from labelled and native membranes were analysed by difference Fourier and direct methods to 8.6 A resolution. The difference Fourier shows the retinal to be located in the centre of the bacteriorhodopsin molecule. The best fit to the data was obtained with the projection of retinal as a 10 A long rod forming an angle of -40 degrees +/- 10 degrees with the x axis centred at x = -0.19 +/- 0.02, y = -0.35 +/- 0.02 in fractional unit cell coordinates. The main peak in the difference Fourier map is at x = -0.17, y = -0.33.

Growth of Lactobacillus casei in deuterated media. Purification of deuterated thymidylate synthase and proton nuclear magnetic resonance studies.

The purification of thymidylate synthase from amethopterin-resistant Lactobacillus casei grown in a deuterated medium is described. The deuterated enzyme and the non-deuterated enzyme appear to be identical with respect to specific activity, pH optimum, electrophoretic mobility on polyacrylamide gels and ability to form ternary complexes with 5-fluoro-2'-deoxyuridine 5'-monophosphate (FdUMP) and 5,10-methylenetetrahydrofolate. The deuterated enzyme remained stable at 25 degrees C for 8 h during the acquisition of 400 MHz 1H-NMR spectra and was stored at 5 degrees C for 3 months without loss of catalytic activity. The incorporation of deuterium was essentially complete as demonstrated by a comparison of the NMR spectra of deuterated and non-deuterated enzyme. Proton NMR data obtained with deuterated thymidylate synthase and 1H-FdUMP are in agreement with the fluorine-19 NMR studies of Lewis et al. (Lewis, C.A., Jr., Ellis, P.D. and Dunlap, R.B. (1980) Biochemistry 19, 116-123) which support the formation of a binary complex where the enzyme is covalently linked to carbon 6 of 5-fluoro-5,6-dihydro-2'-deoxyuridylate.

Resonance Raman spectroscopy of chemically modified and isotopically labelled purple membranes. I. A critical examination of the carbon-nitrogen vibrational modes.

Resonance Raman spectra of bacteriorhodopsin are compared to the spectra of this protein modified in the following ways: (1) selective deuteration at the C-15 carbon atom of retinal, (2) full deuteration of the retinal, (3) the addition of a conjugated double bond in the beta-ionone ring (3-dehydroretinal), (4) full deuteration of the protein and lipid components, (5) 15N enrichment of the entire membrane and (6) deuteration of the entire membrane (including the retinal). A detailed comparison of the 15N-enriched membrane and naturally occurring purple membrane from 800 cm-1 to 1700 cm-1 reveals that 15N enrichment affects the frequency of only two vibrational modes. These occur at 1642 cm-1 and 1620 cm-1 in naturally occurring purple membrane and at 1628 cm-1 and 1615 cm-1 in the 15N-enriched samples. Therefore, this pair of bands reflects the states of protonation of the Schiff base. However, our data also indicate that neither of these modes are simple, localized C=N-H or C=N stretching vibrations. In the case of the 1642 cm-1 band motions of the retinal chain beyond C-15 are not significantly involved. On the other hand, in the 1620 cm-1 band atomic motions in the isoprenoid chain beyond C-15 are involved.

Resonance Raman spectroscopy of chemically modified and isotopically labelled purple membranes. II. Kinetic studies.

Kinetic resonance Raman spectra of native and isotopically labelled purple membranes are compared. Using these data and the assignments of the previous paper in this sequence, we have confirmed that the Schiff base is deprotonated at times that are short in comparison to M412 evolution. In addition, by monitoring the kinetic resonance Raman spectra in 2H2O with 488.0 nm excitation we have been able to characterize in more detail the vibrational features associated with this unprotonated intermediate that precedes M412. Furthermore, the kinetic spectra of fully deuterated purple membrane in H2O have allowed us to assign the 1465 cm-1 band in these spectra to the C=C stretching frequency of BR570 and the 1512 cm-1 band to the C=C stretching frequency of M412. These spectra have also provided an indication of a Raman spectral feature associated with O640 and, finally, our kinetic spectra have provided evidence that there is a significant alteration in the rate constants for the evolution of the various intermediates when the non-exchangeable protons on the membrane are replaced by deuterons.

The subunit positions within RNA polymerase holoenzyme determined by triangulation of centre-to-centre distances.

The complete 'centre-of-subunit structure' of the multisubunit enzyme DNA-dependent RNA polymerase was determined by triangulation of the subunit positions using the intersubunit distances calculated from scattering difference measurements and from the corresponding radii of gyration R. In addition to the centre-to-centre distances d between the core subunits alpha 2, beta and beta' presented in the preceding paper, the values of d between initiation factor sigma and alpha 2 (8.4 +/- 1.6 nm), beta (4.4 +/- 2.2 nm) and beta' (10.7 +/- 1.5 nm) were derived from R of sigma (4.1 +/- 0.3 nm) in situ and of the pairs alpha 2--sigma (6.1 +/- 0.4 nm), beta--sigma (5.6 +/- 0.3 nm) and beta'--sigma (7.5 +/- 0.4 nm) within the holoenzyme (alpha 2 beta beta' sigma). The structural parameters of the subunits within their molecular complex are accessible for neutron small-angle scattering measurements using labelling of the different subunits (deuteration), total reconstitution of isotopic hybrids, scattering length density matching of 'hydrogenated' molecular parts and extended exposure times because of weak scattering effects. The overall shape of sigma bound to core enzyme (alpha 2 beta beta') proved to be identical (within experimental resolution) with sigma in the isolated state measured recently by X-ray small-angle scattering. The refined shape of isolated sigma was reduced to an ellipsoid which was orientated with respect to the core structure (alpha 2--beta--beta') in a 'space-filling' way around the position of the sigma centre obtained by triangulation. The complete subunit arrangement of holoenzyme is shown in a three-dimensional model.

Location of the chromophore in bacteriorhodopsin.

We present a location for the retinylidene chromophore in dark-adapted bacteriorhodopsin based on the differences in neutron scattering between purple membrane preparations reconstituted with retinal and with deuterated retinal. The Fourier difference density map contains more peaks than expected, and additional arguments are introduced to exclude artificial peaks, caused by the reconstitution techniques or the limited resolution of the diffraction data. The membrane preparation used is necessarily dark-adapted and therefore contains 13-cis- and all-trans-retinal isomers in roughly equal amounts. However, we find only a single position for both isomers. Presumably, the difference in conformation caused by isomerization around the C13-C14 double bond is minimized by rotation around other bonds. The retinal is located between alpha-helical segments of the protein and its nearest neighbor (intratrimer) distance is 26 A; the next-nearest neighbor (intertrimer) distance is 38 A.

Neutron scattering measurements with the label triangulation method on the 50 S subunit of E. coli ribosomes.

In the E. coli 50 S ribosomal subunit, proteins L7/L12 and L10 were deuterated by partial reconstitution. The distance between L7/L12 and L10 was measured by the label triangulation method and was found to be approximately 100 A or, with low probability, 60 to 70 A, depending on the concentration.

Biosynthesis of deuterated riboflavin: structure determination by NMR and mass spectrometry.

The riboflavin-producing fungus Eremothecium ashbyii was cultured in various growth media containing high concentrations of deuteriuj, and the product was isolated. The structures of highly deuterated riboflavin, in which at least 13 of 15 nonexchangeable hydrogens were replaced by deuterium, and fully deuterated riboflavin, in which all 15 nonexchangeable sites contained deuterium, were established by NMR and mass spectrometry. The aromatic protons (C-5 and C-8) wer partially substituted in the highly deuterated molecule. Information regarding three areas of the biosynthetic pathway within the microorganism was obtained as a result of the formation of these compounds. Extensive solvent interaction, possibly due to passage of sugar through the transaldolase-transketolase pathway, occurs during formation of the ribityl chain. Limited solvent participation takes place during formation of 6,7-dimethyl-8-ribityllumazine, the immediate precursor of riboflavin. Deuteration of the riboflavin C-6 and C-7 methyl groups indicates significant solvent exchange during the final step of the biosynthetic process.

On the nature of the iron sulfur cluster in a deuterated algal ferredoxin.

A protonated and a completely deuterated two-iron algal ferredoxin from Synechococcus lividus have been studied by optical, electron paramagnetic resonance, electron-nuclear double resonance, proton magnetic resonance and Mossbauer spectroscopies; temperature dependent magnetic susceptibility measurements are reported as well. These studies have confirmed the electron localized model of the active center in the two-iron ferredoxins, as previously deduced from studies of spinach ferredoxin, have yielded much more precise spectroscopic parameters for this center, and have thus greatly increased the confidence in this model.